Advances in Biomedical Research
– selected topics
Advances in Biomedical Research
– selected topics
Editors:
Łukasz Biały
Izabela Młynarczuk-Biały
Lublin 2018
Reviewers:
dr hab. n. med. Elżbieta Rębas, prof. nadzw.
dr hab. Marcin Grąz
dr hab. Roman Paduch
dr n. o zdr. Mariola Janiszewska
dr Maciej Frant
dr n. med. Paweł Kiciński
dr n. o zdr. Magdalena Skórzewska
All published chapters received positive reviews.
Typesetting:
Monika Maciąg
Design:
Marcin Szklarczyk
© Copyright by Wydawnictwo Naukowe TYGIEL sp. z o.o.
ISBN 978-83-65932-55-6
Publisher:
Wydawnictwo Naukowe TYGIEL sp. z o.o.
ul. Głowackiego 35/341, 20-060 Lublin
www.wydawnictwo-tygiel.pl
Table of Contents
Introduction .......................................................................................................................... 7
Application of biosensors in cancer research ...................................................................... 9
Detection of newly synthesized proteins in cancer cells: A practical approach ............... 22
Direct visualization of undegraded protein sequestration within aggresomes
by proteasome inhibitors ................................................................................................... 36
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis,
and Biological Screening ................................................................................................... 48
Immunohistochemical expression of erythropoietin in invasive breast carcinoma
with metastasis to lymph nodes ........................................................................................ 62
The inflammatory components in lung cancer .................................................................. 80
Normalisation of microRNA qPCR results in lung cancer – strategies and problems .... 92
Statistical comparison of deleteriousness prediction methods for single nucleotide
variants from next generation sequencing ....................................................................... 101
Amnion as a potential source of neural and glial cell precursors – from cell culture to
a practical application ...................................................................................................... 109
What makes the bacterial vaginosis difficult to diagnose: our experience in Silesia,
Poland .............................................................................................................................. 119
Detection of sleepiness in narcolepsy .............................................................................. 130
Increased incidence of CNS tumours in Parkinson Disease .......................................... 139
Role of vitamin D in etiology of mood disorders – literature review ............................ 147
Index of authors ............................................................................................................... 154
Introduction
The monograph you hold is the result of the meetings and hard work of young
scientists from various universities and institutes in Poland. These meetings have
been held in Warsaw from 9 years under the auspices of the Department of Histology
and Embryology of the Medical University of Warsaw and for the last two years also
the Warsaw Branch of the Polish Society for Histochemistry and Cytochemistry.
In the monograph we present the most important topics in biomedical research from
2017. All presented works have passed the peer-review process positively. These
works come from various fields of biomedicine from medical biology throughout
biophysics to clinical sciences.
Wishing you enjoyable and productive reading
Editors,
Łukasz Biały
Izabela Młynarczuk-Biały
9
Application of biosensors in cancer research
Anna Sobiepanek*, Tomasz Kobiela
Faculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-664
Warsaw Poland
*Corresponding author: Anna Sobiepanek, Faculty of Chemistry, Warsaw University
of Technology, [email protected]
Abstract
Cancer research is based mainly on identifying many features of cells including mutations in DNA
or RNA, changes in proteins and protein levels as well as on comparing cells’ properties like
morphology, adhesion or elasticity.
The most interesting techniques for biological applications, among different microscopies, are those
using biosensors. As an analytical device, biosensor enables the binding of the selected analyte to the
examined biological material (e.g. tissue, cells, proteins, RNA, DNA). They can be classified according
to their transducer type or the biorecognition elements placed on the sensor. Most analysis performed on
biosensors require the labeling of the analyte with a specific marker. However, there are also many
techniques that allow a direct detection of analytes without prior labeling. The common “label-free”
biosensor technologies are the quartz crystal microbalance (QCM) and the surface plasmon resonance
(SPR).
In this review we would like to focus on the currently described examples of biosensors used
as diagnostic and prognostic tools for cancer development.
Introduction
The first person to use the term ‘biosensor’ was the Professor of Analytical
Chemistry Karl Cammann in 1977. However, the definition of biosensor was not
specified by the International Union of Pure and Applied Chemistry (IUPAC) until
1997. It was designed to combine three areas of science (chemistry, biology and
engineering) in one device to detect bioanalytical molecules from samples (Islam,
Uddin, 2017).
For the construction of a biosensor two components are unavoidable: the transducer
and the biorecognition element placed on the surface of the sensor. Transducer
transforms the biochemical response appearing on the sensor surface to a measurable
output signal. Some commonly used types of transducers are calorimetric (thermal),
electrochemical (amperometric, impedimetric, potentiometric), magnetic (electro-
magnetic, electrodynamic and piezomagnetic), mass sensitive (acoustic, piezo-
electric) and optical (colorimetric, photometric) based systems (Bora, Sett et al.
2013, Lenk, Ballas et al. 2011, Zhang, Yang et al.2013). In table 1 different
techniques using transducer types are gathered. On the other hand, the biorecognition
element (bioreceptor) placed on the sensor ensures capturing the matching analyte
from the solution sample. They may be classified as elements with biocatalytic
properties (enzymes/substrate) or component with specific affinities (antibodies/
antigens, nucleic acids, receptors/ligands, cells/tissues). The detection of some
metabolic or biological components is also possible (Islam, Uddin, 2017, Pihíková,
Kasák et al. 2015, Keshavarz, Behpour et al. 2015).
Anna Sobiepanek, Tomasz Kobiela
10
Table 1. A set of example methods used in bioanalytical techniques T
ech
niq
ues
Calorimetric thermal biosensors
enzyme thermistor (ET)-based
sensors
(Perumal,
Hashim, 2014)
metal oxide thermistors-based
sensors
ceramic semiconductor
thermopile–based sensors
Electrochemical
amperometric first-/second-/third-generation
biosensors
(Borgmann,
Schulte et al.
2011)
voltammetric
cyclic voltammetry (CV)
(Pihíková,
Kasák et al.
2015)
differential pulse voltammetry
(DPV)
square-wave voltammetry
(SWV)
potentiometric ion-selective electrodes (ISE) (Perumal,
Hashim, 2014) ion-sensitive field
Magnetic
magnetic
nanoparticles (MNPs);
magnetic beads
(MBs);
semiconductor
quantum dots (QDs)
magnetoresistance (MR)-based
biosensors
(Llandro,
Palfreyman et
al. 2010)
anisotropic magnetoresistance
(AMR)-based biosensors
giant magnetoresistance
(GMR)-based biosensors
Mass sensitive
bulk acoustic wave
(BAW) devices
shear horizontal acoustic plate
mode (SH-APM) sensor
(Durmuş, Lin
et al. 2008)
quartz crystal microbalance
(QCM)
quartz crystal microbalance in
dissipation mode (QCM-D)
surface acoustic wave
sensors (SAW)
capacitive micromachined
ultrasonic transducers
(CMUTs)
microcantilevers microelectronic mechanical
systems (MEMs)
(Prasad,
Shameem,
2016)
Optical interferometric
changes
resonant mirror (RM)
(Fang, 2006) resonant waveguide grating
(RWG)
surface plasmons resonance
(SPR)
(Zhang, Yang
et al. 2013)
Application of biosensors in cancer research
11
surface plasmons resonance
imaging (SPRi)
fluorescence
fluorescence resonance energy
transfer (FRET)
fluorescent semiconductor-
based sensors
chemiluminescence /
luminescence
luminescent semiconductor-
based sensors
Biosensor measurement methodology determines the type of detection: label-free or
non-label free. Label-free detection is based on binding the original and unmodified
analyte molecule directly to the biorecognition element, whereas in some methods
(amperometric, voltammetric or fluorescent experiments) only analyte molecules
tagged with a label may be recognized by the biorecognition element to obtain an
electroactive signal (Fritz, 2008). Some commonly used labels are fluorophores
(especially for fluorescent microscopy), enzymes (often for Western blot analysis)
and several nanoparticles (essential for magnetic resonance imaging; MRI). For
medical applications the usage of magnetic nanoparticles (MNPs), magnetic beads
(MB), semiconductor quantum dots (QDs) with combined biosensor techniques is
increasingly justified. Metal nanoparticles (mainly gold or silver) have a significant
affinity for cancer cells, that is why they are frequently used in cancer research
(Llandro, Palfreyman et al. 2010, Keshavarz, Behpour et al. 2015). Specific markers
are often introduced into the tested compound using chemical synthesis or genetic
engineering methods. However, the labeling process may require additional sample
preparation or must be followed by a second molecule binding. Unfortunately, the
attachment of the label may significantly alter the properties of the tested molecule;
substances used as markers may attach to other molecules than the target, and when
using living cells, they may interfere with their metabolism. Considering all the
above, label-free methods gain much more attention (ErtuğruL, Uygun, 2013).
Nowadays, popular label-free methods are quartz crystal microbalance (QCM),
surface plasmon resonance (SPR) and microelectronic mechanical (MEM)
cantilevers, where cantilever sensors emerged from the atomic force microscopy
(AFM) (Fritz, 2008) (Fig. 1). These techniques allow tracking and determining the
kinetic/ thermodynamic analysis of the interaction process of two complemental
molecules in real time, where one molecule is immobilized on the surface and the
second one is in flow. Moreover, QCM and MEM-cantilever use changes in resonant
frequency to observe the mass shifts on the sensor (Sobiepanek, Milner-Krawczyk et
al. 2017, Wang, Wang et al. 2014), but SPR utilizes changes in the refractive index
of thin metal layers (like gold surfaces) to quantify the binding process of
biomolecule to the sensor surface (Ahmed, Wiley et al. 2010).
Anna Sobiepanek, Tomasz Kobiela
12
Figure 1. Schematic illustration of the three selected label-free methods: SPR, QCM and AFM
microcantilever (own elaboration)
The development of reliable and sensitive molecular electronic devices is a matter of
great importance in the field of biotechnology, medicine and pharmacy (Pihíková,
Kasák et al. 2015). Especially in case of diagnostics, biosensors may significantly
facilitate and accelerate early detection of several diseases like the Parkinson and
Alzheimer disease, diabetes or various types of cancer. The monitoring of clinical
treatment may also be performed via the usage of properly designed biosensors
(Parolo, Merkoçi, 2013). Body fluids such as urine, blood, saliva, tears or sweat may
be treated also as samples full of specific disease biomarkers for micro/
nanotechnology-based techniques (SPR, QCM, microcantilevers), even though their
concentration in these samples is often very low (Pasinszki, Krebsz et al. 2017,
Wang, Wang et al. 2014). These methods are non-destructive for samples in
comparison to others like Western blot, mass spectrometry (MS), matrix-assisted
laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) or
electrospray ionization mass spectrometry (ESI-MS) (Ahmed, Wiley et al. 2010). On
the other hand, isolation of the selected biomarkers from patient’s tissue or cells (Fig.
2) may simplify their utilization as analytes on a biomolecular-based biosensor
(DNA-, RNA-, aptamer-, protein-, antibody-based biosensors). Identifying mutations
in DNA/RNA or uncovering changes in proteins and protein levels are common
diagnostic strategies for cancers, though they require expensive and time-consuming
tests performed in well-equipped laboratories, as well as good samples quality and an
appropriate sample size. The most popular microarrays for these analyses are the
quantitative polymerase chain reaction (qPCR) and enzyme-linked immunosorbent
assay (ELISA). Nonetheless, experiments executed on whole cells or even tissues are
far more reliable because they reflect the real conditions prevailing in the human
body (Islam, Uddin, 2017). Studying cancerous cell properties with advanced
Application of biosensors in cancer research
13
nanotechnological methods can bring many important biophysical information about
these cells and may improve the knowledge about metastasis. Optical and fluorescent
microscopies are often used for the examination of cell morphology, as well as the
topography imaging with atomic force microscopy. What is more, AFM may be
applied for studies concerning cell elasticity and adhesion, properties that change
during the cancer progression, and even molecular interactions (Sobiepanek, Milner-
Krawczyk et al. 2017). For the adhesion and molecular interaction QCM
investigations can also be adjusted, not only with the molecular-based biosensors, but
also with immobilized cells or even tissues on sensors. Yet, those cell-/tissue-based
biosensors have to be properly prepared (Perumal, Hashim, 2014).
Figure 2. Schematic illustration of various biomarkers as targets for cancer detection with the biosensors
methodology (own elaboration)
New insight
Biosensors may be applied successfully in the medical field, especially in cancer
research. Due to the rising number of cancer cases each year all over the world,
investigations concerning early cancer detection becomes a vital matter. Also, the
possibility of cancer treatment monitoring with biosensor techniques gives hope for
a personalized therapy. That is why providing a more sensitive and simplified
method at lower costs, which brings even more information about the basis of the
disease, is still desired (Abu-Salah, Zourob et al. 2015). For these advanced studies
some special molecular-based or cell-/tissue-based biosensors have already been
developed and will be briefly characterized in the following sections.
Nucleic acid-based (NABs) biosensors
Some typical NABs are deoxyribonucleic acids (DNA), ribonucleic acids (RNA),
peptide nucleic acids (PNA) and aptamers. During the investigation the nucleic acid
should be immobilized on the surface of the sensor by adsorption, biotin-avidin
interaction, covalent bonding, entrapment in a polymer matrix, ionic interaction or
Anna Sobiepanek, Tomasz Kobiela
14
self-assembly (Abu-Salah, Zourob et al. 2015). The most frequently used
immobilization method includes the application of thiolated-NABs for creating
a self-assembly monolayer on a gold sensor surface. The single-stranded DNA and
RNA sequences could anneal to the immobilized complementary sequences and the
occurring interaction depends on the molecules. For DNA and RNA sequences
binding, the Chargaff's rules of base pairing is fulfilled (DNA: A=T, C≡G; RNA:
A=U, C≡G). On these bases, related to cancer mutations in DNA or RNA may be
revealed (Bora, Sett et al. 2013). Very interesting molecules for cell research are
microRNAs (miRNAs), which are naturally existing small non-coding ribonucleic
acids (RNA). They play a significant role in cell development (proliferation, cell
cycle progression, apoptosis) and is related to a number of cancer cases. miRNA may
be extracted from cells or tissues, however, their amount in the cancer cells differs
from that of normal cells (Kilic, Topkaya et al. 2012, Zhang, Chua et al. 2009).
PNAs are synthetic DNA or RNA analogues (sugar-phosphate backbone is replaced
by pseudo-peptide backbone) that bind to their complementary strands with higher
specificity and strength. Aptamers may be classified into two groups: DNA- or RNA-
aptamers (short oligonucleotides) and peptide-aptamers (short peptide domains), but
their detection is more similar to antigen-antibody or receptor-ligand interactions.
They can be easily modified or integrated with a variety of nanomaterials (Bora, Sett
et al. 2013, Sohrabi, Valizadeh et al. 2016).
DNA-DNA binding: Breast cancer is associated with various gene mutations like
breast cancer 1 (BRCA1). Its detection in the concentration range of 10 and 100
μM is possible due to the designed electrochemical biosensor. Short
oligonucleotide DNA was immobilized onto zinc oxide nanowires that was
chemically synthesized onto gold electrode via the hydrothermal technique. The
hybridization of ssDNA was studied by the differential pulse voltammetry
(DPV) (Mansor, Zain et al. 2014).
RNA-RNA binding: The detection of mir21 from the total RNA breast cancer
samples was carried out on a selective and sensitive enzyme-based
electrochemical biosensor. mir21 was covalently attached onto the pencil
graphite electrode (PGE) by coupling agents and the hybridization was achieved
with a biotinylated complementary target. Next an avidin labeled alkaline
phosphatase was introduced to the system for obtaining the biotin-avidin
interaction. Through the enzymatic conversion of the reaction substrate alpha
naphtol phosphate to the reaction product alpha naphtol (α-NAP) the oxidation
signal was detected by Differential Pulse Voltammetry (DPV) (Kilic, Topkaya
et al. 2012).
PNA-RNA binding: In order to detect let-7b in the total RNA extracts from HeLa
cells (human epithelial cervical cancer) via base pairing, the silicon nanowire
field-effect transistors (SiNW-FETs) with immobilized complementary PNA
were used. With the optimized assay the detection limit of 1 fM could be
obtained (Zhang, Chua et al. 2009).
Application of biosensors in cancer research
15
DNA aptamer-antigen binding: One of the well-known biomarkers for prostate
cancer is the prostate-specific antigen (PSA) present in blood samples. The
functionalization of a gold sensor with thiolated-DNA aptamer enabled the
detection of PSA with the quartz crystal microbalance in dissipation mode
(QCM-D) with the affinity constant equal 37 nM. These experiments provided
not only information about the amount of PSA bound to the sensor, but also
information about the aptamer conformation and layer hydration (Formisano,
Jolly et al. 2015).
DNA aptamer-cells-nanoparticles binding: Blood cancer is a very aggressive
type of cancer disease. Leukemia cells may be selectively captured by special
DNA aptamers immobilized on the QCM sensor and the signal may be
increased after the application of gold nanoparticles (AuNPs) on the already
attached cells (Shan, Pan et al. 2014).
The biomolecule-based biosensors
The noncovalent, purely physicochemical binding forces are involved in many
specific interactions like antigen (Ag)-antibody (Ab), enzyme-substrate, lectin-
carbohydrate and ligand-receptor (van Oss, 2000). These interactions may be based
on hydrogen bonds, van der Waals forces and hydrophobic interactions (Coelho,
Silva et al. 2017). To be more precise, some conventional biochemical methods are
the basis of these interactions, especially ELISA or immunofluorescence assay on
Ag-Ab specific binding (Zhou, Wang et al. 2011). However, these methods require
several steps of preparation (proper sample/antibody concentration, washing stages,
sample labeling). Label-free techniques based on biomolecular biosensors (for
example, antibody-/lectin-/protein-based) may not only accelerate the interaction
measurement, but also provide full analysis of the interaction with the affinity
calculation for the examined compound.
Antibody-antigen binding: Three liver cancer antigens: alpha-fetoprotein (AFP),
hepatocyte growth factor (HGF) and gamma-glutamyltransferase-2 (GGT-2);
were detected with high specificity as well as good precision by the novel
microcantilever with immobilized antibodies (Wang, Wang et al. 2014). Also,
the p53 antibody accumulates in human serum for many types of cancer
including breast, lung, prostate, ovarian and melanoma. A quantitative detection
of p53 antibody ranging from 20 ng/ml to 20 μg/ml was obtained for human
serum samples with p53 antigen-coated microcantilever (Zhou, Wang et al.
2011). Moreover, three independent SiNW-FET devices were designed, on
which different antibodies were immobilized for the detection in pg/ml scale of
PSA, carcinoembryonic antigen (CEA) and mucin-1 from blood samples (Chen,
Li et al. 2011).
Antibody-protein binding: Vimentin protein indicates the presence of
Osteosarcoma, a common type of bone cancer. Immobilized anti-vimentin
antibody on the surface of the MEM-cantilever was successfully used in the
early detection of this cancer (Balwir, Sahare et al. 2016). At the same time,
collagen type IV (COLIV) occurs in serum samples of patients with colorectal,
Anna Sobiepanek, Tomasz Kobiela
16
gastric, lung, liver and breast cancers. The immobilization of anti-COLIV
antibody to the Surface Plasmon Resonance imaging (SPRi) gold sensor gave
a dynamic response in molecular binding in the range between 10 and 300 ng/ml
for COLIV (Sankiewicz, Lukaszewski et al. 2016). The laminin-5 protein
functions as a motility factor or as an adhesive factor, depending on the
proteolytic processing state. During the interactions with several cell-surface
receptors it may promote tumor invasion. An antibody-based SPRi biosensor
was used to determine the laminin-5 concentration in blood plasma with the
detection limit 4 pg/ml (Sankiewicz, Romanowicz et al. 2016).
Lectin-carbohydrate binding: Binding kinetics of lectin-carbohydrate
interactions gains attention, due to the fact that cancer cells during their
progression change their glycosylation profile what could be a possible drug
target. Two mannose specific lectins (Lens culinaris and Concanavalin A, Con
A) were immobilized onto gold QCM-D sensors separately via thiol groups and
next carboxypeptidase Y was introduced in the buffer solution. The analysis of
the lectin to carbohydrate affinity may serve as a quick biomarker classification
assay in cancer research (Senkara-Barwijuk, Kobiela et al. 2012). Furthermore,
an interesting application of lectin-based sensors was achieved for cells in
suspension. The modification of QCM sensor with Con A induced the binding
of human leukemia cell line, which then was followed by the attachment of the
second lectin on top of the cells. This approach may lead to the development of
a novel label-free suspension cell-based biosensor (Li, Pei et al. 2013).
Protein-DNA binding: With the SPR method it was also possible to semi-
quantitatively detect the UV-irradiated DNA sequence obtained from human
cell extracts. The DNA sequence was biotinylated and captured onto
a streptavidin-coated sensor chip (Ahmed, Wiley et al. 2010). Protein-DNA
binding can also be investigated by the DNA functionalized (SiNW-FET)
biosensor. The estrogen receptor alpha (ERα, protein) regulates gene expression
by the direct attachment to estrogen receptor sequences (ERE, dsDNA)
immobilized on the sensor, which may be used for detecting these protein-DNA
interactions in nuclear extracts from breast cancer cells. The designed biosensor
was capable of detecting ERα in the 10 fM concentration (Zhang, Huang et al.
2011).
Cell-/tissue-based biosensors
Methods in which you could apply whole cells deliver information which is more
congenial than the techniques using samples isolated from cells or tissues (for
example, DNA/RNA, proteins). This is due to the isolation process, which might
damage or change the conformation of the biomolecules, influence the concentration
of the required analyte or the stability of the sample. What is more, measurements
performed on whole cell-based biosensors may simulate processes taking place in
physiological conditions. Cells seeded onto sensors may be assigned for experiments
with living cells (for compound absorption tests) or fixed cells (like compound
binding ability tests) (Min, Yea et al. 2009). Cells in vitro are typically cultured on
Application of biosensors in cancer research
17
polystyrene or glass surfaces, often on precoated surfaces with proteins from the
extracellular matrix (ECM) like fibronectin, collagen, laminin, vitronectin; but rarely
on metals such as titanium, zirconium or gold (Depprich, Ommerborn et al. 2008,
Ryan, 2008). Bare gold sensors, common for QCM and SPR methods, are not quite
desirable surfaces for cell culture experiments due to non-specific adsorption of
many biological molecules like DNA and proteins that may be used as analytes (Tan,
Lin et al. 2014). That is why gold sensors are frequently coated with polystyrene or
silicon dioxide. Nevertheless, surface modification plays an important role in the up
or down regulation of physiological processes of cells such as adhesion, proliferation
and differentiation (Wang, Ren et al. 2013). Monitoring of the cell attachment and
spreading may be crucial for designing devices that control the behavior of living
cells. Simultaneously, it could bring a more detail knowledge about the metastasis
process. Adhesion of cells is a complex process, which begins with cell
sedimentation followed by the formation of nonspecific cell–substrate interactions
and next the establishment of specific molecular binding of the receptor–ligand type.
With time cell adhesion to the surface alters due to the ongoing physiological
processes. To measure the cell-type-specific interactions, QCM-D technique can be
applied (Iturri, García-Fernández et al. 2015, Wegener, Janshoff et al. 2001). Some
examples of cell-/tissue-based biosensor applications are:
Compound absorption tests on cells: The combination of whole cell sensing and
real-time label-free monitoring of nanoparticle uptake by cells can be obtained
by means of the SPR technique. The uptake kinetic of selected nanoparticles has
already been tested on HeLa cells in the μg/mL concentrations. However, this
process is temperature-dependent: for about 20 °C the uptake is higher, whereas
for 37 °C it is lower (Suutari, Silen et al. 2016).
Compound binding ability tests on cells: Two stages of human colorectal cancer
cells were derived from the same patient (primary and metastases), seeded onto
a gold QCM sensor coated with polystyrene and lectin–carbohydrate interaction
was measured with lectin Helix pomatia agglutinin (HPA). At the end a higher
affinity of HPA to metastatic cells was obtained (Peiris, Markiv et al. 2012).
Also, the glycosylation level of melanocytes and melanoma cells (cultured on
QCM-D gold sensors coated with polystyrene) was investigated by lectin Con
A. The study revealed that mannose and glucose types of oligosaccharides
present on metastatic melanoma cells consists of long and branched structures,
whereas primary tumor cells and normal cells have short and less ramified
oligosaccharides. Furthermore, the affinity of Con A to oligosaccharides on
metastatic melanoma cells was ten times higher than for primary tumor cells and
melanocytes (Sobiepanek, Milner-Krawczyk et al. 2017). Cell-based biosensors
may be also used for cancer drug tests. Antibody-conjugated drug Herceptin
detects the human epidermal growth factor receptor 2 (HER2) protein, which is
over-expressed in 25–30% of breast cancers. It induces a cytostatic effect
associated with the G1 phase cell cycle arrest, as well as antibody dependent
cell-mediated cytotoxity (Peiris, Spector et al. 2017). On the other hand, the G
protein-coupled receptors (GPCRs) are crucial drug targets that may be
Anna Sobiepanek, Tomasz Kobiela
18
activated by histamine, respectively. The SPR study has shown a triphasic
response of HeLa cells to histamine interaction: 1 – GPCRs triggered calcium
release, 2 – alternations of cell-matrix adhesion after the activation of Protein
Kinase C, 3 – dynamic mass redistribution in cells (Lu, Yang et al. 2016).
Surprisingly, up till now only few tissue-based biosensors have been described.
Tonsil, prostate and breast tumor specimens were obtained and immobilized on
the surface of gold QCM sensor. Next the interaction between the rVAR2
protein and placental-like chondroitin sulfate present on most cancer cells was
analyzed and the calculated affinity was in the nanomolar range (Clausen,
Pereira et al. 2016).
Cell adhesion tests: These strategies may be applied for the characterization of
cell membrane receptors activity in cancer cells and for the search of other cell-
specific ligands. For example, cell attachment to the surface is controlled mainly
by the cell transmembrane integrin receptor that binds to Arg-Gly-Asp (RDG)
sequence. The QCM-D sensor was modified with a photo-activatable RGD
peptide for determining the time point of presentation of adhesive ligand from
the human umbilical vein endothelial cells (HUVEC) (Iturri, García-Fernández
et al. 2015). Also, with the use of a novel high-throughput label-free resonant
waveguide grating (RWG) imager the HeLa cells spreading kinetics on the
ligand RGD tripeptide was determined (Orgovan, Peter et al. 2014). On the
other hand, vitronectin protein- as well as antibody (CA-125)-based QCM
biosensors were used for the binding of the suspended melanoma, cervix and
ovarian cancer cells (Ishay, Kapp-Barnea et al. 2015).
Review and discussion
Cancer could develop at a very rapid pace, therefore the simplicity of measurements,
quickness of the test and low costs are in request from the potential new methods that
are to be applied. For this reason, biosensor techniques, especially those with label-
free detection, have gained massive attention recently. Their main assumption is the
specific interaction occurring between the biorecognition element and the selected
analyte. What is important, some of these measurements (like those performed on
QCM-D device) can be made on living or fixed cells and may deliver kinetic and
thermodynamic analysis of the obtained interaction, as well as the information about
the affinity, conformation of the created complex and even viscoelastic properties of
the new appearing biomolecular surface. A wide variety of biosensors is available
among the transducer type and the biorecognition element alike. This is why
biosensors may have a versatile application.
Acknowledgements
This work was supported by the Faculty of Chemistry, Warsaw University of
Technology and by the National Science Centre (Poland), project no
2017/27/N/ST4/01389.
Application of biosensors in cancer research
19
References
Abu-Salah K.M. Zourob M.M., Mouffouk F., Alrokayan S.A., Alaamery M.A., Ansari A.A. DNA-Based
Nanobiosensors as an Emerging Platform for Detection of Disease. Sensors. 2015; 15:14539-14568.
Ahmed F.E., Wiley J.E., Weidner D.A., Bonnerup C., Mota H. Surface Plasmon Resonance (SPR)
Spectrometry as a Tool to Analyze Nucleic Acid–Protein Interactions in Crude Cellular Extracts. Cancer Genomics & Proteomics. 2010; 7:303-310.
Balwir S.P., Sahare M., Raghuwanshi A., Angarwar S., Fulzele R. Design of cantilever based
biosensor for bone cancer detection. International Journal of Research in Engineering and Technology. 2016; 5(special issue 10)NCRTET-2016:22-25.
Bora U., Sett A., Singh D. Nucleic Acid Based Biosensors for Clinical Applications. Biosensors
Journal. 2013; 2(1)104.
Borgmann S., Schulte A., Neugebauer S., Schuhmann W. Amperometric Biosensors. In: Alkire R.C.,
Kolb D.M., Lipkowski J. (Editors) Advances in Electrochemical Science and Engineering. 2011;
WILEY-VCH Verlag GmbH & Co. KGaA (Weinheim); ISBN: 978-3-527-32885-7.
Chen K.I., Li B.R., Chen Y.T. Silicon nanowire field-effect transistor-based biosensors for
biomedical diagnosis and cellular recording investigation. Nano Today. 2011; 6:131-154.
Clausen T.M., Pereira M.A., Oo H.Z., Resende M., Gustavson T., Mao Y., Sugiura N., Liew J., Fazli L.,
Theander T.G., Daugaard M., Salanti A. Real-time and label free determination of ligand binding-
kinetics to primary cancer tissue specimens; a novel tool for the assessment of biomarker
targeting. Sensing and Bio-Sensing Research. 2016; 9:23-30.
Coelho L.C., Silva P.M., Lima V.L., Pontual E.V., Paiva P.M., Napoleão T.H., Correia M.T. Lectins,
Interconnecting Proteins with Biotechnological/ Pharmacological and Therapeutic Applications.
Evidence-Based Complementary and Alternative Medicine. 2017; 2017:1594074.
Depprich R., Ommerborn M., Zipprich H., Naujoks C., Handschel J., Wiesmann H.P., Kübler N.R.,
Meyer U. Behavior of osteoblastic cells cultured on titanium and structured zirconia surfaces.
Head & Face Medicine. 2008; 4:29.
Durmuş N.G., Lin R.L., Kozberg M., Dermici D., Khademhosseini A., Demirci U. Acoustic-Based
Biosensors. In: Li D. (Editor) Encyclopedia of Microfluidics and Nanofluidics. 2008; Springer
(Boston); ISBN: 978-0-387-48998-8.
ErtuğruL H.D., Uygun Z.O. Impedimetric Biosensors for Label-Free and Enzymless Detection. In:
Rinken T. (Editor) State of the Art in Biosensors - General Aspects. 2013; InTech (Rjeka); ISBN:
978-953-51-1004-0.
Fang Y. Label-Free Cell-Based Assays with Optical Biosensors in Drug Discovery. Assay and Drug
Development Technologies. 2006; 4(5):583-595.
Formisano N., Jolly P., Bhalla N., Cromhout M., Flanagan S.P., Fogel R., Limson J.L., Estrela P.
Optimisation of an Electrochemical Impedance Spectroscopy aptasensor by exploiting Quartz Crystal Microbalance with Dissipation Signals. Sensors and Actuators B. 2015; 220:369-375.
Fritz J. Cantilever biosensors. The Analyst. 2008; 133(7):855-863.
Ishay R.B., Kapp-Barnea Y., Grigoriantz I., Teblum E., Lellouche J.P. Real time acoustic profiling of
a live cancerous cell monolayer using QCM. Sensors and Actuators B: Chemical. 2015; 215:373-381.
Islam M.T., Uddin M.A. Biosensors, the Emerging Tools in the Identification and Detection of
Cancer Markers. Journal of Gynecology and Women’s Health. 2017; 5(4): 555667.
Iturri J., García-Fernández L., Reuning U., García A.J., del Campo A., Salierno M.J. Synchronized cell
attachment triggered by photo-activatable adhesive ligands allows QCM-based detection of early integrin binding. Scientific reports. 2015; 5:9533.
Anna Sobiepanek, Tomasz Kobiela
20
Keshavarz M., Mohsen Behpour M., Rafiee-pour H.A. Recent trends in electrochemical microRNA
biosensors for early detection of cancer. RSC Advances. 2015; 5:35651.
Kilic T., Topkaya S.N., Ariksoysal D.O., Ozsoz M., Ballar P., Erac Y., Gozen O. Electrochemical
based detection of microRNA, mir21 in breast cancer cells. Biosensors and Bioelectronics. 2012;
38:195-201.
Lenk A., Ballas R.G., Werthschützky R., Pfeifer G. Magnetic Transducers. In: Lenk A., G. Ballas
R.G., Werthschützky R., Pfeifer G. (Editors) Electromechanical Systems in Microtechnology and
Mechatronics. 2011; Springer (Berlin); ISBN:978-3-642-10806-8.
Li X., Pei Y., Zhang R., Shuai Q., Wang F., Aastrup T., Pei Z. A suspension-cell biosensor for real-
time determination of binding kinetics of protein–carbohydrate interactions on cancer cell
surfaces. Chemical Communications 2013; 49(85):9877-10032.
Llandro J., Palfreyman J.J., Ionescu A., Barnes C. H. W. Magnetic biosensor technologies for medical
applications: a review. Medical & Biological Engineering & Computing. 2010; 48:977-998.
Lu J., Yang Y., Wang W., Li J., Tao N., Wang S. Label-Free Imaging of Histamine Mediated G
Protein-Coupled Receptors Activation in Live Cells. Analytical Chemistry (ACS Publications). 2016;
88(23):11498-11503.
Mansor N.A., Zain Z.M., Hamzah H.H., Noorden M.S.A., Jaapar S.S., Beni V., Ibupoto Z.H. Detection
of Breast Cancer 1 (BRCA1) Gene Using an Electrochemical DNA Biosensor Based on
Immobilized ZnO Nanowires. Open Journal of Applied Biosensor. 2014; 3:9-17.
Min J., Yea C.H., El-Said W.A., Choi J.W. The Application of Cell Based Biosensor and Biochip for
Environmental Monitoring. In: Kim Y.J., Platt U., Gu M.B., Iwahashi H. (Editors) Atmospheric and
Biological Environmental Monitoring. 2009; Springer (Dordrecht Heidelberg London New York);
ISBN: 978-1-4020-9673-0.
Orgovan N., Peter B., Bősze S., Ramsden J.J., Szabó B., Horvath R. Dependence of cancer cell
adhesion kinetics on integrin ligand surface density measured by a high-throughput label-free resonant waveguide grating biosensor. Scientific report. 2014; 4:4034.
Parolo C., Merkoçi A. Paper-based nanobiosensors for diagnostics. Chemical Society Reviews. 2013;
42:450-457.
Pasinszki T., Krebsz M., Tung T.T., Losic D. Carbon Nanomaterial Based Biosensors for Non-
Invasive Detection of Cancer and Disease Biomarkers for Clinical Diagnosis. Sensors. 2017;
17:1919.
Peiris D., Markiv A., Curley G.P., Dwek M.V. A novel approach to determining the affinity of
protein–carbohydrate interactions employing adherent cancer cells grown on a biosensor surface. Biosensors and Bioelectronics. 2012; 35:160-166.
Peiris D., Spector A.F., Lomax-Browne H., Azimi T., Ramesh B., Loizidou M., Welch H., Dwek M.V.
Cellular glycosylation affects Herceptin binding and sensitivity of breast cancer cells to
doxorubicin and growth factors. Scientific reports. 2017; 7:43006.
Perumal V., Hashim U. Advances in biosensors: principle, architecture and applications. Journal of
Applied Biomedicine. 2014; 12(1):1-15.
Pihíková D., Kasák P., Tkac J. Glycoprofiling of cancer biomarkers: Label-free electrochemical
lectin-based biosensors. Open Chemistry. 2015; 13:636-655.
Prasad M.G.G.J., Shameem S. Design and Analysis of Micro-Cantilever Based Biosensor for Swine
Flu Detection. International Journal of Electrical and Computer Engineering. 2016; 6(3):1190-1196.
Ryan J.A. Evolution of Cell Culture Surfaces. BioFiles 2008; 3:8-11.
Application of biosensors in cancer research
21
Sankiewicz A., Lukaszewski Z., Trojanowska K., Gorodkiewicz E. Determination of collagen type IV
by Surface Plasmon Resonance Imaging using a specific biosensor. Analytical Biochemistry. 2016;
515:40-46.
Sankiewicz A., Romanowicz L., Laudanski P., Zelazowska-Rutkowska B., Puzan B., Cylwik B.,
Gorodkiewicz E. SPR imaging biosensor for determination of laminin-5 as a potential cancer
marker in biological material. Analytical and Bioanalytical Chemistry. 2016; 408:5269-5276.
Senkara-Barwijuk E., Kobiela T., Lebed K., Lekka M. Reaction pathway and free energy profile
determined for specific recognition of oligosaccharide moiety of carboxypeptidase Y. Biosensors
and Bioelectronics. 2012; 36:103-109.
Shan W., Pan Y., Fang H., Guo M., Nie Z., Huang Y., Yao S. An aptamer-based quartz crystal
microbalance biosensor for sensitive and selective detection of leukemia cells using silver-enhanced gold nanoparticle label. Talanta. 2014; 126:130-135.
Sobiepanek A., Milner-Krawczyk M., Lekka M., Kobiela T. AFM and QCM-D as tools for the
distinction of melanoma cells with a different metastatic potential. Biosensors and Bioelectronics.
2017; 93:274-281.
Sohrabi N., Valizadeh A., Farkhani S.M., Akbarzadeh A. Basics of DNA biosensors and cancer
diagnosis. Artificial Cells, Nanomedicine, and Biotechnology. 2016; 44(2)654-663.
Suutari T., Silen T., S En Karaman D., Saari H., Desai D., Kerkelä E., Laitinen S., Hanzlikova M.,
Rosenholm J.M., Yliperttula M., Viitala T. Real-Time Label-Free Monitoring of Nanoparticle Cell
Uptake. Small. 2016; 12(45):6289-6300.
Tan L., Lin P., Pezeshkian B., Rehman A., Madlambayan G., Zeng X. Real-Time Monitoring of Cell
Mechanical Changes Induced by Endothelial Cell Activation and their Subsequent Binding with
Leukemic Cell Lines. Biosensors and Bioelectronics. 2014; 56:151-158.
van Oss C.J. Nature of Specific Ligand-Receptor Bonds, in Particular the Antigen-Antibody Bond.
Journal of Immunoassay. 2000; 21(2&3)109-142.
Wang J.L., Ren K.F., Chang H., Jia F., Li B.C., Ji Y., Ji J. Direct Adhesion of Endothelial Cells to
Bioinspired Poly(dopamine) Coating Through Endogenous Fibronectin and Integrin α5β1. Macromolecular Bioscience. 2013; 13(4):483-493.
Wang J., Wang S., Wang X., Zhu Y., Yang J., Yang F. Cantilever Array Sensor for Multiple Liver
Cancer Biomarkers Detection. SENSORS 2014 IEEE. 2014.
Wegener J., Janshoff A., Steinem C. The Quartz Crystal Microbalance as a Novel Means to Study
Cell–Substrate Interactions In Situ. Cell Biochemistry and Biophysics. 2001; 34:121-151.
Zhang G.J., Chua J.H., Chee R.E., Agarwal A., Wong S.M. Label-free direct detection of MiRNAs
with silicon nanowire biosensors. Biosensors and Bioelectronics. 2009; 24:2504-2508.
Zhang G.J., Huang M.J., Ang J.J., Liu E.T., Desai K.V. Self-assembled monolayer-assisted silicon
nanowire biosensor for detection of protein–DNA interactions in nuclear extracts from breast
cancer cell. Biosensors and Bioelectronics. 2011; 26:3233-3239.
Zhang Y., Yang D., Weng L., Wang L. Early Lung Cancer Diagnosis by Biosensors. International
Journal of Molecular Science. 2013; 14:15479-15509.
Zhou Y., Wang Z., Yue W., Tang K., Ruan W., Zhang Q., Liu L. Label-free Detection of p53
Antibody Using A Microcantilever Biosensor With Piezoresistive Readout. IEEE SENSORS 2009
Conference. 2011.
22
Detection of newly synthesized proteins in cancer cells:
A practical approach
Łukasz P. Biały1#
, Łukasz Zaręba2#
, Izabela Młynarczuk-Biały1*
1 Department of Histology and Embryology, Medical University of Warsaw
2 HESA Histology and Embryology Students Association at the Department for
Histology and
Embryology Centre for Biostructure Research, Warsaw Medical University, Poland # Both authors contributed equally to this work
*Corresponding Author: Izabela Młynarczuk-Biały, [email protected]
Abstract
Up to 30% of newly synthesized proteins in cells are DRiPs/RDPs (Ribosomal Defective Products/
Rapidly Degraded Proteins). They are a special subset of nascent proteins that never attain their native
structure and are rapidly degraded after biosynthesis. The major degradation pathway of DRiPs/RDPs
involves classical ubiquitin-26S proteasome pathway but their smaller fraction is eliminated by 20S
proteasome without ubiquitin tagging.
The increasing interest on DRiPs/RDPs biological implications and their kinetic nature results in
invention of several detection methods. Here we provide characterizations and comparison between
three most common pulse-chase radiolabeling icnl. radioisotopes containing amino acids
35S-Methionine 3H-Leucine as well as novel non-radioactive technique using biorthogonal amino acid
tagging and click-IT chemistry. Moreover, we discuss cyto-physiological properties of DRiPs/RDPs as
the main source of MHC class I antigens and contribution of newly synthesized proteins to anticancer
action of proteasome inhibitors.
Finally, in practical approach we show that azidohomoalanine (AHA) metabolic labeling of murine
colon carcinoma cells C26 for 5h followed by ethanol-based fixation and Click-IT detection by Alexa
Fluor 555 alkyne methods can be used for in situ histochemical visualization of synthesized proteins by
means of Laser Scanning Confocal Microscopy. We visualized the forming aggresome in C26 cells
upon proteasome inhibition. This aggresome was composed of the inner dense core and the mantel of
labeled proteins. Thus this method can be a valuable tool in anticancer research on proteasome
inhibitors.
Introduction
Proteins are the main building blocks forming the structure of all living organisms.
They are composed of amino acids and are the products of complicated intracellular
machinery, in which DNA code is transcripted into mRNA and subsequently
translated into polypeptide chain consisted of covalently linked amino acids. This
step takes place on ribosome in ATP-dependent manner. The nascent protein must
undergo proper folding to obtain unique three-dimensional conformation with the
additional help of molecular chaperons (Hartl, Bracher et al. 2011). This multistep
process of protein production makes it susceptible to errors. Any mistake in synthesis
may result in misfolding and loss of macromolecule function. The errors potentially
generated are numerous: mutations, translation errors, age related errors, exposure to
environmental stress conditions such as heat, metal ions, oxidation are the most
prevalent (Berner, Reutter et al. 2018). Therefore, cells adopted various quality
control systems to prevent generation and accumulation of harmful misfolded
proteins in cells (Brandman, Hegde 2016). Two general systems acts parallel in
cellular proteolysis: lysosomal-based degradation incl. authophagy and Ubiquitin-
Detection of newly synthesized proteins in cancer cells: A practical approach
23
Proteasome System (UPS) (Varshavsky 2017). The majority of misfolded proteins
undergoes UPS-degradation (Spinnenhirn, Bitzer et al. 2017).
The special subset of proteins was distinguished in nascent proteins: Defective
Ribosomal Products (DRiPs). DRiPs are defined as proteins that never attain their
native structure and are rapidly degraded after biosynthesis (Qian, Princiotta et al.
2006; Schubert, Anton et al. 2000). The similar term Rapidly Degraded Polypeptides
(RDP) are also used in regard to this pool of short-lived proteins (Qian, Princiotta et al.
2006; Schubert, Anton et al. 2000).
The aim of this article is to discuss and compare the published methods of nascent
protein detection pointing a special attention to their degradation pathways. Moreover
we demonstrate a practical approach of newly synthesized protein visualization during
aggresome formation in C26 cancer cells by Laser Scanning Confocal Microscopy.
Methods
Search strategy and selection criteria
Our research strategy was aimed at evaluating studies for detection and visualization of
nascent protein degradation. The terms “nascent proteins”, ”newly synthesized
proteins", "Short-lived proteins" "Defective Ribosomal Products (DRiPs)", as well as
"protein degradation", "proteasome", "Ubiquitin-Proteasome System (UPS)”, were
searched in PubMed (NCBI) and Google Scholar. Scientific articles from 1971 to 2018
were searched.
Experimental Materials and Methods
Reagents
L-azidohomoalanine (Click-iT® AHA C10102 ThermoFisher) was dissolved with
DMSO to 50mM stock solution. Alexa Fluor® 555 alkyne (A20013 ThermoFisher)
was dissolved in Wather/gicerol (1:1 solution) to the recommended final concentration
of 2.5mM. The proteasome inhibitor MG-132 (Calbiochem/Merck) was dissolved on
DMSO to the final concentration of 10mM. All reagents were Aliquoted and stored at
–20°C until use.
Cell culture
Murine adenocarcinoma C26 cell line was obtained from ATCC. Cells were grown in
RPMI1640 (Biochrom, Germany) with stable glutamine and supplemented with 10%
FCS and standard antibiotics. Cells were cultured in RPMI-1410 medium with stable
glutamine (Biochrom, Germany) supplemented with 10% heat-inactivated FCS
(Biochrom, Germany), 1% Antibiotic–Antimycotic (Thermo Fisher/Life Technologies,
US) in 25 cm2 tissue flasks (Greiner, Germany) and kept at 37°C in a 5% CO2
humidified incubator and passaged every three days with EDTA-Tripsin solution
(Biochrom).
Metabolic labeling and cytochmistry
C26 cells grown onto multi-chamber slides (Becton Dickinson) and before metabolic
labeling were staved for 30 min in RPMI-1410 Methionin-free medium (Sigma). The
medium was replaced 2 times to eliminate the residual methionin before adding
Łukasz P. Biały, Łukasz Zaręba, Izabela Młynarczuk-Biały
24
L-azidohomoalanine. After starvation L-azidohomoalanine was added to the final
concentration of 0.25mM.
Cells were incubated at 37°C, 5% CO2 with L-azidohomoalanine (AHA) for 5h in
the presence of proteasome inhibitor MG-132 in the concentration of 10mM to
prevent degradation of newly synthesized proteins. After metabolic labeling were
fixed using the modified routine procedure for cytopathology with ice-cold buffered
70% ethanol (pH 7.5; phosphate buffer) for 3 minutes and directly rehydrated (pH
7.5; phosphate buffer). Washed once with 3% BSA in PBS. Subsequently, 1 µM
Alexa Fluor® 555 alkyne in Click-iT® cell reaction cocktail with Copper (II) sulfate
(C10269 ThermoFisher) was added or 30 minutes at room temperature in darkness.
Cells were washd three times with 1BSA in PBS and embedded in Vecta Shield
containing 4′,6-diamidino- 2-phenylindole (DAPI) (Vector Laboratories, US).
The analysis and images were made using a confocal laser SP5 microscope equipped
with appropriate lasers (HeNe 633nm, DPSS diode 561nm, and Argon laser: line
488nm) and Las-AF software (Leica, Germany).
The state of Art
The Ubiquitin-Proteasome system in degradation of nascent proteins
UPS is an conservative proteolytic system for protein clearance in eukaryotic cells.
This is composed of ubiquitin conjugation machinery and the proteasome as an
proteolytic complex degrading the targeted protein into oligopeptides. Ubiquitin is
a 76-amino-acid polypeptide, which is covalently attached to the protein in a form of
a polyubiquitin chain serving as a degradation signal. The polyubiquitination is
achieved by cascade of enzymes: the ubiquitin-activating enzyme (E1), ubiquitin-
conjugating enzymes (E2s), and ubiquitin-protein ligases (E3s) The ubiquitin is
activated by E1 enzyme than transferred onto E2. Finally E3 enzyme catalyzes the
conjugation of ubiquitin to the target protein. The activation of ubiquitin catalyzed by
the E1 enzyme is ATP dependent and leads to formation of high-energy thioester
bond (Hershko, Ciechanover 1998).
The ubiquitin is attached to the lysine of the targeted protein but there has been found
that also serine, threonine and cysteine may be modified by ubiquitin (Ciechanover,
Stanhill 2014). Within ubiquitin polypeptide there are 7 internal own lysine residues,
each of which can be used for bonding another ubiquitin molecule and create
polymeric ubiquitin chain. Thus there can be formed various poly ubiquitin chains
but the chain linked by lysine 48 (K48 chains) predominantly serves as a signal for
proteasomal degradation (Chau, Tobias et al. 1989), (Finley, Sadis et al. 1994),
(Pickart, Fushman 2004).
The 26S proteasome is a multisubunit multicatalytic complex composed of 20S catalytic core and two 19S regulatory particles forming the 26S proteasome complex (Voges, Zwickl et al. 1999). The 20Sproteasome is a central cylinder formed from multiple subunits arranged into four heptameric rings with three main proteolytic activities: trypsin-like, chymotrypsin-like and caspase-like attributed to the β1, β2 and β5 subunits respectively (Kruger, Kuckelkorn et al. 2003). The 20S core proteasome displays itself the ability to degrade unfolded polypeptide chains without
Detection of newly synthesized proteins in cancer cells: A practical approach
25
prior polyubiquitination (Ciechanover 2005). The19S cup complexes contains six-subunit protein ring, which occupy each end of the cylinder. The cup structure is a link between arriving ubiquitinated protein and central proteolytic core. It regulates opening of a gate to 20S proteasome complex and unfold proteins in ATP-dependent manner. Each of 19S particles contains ubiquitin receptors, which can bind approaching proteins and a metalloprotease removing polyubiquitin chain before protein degradation (Tomko, Hochstrasser 2013), (Finley, Chen et al. 2016). In the case of defective protein that cannot be successfully folded by molecular chaperons the CHIP co-chaperone performs ubiquitinylation of targeted protein that is transferred to 26S proteasome for degradation (Wojcik 2002). However severe misfolded proteins can be degraded by 20S Proteasome co-translationally (Qian, Princiotta et al. 2006).
Inhibition of the proteasome leads to accumulation of newly synthesized proteins and UPS components in the aggregates, which can be observed within the cell as single round structures termed aggresomes, which initially assembles in the peri-centriolar regions of the cell (Wojcik, Schroeter et al. 1996), (Wojcik 1997), (Johnston, Ward et al. 1998).
Other extralysosomal proteolytic systems
The function of UPS is supplemented by several proteases. TPPII is the biggest known peptidase forming self-compartmentalizing, twisted double stranded cytosolic peptidase complex of 6MDa (Rockel, Peters et al. 2005). It displays both exo- and endo-proteolytic activities . Thus it provides support and even substitutes some of the proteasomal functions. It can supplement the downstream action of the proteasome in sequential protein degradation through trimming of tripeptides from oligopeptides released by the proteasome (Tomkinson, Lindas 2005), (Peters, Schonegge et al. 2011). The products released by 20/26S proteasomes are oligopeptides comprised of 4 to 25 AA (Voges, Zwickl et al. 1999), (Kruger, Kuckelkorn et al. 2003). In order for complete degradation TPPII digests substrates longer than 15 AA (Reits, Neijssen et al. 2004), (York, Bhutani et al. 2006). TPPII was described as being involved in several processes of cell regulation (Mlynarczuk-Bialy 2008), (Rockel, Peters et al. 2005). Recently, TPPII was shown to be recruited into aggresomes in colon cancer C26 cells upon proteasome inhibition (Bialy, Kuckelkorn et al. 2018) and semispecific TPPII inhibitor induces protein aggregation in leukemic U937 cells (Bialy, Fayet et al. 2018).
The UPS cooperates also with an ER-associated degradation (ERAD) in degradation of defective proteins form endoplasmic reticulum. There are 3 types of this pathway: ERAD-L for misfolded luminal Endoplasmic Reticulum proteins, ERAD-M for ER proteins with error in membrane domain and ERAD-C for ER proteins with unfunctional cytosolic domain ERAD systems contain enzymes machinery, which is able to recognize misfolding, prevent leaving invalid proteins from ER to excretory pathway, export it back to the cytosol by retrotranslocation and lead to degradation (Berner, Reutter et al. 2018).
Łukasz P. Biały, Łukasz Zaręba, Izabela Młynarczuk-Biały
26
The origin of newly synthesized proteins and their degradation pathways
From 70th XX century on the basis of degradation kinetics cellular proteins are
divided into two main groups of proteins: “short-lived” (with a fast degradation rate) and “long-lived” (with a much slower degradation rate) (Goldberg, Dice 1974), (Poole, Wibo 1973), (Pine 1972). Both kinetics of degradation were separated by a clear discontinuity and these two groups of proteins have been identified in all cells.
The studies revealed that 30% of newly synthesized proteins are eliminated with a half-life <10 minutes and this part of extremely short-lived proteins was named DRiPs (defective ribosomal products)/RDPs (rapidly degraded proteins) (Qian, Princiotta et al. 2006; Schubert, Anton et al. 2000). 75% of these proteins are eliminated by 26S proteasomes and remaining 25% by 20S proteasome complex independently of 19S regulators (Qian, Princiotta et al. 2006). DRiPs/RDPs can be divided into two classes based on their solubility. The soluble RDPs fraction is degraded by ubiquitin-depended manner, which is linked with 26S proteasome but insoluble fraction seems to be degraded in a different way without involvement of ubiquitin. Qian et al. bring proofs and conclude that 75% of DRiPs/RDPs are degraded by 26S proteasome Hsc70-depended pathway based on ubiquitination. The remaining 25% are degraded without ubiquitin-link by 20S proteasome without involvement of 19S regulatory subunit and are largely unaffected by Hsc70 modulatory activity (Qian, Princiotta et al. 2006). DRiPs/RDPs are also the largest fraction of proteins degraded by proteasomes and constitute 70-75% of all the protease complex substrates (Princiotta, Finzi et al. 2003; Yewdell 2011).
Figure 1. DRiPs/RDP and their degradation pathways
DRIPs/RDPs are degraded co-translationally. The larger fraction of them one is water soluble and undergoes
polyubiquitination prior to degradation by 26S proteasome. The minor fraction (ca 25%) is digested directly
within 20S core of proteasome without previous ubiquitinylation. Most generated oligopeptides is a source of
MHC class I epitopes (vial or cancer antigens). Original Figure.
Detection of newly synthesized proteins in cancer cells: A practical approach
27
DRIPs and immunity
At the beginning DRiPs were defined as ribosomal products that never attain their
mature structure because of defective mRNAs (Shastri, Nguyen et al. 1995; Wang,
Parkhurst et al. 1996), ribosomal frame shifting (Bullock, Eisenlohr 1996; Fetten,
Roy et al. 1991), tRNA- amino acid misacylation (Netzer, Goodenbour et al. 2009),
transcription errors (Gout, Thomas et al. 2013) or all other mistakes during
converting DNA information into protein (Anton, Yewdell 2014). This definition
implicates that RDPs/DRiPs are accidental products of protein synthesis but recent
studies have proven that they’re not only a junk biosynthesis consequence but also
have a crucial contribution to cellular processes. Thus, DRiPs/RDPs was considered
only as abnormal, misfolded proteins but now it is known that they are a main source
of cellular and viral antigens presented by histocompatibility complex class I
molecules: viral and cancer (Reits, Vos et al. 2000; Schubert, Anton et al. 2000).
Yewdell concluded form Dolans studies (Dolan, Li et al. 2011; Yewdell 2011) that
even 70% of MHC class I complexes can originate from DRiPs. This remark is also
supported by interesting observation of immunologic response for virus-infected
cells. CD8+ cytotoxic T-lymphocytes recognize specific antigens on the surface of
infected cell simultaneously to virus is entrance and long before source protein could
be detected within infected cell. Transition of viral protein into antigen should take
much longer and probably is not responsible for activation of early cytotoxic
T-lymphocytes. However antigens for MHC class I presentation can be also derived
from long lived proteins but DRiPs generation seems to be favored for rapid
recognition of infected cells by CD8+ T cells (Anton, Yewdell 2014). DRiPs were
also described as a major contributor to the antigenic presentation associated with
immunological tolerance (Arguello, Reverendo et al. 2016). Noticeable, the role of
rapidly degraded proteins indicates possibility for preconcerted RDP-biosynthesis,
which is not only an error effect. Referring to this hypothesis, scientists have indeed
found unique subset of ribosomes specialized for antigenic relevant DRiPs synthesis
named immunoribosomes (Yewdell 2002; Yewdell 2005). Proteins produced by
those complexes are believed to be rapidly degraded by (immuno)proteasomes into
oligopeptides dedicated to antigen generation. There exists hypothesis for close
relationship of both of them (Dolan, Bennink et al. 2011), which is supported by
discovery of several proteins capable to interact with the both complexes (Panasenko,
Landrieux et al. 2006; Sha, Brill et al. 2009) for example eIF1A (Chuang, Chen et al.
2005).
Newly synthesized proteins and Cancer
Current established therapeutic procedures include inhibition of the proteasome as
anticancer strategy (Teicher, Tomaszewski 2015). Inhibition of proteasome proteo-
lysis leads to misfolded protein response and accumulation of newly synthesized
proteins that induces apoptosis. Thus most sensitive to proteasome inhibition are
cells that perform intensive protein synthesis like plasmmocytoma cells.
Łukasz P. Biały, Łukasz Zaręba, Izabela Młynarczuk-Biały
28
Upon proteasome inhibition newly synthesized ubiquitinated proteins are transported
along microtubules to the pericentriolar proteolytic center of the cell and stored there
as an aggresome (Johnston, Ward et al. 1998; Wojcik 1997; Wojcik, Schroeter et al.
1996). The accumulation of misfolded proteins within the aggresome is one of the
postulated mechanisms of action of proteasome inhibitors as anticancer drugs.
Nonetheless, aggresome disruption by histone deacetylase inhibitors (Nawrocki,
Carew et al. 2006) and microtubule acting agents (Miyahara, Kazama et al. 2016) can
augment the anticancer activity of proteasome inhibitors. Thus, aggresome formation
in cancer cells likely protects them from proteasome inhibitor-induced unfolded
protein stress, and therefore can limit anticancer effects of proteasome inhibitors (see
next chapter).
Comparison of DRIP detection protocols
The biological features and function of DRiPs/RDPs becomes a substantial subject of
studies recently because of their serious impact on cell biology. As mentioned before,
DRiPs/RDPs are nascent proteins with a half-life < 10mins a and can constitute up to
30% of newly synthesized proteins (Qian, Princiotta et al. 2006), (Schubert, Anton et
al. 2000). Thus due to extremely short live time of DRiPs/RDPs, their experimental
detection is difficult, and several studies were focused on acquiring the best way of
their detection. Various protocols were used to achieve this aim.
The most common method is classic radioactive pulse-chase analysis (Qian,
Princiotta et al. 2006). The method involves labeling of newly synthesized proteins
with radioactive amino acids which allows quantitative analysis of the fate of given
proteins in the time-depended manner (Magadan 2014). The detection of radioactive
protein can be subsequently preformed by either liquid scintillation counter or
autoradiography. Alternatively non-radioactive amino acids surrogates such as
L-azidohomoalanine (AHA) can be used for metabolic labeling of proteins followed
by detetection by selective click-chemistry techniques.
Pulse-chase radiolabeling of newly synthesized proteins is a well-known method for
DRiPs/RDPs analysis. First step of the method pulsing of cells for a short time period
with special medium that contains radiolabeled amino acids. The most common
radioactive amino acid used in this method is 35SMethionine. It is used due to its
high specific activity (>800Ci/mmol) and quick detection. However, potential
disadvantage of this amino acids is its low abundance in proteins (~1.8% of the
average amino acid composition) (Bonifacino Juan, Gershlick David et al. 2016) thus
high radioacivity must be used up to 5 mCi/ml (Schubert, Anton et al. 2000).
Alternatively 3HLeucine can be used for radioactive protein labeling. Because of
DRiP high turnover rate pulse-labeling time has to be short enough for appropriate
detection. Qian et al. measured nascent proteins degradation by radiolabeling in
HeLa cells with 35s Met for 5 min (Qian, Princiotta et al. 2006). Furthermore,
Shubert with colleagues did it for only 30 seconds in their studies (Schubert, Anton
et al. 2000).
Detection of newly synthesized proteins in cancer cells: A practical approach
29
The second step of the technique contains chasing of radioactive palsed cells in
complete medium with unlabeled methionine for longer period of time to allow
degradation of newly synthesized proteins. Qian et al. chased cells for up to 4h and
Shubert et al. did it for up to 60 min.
At the end the detection of signal can be made by measuring radioactivity present in
trichloroacetic acid precipitates of whole cell lysates and culture media (Qian,
Princiotta et al. 2006). Measurement can be done by scintillation counter or
alternatively by exposing dried gel from SDS-PAGE to phosphor imager screen or
X-ray film (Qian, Bennink et al. 2005).
Alternatively 3HLeucine radioactive protein labeling was used by Knecht’s group for
pulse-chase analysis of newly synthesized (Fuertes, Villarroya et al. 2003). The
degradation of short-labeled proteins was examined in fibroblasts by release of
trichloroacetic acid-soluble radioactivity. Fuertes at al. estimated DRiP rate to about
30% of newly synthesized. However half-life observed longer up to 1 h probably due
to longer tilme of the radioacive pulse. They also calculated that proteasomes are
responsible for degradation go 60% newly synthesized proteins. The described
methodology has a high accuracy but working with radioactive materials is
potentially dangerous for experimenter and environment especially with long-lasting
3H nuclide.
Details and comparison of the mentioned protocols is displayed in Table 1.
Table 1. Comparison of most common DRiPs/RDPs detection protocols
Shubert et al. Qian et al. Fuertes et al.
Cell line HeLa HeLa Human fibroblasts
Pulse time 30s 5min 15min
Chase time 60min 4h Different chase
time up to 25h
Source of
radioactivity
[35S]methionine
5 mCi/ml
[35S]methionine [3H]leucine or
[3H]valine
1 mCi/ml
Detection SDS-PAGE and
radioactivity
measurement in dried
gel
TCA precipitation and scintillation
counting/ SDS-PAGE and
radioactivity measurement in dried
gel
TCA precipitation
and scintillation
counting
T1/2 10min 7.6min 1.1h
Click-IT chemistry in detection of newly synthesized proteins
Recently new method for detecting on synthesized proteins was described. It allows
to detect proteins of interests without using radioisotopes. Thus, it is safer for
environment, experimenters and furthermore enables in situ histochemical visua-
lization. Bioorthogonal noncanonical amino acid tagging – BONCAT is a novel
Łukasz P. Biały, Łukasz Zaręba, Izabela Młynarczuk-Biały
30
method based on application of amino acid surrogates modified with azide or alkyne
groups, which are incorporated into proteins during biosynthesis (Dieterich, Link et
al. 2006). It is possible because certain tRNA synthetases employed to the process
are unable to recognize natural amino acids and can recognize those with slightly
different structure (Link, Mock et al. 2003). Bioorthogonal molecules are characterized
as synthetic, not biologically synthesized and not interfering with native biochemical
processes. They are frequently analogs of native biomolecules with addition of
special groups required for click chemistry (Hatzenpichler, Scheller et al. 2014).
There have been described a series of biorthogonal amino acids which are able to
successfully compete with native amino acid molecules during translation but there
are only several molecules, which can be used by translational machinery without
genetic modifications of the host cell (Ngo, Tirrell 2011). One of the most efficient,
often described as a member of this subset is an amino acid substitute
L-azidohomoalanine (AHA) – an analog of L-methionine (Kiick, Saxon et al. 2002).
An azide group of AHA incorporated into proteins makes them distinct from other
polypeptides in cell and enables selective click chemistry-mediated detection
(Hatzenpichler, Scheller et al. 2014). The functional azide group reacts with alkyne
through 1,3-dipolar cycloaddition to yield triazole linkage. The reaction is catalyzed
by Cu(II) and here we came across the serious disadvantage – copper catalyst in
higher concentration is toxic to cells. To circumvent the problem several metal-free
azyde-alkyne cycloadditions have been developed (Lim, Lin 2010). The whole
process allows to visualize proteins of interest by many methods e.g. microscope in
vivo, SDS-PAGE due to azide- alkyne fluorescence dye linkage. The 1,2,3- triazole
linkage is extremely strong and resistant to hydrolysis, oxidation, reduction or
ionization in mass spectrometry analysis. Finally, proteins labeled with AHA can be
detected using subsequent analysis by flow cytometry, imaging or standard
biochemistry techniques such as gel electrophoresis. Detection sensitivity obtained in
this process is similar to that in radioactive 35S Met method and is compatible with
downstream LC-MS/MS and MALDI MS analysis ("Molecular probes the
handbook").
The Practical Approach of newly synthesized protein visualization in cells
We show localization of newly synthesized proteins using amino acid analogue AHA
visualized by Click chemistry with fluorochrome within murine colon
adenocarcinoma C26 cells.
C26 cells were chosen for this practical approach due to their properties. In our
laboratory we investigated anticancer properties of various proteasome inhibitors
against many cancer cell lines and among them C26 cells respond to proteasome
inhibition by formation of the giant aggresomes and can survive long-term
proteasome inhibition making them an ideal candidate for protein aggregation study.
The aggregate is formed in perinuclear region and fully mature easy to detect large
and round is formed within 12-24 hours of proteasome inhibition (unpublished labor
data).
Detection of newly synthesized proteins in cancer cells: A practical approach
31
Results
As it is shown in Fig 1 after 5h of proteasome inhibition by 10µM MG-132
synthesized proteins (red) accumulate in the cytoplasm of C26 cells. They
predominately accumulate in a one single round structure in perinuclear region
corresponding to forming aggresome (white arrows in panel B). The forming
aggregate is composed of a dense solid core (arrows in panel C) with a mantle
containing less dense material (arrowheads in Panel C). Moreover, the residual part
of cytoplasm is stained less intensively. There was no signal within nuclei observed.
Figure 2: Visualization of newly synthesized protein in C26 mouse adenocarcinoma
C26 cells were treated for 6h with 10uM Mg132 to inhibit protein degradation. Synthesized proteins were
detected by AHA incorporation (for 6h) and visualized by Alexa Fluor 555 alkyne after ethanol-based fixation
using confocal microscopy. Nuclei – Blue – nuclei stained with DAPI; Synthesized proteins – Red – proteins
with incorporated AHA; Merge; Scale bar: 7.5 µM.
Łukasz P. Biały, Łukasz Zaręba, Izabela Młynarczuk-Biały
32
Short Discussion and Conclusion
Non-radioactive by Click-IT a chemistry-based method of newly synthesized protein
detection gives an interesting alternative for their radioactive detecting. Moreover
Click-IT methods have an additional advantage because they can be used in situ
histochemical visualization of synthesized proteins. Moreover, our validation
endorsed the ethanol based fixation suitable for this application. We visualized the
forming aggresome in C26 cells upon proteasome inhibition. This aggresome was
composed of the inner dense core and the mantel of labeled proteins.
Concluding this method can be a valuable tool in anticancer research on proteasome
inhibitors incl. aggresome formation studies.
References
Anton, L. C., Yewdell, J. W. Translating DRiPs: MHC class I immunosurveillance of pathogens
and tumors. J Leukoc Biol. 2014; 95(4):551-562.
Arguello, R. J., Reverendo, M., Gatti, E., Pierre, P. Regulation of protein synthesis and autophagy in
activated dendritic cells: implications for antigen processing and presentation. Immunol Rev. 2016;
272(1):28-38.
Berner, N., Reutter, K. R., Wolf, D. H. Protein Quality Control of the Endoplasmic Reticulum and
Ubiquitin-Proteasome-Triggered Degradation of Aberrant Proteins: Yeast Pioneers the Path. Annu Rev Biochem. 2018.
Bialy, L.P., Fayet, J., Wilczynski, G.M., Mlynarczuk-Bialy, I. Semispecific TPPII inhibitor Ala-Ala-
Phe-chloromethylketone (AAF-cmk) displays cytotoxic activity by induction of apoptosis, autophagy and protein aggregation in U937 cells, Folia Histochem Cytobiol in press (2018).
Bialy, L.P., Kuckelkorn, U., Henklein, P., J. Fayet, Wilczynski G. M., Kaminski, A., Mlynarczuk-Bialy,
I, Changes in spatio-temporal localization of tripeptidyl peptidase II (TPPII) in murine colon
adenocarcinoma cells during aggresome formation: a microscopy study based on a novel fluorescent proteasome inhibitor, Histol Histopathol, (2018), s. 18042.
Bonifacino Juan, S., Gershlick David, C., Dell'Angelica Esteban, C. Immunoprecipitation. Current
Protocols in Cell Biology. 2016; 71(1).
Brandman, O., Hegde, R. S. Ribosome-associated protein quality control. Nat Struct Mol Biol. 2016;
23(1):7-15.
Bullock, T. N., Eisenlohr, L. C. Ribosomal scanning past the primary initiation codon as
a mechanism for expression of CTL epitopes encoded in alternative reading frames. J Exp Med.
1996; 184(4):1319-1329.
Chau, V., Tobias, J. W., Bachmair, A., Marriott, D., Ecker, D. J., Gonda, D. K., Varshavsky, A. A
multiubiquitin chain is confined to specific lysine in a targeted short-lived protein. Science. 1989;
243(4898):1576-1583.
Chuang, S. M., Chen, L., Lambertson, D., Anand, M., Kinzy, T. G., Madura, K. Proteasome-mediated
degradation of cotranslationally damaged proteins involves translation elongation factor 1A. Mol
Cell Biol. 2005; 25(1):403-413.
Ciechanover, A. Proteolysis: from the lysosome to ubiquitin and the proteasome. Nat Rev Mol Cell
Biol. 2005; 6(1):79-87.
Ciechanover, A., Stanhill, A. The complexity of recognition of ubiquitinated substrates by the 26S
proteasome. Biochim Biophys Acta. 2014; 1843(1):86-96.
Detection of newly synthesized proteins in cancer cells: A practical approach
33
Dieterich, D. C., Link, A. J., Graumann, J., Tirrell, D. A., Schuman, E. M. Selective identification of
newly synthesized proteins in mammalian cells using bioorthogonal noncanonical amino acid
tagging (BONCAT). Proc Natl Acad Sci U S A. 2006; 103(25):9482-9487.
Dolan, B. P., Bennink, J. R., Yewdell, J. W. Translating DRiPs: progress in understanding viral and
cellular sources of MHC class I peptide ligands. Cell Mol Life Sci. 2011; 68(9):1481-1489.
Dolan, B. P., Li, L., Veltri, C. A., Ireland, C. M., Bennink, J. R., Yewdell, J. W. Distinct pathways
generate peptides from defective ribosomal products for CD8+ T cell immunosurveillance. J
Immunol. 2011; 186(4):2065-2072.
Fetten, J. V., Roy, N., Gilboa, E. A frameshift mutation at the NH2 terminus of the nucleoprotein
gene does not affect generation of cytotoxic T lymphocyte epitopes. J Immunol. 1991; 147(8):2697-
2705.
Finley, D., Chen, X., Walters, K. J. Gates, Channels, and Switches: Elements of the Proteasome
Machine. Trends Biochem Sci. 2016; 41(1):77-93.
Finley, D., Sadis, S., Monia, B. P., Boucher, P., Ecker, D. J., Crooke, S. T., Chau, V. Inhibition of
proteolysis and cell cycle progression in a multiubiquitination-deficient yeast mutant. Mol Cell
Biol. 1994; 14(8):5501-5509.
Fuertes, G., Villarroya, A., Knecht, E. Role of proteasomes in the degradation of short-lived proteins in
human fibroblasts under various growth conditions. Int J Biochem Cell Biol. 2003; 35(5):651-664.
Goldberg, A. L., Dice, J. F. Intracellular protein degradation in mammalian and bacterial cells.
Annu Rev Biochem. 1974; 43(0):835-869.
Gout, J. F., Thomas, W. K., Smith, Z., Okamoto, K., Lynch, M. Large-scale detection of in vivo
transcription errors. Proc Natl Acad Sci U S A. 2013; 110(46):18584-18589.
Hartl, F. U., Bracher, A., Hayer-Hartl, M. Molecular chaperones in protein folding and proteostasis.
Nature. 2011; 475(7356):324-332.
Hatzenpichler, R., Scheller, S., Tavormina, P. L., Babin, B. M., Tirrell, D. A., Orphan, V. J. In situ
visualization of newly synthesized proteins in environmental microbes using amino acid tagging
and click chemistry. Environ Microbiol. 2014; 16(8):2568-2590.
Hershko, A., Ciechanover, A. The ubiquitin system. Annu Rev Biochem. 1998; 67:425-479.
Johnston, J. A., Ward, C. L., Kopito, R. R. Aggresomes: a cellular response to misfolded proteins. J
Cell Biol. 1998; 143(7):1883-1898.
Kiick, K. L., Saxon, E., Tirrell, D. A., Bertozzi, C. R. Incorporation of azides into recombinant
proteins for chemoselective modification by the Staudinger ligation. Proc Natl Acad Sci U S A.
2002; 99(1):19-24.
Kruger, E., Kuckelkorn, U., Sijts, A., Kloetzel, P. M. The components of the proteasome system and
their role in MHC class I antigen processing. Rev Physiol Biochem Pharmacol. 2003; 148:81-104.
Lim, R. K., Lin, Q. Bioorthogonal chemistry: recent progress and future directions. Chem Commun
(Camb). 2010; 46(10):1589-1600.
Link, A. J., Mock, M. L., Tirrell, D. A. Non-canonical amino acids in protein engineering. Curr Opin
Biotechnol. 2003; 14(6):603-609.
Magadan, J. G. Radioactive Pulse-Chase Analysis and Immunoprecipitation. Bio Protoc. 2014; 4(8).
Miyahara, K., Kazama, H., Kokuba, H., Komatsu, S., Hirota, A., Takemura, J., Hirasawa, K., Moriya,
S., Abe, A., Hiramoto, M., Ishikawa, T., Miyazawa, K. Targeting bortezomib-induced aggresome
formation using vinorelbine enhances the cytotoxic effect along with ER stress loading in breast cancer cell lines. Int J Oncol. 2016; 49(5):1848-1858.
Łukasz P. Biały, Łukasz Zaręba, Izabela Młynarczuk-Biały
34
Mlynarczuk-Bialy, I. Enigmatic tripeptidylpeptidase II - protease for special tasks. Postepy Biologii
Komorki (Advances in Cell Biology). 2008; 35(4):427-439.
Molecular probes the handbook. In Termofisher (Ed.),
https://www.thermofisher.com/pl/en/home/references/molecular-probes-the-handbook/reagents-for-
modifying-groups-other-than-thiols-or-amines/click-chemistry.html.
Nawrocki, S. T., Carew, J. S., Pino, M. S., Highshaw, R. A., Andtbacka, R. H., Dunner, K., Jr., Pal, A.,
Bornmann, W. G., Chiao, P. J., Huang, P., Xiong, H., Abbruzzese, J. L., McConkey, D. J. Aggresome
disruption: a novel strategy to enhance bortezomib-induced apoptosis in pancreatic cancer cells. Cancer Res. 2006; 66(7):3773-3781.
Netzer, N., Goodenbour, J. M., David, A., Dittmar, K. A., Jones, R. B., Schneider, J. R., Boone, D.,
Eves, E. M., Rosner, M. R., Gibbs, J. S., Embry, A., Dolan, B., Das, S., Hickman, H. D., Berglund, P.,
Bennink, J. R., Yewdell, J. W., Pan, T. Innate immune and chemically triggered oxidative stress
modifies translational fidelity. Nature. 2009; 462(7272):522-526.
Ngo, J. T., Tirrell, D. A. Noncanonical amino acids in the interrogation of cellular protein
synthesis. Acc Chem Res. 2011; 44(9):677-685.
Panasenko, O., Landrieux, E., Feuermann, M., Finka, A., Paquet, N., Collart, M. A. The yeast Ccr4-
Not complex controls ubiquitination of the nascent-associated polypeptide (NAC-EGD) complex. J
Biol Chem. 2006; 281(42):31389-31398.
Peters, J., Schonegge, A. M., Rockel, B., Baumeister, W. Molecular ruler of tripeptidylpeptidase II:
mechanistic principle of exopeptidase selectivity. Biochem Biophys Res Commun. 2011; 414(1):209-214.
Pickart, C. M., Fushman, D. Polyubiquitin chains: polymeric protein signals. Curr Opin Chem Biol.
2004; 8(6):610-616.
Pine, M. J. Turnover of intracellular proteins. Annu Rev Microbiol. 1972; 26:103-126.
Poole, B., Wibo, M. Protein degradation in cultured cells. The effect of fresh medium, fluoride,
and iodoacetate on the digestion of cellular protein of rat fibroblasts. J Biol Chem. 1973;
248(17):6221-6226.
Princiotta, M. F., Finzi, D., Qian, S. B., Gibbs, J., Schuchmann, S., Buttgereit, F., Bennink, J. R.,
Yewdell, J. W. Quantitating protein synthesis, degradation, and endogenous antigen processing.
Immunity. 2003; 18(3):343-354.
Qian, S. B., Bennink, J. R., Yewdell, J. W. Quantitating defective ribosome products. Methods Mol
Biol. 2005; 301:271-281.
Qian, S. B., Princiotta, M. F., Bennink, J. R., Yewdell, J. W. Characterization of rapidly degraded
polypeptides in mammalian cells reveals a novel layer of nascent protein quality control. J Biol
Chem. 2006; 281(1):392-400.
Reits, E., Neijssen, J., Herberts, C., Benckhuijsen, W., Janssen, L., Drijfhout, J. W., Neefjes, J. A major
role for TPPII in trimming proteasomal degradation products for MHC class I antigen
presentation. Immunity. 2004; 20(4):495-506.
Reits, E. A., Vos, J. C., Gromme, M., Neefjes, J. The major substrates for TAP in vivo are derived
from newly synthesized proteins. Nature. 2000; 404(6779):774-778.
Rockel, B., Peters, J., Muller, S. A., Seyit, G., Ringler, P., Hegerl, R., Glaeser, R. M., Baumeister, W.
Molecular architecture and assembly mechanism of Drosophila tripeptidyl peptidase II. Proc Natl
Acad Sci U S A. 2005; 102(29):10135-10140.
Schubert, U., Anton, L. C., Gibbs, J., Norbury, C. C., Yewdell, J. W., Bennink, J. R. Rapid
degradation of a large fraction of newly synthesized proteins by proteasomes. Nature. 2000;
404(6779):770-774.
Detection of newly synthesized proteins in cancer cells: A practical approach
35
Sha, Z., Brill, L. M., Cabrera, R., Kleifeld, O., Scheliga, J. S., Glickman, M. H., Chang, E. C., Wolf, D.
A. The eIF3 interactome reveals the translasome, a supercomplex linking protein synthesis and
degradation machineries. Mol Cell. 2009; 36(1):141-152.
Shastri, N., Nguyen, V., Gonzalez, F. Major histocompatibility class I molecules can present cryptic
translation products to T-cells. J Biol Chem. 1995; 270(3):1088-1091.
Spinnenhirn, V., Bitzer, A., Aichem, A., Groettrup, M. Newly translated proteins are substrates for
ubiquitin, ISG15, and FAT10. FEBS Lett. 2017; 591(1):186-195.
Teicher, B. A., Tomaszewski, J. E. Proteasome inhibitors. Biochem Pharmacol. 2015; 96(1):1-9.
Tomkinson, B., Lindas, A. C. Tripeptidyl-peptidase II: a multi-purpose peptidase. Int J Biochem
Cell Biol. 2005; 37(10):1933-1937.
Tomko, R. J., Jr., Hochstrasser, M. Molecular architecture and assembly of the eukaryotic
proteasome. Annu Rev Biochem. 2013; 82:415-445.
Varshavsky, A. The Ubiquitin System, Autophagy, and Regulated Protein Degradation. Annu Rev
Biochem. 2017; 86:123-128.
Voges, D., Zwickl, P., Baumeister, W. The 26S proteasome: a molecular machine designed for
controlled proteolysis. Annu Rev Biochem. 1999; 68:1015-1068.
Wang, R. F., Parkhurst, M. R., Kawakami, Y., Robbins, P. F., Rosenberg, S. A. Utilization of an
alternative open reading frame of a normal gene in generating a novel human cancer antigen. J
Exp Med. 1996; 183(3):1131-1140.
Wojcik, C. On the spatial organization of ubiquitin-dependent proteolysis in HeLa cells. Folia
Histochem Cytobiol. 1997; 35(2):117-118.
Wojcik, C. Regulation of apoptosis by the ubiquitin and proteasome pathway. J Cell Mol Med.
2002; 6(1):25-48.
Wojcik, C., Schroeter, D., Wilk, S., Lamprecht, J., Paweletz, N. Ubiquitin-mediated proteolysis
centers in HeLa cells: indication from studies of an inhibitor of the chymotrypsin-like activity of the proteasome. Eur J Cell Biol. 1996; 71(3):311-318.
Yewdell, J. To DRiP or not to DRiP: generating peptide ligands for MHC class I molecules from
biosynthesized proteins. Mol Immunol. 2002; 39(3-4):139-146.
Yewdell, J. W. The seven dirty little secrets of major histocompatibility complex class I antigen
processing. Immunol Rev. 2005; 207:8-18.
Yewdell, J. W. DRiPs solidify: progress in understanding endogenous MHC class I antigen
processing. Trends Immunol. 2011; 32(11):548-558.
York, I. A., Bhutani, N., Zendzian, S., Goldberg, A. L., Rock, K. L. Tripeptidyl peptidase II is the
major peptidase needed to trim long antigenic precursors, but is not required for most MHC class
I antigen presentation. J Immunol. 2006; 177(3):1434-1443.
36
Direct visualization of undegraded protein
sequestration within aggresomes
by proteasome inhibitors
Łukasz P. Biały1, Przemysław Szymański
2, Izabela Młynarczuk-Biały
1*
1. Department for Histology and Embryology Centre for Biostructure Research,
Warsaw Medical University
2. HESA Histology and Embryology Students Association at the Department for
Histology and Embryology Centre for Biostructure Research, Warsaw Medical
University
* Corresponding author: Izabela Młynarczuk-Biały, Chałubińskiego 5, 02-004
Warszawa, Poland, [email protected]
Abstract
The ubiquitin-proteasome system (UPS) plays a crucial role in maintenance of cellular homeostasis.
It controls cell cycle progression, signal response, apoptosis, transcription, NF-κB activation. Inhibition
of the UPS induces accumulation of proteasomal substrates in a form of cellular aggregates that are
localized mainly in the perinuclear region. Tumour cells with deregulated protein turnover are especially
sensitive to inhibition of the UPS, thus proteasome inhibitors are applied in treatment of selected
malignancies like multiple myeloma. However, proteasome-inhibitor resistance of tumour cells is
a serious treatment perturbation and research on overcoming the resistance can improve therapy
effectiveness. One of possibilities for this resistance is formation of big aggresomes, in which
undegraded proteins are separated from important cellular processes by sequestration. Disruption of the
big aggresome made resistant tumour cells sensitive to death induction by proteasome inhibitors.
Herein, we show quick and direct inhibitor-based fluorescent staining proteasomes that enables fast and
effective analysis of aggresome structure and formation dynamics. This method can improve studies on
the reasons of tumour cell insensitivity towards proteasome inhibitors and can be a valuable screening
for novel proteasome inhibitor effects. Moreover, within this work we discuss the function of aggresome
and its impact on tumor resistance towards proteasome inhibition. In addition we present clinical
application of proteasome inhibitors with data from clinical trials.
Keywords: proteasome inhibitors, aggresome, inhibitor-based staining, UPS, HDAC6,
Introduction
The ubiquitin-proteasome system (UPS) is indispensable for cellular homeostasis.
The UPS is directly responsible for most of cytoplasmic protein degradation and
participates in a wide array of biological functions such as regulation of the cell
cycle, gene transcription, signal transduction, antigen processing and activation of
NF-κB (Adams, Palombella et al. 2000; Wojcik 1999, 2002).
The ubiquitin-proteasome system- general aspects
Intracellular protein degradation and turnover is regulated through the action of
ubiquitinylation enzymes designating proteins for degradation by the proteasome.
The proteasome is a multicatalytic enzyme complex localized in the in the cytoplasm
and in the nucleus of all eukaryotic cells. The fully active 26S proteasome is
composed of 20S barrel-shaped core particle and two 19S regulatory subunits,
localized at both ends of the 20S core particle. Each 20S core consists of 4 rings: two
outer α rings and two inner β rings. Each ring contains 7 globular subunits, α1-α7
Direct visualization of undegraded protein sequestration within aggresomes
by proteasome inhibitors
37
within α-ring and β1-β7 within β-ring, respectively. Within neighbouring β-rings
there are 3 subunits per each ring displaying proteolytic activity. The distinct
activities are: trypsin-like (within the β1 subunit), caspase-like (within the β2
subunit) and chymotrypsin-like (within the β5 subunit) of the 20S core (Adams 2003;
Tanaka 2009).
Substrates for degradation by 26S proteasome are tagged with polyubiquitin chain in
the process named ubiquitination (Haglund, Dikic 2005). The ubiquitin (Ub) is
a highly conserved polypeptide, 76 acids in length, which acts as a molecular label in
cellular protein trafficking. Within each Ub polypeptide there is seven lysine residues
localized. These lysine residues are involved in linking multiple Ub polypeptides
together to form polyubiquitin chain. Ubiquitination alters proteins in many ways: it
can mark them for proteasomal degradation, alter their cellular localization, affect
their activity and promote or prevent protein interactions (Miranda, Sorkin 2007).
Three types of ubiquitination are distinguished in the process of cellular protein
modification. Mono-ubiquitination – is an attachment of single Ub molecule to the
substrate that signals for endocytosis, endosomal sorting, histone regulation, DNA
repair, virus budding, nuclear export. Multi-ubiquitination occurs when several single
Ub molecules are attached to several lysine residues on the substrate that signals for
endocytosis, or finally polyubiquitination that means formation of long ubiquitin
chain that is bound to the substrate via first ubiquitin in the poly-chain. Ub chains,
which are formed via Lys48 (K48) of Ub, target the modified protein for proteasomal
degradation (Pickart, Fushman 2004). Chains linked via Lys63 (K63) are implicated
in DNA repair and activation of protein kinases or lysosomal pathway (Acconcia,
Sigismund et al. 2009; Haglund, Dikic 2005).
The ubiquitination process serving as proteasomal degradation signal is regulated by
a cascade of E1, E2, E3 enzymes (Adams 2003). The ubiquitin-activating E1 enzyme
binds to the ubiquitin protein and passes it to the ubiquitin-conjugating E2 enzyme.
Subsequently, E2 enzyme and ubiquitin-ligase E3 enzyme are responsible for linking
ubiquitin molecule with protein targeted to the proteasomal degradation. This
connection occurs due to binding C-terminal glycine residue of ubiquitin to lysine
residue of the ubiquitinated protein. Consecutive ubiquitin molecules are bound to
one of the lysine residues of the previously linked ubiquitin molecule (Frankland-
Searby, Bhaumik 2012). In this way, polyubiquitin chain is formed, attached to the
substrate protein. The polyubiquitinated proteins are recognized by the 19S
regulatory particle and fed to the 20S core particle where they are degraded through
an ATP-dependent proteolysis (see Fig 1).
The role of the UPS system in cancer
Proper activity of the UPS system is crucial in maintaining the homeostasis in cells.
This system is engaged in regulation of prominent cellular processes such as growth,
mitosis and angiogenesis. Hence, the imbalance in degradation of some proteins may
result in serious disturbance to proliferation.
One of the key proteins, which activity is regulated by the UPS, is the nuclear factor
kappa-light-chain-enhancer of activated B cells (NF-κB). It was discovered for the
Łukasz P. Biały, Przemysław Szymański, Izabela Młynarczuk-Biały
38
first time in murine B leukocytes (Sen, Baltimore 1986). In fact, NF-κB is a group of
transcriptional factors which regulate expression of genes encoding cytokines,
chemokines, immunomodulatory proteins and growth factors (Pahl 1999). In context
of tumor development, permanently activated transcription factor NF-κB, triggers
pro-survival and anti-apoptotic phenotype of cancer cells. These conditions make
tumor cells resistant to apoptosis-inducing chemotherapeutics and enable their
survival despite of difficulties like hypoxia, acidosis, malnutrition, lack of external
stimuli especially in myeloma multiplex cells (Tornatore, Sandomenico et al. 2014).
Figure 1: Schematic presentation of the ubiquitin-proteasome system, deubiquitination process and its key
enzymes participating the process of cytosolic degradation of proteins. (Original Figure)
Cellular effects of proteasome inhibition
Most of cytosolic proteins are ubiquitinated and subsequently degraded within the 26S proteasome. If the system is impaired, for instance under action of proteasome inhibitors, polyubiquitinated proteins accumulate and sequestrate in perinuclear region in a form of protein-rich aggregate named aggresome. (Wojcik, Schroeter et al. 1996). Undegraded protein particles accumulate locally around microtubules and by an unknown mechanism are transported along microtubules to the minus end of microtubules. In the pericentriolar region numerous of such particles form a huge structure of aggresome, which diameter can be similar to the diameter of the nucleus (Paweletz, Wojcik et al. 1996), (Johnston, Ward et al. 1998).
Ub
Ub
E3
Ub
UbUbUbUb
Ub
Ub
Deubiquitinating
enzymes
Ub
E1Ub
Activating enzyme
E2Ub
Conjugating
enzyme
E3
Ub
Ligase
20S {
Proteasome 26S
ββα
α
19S
19S
Direct visualization of undegraded protein sequestration within aggresomes
by proteasome inhibitors
39
Healthy, non-tumour cells are relatively resistant to action of proteasome inhibitors, while tumour cells, which display intensive protein turnover, are impaired by the inhibition of proteasome, resulting in accumulation of undegraded substrates, NF-κB inhibition and subsequently cell death (Mlynarczuk-Bialy, Doeppner et al. 2014; Mlynarczuk-Bialy, Roeckmann et al. 2006; Pleban, Bury et al. 2001).
The knowledge of the special feature that tumour cells are selectively sensitive to action of proteasome inhibitors was an inspiration for design and development of these compounds as anticancer drugs. After approval of the first 20S inhibitor, Bortezomib, second generation inhibitors (Carfilzomib, Marizomib) were studied and subsequently involved in clinical use (Kubiczkova, Pour et al. 2014). Presently, proteasome inhibitors are used in the treatment of multiply myeloma, mantle cell lymphoma and gastrointestinal stromal tumors (Adams 2001; Adams, Kauffman 2004)
In contrast to classical chemotherapeutics (like cisplatin, doxorubicin, etoposide), clinically used proteasome inhibitors display less systemic toxicity (Mlynarczuk-Bialy, Doeppner et al. 2014), however development of tumor cell resistance is a serious limitation of this group of drugs represented by proteasome inhibitors (Adams, Palombella et al. 2000; Merin, Kelly 2015; Richardson, Zimmerman et al. 2016).
Proteasome inhibitors at clinical applications
Bortezomib is the first proteasome inhibitor that was studied in clinical trials, it has shown in vitro and in vivo activity against a variety of malignancies, including myeloma, chronic lymphocytic leukaemia, prostate cancer, pancreatic cancer, and colon cancer. Bortezomib is a dipeptide boronic acid analogue whose trade name is Velcade (formerly known as PS-341, LDP-341 and MLM341). The drug is rapidly cleared from the vascular compartment, however according to pharmacodynamics data bortezomib-mediated proteasome blockade is dose-dependent and reversible. Basing on phase I studies it was demonstrated that bortezomib has manageable toxicities in patients with advanced cancers (Papandreou, Daliani et al. 2004). According to clinical trials determining the dose-limiting toxicity and maximum-tolerated dose, a dose-related 20S proteasome inhibition was seen, with dose-limiting toxicity at 2.0 mg/m
2 (diarrhoea, hypotension) occurring at an average 1-hour post-
dose of >/= 75% 20S PI. Other side effects were fatigue, hypertension, constipation, nausea, and vomiting. No relationship was seen between body-surface area and bortezomib clearance in that study (Papandreou, Daliani et al. 2004). There was evidence of biologic activity (decline in serum prostate-specific antigen and interleukin-6 levels) at >/= 50% 20S inhibition (Papandreou, Daliani et al. 2004).
Carfilzomib (Kyprolis) is a proteasome inhibitor used in treatment of the relapsed or refractory multiple myeloma. Firstly, it was approved as monotherapy, but now it is indicated as a part of combined therapy with dexamethasone or with lenalidomide (Revlimid) and dexamethasone. Carfilzomib is a modified epoxyketone which selectively and irreversibly binds with the 20S proteasome causing its inhibition. The medication is administered via intravenous infusion. The recommended starting dose is 20 mg/m
2. Carfilzomib is metabolized extrahepatically, mainly through peptidase
cleavage and epoxide hydrolysis. It is quickly eliminated by biliary and renal
Łukasz P. Biały, Przemysław Szymański, Izabela Młynarczuk-Biały
40
excretion. Clinical trials showed that the most often adverse effects involve anaemia, diarrhoea, neutropenia, fatigue, thrombocytopenia, pyrexia, hypokalaemia, hypertension, acute renal failure and cardiac failure (Perel, Bliss et al. 2016).
Marizomib (salinosporamide A or NPI-0052) is a novel second generation irreversible proteasome inhibitor which is currently clinically tested. This compound was discovered in Salinispora tropica, marine bacterium, which naturally synthesize it. Marizomib belongs to β-lactone-γ-lactam superfamily of proteasome inhibitors (Potts, Albitar et al. 2011). In comparison to bortezomib and carfilzomib which inhibit mainly β5 (chymotrypsin-like) subunit of the 20S proteasome, marizomib also significantly affects β1 and β2 subunits (trypsin-like and caspase-like, respectively). Thus, it is expected to overcome cancer cells’ resistance to other PIs due to its pan-subunit activity and irreversible binding. Moreover, marizomib showed less side effects than other PIs during phase I clinical trials. The most frequent of them are fatigue, headache, nausea, diarrhoea, dizziness, and vomiting. Additionally, significant peripheral neuropathy, which occurs during treatment by bortezomib, was not observed. In phase 1 clinical trials the patients received marizomib via IV injection and the maximum dose was 0.7 mg/m
2 or 0.5 mg/m
2, depending on the
treatment schedule (Richardson, Zimmerman et al. 2016). Marizomib has also more lipophilic structure than bortezomib and carfilzomib. Therefore, it is expected to penetrate blood-brain barrier and may be efficient in treatment of brain tumours (Di, Lloyd et al. 2016).
MG132 is a peptide-aldehyde proteasome inhibitor obtained from a chinese medicinal plant. Its use is limited to the laboratory research (Guo, Peng 2013).
BSc2118 is another proteasome inhibitor, designed as an analogue of MG132, with very similar inhibitory profile to bortezomib and with relatively low toxicity. Its activity involve inhibition of all three hydrolytic subunits of the 20S proteasome (Braun, Umbreen et al. 2005). The binding of BSc2118 is reversible (Doeppner, Mlynarczuk-Bialy et al. 2012). BSc2118 proved an effective anti-tumour agent against melanoma in vitro (Mlynarczuk-Bialy, Roeckmann et al. 2006) and in a mouse melanoma model (Mlynarczuk-Bialy, Doeppner et al. 2014) and also against multiple myeloma in vitro (Sterz, Jakob et al. 2010).
Proteasome inhibitor induced ER-stress
The inhibition of UPS affects general protein turnover producing proteotoxic conditions and may also influence protein folding in the ER. Conditions affecting protein folding in the ER, such as heat shock, acidosis, action of tunicamycin activate cell adaptive pathway known as the unfolded protein response (UPR) (Lindholm, Korhonen et al. 2017). The UPR is composed of two main elements: on one hand general protein synthesis is decreased to reduce nascent protein delivery to the ER, on the other hand the level of ER resident conformation guards like chaperones, folding enzymes is elevated by up-regulated transcription of adaptive proteins. If these regulatory responses fail to prevent accumulation of misfolded proteins in the ER, these proteins are recognized by ER quality control and restored in the ER, stopped in further maturation. Finally protein which is not properly refolded, is targeted for ER-associated protein degradation (ERAD) (Berner, Reutter et al. 2018). ERAD include retrograde delivery of misfolded proteins to the cytoplasm and
Direct visualization of undegraded protein sequestration within aggresomes
by proteasome inhibitors
41
subsequent degradation by the ubiquitin-proteasome system. Thus the condition and activity of ubiquitin-proteasome system keeps the maintenance of proper ER function especially in stress conditions. In multiple myeloma (MM) cells it was found that treatment of MM cells with proteasome inhibitors (PIs) initiates the UPR by inhibiting the retrograde translocation of misfolded proteins from the ER and that MM cells are highly sensitive to these agents because they produce large amounts of protein, namely immunoglobulin, which must be processed within the ER. Interestingly, it was found that MM cells constitutively express high levels of UPR survival components, but that PI treatment leads to the rapid induction of proapoptotic UPR genes (Obeng, Carlson et al. 2006) .
Clearance of aggresomes by aggrephagy
Another mechanism for clearance of misfolded proteins is aggresome-macro-autophagy/autophagy pathway called aggrephagy (Lamark, Johansen 2012). Unfolded proteins tend to form aggregates that are potentially harmful for the cell, like in the case of numerous neurodegenerative diseases (for instance Alzheimer disease, Parkinson Disease). To prevent such conditions cells developed resistance mechanisms preventing from harmful effects aggregation. Soluble proteins are mainly degraded by the UPS, while large insoluble aggregated proteins can be isolated via isolation membranes within autophagosome for further destruction. The distinguishing of both clearance pathways depends on different tagging of substrates by ubiquitin. Namely the same molecule but differentially attached to the substrate leads either to the UPS or to the lysosomal/autophagy pathway. Which mechanism lead to different ubiquitination of the substrate, remains unknown. It is known that Lys48-linked chains designate proteins for proteasomes whereas Lys63-linked poly-ubiquitin chains facilitate degradation by lysosome/autophagy. However, autophagy receptors show no obvious preference binding Lys63-linked poly-ubiquitin chains over Lys48-linked chains. Thus, how substrates are sorted to their designated degradation pathway remained unclear (Acconcia, Sigismund et al. 2009; Haglund, Dikic 2005; Miranda, Sorkin 2007).
Aggresome disruption as re-sensitization to proteasome inhibitor resistant cancer cells
Nawrocki et al. observed and described the phenomenon of protective role of big aggresomes from death induced by proteasome inhibitors. The spatial cytoplasmic sequestration of undegraded protein in tumour cells had pro-survival effect while aggresome disruption restored tumour cells sensitivity to the action of proteasome inhibitors (Nawrocki, Carew et al. 2006). Sequestration of potentially harmful proteins within aggresome can reduce unfolded protein response and ER-stress thus by incubation of cells with histone deacetylase inhibitors and thereby aggresome disruption turns off this protective effect (see Figure 2).
Histone deacetylases HDAC6 and aggresome
Histone deacetylases (HDACs) is a superfamily of enzymes, which name should be considered rather historically. They take part not only in deacetylation of histones, but their targets comprise plenty of non-histone proteins. Thus, they are also described as lysine deacetylases to show their function more accurately (Dokmanovic, Clarke et al. 2007). HDAC6 belongs to class II HDACs and it is predominantly localized in
Łukasz P. Biały, Przemysław Szymański, Izabela Młynarczuk-Biały
42
the cytoplasm. One of its known functions is deacetylation of α-tubulin which improves cell mobility. Moreover, it displays ability to bind both polyubiquitinated proteins and dynein motors. Hence, HDAC6 plays a crucial role in transport of misfolded protein with use of dynein motors, which results in formation of protein-rich structures called aggresomes. (Kawaguchi, Kovacs et al. 2003) This process has an adverse effect to treatment by proteasome inhibitors. Through sequestration of misfolded proteins and reduced exposition of cell molecules to their harmful action, aggresomes make cancer cells less sensitive to cytotoxic activity of proteasome inhibitors (Rodriguez-Gonzalez, Lin et al. 2008). However, the cytoprotective process of aggresome formation may be disrupted by HDAC inhibitors or siRNA and, consequently, effectiveness of treatment with proteasome inhibitors could be enhanced (Figure 2). Therefore, HDAC inhibitors was tested in preclinical and clinical studies, e.g. in multiple myeloma and chronic myelogenous leukaemia (Garcia-Manero, Yang et al. 2008), (Blum, Advani et al. 2009). Different new selective HDAC6 inhibitors are being developed to increase efficiency of their activity (Boumber, Younes et al. 2011).
Figure 2: The proposed model of aggresome role in cell death and survival upon proteasome inhibition
Big aggresome displays cytoptotective action hence aggresome disruption promotes apoptosis.
The vital inhibitor-based staining procedure of proteasomes
In order to analyze aggresome formation dynamics we develop vital inhibitor-based staining using a novel inhibitor BSc2118 fluorescent derivative.
Material and Methods Proteasome inhibitor
The novel proteasome inhibitor BSc2118 was synthesized as described previously (Braun, Umbreen et al. 2005). The fluorescent variant of BSc2118 (further named as BSc2118-FL) was synthesized as described elsewhere (Mlynarczuk-Bialy, Doeppner
Cell
nucleus
Big
aggre
gate
Inhibitors of
HDAC
Incubation with
proteasome inhibitors
Apoptotic
bodies
No
apoptosis
Aggregate
disruption
Direct visualization of undegraded protein sequestration within aggresomes
by proteasome inhibitors
43
et al. 2014). BSc2118 was dissolved in DMSO at 20 mM and fluorescent Bodipy-BSc2118 was dissolved in DMSO at 1 μg/μl and kept at -20°C prior to use.
Cell culture
NIH-3T3 Cells (ATCC) were grown in DMEM containing stable L-glutamine and supplemented with 10% FCS and antibiotics. Cells were grown in cell culture incubator in a humidified atmosphere of 5% CO2 and passaged every three days.
Confocal microscopy
NIH-3T3 cells were directly incubated with BSc2118-FL at the final concentration of 1μM for indicated time. Subsequently cells were fixed with ice-cold buffered 70% ethanol (pH 7.5; phosphate buffer) for 3 minutes and directly rehydrated (pH 7.5; phosphate buffer). Thereafter cell nuclei were stained with DAPI (Sigma). Specimens were embedded in Vecta shield medium (Vector Laboratories). The specimens were examined with a Leica confocal laser scanning microscope SP5 (Leica Microsystems) equipped with a krypton-argon laser. Sequential scans at a series of optical planes were performed with a 63x oil immersion objective lens through specimens.
Figure 3. The dynamics of aggresome formation in NIH3T3 cells
Cells were incubated with BSc2118-FL (1μM) for 5 min (A) of for 2 h (B), 6 h (C) 24 h (D).
Scale Bars 10μM
Łukasz P. Biały, Przemysław Szymański, Izabela Młynarczuk-Biały
44
Results
The dynamics of aggresome formation in NIH3T3 cells show time-dependent kinetic
spectrum of alternations. In NIH3T3 cells stained for 5 minutes with BSc2118-FL
(FIG 3 A) the cytoplasm is homogenously traced with the fluorescent inhibitor. The
cellular processes are stained as well. The nuclei remain unstained at this time point.
In comparison to shortly stained group, cells incubated with BSc2118-FL for 2 hours
demonstrate increase in signal in perinuclear region. This enhancement of the signal
in perinuclear region corresponds with forming aggresomes as was shown elsewhere
(Mlynarczuk-Bialy, Doeppner et al. 2014) using counterstaining with antibody for
20S and polyubiquitin. This early aggregates are visible after 2 hours of 20S
inhibition. There is a decrease in signal intensity within cellular processes, however
the processes still remain visible. Within cell nucleus there are visible clusters which
are stained with BSc2118-FL at 2h time point (FIG 3 B). At 6 h the bright
fluorescent signal is localized within aggresomes at perinuclear region (FIG 3 C)
whereas relative decrease in fluorescence intensity is observed within NIH3T3
cytoplasmic processes in comparison to the previous group. Starting form 24 h
period of proteasome inhibition, the aggresomes in perinuclear region form bright
fluorescent rounded structures, in nuclei of some cells there is a reduction of
fluorescent clusters that might be a sign of beginning degeneration (FIG 3 D).
The degeneration includes: condensation or even defragmentation of the cell nucleus,
alternations of the cytoplasmic region: vacuolisation due to endoplasmic reticulum
stress, cell membrane blebbing as a sign of apoptosis induction or cell membrane
rupture with release of the aggresome.
Short discussion and conclusion
Fluorescent-inhibitor based staining of enzymes give and interesting alternative for
classical immunolabeling in both microscopy and flow cytometry by reducing of
possible antibody cross-reactivity esp. with multi-color staining (Darzynkiewicz,
Pozarowski et al. 2011), (Marchetti, Gallego et al. 2004) . Inhibitor based detection
of proteasomes in biosensors enables proteasome and immunoproteasome
quantitative measurement in the serum (Patent Pending PL: P.396170; P.396171;
P.417435). The vital inhibitor-based staining procedure of proteasomes (BSc2118-FL
prior to fixation) combined with subsequent ethanol based fixation enables efficient
proteasome staining (Mlynarczuk-Bialy, Doeppner et al. 2014), (Bialy, Kuckelkorn
et al. 2018).
We found this method as promising for fast and effective aggresome formation
studies. Studies on aggresome formation process and development of methods for
their visualisation can lead to invention of diagnostic tools enabling prediction if
proteasome inhibitors can be effective in particular types of malignancy. In times of
more and more personalized medicine and due to facts that proteasome inhibitors are
expensive and not always effective such diagnostic tool would be a great benefit for
patients and the public health system.
Direct visualization of undegraded protein sequestration within aggresomes
by proteasome inhibitors
45
References
Acconcia, F., Sigismund, S., Polo, S. Ubiquitin in trafficking: The network at work. Exp Cell Res.
2009; 315(9):1610-1618.
Adams, J. Proteasome inhibition in cancer: development of PS-341. Semin Oncol. 2001; 28(6):613-619.
Adams, J. The proteasome: structure, function, and role in the cell. Cancer Treat Rev. 2003; 29
Suppl 1:3-9.
Adams, J., Kauffman, M. Development of the proteasome inhibitor Velcade (Bortezomib). Cancer
Invest. 2004; 22(2):304-311.
Adams, J., Palombella, V. J., Elliott, P. J. Proteasome inhibition: a new strategy in cancer treatment.
Invest New Drugs. 2000; 18(2):109-121.
Berner, N., Reutter, K. R., Wolf, D. H. Protein Quality Control of the Endoplasmic Reticulum and
Ubiquitin-Proteasome-Triggered Degradation of Aberrant Proteins: Yeast Pioneers the Path. Annu Rev Biochem. 2018.
Bialy, L.P., Kuckelkorn, U., Henklein, P., J. Fayet, Wilczynski G. M., Kaminski, A., Mlynarczuk-Bialy,
I, Changes in spatio-temporal localization of tripeptidyl peptidase II (TPPII) in murine colon
adenocarcinoma cells during aggresome formation: a microscopy study based on a novel fluorescent proteasome inhibitor, Histol Histopathol, (2018), s. 18042.
Blum, K. A., Advani, A., Fernandez, L., Van Der Jagt, R., Brandwein, J., Kambhampati, S., Kassis, J.,
Davis, M., Bonfils, C., Dubay, M., Dumouchel, J., Drouin, M., Lucas, D. M., Martell, R. E., Byrd, J. C.
Phase II study of the histone deacetylase inhibitor MGCD0103 in patients with previously treated chronic lymphocytic leukaemia. Br J Haematol. 2009; 147(4):507-514.
Boumber, Y., Younes, A., Garcia-Manero, G. Mocetinostat (MGCD0103): a review of an isotype-
specific histone deacetylase inhibitor. Expert Opin Investig Drugs. 2011; 20(6):823-829.
Braun, H. A., Umbreen, S., Groll, M., Kuckelkorn, U., Mlynarczuk, I., Wigand, M. E., Drung, I.,
Kloetzel, P. M., Schmidt, B. Tripeptide mimetics inhibit the 20 S proteasome by covalent bonding
to the active threonines. J Biol Chem. 2005; 280(31):28394-28401.
Darzynkiewicz, Z., Pozarowski, P., Lee, B. W., Johnson, G. L. Fluorochrome-labeled inhibitors of
caspases: convenient in vitro and in vivo markers of apoptotic cells for cytometric analysis. Methods Mol Biol. 2011; 682:103-114.
Di, K., Lloyd, G. K., Abraham, V., MacLaren, A., Burrows, F. J., Desjardins, A., Trikha, M., Bota, D.
A. Marizomib activity as a single agent in malignant gliomas: ability to cross the blood-brain
barrier. Neuro Oncol. 2016; 18(6):840-848.
Doeppner, T. R., Mlynarczuk-Bialy, I., Kuckelkorn, U., Kaltwasser, B., Herz, J., Hasan, M. R.,
Hermann, D. M., Bahr, M. The novel proteasome inhibitor BSc2118 protects against cerebral
ischaemia through HIF1A accumulation and enhanced angioneurogenesis. Brain. 2012; 135(Pt
11):3282-3297.
Dokmanovic, M., Clarke, C., Marks, P. A. Histone deacetylase inhibitors: overview and
perspectives. Mol Cancer Res. 2007; 5(10):981-989.
Frankland-Searby, S., Bhaumik, S. R. The 26S proteasome complex: an attractive target for cancer
therapy. Biochim Biophys Acta. 2012; 1825(1):64-76.
Garcia-Manero, G., Yang, H., Bueso-Ramos, C., Ferrajoli, A., Cortes, J., Wierda, W. G., Faderl, S.,
Koller, C., Morris, G., Rosner, G., Loboda, A., Fantin, V. R., Randolph, S. S., Hardwick, J. S., Reilly, J.
F., Chen, C., Ricker, J. L., Secrist, J. P., Richon, V. M., Frankel, S. R., Kantarjian, H. M. Phase 1 study
of the histone deacetylase inhibitor vorinostat (suberoylanilide hydroxamic acid [SAHA]) in patients with advanced leukemias and myelodysplastic syndromes. Blood. 2008; 111(3):1060-1066.
Łukasz P. Biały, Przemysław Szymański, Izabela Młynarczuk-Biały
46
Guo, N., Peng, Z. MG132, a proteasome inhibitor, induces apoptosis in tumor cells. Asia Pac J Clin
Oncol. 2013; 9(1):6-11.
Haglund, K., Dikic, I. Ubiquitylation and cell signaling. The EMBO Journal. 2005; 24(19):3353-3359.
Johnston, J. A., Ward, C. L., Kopito, R. R. Aggresomes: a cellular response to misfolded proteins. J
Cell Biol. 1998; 143(7):1883-1898.
Kawaguchi, Y., Kovacs, J. J., McLaurin, A., Vance, J. M., Ito, A., Yao, T. P. The deacetylase HDAC6
regulates aggresome formation and cell viability in response to misfolded protein stress. Cell.
2003; 115(6):727-738.
Kubiczkova, L., Pour, L., Sedlarikova, L., Hajek, R., Sevcikova, S. Proteasome inhibitors – molecular
basis and current perspectives in multiple myeloma. J Cell Mol Med. 2014; 18(6):947-961.
Lamark, T., Johansen, T. Aggrephagy: selective disposal of protein aggregates by macroautophagy.
Int J Cell Biol. 2012; 2012:736905.
Lindholm, D., Korhonen, L., Eriksson, O., Koks, S. Recent Insights into the Role of Unfolded Protein
Response in ER Stress in Health and Disease. Front Cell Dev Biol. 2017; 5:48.
Marchetti, C., Gallego, M. A., Defossez, A., Formstecher, P., Marchetti, P. Staining of human sperm
with fluorochrome-labeled inhibitor of caspases to detect activated caspases: correlation with
apoptosis and sperm parameters. Hum Reprod. 2004; 19(5):1127-1134.
Merin, N. M., Kelly, K. R. Clinical Use of Proteasome Inhibitors in the Treatment of Multiple
Myeloma. Pharmaceuticals. 2015; 8(1):1-20.
Miranda, M., Sorkin, A. Regulation of receptors and transporters by ubiquitination: new insights
into surprisingly similar mechanisms. Mol Interv. 2007; 7(3):157-167.
Mlynarczuk-Bialy, I., Doeppner, T. R., Golab, J., Nowis, D., Wilczynski, G. M., Parobczak, K.,
Wigand, M. E., Hajdamowicz, M., Bialy, L. P., Aniolek, O., Henklein, P., Bahr, M., Schmidt, B.,
Kuckelkorn, U., Kloetzel, P. M. Biodistribution and Efficacy Studies of the Proteasome Inhibitor
BSc2118 in a Mouse Melanoma Model. Transl Oncol. 2014; 7(5):570-579.
Mlynarczuk-Bialy, I., Roeckmann, H., Kuckelkorn, U., Schmidt, B., Umbreen, S., Golab, J., Ludwig,
A., Montag, C., Wiebusch, L., Hagemeier, C., Schadendorf, D., Kloetzel, P. M., Seifert, U. Combined
effect of proteasome and calpain inhibition on cisplatin-resistant human melanoma cells. Cancer
Res. 2006; 66(15):7598-7605.
Nawrocki, S. T., Carew, J. S., Pino, M. S., Highshaw, R. A., Andtbacka, R. H., Dunner, K., Jr., Pal, A.,
Bornmann, W. G., Chiao, P. J., Huang, P., Xiong, H., Abbruzzese, J. L., McConkey, D. J. Aggresome
disruption: a novel strategy to enhance bortezomib-induced apoptosis in pancreatic cancer cells. Cancer Res. 2006; 66(7):3773-3781.
Obeng, E. A., Carlson, L. M., Gutman, D. M., Harrington, W. J., Lee, K. P., Boise, L. H. Proteasome
inhibitors induce a terminal unfolded protein response in multiple myeloma cells. Blood. 2006;
107(12):4907-4916.
Pahl, H. L. Activators and target genes of Rel/NF-kappaB transcription factors. Oncogene. 1999;
18(49):6853-6866.
Papandreou, C. N., Daliani, D. D., Nix, D., Yang, H., Madden, T., Wang, X., Pien, C. S., Millikan, R.
E., Tu, S. M., Pagliaro, L., Kim, J., Adams, J., Elliott, P., Esseltine, D., Petrusich, A., Dieringer, P.,
Perez, C., Logothetis, C. J. Phase I trial of the proteasome inhibitor bortezomib in patients with
advanced solid tumors with observations in androgen-independent prostate cancer. J Clin Oncol.
2004; 22(11):2108-2121.
Paweletz, N., Wojcik, C., Schroeter, D., Finze, E. M. Are proteasomes involved in the formation of
the kinetochore? Chromosome Res. 1996; 4(6):436-442.
Direct visualization of undegraded protein sequestration within aggresomes
by proteasome inhibitors
47
Perel, G., Bliss, J., Thomas, C. M. Carfilzomib (Kyprolis): A Novel Proteasome Inhibitor for
Relapsed And/or Refractory Multiple Myeloma. P T. 2016; 41(5):303-307.
Pickart, C. M., Fushman, D. Polyubiquitin chains: polymeric protein signals. Curr Opin Chem Biol.
2004; 8(6):610-616.
Pleban, E., Bury, M., Mlynarczuk, I., Wojcik, C. Effects of proteasome inhibitor PSI on neoplastic
and non-transformed cell lines. Folia Histochem Cytobiol. 2001; 39(2):133-134.
Potts, B. C., Albitar, M. X., Anderson, K. C., Baritaki, S., Berkers, C., Bonavida, B., Chandra, J.,
Chauhan, D., Cusack, J. C., Jr., Fenical, W., Ghobrial, I. M., Groll, M., Jensen, P. R., Lam, K. S., Lloyd,
G. K., McBride, W., McConkey, D. J., Miller, C. P., Neuteboom, S. T., Oki, Y., Ovaa, H., Pajonk, F.,
Richardson, P. G., Roccaro, A. M., Sloss, C. M., Spear, M. A., Valashi, E., Younes, A., Palladino, M.
A. Marizomib, a proteasome inhibitor for all seasons: preclinical profile and a framework for
clinical trials. Curr Cancer Drug Targets. 2011; 11(3):254-284.
Richardson, P. G., Zimmerman, T. M., Hofmeister, C. C., Talpaz, M., Chanan-Khan, A. A., Kaufman, J.
L., Laubach, J. P., Chauhan, D., Jakubowiak, A. J., Reich, S., Trikha, M., Anderson, K. C. Phase 1
study of marizomib in relapsed or relapsed and refractory multiple myeloma: NPI-0052-101 Part 1. Blood. 2016; 127(22):2693-2700.
Rodriguez-Gonzalez, A., Lin, T., Ikeda, A. K., Simms-Waldrip, T., Fu, C., Sakamoto, K. M. Role of the
aggresome pathway in cancer: targeting histone deacetylase 6-dependent protein degradation. Cancer Res. 2008; 68(8):2557-2560.
Sen, R., Baltimore, D. Multiple nuclear factors interact with the immunoglobulin enhancer
sequences. Cell. 1986; 46(5):705-716.
Sterz, J., Jakob, C., Kuckelkorn, U., Heider, U., Mieth, M., Kleeberg, L., Kaiser, M., Kloetzel, P. M.,
Sezer, O., von Metzler, I. BSc2118 is a novel proteasome inhibitor with activity against multiple
myeloma. Eur J Haematol. 2010; 85(2):99-107.
Tanaka, K. The proteasome: Overview of structure and functions. Proceedings of the Japan
Academy. Series B, Physical and Biological Sciences. 2009; 85(1):12-36.
Tornatore, L., Sandomenico, A., Raimondo, D., Low, C., Rocci, A., Tralau-Stewart, C., Capece, D.,
D'Andrea, D., Bua, M., Boyle, E., van Duin, M., Zoppoli, P., Jaxa-Chamiec, A., Thotakura, A. K.,
Dyson, J., Walker, B. A., Leonardi, A., Chambery, A., Driessen, C., Sonneveld, P., Morgan, G.,
Palumbo, A., Tramontano, A., Rahemtulla, A., Ruvo, M., Franzoso, G. Cancer-selective targeting of
the NF-kappaB survival pathway with GADD45beta/MKK7 inhibitors. Cancer Cell. 2014;
26(4):495-508.
Wojcik, C. Proteasomes in apoptosis: villains or guardians? Cell Mol Life Sci. 1999; 56(11-12):908-917.
Wojcik, C. Regulation of apoptosis by the ubiquitin and proteasome pathway. J Cell Mol Med.
2002; 6(1):25-48.
Wojcik, C., Schroeter, D., Wilk, S., Lamprecht, J., Paweletz, N. Ubiquitin-mediated proteolysis
centers in HeLa cells: indication from studies of an inhibitor of the chymotrypsin-like activity of the proteasome. Eur J Cell Biol. 1996; 71(3):311-318.
48
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
Piotr Strus1, Kamil Lisiecki
2, Zbigniew Czarnocki
2, Izabela Młynarczuk-Biały
3*,
Łukasz P. Biały3
1. HESA Histology and Embryology Students Association at the Department for Histology and Embryology Centre for Biostructure Research, Warsaw Medical University, Poland
2. Faculty of Chemistry, University of Warsaw, Pasteura 1, Warsaw 02-093, Poland 3. Department for Histology and Embryology Centre for Biostructure Research,
Warsaw Medical University, Poland * Corresponding author: [email protected], Chalubinskiego 5, 02-004 Warszawa, Poland
Abstract
The systemic treatment of advanced cancer, where local surgery is ineffective, still offers numerous difficulties due to therapy resistance, side effects and relapse. Thus, novel therapeutics are ongoing. The scaffolds for novel drugs often derive from naturally occurring substances. One of such example is podophyllotoxin – a plant derived substance. Podophyllotoxin itself is indicated in local anticancer treatment and due to its’ substantial toxicity is not recommend for systemic therapy. However, its analogues like etoposide is an element of systemic anticancer therapy administered in many malignancies. Limiting for etoposide is bone marrow depression and secondary leukemia induced in susceptible individuals. Thus, less toxic and more effective substances are needed for cancer therapy. Due to complexity of podophyllotoxin molecule it’s scaffold is nowadays intensively studied as a source for novel therapeutic substances. Within this project our team aimed to modify the molecule of podophyllotoxin to obtain novel derivatives that were screened for anticancer potential. Furthermore, one of new compound was conjugate of podophyllotoxin and benzothiazole. Benzothiazole is widely used in research because of antitumor, antibacterial, anticonvulsant, anti-inflammatory and other activities of its derivatives. The obtained derivates turned out to be less toxic for normal fibroblasts in comparison to parental podophyllotoxin.
Keywords: podophyllotoxin, benzothiazole, cancer, cell viability,
Introduction
A number of studies and many clinical examples demonstrate that the treatment of cancer often encounters numerous difficulties. The lack of therapy response, relapse, or even toxic side effects are a few of such examples (Miller, Siegel et al. 2016; Siegel, Miller et al. 2016). Moreover, in developed countries, the prevalence of cancer shows increasing tendency and application of cancer medicals makes its treatment one of the most costly (Miller, Siegel et al. 2016; Siegel, Miller et al. 2016). In advanced stages of malignant disease the existing treatment remains often ineffective or relapse is diagnosed (Miller, Siegel et al. 2016; Siegel, Miller et al. 2016). Hence, to overcome these difficulties new drugs that are effective, specific, less toxic and relatively cheap are ongoing.
Thus, the ideal anti-cancer drug – would be one that combined two features: nontoxic for normal cells and effectiveness for tumor cells (Gewirtz, Holt et al. 2007; Mattson, Calabrese 2009). It also should be cost-effective and have uncomplicated application.
Over last years, natural products continued to play a highly significant role in the drug discovery and development process. Thus in the area of anticancer drugs, according to published data including 25-year period of drug development, 47% of
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
49
all approved antitumor drugs worldwide were either natural products or directly derived therefrom (Newman, Cragg 2007).
Our team was inspired by the ideas above. Therefore, we designed and synthesized derivatives of podophyllotoxin (PTOX) which is the anti-cancer, plant-derived drug (Ardalani, Avan et al. 2017; Zhang, Rakesh et al. 2018). PTOX belongs to the class of aryltetralinlactone cyclolignans and it is purer and more stable form of podophyllin (Gordaliza, Garcia et al. 2004; Saraiva, Vega et al. 2007; Silva, Coimbra et al. 2007; Silva, Martins et al. 2009; Srivastava, Negi et al. 2005; Yousefzadi, Sharifi, Behmanesh et al. 2010; Yousefzadi, Sharifi, Chashmi et al. 2010).
The crude Podophyllum peltatum plant extract is named podophyllin. However, podophyllin as a mixture of different active substances contains little active compound and numerous harmful ingredients of high toxicity, thus does not comply with the WHO guidelines for plant derived treatments and is not recommended for clinical treatment protocols (von Krogh, Longstaff 2001).
The purified form of podophyllin-derived active compound is named podophyllotoxin. This polycyclic compound was first isolated in 1880 from the Podophyllum peltatum species. However, there are few other species, such as Berberidaceae, Apocynaceae, Polygalaeae, Apiaceae, Linaceae and other, which contain PTOX and analogs. PTOX contains five rings, which are methylenedioxy, two tetrahydronaphthalene, lactone and aryl rings (Scheme 1). PTOX is common and the most effective cure for anogenital warts; especially for condyloma acuminatum caused by human papilloma virus (HPV) (Gilson, Ross et al. 2009; Keri, Patil et al. 2015; Seth 2015).
The exact mechanism of antineoplastic action of PTOX is unknown. However, it has been shown that PTOX due to binding tubulin, a subunit of microtubules, prevents the polymerization of tubulin into microtubules. Affinity and site of binding is similar to colchicine, well-known plant-derived (Colchicum autumnale) drug with similar mechanism (Spasevska, Ayoub et al. 2017), although PTOX binds more rapidly and reversibly whereas colchicine bind irreversibly. Thereby, this mechanism could include the cell cycle arrest at mitosis and impede the formation of the mitotic-spindle (Ardalani, Avan et al. 2017). Interestingly, PTOX competitively inhibits the binding of colchicine (Lu, Chen et al. 2012). As mentioned above, PTOX alone is used in a local treatment of anogenital warts (Lewis, Goldmeier 1995; Longstaff, von Krogh 2001). Clinical results with systemic application of PTOX were disappointing due to severe gastrointestinal side effects (Imbert 1998), therefore PTOX won’t be approved for systemic treatment. However, continuous efforts concerning the synthesis of PTOX analogues led to the discovery of new anticancer drugs. For example its derivative, etoposide is currently used in the clinic for the treatment of a variety of malignancies including lung and testicular cancers, glioblastoma multiforme, lymphoma and nonlymphocytic leukemia (Bohlin, Rosen 1996). Teniposide another PTOX derivate, is applied for the treatment of childhood acute lymphocytic leukemia (Bohlin, Rosen 1996). In contrast to the parent podophyllotoxin, which binds to tubulin and inhibits microtubule assembly, etoposide has a distinct mechanism of action (Kumar, Kumar et al. 2011). In fact, molecules such as etoposide, amsacrine, and mitoxantrone are topoisomerase II inhibitors that induce cell death by enhancing topoisomerase II-mediated DNA cleavage through stabilization of the transient DNA/topoisomerase II cleavage complex (Kumar, Kumar et al. 2011; Wang 1996).
Piotr Strus, Kamil Lisiecki, Zbigniew Czarnocki, Izabela Młynarczuk-Biały, Łukasz P Biały
50
Like numerous anticancer drugs, also etoposide is not free of toxic side effects. Bone marrow depression is a serious, dose-limiting side effect diagnosed in patients receiving etoposide (Leroy, Kajava et al. 2001). The use of effective doses of etoposide is also associated with an increased risk of secondary acute myelogenous leukemia. For this reason, there exist an urgent need for development of more potent analogues of podophyllotoxin characterized by less toxic side effects.
Within this study we aimed to modify the molecule of phodophyllotoxin by binding it with functional groups to obtain novel compounds with potential for systemic application.
Therefore, three new cyclolignans were synthesized using photocyclization or acid-catalyzed cyclization strategy
(Lisiecki, Krawczyk et al. 2017; Lisiecki, Krawczyk et
al. 2016). One of the new compounds, named KL3, is a conjugate of two molecules with anti-tumor activity, podophyllotoxin and benzothiazole. The analogues of benzothiazole and its derivates are widely used in pharmaceutical research, because of their biological and pharmacological properties. Benzothiazole is a class of heterocyclic compounds having 2 hetero atoms: sulphur and nitrogen (Seth 2015). Benzothiazole is a privileged bicyclic ring system with multiple applications. In the 1950s, 2-aminobenzothiazoles were intensively studied as central muscle relaxants. Several years later riluzole was found to interfere with glutamate neurotransmission (6-trifluoromethoxy-2-benzothiazolamine, PK-26124, RP-25279, Rilutek). After that benzothiazole derivatives have been also studied as anticancer agents (Youssef, Noaman 2007). The potential of benzothiazole and related compounds was examined in breast tumors, regardless of estrogen receptor status, and against ovarian, renal, lung, and colon cancer cells (Kashiyama, Hutchinson et al. 1999).
Derivatives of benzothiazole serve as scaffolds for experimental drug design without technical difficulties (Ahmed, Yellamelli Valli Venkata et al. 2012; Ali, Siddiqui 2013; Bhuva, Kini 2010; Bolelli, Musdal et al. 2017; Henriksen, Hauser et al. 2007; Kok, Gambari et al. 2008; Liu, Lin et al. 2008). This versatile and unique compounds have received remarkable attention because of antitumor activity against breast (both estrogen receptor-positive and estrogen receptor-negative cell lines), ovarian, colon lung and renal cell lines and their interesting pharmacological activities, including anticonvulsant, analgesic, anti-tumor, antibacterial, anti-microbial, skeletal muscle relaxant and other activities such as: antidiabetic, anti-inflammatory, anticonvulsant, antiviral, antioxidant, antitubercular, antimalarial, antiasthmatic, antihelmintic, photosensitizing and diuretic (Amnerkar, Bhusari 2010; Bondock, Fadaly et al. 2009, 2010; Bradshaw, Wrigley et al. 1998; Caleta, Grdisa et al. 2004; Charris, Monasterios et al. 2002; Duggan, Lewis et al. 2009; Gurupadayya, Gopal et al. 2008; Hutchinson, Jennings et al. 2002; Kashiyama, Hutchinson et al. 1999; Kumbhare, Kosurkar et al. 2012; Sharma, Sinhmar et al. 2013; Shi, Bradshaw et al. 1996; Yoshida, Hayakawa et al. 2005).
In 2016 we have developed methodology for the stereoselective synthesis of cyclolignans related to podophyllotoxin (Lisiecki, Krawczyk et al. 2016). It involves the use of L-prolinol as a chiral auxiliary and continuous flow irradiation of a chiral atropoisomeric 1,2-bisbenzylidenesuccinate amide ester. Based on this discovery, a formal synthesis of (-)-podophyllotoxin and total synthesis of (+)-epigalcatin were
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
51
completed (Lisiecki, Czarnocki 2018; Lisiecki, Krawczyk et al. 2017; Lisiecki, Krawczyk et al. 2016).
New compounds should be soluble and stable in water solution. They need to have acceptable bioavailability, easily pass through cancer cells membrane and they should provide the evidence of high selectivity against cancer cells. An unsophisticated and cheap way to fast screening of new compounds is their incubation with selected cell lines in a cell culture. We can observe the effects of novel compounds on proliferation of cells, cell morphology and identify signs for degeneration using microscopic imaging as well as we can study cytotoxic/cytostatic effects within viability assays (Hejchman, Taciak et al. 2015; Mlynarczuk-Bialy, Doeppner et al. 2014; Mlynarczuk-Bialy, Roeckmann et al. 2006).
An assessment of cell survival is essential to test if compounds have cytotoxic/cytostatic effects. Cytotoxic effect results from direct harmful effect of examined compound on cells and cytostatic effect is a consequence of decreased cell proliferation that secondly may also lead to cell death. One of the best established viability assays is crystal violet decolorization assay (CVDA). In this method staining of attached, living cells with crystal violet dye, which binds to proteins and DNA, allow to detect the amount of remaining alive cells. Dead cells lose their adherence and by washing with PBS are eliminated from the population of cells, reducing the amount of bound crystal violet in a culture in comparison to the control, vehicle-treated group. This is a quick and if validated, adequate screening method that can be used for the estimation of cytostatic/cytotoxic effects after treatment with potential anticancer compounds (Hejchman, Taciak et al. 2015; Mlynarczuk-Bialy, Doeppner et al. 2014; Mlynarczuk-Bialy, Roeckmann et al. 2006). Therefore, to estimate the number of living cells we used CVDA.
To improve our results and determine cytopathic effects of novel compounds we also used light microscope. Due to specification of our Juli-Stage (NanoEnTek) microscope we were able to verify proliferation and amount of cells and also take photographs of cells incubated with new compounds.
Results and discussion Design of PTOX derivates
As shown in the available literature, cyclolignans are compounds with remarkable potential. However, most of derivatives are obtained by modifying natural compound (mainly podophyllotoxin). This does not allow for accessing other stereoisomers because some stereochemical features in the main carbon skeleton of podophyllotoxin cannot be changed during synthesis. What is more, the structural complexity of podophyllotoxin, derived from the presence of four stereogenic carbons in ring C has limited most of the structural activity relationship obtained by derivatization of the parent natural product rather than by the total synthesis. These features make it necessary to search for derivatives of PTOX with simplified structures, which can be obtained via short synthetic pathway from simple starting materials. Such are needed, since the therapeutic potential of PTOX and its derivatives is often limited by problems of drug resistance (Gupta 1983), hydrophobicity and low selectivity (Gordaliza, Garcia et al. 2004). Although many methods of stereoselective synthesis of cyclolignans are known, their main drawback is the lack of generality and often
Piotr Strus, Kamil Lisiecki, Zbigniew Czarnocki, Izabela Młynarczuk-Biały, Łukasz P Biały
52
high cost of synthesis. To meet this, we developed new methodology (Yu, Che et al. 2017) based on the use of cheap and readily available starting materials such as simple aromatic aldehydes, succinic acid ester, and L-prolinol as a chiral auxiliary. The key step of the synthesis is photocyclization. It has been shown that this process occurs at much higher yield when carried out under flow conditions. This approach has an additional advantage, it allows for unlimited scale of synthesis, which is an important issue in the multi-step syntheses. In addition, avoiding the use of sophisticated organocatalysts or catalysts containing heavy metals, will be extremely important because of the elimination of the possibility of contamination with toxic metals. Our method is therefore extremely advantageous in the context of the synthesis of compounds with potential pharmaceutical application.
In this study, we investigated anticancer properties of three new derivatives of PTOX - 2, 3, 4. Compound 2 is a derivative of 1-arylnaphthalene which poses anti-cancer activity (Hui, Zhao et al. 2011). Derivative 3 contains in its structure L-prolinol moiety which is a H-bond donor and its conformationally restricted main chain may improve receptor binding. Compound 4 is a conjugate of podophyllotoxin (red) and benzothiazole moiety (blue). In recent years, benzothiazoles have been recognized as promising pharmacophores with diverse biological properties including anti-cancer activity (Sharma, Sinhmar et al. 2013). All of those compounds are in good accordance with Lipinski's rule of five (Lipinski, Lombardo et al. 2001) – less than 5 hydrogen bond donors, less than 10 hydrogen bond acceptors and molecular mass less than 500 daltons or not much above. Compounds 2 and 3 were obtained by us previously
(Lisiecki, Krawczyk et al. 2017; Lisiecki, Krawczyk et al.
2016) and compound 4 was synthetized from podophyllic aldehyde (total synthesis of this compound was already reported by us (Lisiecki, Krawczyk et al. 2016)) and 2-aminothiophenol (Scheme 1). The synthesis of presented compounds is highly efficient and starts from cheap and readily available substrates.
Scheme 1. Chemical structures of podophyllotoxin (PTOX), compounds 2 -KL1 and 3 -KL2 (top) and synthesis of compound 4 -KL3 (bottom).
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
53
Cytotoxicity assay
We performed crystal violet viability assay to examine the sensitivity of tumor and non-tumor cells towards novel derivates in comparison to the parental one PTOX.
Compound KL1 did not induce any significant cytotoxic/cytostatic effect against HeLa, MDA-MB, DU145 and CFPAC. Also, in comparison to PTOX – a model toxic drug for NIH-3T3 cells, the novel KL1 compound was 100 times less toxic. Similar results were obtained for KL2 compound. The values of IC50 for this substance on all but one cell line (MCF7) was good above 140 µM and ranged from 140 µM to 490 µM (Table 1). In addition, in PC3 cell line the IC50 for KL1 was 48 µM. Also, non-tumor NIH-3T3 cells turned out to be 90 times less susceptible to the action to KL2 in comparison to the parental PTOX.
These results showed that compounds KL1 and KL2 by chemical modifications of parental PTOX exhibited lower cytotoxicity than PTOX itself. Moreover, both compounds turned out to be highly effective in induction of cytotoxic/cytostatic effects on MCF7 cells and in comparison to non tumor mouse fibroblasts they were tumor cell specific with selectivity index (SI) of 7.9 and 4.5 for KL1 and KL2, respectively (Table 2).
Compound KL3 turned out to be most active in induction of cytotoxic/cytostatic effects from all of three novel compounds. The IC50 values for tumor cells were form 0.5 µM to 10 µM. Only MCF7 breast cancer cells remained relatively resistant to KL3 with IC50 of 116 µM.
Such a good anti tumor profile of KL3 was accompanied with less toxicity towards NIH-3T3 cells in comparison to the parental PTOX.
As suspected – combination of two anti tumor agents within one molecule (KL3) improved the antitumor activity, in comparison to KL1 and KL2, and what is also very desiderated this drug was more toxic for most cancer cell lines than to no-tumor fibroblasts.
Table 1 Compilation of IC50 value of novel compounds and of the parental POTX estimated in 7 cell lines. IC50 – Concentration of compound corresponding to 50% growth inhibition after 48-hour incubation.
Inhibitory concentration (IC 50) [μM]
Compound
Immortalized non tumorogenic cells
Tumor derived cells
Fibroblast Cervix cancer
Breast cancer Prostate cancer Pancreas cancer
NIH-3T3 HeLa MDA-MB MCF7 PC3 DU145 CFPAC
KL1 55 165 300 7 48 425 370
KL2 45 140 490 10 144 290 400
KL3 1 0.5 0.7 116 0.7 6.8 10
PTOX 0.5 0.5 125 123 10 1 100
Piotr Strus, Kamil Lisiecki, Zbigniew Czarnocki, Izabela Młynarczuk-Biały, Łukasz P Biały
54
The standard error bars did not extend 10% and for more transparency are not
included in the table. Results are from 3 independent experiments. The selectivity index is helpful for comparison between different tumor cell lines.
However, when IC50 values are high, the comparison between substances does not
matter. Obtained substances KL1 and KL2 are less toxic. Thus, we have obtained
leading structures of lower cytotoxicity than parental PTOX. In turn, KL3 as
compound represented by covalently bound of two anti tumor substances turned out
to be the most effective with the highest anti tumor activity from all studied
substances. The IC50 value for KL3 was below 1 μM for three cancer cell lines.
Table 2 Tumor cell selectivity
Selectivity Index (SI)
Compound Tumor derived cells
Cervix cancer Breast cancer Prostate cancer Pancreas cancer
HeLa MDA-MB MCF7 PC3 DU145 CFPAC
KL1 0.3 0.2 7.9 1.1 0.1 0.1
KL2 0.3 0.1 4.5 0.3 0.2 0.1
KL3 2.0 1.4 0.009 1.4 0.1 0.1
PTOX 1.0 0.004 0.004 0.1 0.5 0.005
Microscopy analysis
A series of images was recorded and representative regions of particular groups is
shown at Fig 1. Microscopic experiments were performed on NIH-3T3 cells conducted
from disaggregated BALB/c mouse embryos. They are extremely sensitive to contact
inhibition and are highly susceptible to transformation by SV40 VIRUS and murine
sarcoma virus (Labruère, Gautier et al. 2010).
In comparison to the control group, demonstrating spindle-shaped morphology of
NIH-3T3 cells that are elongated with long processes and share 95% of confluence,
both treated groups vary in cell count and shape from the control group. Cells treated
with 10 µM or 100 µM PTOX demonstrate 60 and 50% of confluence, respectively.
In comparison to the controls, their processes are shortened and thickened, the
perinuclear region is enlarged, the total surface occupied by a single cell is enlarged.
We can observe also signs of cell degeneration like small rounded cells and pyknotic
cells that’s number increases by increasing concentration of PTOX.
Compound KL3 induced shortness of NIH-3T3 processes similarly to parental
PTOX, also the perinuclear region was enlarged. Moreover, this enlargement was
more pronounced than in PTX treated group. The ratio of degenerating cells was also
higher than in PTX treated cells. We observed numerous pyknotic cells in KL3 100
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
55
µM treated group and some amount of big rounded cells – suggesting mitotic
catastrophe as a reason of such perturbations (Figure 1). Even if such initial compounds might have diminished cytotoxic potencies compared
with the parent PTOX, the ease of preparation of carefully designed libraries of
analogues would lead to more informative studies and expeditious structure
optimization. In this regard we have obtained less toxic scaffolds represented by KL1
and KL2 and a highly effective KL3 that is twice less toxic for NIH-3T3 cells in
comparison to PTOX and turned out to be selective for tumor cells. To improve the
molecules, further modifications should be designed and performed combined with
studies on detailed mechanisms of action of these compounds.
Materials and methods
Synthesis
Methyl 4-(benzo[d][1,3]dioxol-5-yl)-5,6,7-trimethoxy-3-methyl-2-naphthoate (2) was
prepared in accordance with the procedure previously described.3
(5R,6R)-Methyl 7-((S)-2-(hydroxymethyl)pyrrolidine-1-carbonyl)-5-(3,4,5-trimethoxy-
phenyl)-5,6-dihydronaphtho[2,3-d][1,3]dioxole-6-carboxylate (3) was prepared in
accordance with the procedure previously described.4
(5R,6R)-Methyl 7-(benzo[d]thiazol-2-yl)-5-(3,4,5-trimethoxyphenyl)-5,6-dihydronaphtho
[2,3-d][1,3]dioxole-6-carboxylate (4). A mixture of podophyllic aldehyde (prepared
in accordance with the procedure previously described4) (64 mg, 0.15 mmol, 1
equiv.), 2-aminothiophenol (19.7 mg, 0.16 mmol, 1.05 equiv.) and Na2S2O5 (30 mg,
0.16 mmol, 1.05 equiv.) in 1 mL of DMSO was heated at 120˚C for 2 h. The reaction
mixture was allowed to cool to room temperature, excess water was added, and
yellow solid precipitate was collected by filtration. The precipitate was washed with
water, dried and purified on a silica gel column, using ethyl acetate in n-hexane
(gradient, from 0% to 10%) as an eluent. A yellow solid was obtained (52 mg, 0.098
mmol, 65%). M.p. 165–166°C. Rf (50% AcOEt/n-hexane) 0.54. [α]D25 = 281.0 (c
0.1, CHCl3). 1H NMR (CDCl3, 300 MHz): δ 7.95 (dd, J1 = 8.1 Hz, J2 = 0.6 Hz,
1H), 7.84 (dd, J1 = 7.8 Hz, J2 = 0.6 Hz, 1H), 7.43 (m, 2H), 7.35 (dt, J = 7.8, 1.2 Hz,
1H), 6.86 (s, 1H), 6.63 (s, 1H), 6.56 (s, 2H), 5.98 (d, J = 1.5 Hz, 1H), 5.96 (d, J = 1.5
Hz, 1H), 4.61 (d, J = 7.5 Hz, 1H), 4.44 (d, J = 7.8 Hz, 1H), 3.89 (s, 3H), 3.84 (s, 6H),
3.43 (s, 3H). 13C NMR (CDCl3, 75 MHz): δ 171.7, 167.1, 153.8, 153.3, 148.6,
146.5, 137.4, 134.8, 134.4, 133.4, 131.7, 128.5, 127.0, 126.2, 125.3, 123.1, 121.4,
108.8, 108.8, 106.6, 101.4, 77.4, 77.0, 76.6, 60.9, 56.1, 51.7, 49.1, 48.5. HRMS (ESI)
m/z: calcd for C29H25NO7SH [M+H]+, 532.1424; found, 532.1440.
Reagents
PROX (SIGMA) and Novel compounds were dissolved in DMSO (SIGMA) at 100
mM and kept in the fridge until examination.
Piotr Strus, Kamil Lisiecki, Zbigniew Czarnocki, Izabela Młynarczuk-Biały, Łukasz P Biały
56
Cell Culture
Mouse NIH-3T3 fibroblasts (ATCC, USA) were cultured in Dulbeco (Biochrom,
Berlin, Germany) supplemented with 10% heat-inactivated FCS, penicillin (100
U/ml), streptomycin 100 µg/ml.
For analysis of antitumor activity of examined compounds, following cell lines were
used: MDA-MB-431, HeLa, PC3 and DU145. All of cell lines was obtained from
ATCC. All cells were cultured in media recommended by suppliers. Then
supplemented with standard antibiotics and 10% FCS all from Sigma-Aldrich. Cells
were kept in 25 cm2 tissue flasks (Greiner, Berlin, Germany) and passaged every 2-3
days.
Cell Proliferation Assay
The cytotoxic/cytostatic effects of novel compounds on culture cells were examined
in vitro using the crystal violet assay, as previously described (Mlynarczuk-Bialy et
al., 2006). Briefly, cells (5 x 103 cells/well) were seeded in 96-well microtiter plates
(BD, Biosciences, San Jose, California, USA) and incubated with serial dilutions of
examined compounds. Inhibitors were added in quadruplicate to a final volume of
200 µL. Appropriate volumes of culture medium, supplemented with DMSO
(<0.1%) were added as controls. After an incubation period of 24, 48 or 72 hours,
cells were washed once with PBS, fixed with 70% ethanol for 30 min and finally
stained with 0,1% crystal violet in PBS for 30 min and washed carefully with water
to remove unbound dye. The remaining dye was eluted by 1% SDS in water and
determined at 550 nM. Cytostatic/cytotoxic effect was expressed as relative viability
of treated cells (% of control cells incubated with medium only) and was calculated
as follows: relative viability = (Ae-Ab) x 100/(Ac-Ab), where Ab is background
absorbance, Ae is experimental absorbance and Ac is the absorbance of untreated
controls.
Light microscopy
In order to assess the influence of novel derivates on morphology of examined cells,
after 48 h of incubation period, cells form cell culture were directly imaged by
a phase contrast microscope (Juli, NanoEnTek) at magnification of 100x.
In comparison to the controls (A,B) we observe dose-dependent cytopathic effects
including shortness of cytoplasmic processes and thickness of perinuclear region. In
the higher drug concentration, we observe also cell rounding and detachment from
the bottom. In comparison to PTOX (C,D) the effects induced by KL3 (E,F) are
similar but more pronounced (advanced).
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
57
Figure 1: Photomicrographs of NIH-3T3 cells:
Untreated cells (A, B smaller and bigger magnification, respectively), C,D: cells treated with PTOX 10 µM or
100 µM – respectively, or E,F groups treated with KL3 10 µM or 100 µM – respectively. Scale bars 50 µM
Summary and conclusion
We aimed to create novel compounds of anti-cancer drug – PTOX. We planned and
designed the synthesis of more effective as well as less toxic compounds.
Furthermore, we were also focused on economic aspects and would like to reduce
costs of synthesis. Due to these assumptions we planned to use cheap and easily
available substrates. Moreover, in the process of synthesizing we don’t need to use
heavy metal catalysts, and thus there is no risk of contamination of compounds
A B
C D
E F
Piotr Strus, Kamil Lisiecki, Zbigniew Czarnocki, Izabela Młynarczuk-Biały, Łukasz P Biały
58
obtained with heavy metals (whose content in the drug must be at a very low level).
The next advantage form the used methodology, thanks to the use of flow conditions
in the photocyclic process, it was easy to conduct synthesis on a large scale. This
methodology also allowed obtaining numerous derivatives with differently decorated
rings, thanks to which it was possible to influence parameters such as: solubility,
bioavailability, stability. What is interesting, when it comes to the KL3 compound
itself, it was the first synthesis of the PTOX and benzothiazole congeners with the cis
configuration at the C1 and C2 carbon atoms.
For in vitro tests of new substances, we applied low-cost and well-standardized
viability assays in combination with bio-imaging of the tested cells. We showed that
the applied methods are able to verify in an unambiguous manner whether the
examined chemical structures exhibit biological effects.
Result of cytotoxicity assay and microscopic analysis proved, novel compounds are
less toxic and more effective in comparison to parental PTOX. That is a promising
result because the toxicity of PTOX excludes its systemic application. Anti cancer
KL3 compound with better effectiveness and lover toxicity can be useful in more
applications as the initial substances.
In summary, we obtained KL3 derivative, that is less toxic than the parental one, in
an innovative synthesis process that allows to synthesize large quantities of the
product in its pure form - this work is therefore a comprehensive description of the
compound in cell culture-based model. Its future testing in animal models can verify
pre-clinical potential of KL3 compound.
Bibliography
Ahmed, K., Yellamelli Valli Venkata, S., Mohammed, N. A., Sultana, F., Methuku, K. R. Recent
advances on structural modifications of benzothiazoles and their conjugate systems as potential chemotherapeutics. Expert Opin Investig Drugs. 2012; 21(5):619-635.
Ali, R., Siddiqui, N. Biological Aspects of Emerging Benzothiazoles: A Short Review. Journal of Chemistry. 2013; 2013:12.
Amnerkar, N. D., Bhusari, K. P. Synthesis, anticonvulsant activity and 3D-QSAR study of some prop-2-eneamido and 1-acetyl-pyrazolin derivatives of aminobenzothiazole. Eur J Med Chem. 2010; 45(1):149-159.
Ardalani, H., Avan, A., Ghayour-Mobarhan, M. Podophyllotoxin: a novel potential natural anticancer agent. Avicenna J Phytomed. 2017; 7(4):285-294.
Bhuva, H. A., Kini, S. G. Synthesis, anticancer activity and docking of some substituted benzothiazoles as tyrosine kinase inhibitors. J Mol Graph Model. 2010; 29(1):32-37.
Bohlin, L., Rosen, B. Podophyllotoxin derivatives: Drug discovery and development. Drug Discov. Today. 1996; 1:343-351.
Bolelli, K., Musdal, Y., Aki-Yalcin, E., Mannervik, B., Yalcin, I. Synthesis and activity mechanism of some novel 2-substituted benzothiazoles as hGSTP1-1 enzyme inhibitors. SAR QSAR Environ Res. 2017; 28(11):927-940.
Bondock, S., Fadaly, W., Metwally, M. A. Enaminonitrile in heterocyclic synthesis: synthesis and
antimicrobial evaluation of some new pyrazole, isoxazole and pyrimidine derivatives incorporating a benzothiazole moiety. Eur J Med Chem. 2009; 44(12):4813-4818.
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
59
Bondock, S., Fadaly, W., Metwally, M. A. Synthesis and antimicrobial activity of some new thiazole, thiophene and pyrazole derivatives containing benzothiazole moiety. Eur J Med Chem. 2010; 45(9):3692-3701.
Bradshaw, T. D., Wrigley, S., Shi, D. F., Schultz, R. J., Paull, K. D., Stevens, M. F. 2-(4-Aminophenyl) benzothiazoles: novel agents with selective profiles of in vitro anti-tumour activity. Br J Cancer. 1998; 77(5):745-752.
Caleta, I., Grdisa, M., Mrvos-Sermek, D., Cetina, M., Tralic-Kulenovic, V., Pavelic, K., Karminski-Zamola, G. Synthesis, crystal structure and antiproliferative evaluation of some new substituted benzothiazoles and styrylbenzothiazoles. Farmaco. 2004; 59(4):297-305.
Charris, J., Monasterios, M., Dominguez, J., Infante, W., De Castro, N. (2002). Synthesis of some 5-nitro-2-furfurylidene derivatives and their antibacterial and antifungal activities, Heterocyclic Communications (Vol. 8, pp. 275).
Duggan, P. J., Lewis, R. J., Phei Lok, Y., Lumsden, N. G., Tuck, K. L., Yang, A. Low molecular weight non-peptide mimics of omega-conotoxin GVIA. Bioorg Med Chem Lett. 2009; 19(10):2763-2765.
Gewirtz, D. A., Holt, S. E., Grant, S. Apoptosis, Senescence and Cancer 2007; Humana Press.
Gilson, R. J., Ross, J., Maw, R., Rowen, D., Sonnex, C., Lacey, C. J. A multicentre, randomised,
double-blind, placebo controlled study of cryotherapy versus cryotherapy and podophyllotoxin cream as treatment for external anogenital warts. Sex Transm Infect. 2009; 85(7):514-519.
Gordaliza, M., Garcia, P. A., del Corral, J. M., Castro, M. A., Gomez-Zurita, M. A. Podophyllotoxin: distribution, sources, applications and new cytotoxic derivatives. Toxicon. 2004; 44(4):441-459.
Gupta, R. S. Podophyllotoxin-resistant mutants of Chinese hamster ovary cells: cross-resistance studies with various microtubule inhibitors and podophyllotoxin analogues. Cancer Res. 1983; 43(2):505-512.
Gurupadayya, B. M., Gopal, M., Padmashali, B., Manohara, Y. N. Synthesis and Pharmacological Evaluation of Azetidin-2-ones and Thiazolidin-4-ones Encompassing Benzothiazole. Indian J Pharm Sci. 2008; 70(5):572-577.
Hejchman, E., Taciak, P., Kowalski, S., Maciejewska, D., Czajkowska, A., Borowska, J., Sladowski, D., Mlynarczuk-Bialy, I. Synthesis and anticancer activity of 7-hydroxycoumarinyl gallates. Pharmacol Rep. 2015; 67(2):236-244.
Henriksen, G., Hauser, A. I., Westwell, A. D., Yousefi, B. H., Schwaiger, M., Drzezga, A., Wester, H. J. Metabolically stabilized benzothiazoles for imaging of amyloid plaques. J Med Chem. 2007; 50(6):1087-1089.
Hui, J., Zhao, Y., Zhu, L. Synthesis and in vitro anticancer activities of novel aryl-naphthalene lignans 2011; Vol. 21.
Hutchinson, I., Jennings, S. A., Vishnuvajjala, B. R., Westwell, A. D., Stevens, M. F. Antitumor
benzothiazoles. 16. Synthesis and pharmaceutical properties of antitumor 2-(4-aminophenyl)benzothiazole amino acid prodrugs. J Med Chem. 2002; 45(3):744-747.
Imbert, T. F. Discovery of podophyllotoxins. Biochimie. 1998; 80(3):207-222.
Kashiyama, E., Hutchinson, I., Chua, M. S., Stinson, S. F., Phillips, L. R., Kaur, G., Sausville, E. A., Bradshaw, T. D., Westwell, A. D., Stevens, M. F. Antitumor benzothiazoles. 8. Synthesis, metabolic
formation, and biological properties of the C- and N-oxidation products of antitumor 2-(4-aminophenyl)benzothiazoles. J Med Chem. 1999; 42(20):4172-4184.
Keri, R. S., Patil, M. R., Patil, S. A., Budagumpi, S. A comprehensive review in current developments of benzothiazole-based molecules in medicinal chemistry. Eur J Med Chem. 2015; 89:207-251.
Kok, S. H., Gambari, R., Chui, C. H., Yuen, M. C., Lin, E., Wong, R. S., Lau, F. Y., Cheng, G. Y., Lam, W. S., Chan, S. H., Lam, K. H., Cheng, C. H., Lai, P. B., Yu, M. W., Cheung, F., Tang, J. C., Chan, A.
Piotr Strus, Kamil Lisiecki, Zbigniew Czarnocki, Izabela Młynarczuk-Biały, Łukasz P Biały
60
S. Synthesis and anti-cancer activity of benzothiazole containing phthalimide on human carcinoma cell lines. Bioorg Med Chem. 2008; 16(7):3626-3631.
Kumar, A., Kumar, V., Alegria, A. E., Malhotra, S. V. Synthetic and application perspectives of azapodophyllotoxins: alternative scaffolds of podophyllotoxin. Curr Med Chem. 2011; 18(25):3853-3870.
Kumbhare, R. M., Kosurkar, U. B., Janaki Ramaiah, M., Dadmal, T. L., Pushpavalli, S. N., Pal-Bhadra, M. Synthesis and biological evaluation of novel triazoles and isoxazoles linked 2-phenyl benzothiazole as potential anticancer agents. Bioorg Med Chem Lett. 2012; 22(17):5424-5427.
Labruère, R., Gautier, B., Testud, M., Seguin, J., Lenoir, C., Desbène‐Finck, S., Helissey, P., Garbay, C., Chabot, G. G., Vidal, M., Giorgi‐Renault, S. Design, Synthesis, and Biological Evaluation of the First Podophyllotoxin Analogues as Potential Vascular‐Disrupting Agents. ChemMedChem. 2010; 5(12):2016-2025.
Leroy, D., Kajava, A. V., Frei, C., Gasser, S. M. Analysis of etoposide binding to subdomains of human DNA topoisomerase II alpha in the absence of DNA. Biochemistry. 2001; 40(6):1624-1634.
Lewis, D. A., Goldmeier, D. The cost effectiveness of hospital-based 25% podophyllin vs home-based 0.5% podophyllotoxin in the treatment of anogenital warts. Int J STD AIDS. 1995; 6(3):224-225.
Lipinski, C. A., Lombardo, F., Dominy, B. W., Feeney, P. J. Experimental and computational
approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev. 2001; 46(1-3):3-26.
Lisiecki, K., Czarnocki, Z. Flow Photochemistry as a Tool for the Total Synthesis of (+)-Epigalcatin. Org Lett. 2018; 20(3):605-607.
Lisiecki, K., Krawczyk, K. K., Roszkowski, P., Maurin, J. K., Budzianowski, A., Czarnocki, Z. Unusual visible-light photolytic cleavage of tertiary amides during the synthesis of cyclolignans related to podophyllotoxin. Tetrahedron. 2017; 73(44):6316-6328.
Lisiecki, K., Krawczyk, K. K., Roszkowski, P., Maurin, J. K., Czarnocki, Z. Formal synthesis of (-)-
podophyllotoxin through the photocyclization of an axially chiral 3,4-bisbenzylidene succinate amide ester--a flow photochemistry approach. Org Biomol Chem. 2016; 14(2):460-469.
Liu, C., Lin, J., Pitt, S., Zhang, R. F., Sack, J. S., Kiefer, S. E., Kish, K., Doweyko, A. M., Zhang, H., Marathe, P. H., Trzaskos, J., McKinnon, M., Dodd, J. H., Barrish, J. C., Schieven, G. L., Leftheris, K. Benzothiazole based inhibitors of p38alpha MAP kinase. Bioorg Med Chem Lett. 2008; 18(6):1874-1879.
Longstaff, E., von Krogh, G. Condyloma eradication: self-therapy with 0.15-0.5% podophyllotoxin versus 20-25% podophyllin preparations--an integrated safety assessment. Regul Toxicol Pharmacol. 2001; 33(2):117-137.
Lu, Y., Chen, J., Xiao, M., Li, W., Miller, D. D. An overview of tubulin inhibitors that interact with the colchicine binding site. Pharm Res. 2012; 29(11):2943-2971.
Mattson, M. P., Calabrese, E. J. Hormesis: A Revolution in Biology, Toxicology and Medicine 2009; Humana Press.
Miller, K. D., Siegel, R. L., Lin, C. C., Mariotto, A. B., Kramer, J. L., Rowland, J. H., Stein, K. D., Alteri, R., Jemal, A. Cancer treatment and survivorship statistics, 2016. CA Cancer J Clin. 2016; 66(4):271-289.
Mlynarczuk-Bialy, I., Doeppner, T. R., Golab, J., Nowis, D., Wilczynski, G. M., Parobczak, K., Wigand, M. E., Hajdamowicz, M., Bialy, L. P., Aniolek, O., Henklein, P., Bahr, M., Schmidt, B., Kuckelkorn, U., Kloetzel, P. M. Biodistribution and Efficacy Studies of the Proteasome Inhibitor BSc2118 in a Mouse Melanoma Model. Transl Oncol. 2014; 7(5):570-579.
Mlynarczuk-Bialy, I., Roeckmann, H., Kuckelkorn, U., Schmidt, B., Umbreen, S., Golab, J., Ludwig, A., Montag, C., Wiebusch, L., Hagemeier, C., Schadendorf, D., Kloetzel, P. M., Seifert, U. Combined effect of proteasome and calpain inhibition on cisplatin-resistant human melanoma cells. Cancer Res. 2006; 66(15):7598-7605.
Novel podophyllotoxin derivatives as Anticancer Agents: Design, Synthesis, and Biological Screening
61
Newman, D. J., Cragg, G. M. Natural products as sources of new drugs over the last 25 years. J Nat Prod. 2007; 70(3):461-477.
Saraiva, J., Vega, C., Rolon, M., da Silva, R., ML, E. S., Donate, P. M., Bastos, J. K., Gomez-Barrio, A., de Albuquerque, S. In vitro and in vivo activity of lignan lactones derivatives against Trypanosoma cruzi. Parasitol Res. 2007; 100(4):791-795.
Seth, S. A Comprehensive Review on Recent advances in Synthesis & Pharmacotherapeutic potential of Benzothiazoles. Anti-Inflammatory & Anti-Allergy Agents In Medicinal Chemistry. 2015; 14(2):98-112.
Sharma, P. C., Sinhmar, A., Sharma, A., Rajak, H., Pathak, D. P. Medicinal significance of benzothiazole scaffold: an insight view. J Enzyme Inhib Med Chem. 2013; 28(2):240-266.
Shi, D. F., Bradshaw, T. D., Wrigley, S., McCall, C. J., Lelieveld, P., Fichtner, I., Stevens, M. F. Antitumor benzothiazoles. 3. Synthesis of 2-(4-aminophenyl)benzothiazoles and evaluation of their activities against breast cancer cell lines in vitro and in vivo. J Med Chem. 1996; 39(17):3375-3384.
Siegel, R. L., Miller, K. D., Jemal, A. Cancer statistics, 2016. CA Cancer J Clin. 2016; 66(1):7-30.
Silva, M. L., Coimbra, H. S., Pereira, A. C., Almeida, V. A., Lima, T. C., Costa, E. S., Vinholis, A. H., Royo, V. A., Silva, R., Filho, A. A., Cunha, W. R., Furtado, N. A., Martins, C. H., Carvalho, T. C., Bastos, J. K. Evaluation of piper cubeba extract, (-)-cubebin and its semi-synthetic derivatives against oral pathogens. Phytother Res. 2007; 21(5):420-422.
Silva, M. L., Martins, C. H., Lucarini, R., Sato, D. N., Pavanb, F. R., Freitas, N. H., Andrade, L. N., Pereira, A. C., Bianco, T. N., Vinholis, A. H., Cunha, W. R., Bastos, J. K., Silva, R., Da Silva Filho, A. A. Antimycobacterial activity of natural and semi-synthetic lignans. Z Naturforsch C. 2009; 64(11-12):779-784.
Spasevska, I., Ayoub, A. T., Winter, P., Preto, J., Wong, G. K., Dumontet, C., Tuszynski, J. A. Modeling the Colchicum autumnale Tubulin and a Comparison of Its Interaction with Colchicine to Human Tubulin. Int J Mol Sci. 2017; 18(8).
Srivastava, V., Negi, A. S., Kumar, J. K., Gupta, M. M., Khanuja, S. P. Plant-based anticancer molecules: a chemical and biological profile of some important leads. Bioorg Med Chem. 2005; 13(21):5892-5908.
von Krogh, G., Longstaff, E. Podophyllin office therapy against condyloma should be abandoned. Sex Transm Infect. 2001; 77(6):409-412.
Wang, J. C. DNA topoisomerases. Annu Rev Biochem. 1996; 65:635-692.
Yoshida, M., Hayakawa, I., Hayashi, N., Agatsuma, T., Oda, Y., Tanzawa, F., Iwasaki, S., Koyama, K., Furukawa, H., Kurakata, S., Sugano, Y. Synthesis and biological evaluation of benzothiazole derivatives as potent antitumor agents. Bioorg Med Chem Lett. 2005; 15(14):3328-3332.
Yousefzadi, M., Sharifi, M., Behmanesh, M., Ghasempour, A., Moyano, E., Palazon, J. Salicylic acid
improves podophyllotoxin production in cell cultures of Linum album by increasing the expression of genes related with its biosynthesis. Biotechnol Lett. 2010; 32(11):1739-1743.
Yousefzadi, M., Sharifi, M., Chashmi, N. A., Behmanesh, M., Ghasempour, A. Optimization of podophyllotoxin extraction method from Linum album cell cultures. Pharm Biol. 2010; 48(12):1421-1425.
Youssef, A. M., Noaman, E. Synthesis and evaluation of some novel benzothiazole derivatives as potential anticancer and antimicrobial agents. Arzneimittelforschung. 2007; 57(8):547-553.
Yu, X., Che, Z., Xu, H. Recent Advances in the Chemistry and Biology of Podophyllotoxins. Chemistry. 2017; 23(19):4467-4526.
Zhang, X., Rakesh, K. P., Shantharam, C. S., Manukumar, H. M., Asiri, A. M., Marwani, H. M., Qin, H. L. Podophyllotoxin derivatives as an excellent anticancer aspirant for future chemotherapy: A key current imminent needs. Bioorg Med Chem. 2018; 26(2):340-355.
62
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis
to lymph nodes
Maksimiuk M.2, Sobiborowicz A.
2, Sobieraj M.
2, Liszcz A.
2, Sobol M.
1, Patera J.
3,
Badowska-Kozakiewicz A.M.1
1 Department of Biophysics and Human Physiology, Medical University of Warsaw
Chałubińskiego 5, 02-004 Warsaw 2 Students' Scientific Organization at the Medical University of Warsaw
Oczki 1a, 02-007 Warsaw 3 Department of Pathomorphology, Military Institute of Health Services, Warsaw, Poland
corresponding author: Marta Maksimiuk
Department of Biophysics and Human Physiology,
Medical University of Warsaw
Chałubińskiego 5, 02-004 Warsaw
tel: 691-230-268
ABSTRACT
Introduction: Tumor characteristics, such as size, lymph node status, histological type of the neoplasm
and its grade, are well known prognostic factors in breast cancer. The ongoing search for new
prognostic factors include Bcl2, Bax, Cox-2 or HIF1-alpha, which plays a key role in phenomenon of
tumor hypoxia and might induce transcription of the EPO gene. Erythropoietin may influence lymph
node metastasis or stimulate tumor progression, thus it seemed interesting to determine its expression in
invasive breast cancers with lymph node metastases presenting different basic immunohistochemical
profiles (ER, PR, HER2).
Aim: To evaluate the relationship between histological grade, tumor size, lymph node status, expression
of ER, PR, HER2 and immunohistochemical expression of erythropoietin in invasive breast cancer with
metastasis to lymph nodes.
Materials and methods: The material consisted of histological preparations derived from patients
treated for invasive breast cancer with metastasis to lymph nodes. Examined samples were stained using
standard methods. Routine tests were additionally performed in order to determine immuno-
histochemical expression of basic profile of diagnostic markers, such as estrogen receptor (ER),
progesterone receptor (PR) and HER2. Expression of erythropoietin (EPO) was assessed through using
of an appropriate antibody against its antigen.
Results: Among the studied group of cancers we selected four subgroups with different basic
immunohistochemical profiles. Additionally we studied the relationship between histological type of
breast cancer and basic immunohistochemical profile and a statistically significant correlation was noted
in all cases. Expression of erythropoietin was determined in all histological breast cancer types and it
was most frequently identified independently in: invasive ductal carcinoma of no special type (IDC
– NST), cancers with the highest histological grade (G3) and cancers evaluated as pT2 and pN1. The
largest group expressing EPO consisted of cancers presenting the ER-/PR-/HER2+ basic
immunohistochemical profile, while no EPO expression was most often demonstrated in cancers with
ER+/PR+/HER2- immunohistochemical profile. Furthermore our study revealed EPO expression in
over 40% of cases in the group of TNBC (ER-/PR-/HER2-).
Conclusions: In our study we demonstrated a statistically significant correlation between markers included in the basic immunohistochemical profile (ER, PR, HER2) with histological type of invasive breast cancer. There was no relationship between markers of basic immunohistochemical profile and
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
63
expression of EPO as well as no statistically significant dependence was found between expression of EPO and basic clinical features. However, conducted studies allow for formulating important conclusions for pathomorphological diagnostics. Moreover, expression of EPO in almost 40% of cancers with ER-/PR-/HER2- immunohistochemical profile (TNBC) suggests that in TNBC erythropoietin might be a prognostic marker.
Introduction
Tumor characteristics, such as size, lymph node status and histological type of the neoplasm, are well known prognostic factors in breast cancer. Certain microscopic features of the neoplasm, such as histological grade of malignancy or presence of neoplastic cells in blood and lymphatic vessels, also deserve special attention [1, 2]. Immunohistochemical techniques constitute an excellent tool for determination of new factors characteristic for neoplasms that may be considered prognostic or predictive factors and therefore, enabling determination of immunohistochemical profiles of various histological types of breast cancer. Standardization of methods of immunohistochemical detection of estrogen receptor (ER), progesterone receptor (PR) and HER2 protein overexpression contributed to inclusion of ER, PR and HER2 status assessment into routine evaluation of prognostic and predictive factors in breast cancer. Assessment of ER, PR and HER2 is currently essential to the therapeutic process. There is a number of factors, such as: Bcl2, Bax, or Cox-2 playing a key role in the process of carcinogenesis and cancer progression, but due to lack of standardized assessment methods, as well as lack of evidence for their clinical value, they are not used in routine pathomorphological evaluation of breast cancer and remain subject of intense research [3]. Since achieving the best antineoplastic effect and providing the patient with optimal quality of life are primary goals of treatment of neoplastic diseases, including breast cancer, there is an ongoing search for prognostic and predictive factors based on morphological features and knowledge of other characteristics of neoplastic cells enabling proper assessment of disease course, malignancy risk and susceptibility to treatment. In cases of neoplasms without lymph node metastases the following factors were proven important for the prognosis: patient age, histological type of neoplasm, histological grade of malignancy, proliferative activity and DNA ploidy. Therefore, the factors currently taken into consideration in the assessment of breast cancers include: markers of proliferation (mitotic index, proportion of cells in the S phase of the cell cycle, expression of nuclear antigen Ki67 and PCNA), DNA ploidy, expression of estrogen and progesterone receptors as well as HER2 protein [4].
Recent years have brought about growing interest in the phenomenon of tumor hypoxia and augmented activity of HIF1-alpha, a marker of hypoxia, which is considered one of the most important factors responsible for activation of tumor angiogenesis. It is believed that HIF1-alpha might induce transcription of the EPO gene. Its product, erythropoietin, is a physiological regulator of erythropoiesis sensitive to hypoxia. Biological action of erythropoietin involves induction of erythropoiesis by stimulating proliferation and differentiation of red blood cell precursors into mature erythrocytes. However, the effects of erythropoiesis are not limited to the hematopoietic system. Numerous reports published in the recent years indicate that erythropoietin is a cytokine with autocrine and paracrine properties and affects many non-hematological tissues [5]. Local production of erythropoietin was
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
64
demonstrated in nervous tissue, genital tract and placenta, while expression of erythropoietin receptor was identified, among other locations, in kidneys, lungs and muscle tissue [5, 6]. Erythropoietin receptors were also found in many neoplasms: colon adenocarcinoma, stomach cancer, lung cancer, brain cancer, neoplasms of the head and neck, renal tumors, and in prostate cancer [7]. Ability of erythropoietin to stimulate angiogenesis in both normal and neoplastic cells together with a finding of erythropoietin receptor expression on neoplastic cells and cells of vascular endothelium suggest that erythropoietin might directly influence tumor growth and inhibit apoptosis, stimulating tumor progression or metastasis. It could modify tumor cells’ sensitivity to chemo- and radiotherapy [5, 6, 7].
The influence of erythropoietin on lymph node metastasis is an interesting, yet scarcely investigated, issue. Therefore, it seemed interesting to determine erythropoietin expression in invasive breast cancers with lymph node metastases presenting different basic immunohistochemical profiles (ER, PR, HER2). From a time perspective, results of this research might become an important parameter in the assessment of the risk of metastasis to lymph nodes and other organs.
Aim
The aim of this study was to evaluate the relationship between histological grade, tumor size, lymph node status, expression of ER, PR, HER2 and immuno-histochemical expression of erythropoietin in invasive breast cancer with metastasis to lymph nodes.
Materials and methods
The material consisted of histological preparations derived from patients treated for invasive breast cancer with metastasis to lymph nodes. Histological and immunohistochemical studies were performed at the Department of Pathology, Military Medical Institute in Warsaw. Material for the study came from biopsies, excisional biopsies and modified radical mastectomies. Samples of tumors were fixed in 10% buffered formalin phosphate. After 24-hour fixation, material was dehydrated using alcohol in gradually increasing concentrations and embedded in paraffin. Paraffin blocks were cut into serial sections 4µm in thickness. They were then stained using standard methods. The tumors were classified and graded according to the WHO and the Nottingham modification of the Scarff-Bloom-Richardson systems. In the sections stained with routine H&E method, the following determinations were carried out: type of neoplasm (WHO classification), tumor grade including tubule formation, and intensity of division as well as the degree of neoplastic cell differentiation and mitotic index defined as a mean number of mitotic figures in neoplastic cells counted in 10 fields of vision at a x400 magnification (surface field 0.17 mm
2).
Immunohistochemical Staining Steroid Receptor (ER, PgR) and HER2
Paraffin sections on slides covered with 2% saline solution in acetone at a tempe-rature of 42
oC were used for immunohistochemical examination. Routine tests were
also performed in order to determine immunohistochemical expression of basic profile of diagnostic markers, such as estrogen receptor (ER), progesterone receptor (PR) and HER2. Immunohistochemistry was performed using the EnVision
TM
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
65
+ HRP DakoCytomation (EnVision TM
Dual Link System-HRP, DAB+, Code: K4065). Monoclonal antibodies against receptors for estrogen (Monoclonal Mouse Anti-Human Estrogen Receptor alpha, 1:50 dilution, Clone: 1D5, Code: IR654, DAKO) and progesterone (Monoclonal Mouse Anti-Human Progesterone Receptor, 1:400 dilution, Clone: PgR636, Code: IR068, DAKO) were used in order to determine the expression of steroid receptors. The study was conducted as follows: sections were incubated at 60°C overnight and subsequently dewaxed. The next step involved revealing the epitope by heating the slides in a buffer for 40 minutes. Subsequently, preparations were left at a room temperature for 20 minutes. Preparations were rinsed in buffer and endogenous peroxidase was blocked by washing in 3% H2O2 for 10 minutes. In the next step, preparations were incubated with an appropriate antibody for 30 minutes. After incubation, preparations were rinsed in buffer for 10 minutes, and then incubated with the reagent (Visualization Reagent) for 30 minutes. After incubation with the reagent, preparations were washed in TBS (Tris - Buffered Saline, Code: S1968) with pH 7.6 for 10 minutes, and then incubated with 3,3’-diaminobenidine (DAB) (Substrate Chromogen Solution) for 10 minutes to visualize the color of the reaction. At the end of the procedure, preparations were stained with hematoxylin. Evaluation of the immunohistochemical markers was performed by two pathologists as follows: ER and PR were categorized as negative – (0%), low positive – (1-10%); nuclear staining in >10% of tumor cells was considered positive for ER and PR. HER2 expression was determined using HerceptTest
TM DAKO test (Code: K5204). It enabled
detection of HER2 expression using a polyclonal antibody against this protein (Rb A – Hu HER2 – Rabbit Anti-Human HER2 Protein). Antigen retrieval for HER2 using HerceptTest was performed by immersing and incubating the slides in 10-mmol/L citrate buffer in a calibrated water bath (95-99°C) for 40 minutes (+/- 1 minute). After decanting the epitope- retrieving solution, sections were rinsed in the wash buffer and later, soaked in the buffer for 5 to 20 minutes before staining. The slides were loaded onto the autostainer using the HercepTest program, as described in the manufacturer’s insert. In the autostainer, the slides were rinsed, placed in 200μL of peroxidase-blocking reagent for 5 minutes, rinsed, placed in 200μL of primary anti-HER2 protein (or negative control reagent) for 30 minutes, rinsed twice and immersed in 200μL of substrate chromogen solution – DAB for 10 minutes. The slides were counterstained with hematoxylin and finally coverslipped. HER2 results were determined based on the maximum area of staining intensity according to the instruction in the package insert and the ASCO/CAP guidelines as follows: strong, circumferential membranous, staining in >30% of invasive carcinoma cell was scored 3+, moderate, circumferential, membranous staining in ≥10% of invasive tumor cells or 3+ staining in ≤30% of cells was designated as 2+ staining, weak and incomplete membranous staining in invasive tumor cells was scored 1+ and no staining was scored 0. Tumors with 0 and 1+ staining were considered negative.
Immunohistochemical staining of EPO
Expression of erythropoietin (EPO) was also assessed in all studied invasive breast cancers through use of an appropriate antibody against EPO antigen (Polyclonal Rabbit Anti-Human EPO, 1:100 dilution, Clone: H-162, Santa Cruz Biotechnology ®, Inc.) and subsequent application of the ImmunoCruz
TM Rabbit ABC Staining
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
66
System for visualization (Santa Cruz Biotechnology ®, Inc.). EPO staining results were scored according to the percentage of cytoplasmic positive cells as follows: (-), <10%; (+), 10%-20%; (++), >20%. Moderate expression EPO was defined as >20% tumor cells with positive staining whereas low expression was <20% [8].
Statistical analysis
All statistical analyses were performed with SPSS software v12.0 for Windows. The frequency of HER2 expression according to joint ER/PR status and the distribution of the hormone receptor status (ER, PR, and joint ER/PR) according to HER2 were
also calculated. The Chi-square (χ2) was used to assess the relationship between EPO and expression of steroid receptors, HER2, histological type of tumor, degree of histological malignancy and clinical staging. The Fisher exact test was used when the expected cell counts were less than 5. Differences were considered statistically significant at P < 0.05.
Results
Mean age of studied women was 60.5±11.7 years (median 61 years, range: 43-78 years). Histopathological examination revealed the following percentages of tumors: 68.96% IDC-NST, 18.96% IDC, 3.44% ILC, 6.9% metaplastic carcinoma and 1.72% mixed ductal and lobular carcinoma. Among the studied group of cancers we selected four subgroups with different basic immunohistochemical profiles, such as
PR+/ER+/HER2+ (22.4%); PR-/ER-/HER2+ (32.76%); PR-/ER-HER2- (12.07%); PR+/ER+/HER2- (32.76%). In all four subgroups of cancers presenting various basic immunohistochemical profiles IDC-NST was most numerous (68.96%) (Table 1).
Table 1. Correlation between histological type of invasive breast cancer and the basic immunohistochemical profile (ER, PR, HER2)
Immunohisto-chemistry basal panel for diagnosis of breast cancer
Frequency
n=58
Histopathological type of invasive breast cancer P-value*
IDC-NST
IDC ILC Metaplastic carcinoma
Mixed
ductal and lobular
PR+/ER+/HER2+
13 9 3 0 1 0 0.00001132*
PR-/ER-/HER2+
19 10 7 0 1 1 0.00005*
PR-/ER-/HER2-
7 6 0 0 1 0 0.012051*
PR+/ER+/HER2-
19 15 1 2 1 0 <0.001*
*Statistically significant results (P < 0.05) IDC-NST – invasive ductal carcinoma of no special type IDC – invasive ductal carcinoma ILC – invasive lobular carcinoma
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
67
In our material we studied the relationship between histological type of breast cancer
and basic immunohistochemical profile and a statistically significant correlation was
noted in all cases (p<0.05) (Table 1). In our study basic immunohistochemical profile
was extended to include a new marker – erythropoietin (EPO) – its expression was
determined in all histological breast cancer types (Figure 1). Depending on
expression of EPO in invasive breast cancers we differentiated eight subgroups of
cancers presenting different immunohistochemical profiles taking into consideration
this novel marker – EPO: PR+/ER+/HER2+/EPO+ (10.35%); PR-/ER-/HER2+/EPO+
(12.06%); PR-/ER-/HER2-/EPO+ (5.17%); PR+/ER+/HER2-/EPO+ (8.62%);
PR+/ER+/HER2+/EPO- (12.07%); PR-/ER-/HER2+/EPO- (20.69%); PR-/ER-
/HER2-/EPO- (6.90%); PR+/ER+/HER2-/EPO- (24.14%) (Table 2). In our study,
among all breast cancers 36.2% exhibited EPO expression in neoplastic cells, while
in 63.8% EPO expression was not found. IDC-NST was the most numerous group
among all histological types of cancers positive for EPO expression (24.1%) (Table
2). No expression of EPO was found in 43.1% of IDC-NST. Among the remaining
cancers EPO expression was found in 3.5% of IDC and in 8.6% of metaplastic
carcinomas, while no expression of EPO was demonstrated in 15.5% of IDC, 2.5%
of ILC and 1.7% of mixed ductal and lobular carcinomas (Table 2). A correlation
between histological type of cancer and immunohistochemical profile that included
EPO expression was studied and no statistically significant relationships were found
(Table 2).
Table 2. Correlation between the histological type of invasive breast cancer and the basic profile of
immunohistochemical, including EPO expression
Immunohistochemistry
basal panel for diagnosis
of breast cancer and
expression of EPO
Frequen-
cy
n=58
Histopathological type of invasive breast
cancer
P-value*
IDC-
NST
IDC ILC Meta-
plastic
carci-
noma
Mixed
ductal
and
lobular
PR+/ER+/HER2+/EPO+ 6 5 0 0 1 0 0.948854
PR-/ER-/HER2+/EPO+ 7 5 1 0 1 0 0.95319379
PR-/ER-/HER2-/EPO+ 3 1 0 0 2 0 0.5119721
PR+/ER+/HER2-/EPO+ 5 3 1 0 1 0 0.9427068
PR+/ER+/HER2+/EPO- 7 4 3 0 0 0 0.764538
PR-/ER-/HER2+/EPO- 12 5 6 0 0 1 0.1861714
PR-/ER-/HER2-/EPO- 4 4 0 0 0 0 0.4695367
PR+/ER+/HER2-/EPO- 14 12 0 2 0 0 0.153851
*Statistically significant results (P < 0.05)
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
68
Fig.1. Immunohistochemical image of invasive breast cancer – positive immunohistochemical staining for
EPO (original magnification, x200; scale bar, 50µm)
In our study we also assessed the dependence between the degree of histological
malignancy (G1-G3), tumor size (pT) or presence of lymph node metastases (pN),
and basic immunohistochemical profile that included EPO expression, but no
statistically significant correlations were found (Tables 3-5). Among G3 cancers 47%
exhibited expression of EPO, while no EPO expression was found in 53% of cases.
In a group of G2 cancers 24% were positive for EPO, while 76% of them did not
exhibit EPO expression. Among all cancers, the largest group expressing EPO were
the G3 cancers (67%) (Table 3).
Table 3. Correlation between the basic profile of immunohistochemical, including EPO expression, and
histological grading of invasive breast cancer
Immunohistochemistry basal
panel for diagnosis of breast
cancer and expression of EPO
Frequency
n=58
Histological grade
G1 G2 G3 P-value*
PR+/ER+/HER2+/EPO+ 6 0 2 4 0.83946
PR-/ER-/HER2+/EPO+ 7 0 1 6 0.76498
PR-/ER-/HER2-/EPO+ 3 0 0 3 0.50667
PR+/ER+/HER2-/EPO+ 5 1 3 1 0.25208
PR+/ER+/HER2+/EPO- 7 0 2 5 0.45285
PR-/ER-/HER2+/EPO- 12 1 7 4 0.97190
PR-/ER-/HER2-/EPO- 4 0 1 3 0.89404
PR+/ER+/HER2-/EPO- 14 1 9 4 0.55822
*Statistically significant results (P < 0.05)
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
69
We also evaluated the relationship between tumor size (pT) and basic immuno-
histochemical profile that included expression of EPO and it was demonstrated that
among cancers exhibiting EPO expression the pT2 cancers comprised the largest
group (52.3%) (Table 4). As much as 36.2% of cancers with metastases to lymph
nodes expressed EPO, while no EPO expression was found in 63.8% of cancers with
lymph node metastases (Table 5).
Table 4. Correlation between the basic profile of immunohistochemical, including EPO expression, and tumor
stage of invasive breast cancer
Immunohistochemistry basal
panel for diagnosis of breast
cancer and expression of EPO
Frequency
n=58
Tumor stage
pT1 pT2 pT3 pT4 P-value*
PR+/ER+/HER2+/EPO+ 6 1 3 0 2 0.80270
PR-/ER-/HER2+/EPO+ 7 0 5 1 1 0.48037
PR-/ER-/HER2-/EPO+ 3 1 2 0 0 0.74510
PR+/ER+/HER2-/EPO+ 5 4 1 0 0 0.16947
PR+/ER+/HER2+/EPO- 7 5 1 0 1 0.45243
PR-/ER-/HER2+/EPO- 12 7 4 0 1 0.81794
PR-/ER-/HER2-/EPO- 4 1 2 1 0 0.55984
PR+/ER+/HER2-/EPO- 14 3 10 0 1 0.23777
*Statistically significant results (P < 0.05)
Table 5. Correlation between the basic profile of immunohistochemical, including EPO expression, and nodal
stage of invasive breast cancer
Immunohistochemistry basal
panel for diagnosis of breast
cancer and expression of EPO
Frequency
n=58
Nodal stage
pN1 pN2 pN3 P-value*
PR+/ER+/HER2+/EPO+ 6 2 2 2 0.91074
PR-/ER-/HER2+/EPO+ 7 2 3 2 0.91759
PR-/ER-/HER2-/EPO+ 3 2 1 0 0.91308
PR+/ER+/HER2-/EPO+ 5 3 2 0 0.78999
PR+/ER+/HER2+/EPO- 7 4 3 0 0.62719
PR-/ER-/HER2+/EPO- 12 6 3 3 0.81628
PR-/ER-/HER2-/EPO- 4 2 0 2 0.46813
PR+/ER+/HER2-/EPO- 14 10 3 1 0.70751
*Statistically significant results (P < 0.05)
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
70
A relationship between expression of EPO and basic immunohistochemical profile
(ER, PR, HER2) was also examined in the studied group of cancers and it was shown
that the largest group expressing EPO consisted of cancers presenting the ER-/PR-
/HER2+ basic immunohistochemical profile (33.3%), while no EPO expression was
most often demonstrated in cancers with ER+/PR+/HER2- immuno-histochemical
profile (37.8%) (Table 6). Our study also showed that in the group of “triple-negative
cancers” (TNBC) (ER-/PR-/HER2-) EPO expression was found in over 40% of
cases.
Table 6. The relationship between the basic profile immunohistochemistry and expression of EPO
Immunohistochemistry basal
panel for diagnosis of breast
cancer
Frequency
n=58
Expression of EPO P-value*
Positive1 Negative2
PR+/ER+/HER2+ 13 6 7 0.3968007
PR-/ER-/HER2+ 19 7 12 0.943628
PR-/ER-/HER2- 7 3 4 0.97477
PR+/ER+/HER2- 19 5 14 0.421906
*Statistically significant results (P < 0.05) 1 (+) 10%-20%; (++) >20% 2 (-) <10%
Discussion
It is presently known that erythropoietin stimulates the process of angiogenesis, both
in normal and neoplastic cells, results of some studies suggesting that erythropoietin
might influence tumor size and inhibit apoptotic mechanisms [9], which can induce
metastasis [10,11]. Therefore, such an effect may change neoplastic cells’
susceptibility to chemo- and radiotherapy, although this suggestion requires numerous
further studies. Data regarding expression of erythropoietin in invasive breast cancers
with lymph node metastasis in the context of various histological types and basic
immunohistochemical profiles (ER, PR, HER2) is scarce. There is also little data on
the influence of erythropoietin on lymph node metastasis thus, it seemed interesting
to determine expression of erythropoietin in invasive breast cancers depending on the
histological type of cancer, presence of lymph node metastasis, expression of
estrogen and progesterone receptors, as well as expression of HER2 – markers
determined routinely in immunohistochemical diagnostics. Our study was conducted
on a group of 58 patients with various histological types of invasive breast cancer.
Expression of basic markers used in routine immunohistochemical diagnostics, such
as ER, PR and HER2, was assessed and a statistically significant relationship was
demonstrated between all histological types of invasive breast cancers and expression
of estrogen and progesterone receptors, as well as HER2 expression (p<0.05) (Table
1). Since the main goal of this research was to assess the expression of EPO in all
studied histological types of invasive breast cancers, it seemed important to distinct
four subgroups of cancers exhibiting different immunohistochemical profiles:
PR+/ER+/HER2+ (22.4%); PR-/ER-/HER2+ (32.76%); PR-/ER-HER2- (12.07%);
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
71
PR+/ER+/HER2- (32.76%). In all four subgroups of invasive cancers presenting
various basic immunohistochemical profiles expression of erythropoietin was
examined using immunohistochemical techniques. In invasive breast cancer, we
found weak-to-moderate, granular, cytoplasmic staining for EPO in 36.2% of
specimens. In their studies Acs, Zhang et al. (2002) found the presence of weak-to-
moderate, granular, cytoplasmic EPO immunostaining in benign mammary epithelial
cells in 91.8% and 95.6% of the specimens, respectively [12]. Taking into
consideration histological type of invasive breast cancer the largest EPO-positive
group in our material comprised the IDC-NST (24.1%). Among the remaining
cancers EPO expression was demonstrated in 3.5% of IDC and in 8.6% of
metaplastic carcinomas, while in 15.5% of IDC, 3.5% of ILC and in 1.7% of mixed
ductal and lobular carcinomas expression of EPO was not found. No statistically
significant relationship between histological type of invasive cancers and expression
of EPO was found (p>0.05). In studies Acs, Zhang et al. (2002) EPO
immunostaining was similar in invasive ductal and lobular carcinomas and in
carcinomas with mixed ductal and lobular features. Acs, Zhang et al. (2002) found no
differences in EPO immunostaining between DCIS and LCIS. Also, the authors
found no statistically significant relationship between the expression of EPO and
histological type of breast cancer [12]. In our study we assessed the dependence
between histological grade of malignancy (G1-G3) and basic immunohistochemical
profile, including EPO expression, but no statistically significant correlations were
found (p>0.05) (Table 3). Among all cancers the most numerous group exhibiting
EPO expression were the G3 cancers (67%) (Table 3). We also assessed the
relationship between tumor size (pT), presence of lymph node metastasis and
expression of EPO, finding that among all cancers expressing EPO, cancers assessed
as pT2 comprised the largest group (52.3%) (Table 4). As much as 36.2% of cancers
with lymph node metastasis exhibited EPO expression., while in 63.8% of cancers,
expression of EPO was not demonstrated (Table 5). We also assessed the relationship
between expression of EPO and basic immunohistochemical profile (ER, PR, HER2)
in studied cancers and found that the most numerous group exhibiting EPO
expression were the cancers presenting ER-/PR-/HER2+ basic immunohistochemical
profile (33.3%), while no EPO expression was most often demonstrated in cancers of
ER+/PR+/HER2- basic immunohistochemical profile (37.8%) (Table 6). No
statistically significant correlation was demonstrated between EPO expression and
tumor size (pT), lymph node metastasis (pN) or expression of estrogen and
progesterone receptors as well as HER2 (p>0.05). Similarly, Acs, Zhang et al. (2002)
found in their studies no correlation between EPO immunostaining and tumor size,
tumor grade, lymph node status, hormone receptor status, or HER2/neu
overexpression [12]. In a study by Acs, Zhang et al. (2002), when the analysis was
performed using the differential tumor score for EPO, we found a significant
correlation between increased EPO immunostaining in the tumors and the presence
of lymph node metastases [12]. From a clinical point of view it is important to
determine whether adverse effects of rhEPO therapy correlate with the expression of
EpoR in neoplastic cells, since rhEPO remains an important element of treatment
scheme for anemia of malignancy. EpoR protein has been proposed to be
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
72
nonfunctional in tumor cells due to a non-cell surface location, and therefore,
presumably, is not available for activation by rhEPO [13]. Todaro, Turdo et al.
(2013) found that breast cancer stem–like cells (BCSCs) isolated from patient tumors
express the EpoR. Among all types of breast cancer basal-like subtype demonstrated
the greatest expression of EpoR. This trial also showed that BCSCs respond to EPO
treatment with increased cell proliferation and self-renewal rate. Importantly, EPO
stimulation increased BCSCs resistance to chemotherapeutic agents and activated
cellular pathways responsible for survival and drug resistance [14]. Specifically, the
Akt and ERK pathways were activated in BCSCs at early time points following EPO
treatment, whereas Bcl-xL levels increased at later times. In vivo, EPO admini-
stration counteracted the effects of chemotherapeutic agents on BCSCs-derived
orthotopic tumor xenografts and promoted metastatic progression both in the
presence and in the absence of chemotherapy treatment [14]. All this data shows that
EPO acts directly on the breast cancer stem-like cells through activation of specific
signal transduction pathways responsible for tumor protection from chemotherapy
and accelerates disease progression. According to the latest results of research, the
population of neoplastic stem cells seems to be responsible for progression of
neoplastic disease, its recurrence and metastasis. In light of those discoveries it is
necessary to develop treatment that would target this population of cells. It has been
known for several years now that use of rhEPO negatively influences survival among
breast cancer patients. Studies by Todaro, Turdo et al. (2013) were the first to show
the effect of EPO and EpoR on breast cancer stem cells. These results confirmed that
EPO contributes to development of chemotherapy resistance and acceleration of
tumor growth [14]. Several years before, Phillips, Kim et al. (2007) investigated the
effect of erythropoietin on cancer stem cells in breast cancer cell lines. They found
that pharmacological concentrations of rhEPO increased the number of putative
breast cancer initiating cells (BCICs) in established breast cancer cell lines. This
increase was mediated by the activation of the Notch signaling pathway. Primarily,
the Notch pathway is important for cell-cell communication, which involves gene
regulation mechanisms that control multiple cell differentiation processes. The
increase in the number of BCICs observed after rhEPO treatment was significant and
the cells were not only viable but, what is more important, exhibited an increased
self-renewal capacity as demonstrated by primary in vitro sphere formation [15].
Phillips, McBride et al. (2006) have previously proved that activation of the Notch
signaling pathway is a part of the cellular stress response to clinical doses of ionizing
radiation [16]. This effect was mediated by increased expression of the Notch
receptor ligand Jagged-1 in the non-BCIC population that activated Notch signaling
in BCICs. As for radiation, Phillips, Kim et al. (2007) have demonstrated that rhEPO
treatment activated Notch signaling pathway in BCICs [15]. Research conducted for
the past several years enabled determining function of EPO and EpoR in various
types of cells, including demonstration of high activity within neoplastic cells. Over
a decade ago Acs, Chen et al. (2004) described autocrine regulation of apoptosis by
these proteins in breast cancer. MCF-7 breast cancer cell lines were incubated in
increasing concentrations of O2 and it was shown that increased expression of EPO
mRNA correlates tightly with increasing degree of hypoxia in cell culture [17].
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
73
Similarly, escalation of acute hypoxic conditions induced transcription of EpoR
mRNA. Verification with Western blot corroborated the differences in EPO and
EpoR protein expression in a manner corresponding to mRNA transcripts. The study
also confirmed that autocrine EPO signaling induced by moderate levels of hypoxia
inhibits hypoxia-induced apoptosis and promotes survival in MCF-7 human breast
cancer cells. Anti-apoptotic effect of EPO correlated with upregulation of Bcl-2 and
Bcl-XL, thus its mechanisms appear to be similar to those described in hematopoietic
cells. The above-described mechanism, colloquially known as “apoptosis escape of
tumors,” plays a key role in development and progression of neoplastic process.
Evasion of apoptosis is also one of the most important features enabling deter-
mination of the degree of tumor malignancy. Acs, Zhang et al. (2003) previously
shown that EPO can inhibit chemotherapeutic drug-induced apoptosis and
cytotoxicity [18]. In the next study their results suggested that the increased EPO
signaling induced by tumor hypoxia can play a significant role in resistance to
therapy of hypoxic tumors [19]. Hypoxia is a common phenomenon accompanying
solid tumors, such as breast cancer. Inadequate oxygen distribution, which most often
affects central regions of rapidly growing lesions, leads to rearrangement of gene
expression in tumor tissues subjected to hypoxia. In effect, there is increased
synthesis of proteins directed at protecting the cell from adverse conditions. It is
achieved by, among other things, blocking cell’s ability to initiate apoptosis. This
phenomenon is advantageous in most cells, but very dangerous from a point of view
of neoplastic process. Characteristic proteins that exhibit increased expression under
hypoxic conditions include EPO, EpoR and HIF-1α. Increased intracellular
concentration of STAT3 (signal transducer and activator of transcription 3) may be
also one of the markers of hypoxia. This transcription factor is activated through
phosphorylation of tyrosine 705. Despite the fact that this protein becomes
overexpressed in neoplasms and was even considered a protooncogene, accumulation
of STAT3 can be detected under various non-neoplastic conditions of increased cell
turnover and enhanced biosynthesis of various proteins [20, 21]. Links between
expressions of the proteins suggests functional dependences between STAT, HIF-1α,
EPO and EpoR in cell-to-cell signaling in breast cancer [22]. Breast cancer also
involves upregulation of transcriptional agents like STAT3, STAT3 activator – EpoR
(erythropoietin receptor), and a HIF-1 downstream protein – EPO [23]. STAT3
contributes to increase of EPO expression, which is also HIF-1α dependent. Tyrosine
phosphorylation of STAT3 is triggered by EPO. HIF-1α overexpression is an
indicator of poor prognosis and significantly decreased survival among patients with
breast cancer [24]. HIF-1 upregulates transcription of angiogenic genes like EPO and
vascular endothelial growth factor (VEGF), which induce sprouting of new vessels
and result in increased risk of metastasis as they increase contact surface between
tumor cells and vasculature. HIF-1 is responsible mainly for cellular adaptation to
hypoxic conditions. Genes triggered by this factor are responsible mainly for the
improvement in oxygen supply, adaptation of cells to anaerobic metabolism
conditions as well as for other changes facilitating cell survival in insufficient oxygen
availability. HIF-1 induces transcription of cytoprotective proteins in malignant cells
in hypoxic conditions [25, 26]. HIF-1 and STAT3 actions have been associated
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
74
indirectly with each other. This relationship was highlighted mainly by interference
of STAT3 transcription with small-molecule inhibitor and resultant downregulation
of HIF-1 and VEGF that delayed tumor growth and angiogenesis [27].
EPO and EpoR are induced by hypoxia in breast cancer and could contribute to
increased survival rate of tumor cells via counteraction to hypoxic injury [28]. EPO
counteracts outflow of cytochrome c from mitochondrion by upregulation of Bcl-xL.
EPO prevents apaf-1 complex dependent activation of caspase 9 and 3 by inhibition of
binding cytochrome c to apaf-1 and cyt-c in cytoplasm [23]. Wincewicz, Sulkowska et
al. (2007) described the above correlations in ductal breast carcinoma. STAT3 was
detected in 50% of cancers, HIF-1α in 72% of cases, EPO in 89% of all the cancers and
EpoR in 72%. There were significant associations between expression of STAT3 and
HIF-1α. STAT3 was significantly associated with expressions of EPO and EpoR in
cancers of all patients. HIF-1α was correlated with EPO and EpoR in most of analyzed
groups. This data indicates tight interrelationships between all those proteins in breast
cancer. They all become overexpressed under hypoxic conditions and they all
significantly influence the biology of the tumor. Occurrence of HIF-1α was
significantly increased in chemotherapy spared tumors compared to chemotherapy
treated cancers because chemotherapy could destructively affect cancer cells via
inhibition of protein expression [23]. Correlations between STAT3 and EPO suggest
their action in accord to support survival of breast cancer cells in human tumors in the
same fashion as in cell lines [29].
Results of studies by Reinbothe, Larsson et al. (2014) somewhat revolutionized the
perception of the role of EPO and EpoR in breast cancer. According to these results,
rhEPO stimulation of cultured EpoR-expressing breast cancer cells did not result in
increased proliferation, overt activation of EpoR (receptor phosphorylation) or
a consistent activation of canonical EpoR signaling pathway mediators such as
JAK2, STAT3, STAT5, or AKT. However, EpoR knockdown experiments suggested
functional EPO receptors in estrogen receptor positive (ERα+) breast cancer cell
lines, as reduced EpoR expression resulted in decreased proliferation, but not cell
death. This effect on proliferation was not seen in estrogen receptor negative (ERα-)
cells. These observations suggest that decreased expression of EpoR reduced the
ERα-dependent proliferation in breast cancer [30]. EpoR expression seems to play
a role in proliferation control of ERα+
breast cancers while survival seems to be
unaffected by reduction of EpoR expression. Studies by Reinbothe, Larsson et al.
(2014) show that in neoplasms of the breast exhibiting EpoR expression proliferation
is induced in conjunction with this receptor, but by an EPO-independent mechanism
in ERα+ breast cancers. Molecular mechanisms affecting the interactions between
EpoR and ERα as well as their effect on the biology of tumor cells require further
studies [30]. It is currently unknown how these proteins cooperate to regulate cell
proliferation in breast cancer. Questions need to be addressed in order to find out
how EpoR should be targeted to modulate the potent ERα signaling pathways.
Answers to this query may reveal the new potent therapeutic methods in breast
cancer. Conclusions from the studies by Reinbothe, Larsson et al. (2014) that use of
rhEPO does not influence proliferative activity or survival in the five tested breast
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
75
cancer cell lines are consistent with the results of studies by LaMontagne, Butler et
al. (2006) [30,31]. On the other hand, contradictory findings were reported in a study
by Acs, Acs et al. (2001) [28]. There are also discrepancies in the published data that
demonstrate that rhEPO stimulation of breast cancer cells results in changes in cell
signaling mediators, such as AKT, ERK1/2 and STATs [32,33].
Despite the fact that initially EPO was only attributed a role in regulation of
erythropoiesis, this protein turned out to be an important link connecting numerous
signal transduction pathways, both in many normal as well as cancerous non-
hematopoietic tissues. Moreover, in the recent years there have been reports
describing functional autocrine/paracrine EPO/EpoR systems in human neoplastic
cells derived from breast cancer, cervical cancer, melanoma and prostate cancer. This
data suggests that EPO/EpoR axis may affect tumor growth, disease progression and
metastasis [17, 34].
Liang, Qiu et al. (2014) described the autocrine/paracrine functions of EPO produced
by breast cancer cells under both normoxic and hypoxic conditions. They found that
the level of EPO produced by these cells was higher in hypoxia than in normoxia.
This observation is consistent with the knowledge that EPO is a product of hypoxia-
inducible gene expression. Unfortunately, this study did not allow for determining
whether the effects of EPO are produced as a result of auto- or paracrine signaling.
However, it was determined with high certainty that both types of intercellular
communication influenced the observed effects in cultured cell lines. Results
presented in the publication demonstrated also that EPO/EpoR autocrine/paracrine
signaling could mediate migration and influence invasion potential of breast cancer
cells. It suggests that autocrine/paracrine EPO signaling could be one of the effector
mechanisms, through which HIF-1 factor influences the invasiveness of breast cancer
[35]. Targeting of HIF-1, which is being actively investigated as a potential strategy
for cancer therapy, could inhibit the effects of autocrine/paracrine EPO signaling on
cell migration and invasiveness. Inhibition of EPO autocrine/paracrine signaling
pathways in cancer cells could be one of the mechanisms explaining the anticancer
effects of several anti-HIF-1 agents reported in the literature [36]. Furthermore, we
found that autocrine/paracrine production of EPO also played a role in stimulating
tumorsphere growth of breast cancer cells.
In recent studies Zhou, Damrauer et al. (2015) corroborated the hypothesis that
exogenous EPO does not significantly influence cell proliferation and does not
protect them from chemotherapy-induced apoptosis in vitro. It is somewhat different
under in vivo conditions, where EPO distinctly promotes progression of breast
cancer. According to these results, researches hypothesized that EPO’s tumor
promoting effects are seen in vivo but not with in vitro assays because it affects
a limited fraction of cells, such as breast tumor initiating cells. Thus, its effects might
only be seen with a longer period of EPO administration, such as those achieved in
vivo [37].
Moreover, it was observed that treatment of breast tumor-initiating cells (TICs) with
EPO activated JAK/STAT signaling as well as promoted their self-renewal. Studies
also confirmed the protumorigenic role of endogenous EPO in breast tumorigenesis.
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
76
It turned out that both breast cancer cells, as well as tumor-associated endothelial
cells can produce and release EPO into the microenvironment of the tumor.
Endogenous EPO expression was hypoxia-inducible in breast cancer cell lines, but
not in human mammary epithelial cells. EPO overexpression in breast cancer cells
correlates negatively with progression-free survival [37].
Research conducted in the recent years ascribes EPO another, clinically important
role. Observations allow saying that EPO antagonizes treatment with the anti-HER2
antibody, trastuzumab, by activating EpoR/JAK2 downstream effectors, effectively
bypassing HER2 signaling [38]. Moreover, the antagonizing effect should be only
apparent in patients treated with anti-HER2 antibodies. The fact that adverse effects
of rhEPO were observed among various subtypes of breast cancer as well as patients
with anemia of malignancy who had not been treated with trastuzumab strongly
suggests that there are other pathways not related to HER2, through which EPO
promotes tumor progression. The above observations demonstrate an important role
of both exogenous and endogenous EPO in the course of breast cancer. Results of
studies by Liang, Esteva et al. (2010) show that JAK2 inhibition in combination with
chemotherapy is a promising therapeutic strategy rather than EPO/EpoR inhibition in
the treatment of breast cancer patients [33].
Conclusions
Literature data suggest that EPO might influence tumor growth, disease progression
and metastasis. However, numerous studies are conducted on cell lines, but there is
scarce data from studies conducted on tissues using immuno-histochemical methods
routinely performed in pathomorphological diagnostics. There is an ongoing search
for prognostic markers in the diagnostics of breast cancer in women and researchers
strive to expand the basic immunohistochemical profile in the diagnostics of breast
cancer. In our study we demonstrated a statistically significant correlation between
markers included in the basic immunohistochemical profile (ER, PR, HER2) with
histological type of invasive breast cancer. There was no relationship between
markers of basic immunohistochemical profile and expression of EPO as well as no
statistically significant dependence was found between expression of EPO and basic
clinical features, such as tumor size, grade of histological malignancy or lymph node
status.
However, conducted studies allow for formulating important conclusions for
pathomorphological diagnostics. Based on the analysis, it may be concluded that:
1. Expression of erythropoietin (EPO) in the cytoplasm of cancer cells, was most
frequently identified in invasive ductal carcinoma of no special type (IDC
– NST);
2. Expression of EPO was most frequently identified in invasive breast cancers
evaluated as pT2 (52.3%);
3. Expression of EPO was most frequently identified in invasive breast cancers
evaluated as pN1;
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
77
4. Expression of erythropoietin (EPO) in the cytoplasm of cancer cells was most
often identified among cancers with the highest histological grade (G3). Our
research results suggest that increased EPO signaling may represent a novel
mechanism modulating differentiation of cancer cells and is connected with low
degree of differentiation of tumor cells;
5. Cancers with ER-/PR-/HER2- immunohistochemical profile, the so-called
“triple-negative” (TNBC) are characterized by, among other things: greater
clinical advancement of the disease at the moment of diagnosis, poor
histological differentiation (acc. to Bloom-Richardson classification; G3). In our
study the G3 cancers most often expressed EPO; also, among the TNBC
expression of EPO was confirmed in over 40% of cases, suggesting that in
TNBC erythropoietin might be a prognostic marker.
References
1. Badowska-Kozakiewicz A.M., Patera J., Sobol M., Przybylski J. The role of estrogen and
progesterone receptors in breast cancer – immunohistochemical evaluation of estrogen and
progesterone receptor expression in invasive breast cancer in women. Contemporary Oncology.
2015; 3(19): 1-6.
2. Badowska-Kozakiewicz A.M., Sobol M., Patera J., Kozłowski W. Immunohistochemical evaluation
of human epidermal growth factor receptor 2 and estrogen and progesterone receptors in
invasive breast cancer in women. Archives of Medical Sciences. 2013; 9(3): 466-471.
3. Trost N., Stepisnik T., Berne S., Pucer A., Petan T., Komel R., Debeljak N.
Recombinant human erythropoietin alters gene expression and stimulates proliferation of MCF-7 breast cancer cells. Radiology and Oncology. 2013; 47(4): 382-389.
4. Lopez T.V., Lappin T.R., Maxwell P., Shi Z., Lopez-Marure R., Aguilar C., Rocha-Zavaleta L.
Autocrine/paracrine erythropoietin signalling promotes JAK/STAT-dependent proliferation
of human cervical cancer cells. International Journal of Cancer. 2011; 129: 2566-2576.
5. Arcasoy M.O., Amin K., Vollmer R.T., Jiang X., Demark-Wahnefried W., Haroon Z.A.
Erythropoietin and erythropoietin receptor expression in human prostate cancer. Modern
Pathology. 2005; 18: 421-430.
6. Kaushansky K. Lineage-specific hematopoietic growth factors. The New England Journal of
Medicine. 2006; 354: 2034-2045.
7. Lai S.Y., Grandis J.R. Understanding the presence and function of erythropoietin receptors on
cancer cells. Journal of Clinical Oncology. 2006; 24: 4675-4676.
8. Wang L., Li H.G., Xia Z.S. Prognostic significance of erythropoietin and erythropoietin
receptor in gastric adenocarcinoma. World Journal of Gastroenterology. 2011; 17(34): 3933-
3940.
9. Aapro M., Leonard R.C., Barnadas A., Marangolo M., Untch M., Malamos N., Mayordomo J.,
Reichert T., Pedrini J.L., Ukarma L., Scherhag A., Burger H. Effects of once-weekly epoetin beta
on survival in patients with metastatic breast cancer receiving anthracycline- and/or taxane-
based chemotherapy. Results of the Breast Cancer-Anemia and the Value of Erythropoietin (BRAVE) Study. Journal of Clinical Oncology. 2008; 26: 592-598.
10. Volgger B., Kurz K., Zoschg K., Theurl I., Ciresa-Konig A., Marth C., Weiss G. Importance of
erythropoietin receptor expression in tumour. Anticancer Research. 2010; 30: 3721-3726.
Maksimiuk M., Sobiborowicz A., Sobieraj M., Liszcz A., Sobol M., Patera J., Badowska-Kozakiewicz A.M.
78
11. Hershman D.L., Buono D.L., Malin J., McBride R., Tsai W.Y., Neugut A.I. Patterns of use and
risks associated with erythropoiesis-stimulating agents among Medicare patients with cancer. Journal of the National Cancer Institute. 2009; 101: 1633-41.
12. Acs G., Zhang P.J., Rebbeck T.R., Acs P., Verma A. Immunohistochemical expression of
erythropoietin receptor in breast carcinoma. Cancer. 2002; 95(5): 969-981.
13. Kokhaei P., Abdalla A.O., Hansson L., Mikaelsson E., Kubbies M., Haselbeck A., Jernberg-Wiklund
H., Mellstedt H., Osterborg A. Expression of erythropoietin receptor and in vitro functional effects
of epoetins in B-cell malignancies. Clinical Cancer Research. 2007; 13: 3536-44.
14. Todaro M., Turdo A., Bartucci M., Iovino F., Dattilo R., Biffoni M., Stassi G., Federici G., De Maria
R., Zeuner A. Erythropoietin activates cell survival pathways in breast cancer stem–like cells to
protect them from chemotherapy. Cancer Research. 2013; 73(21): 6393-6400.
15. Phillips T.M., Kim K., Vlashi E., McBride W.H., Pajonk F. Effects of recombinant erythropoietin
on breast cancer–initiating cells. Neoplasia. 2007; 9(12): 1122-1129.
16. Phillips T.M., McBride W.H., and Pajonk F. The response of CD24(-/low)/CD44+ breast cancer
– initiating cells to radiation. Journal of the National Cancer Institute. 2006; 98: 1777-1785.
17. Acs G., Chen M., Xu X., Acs P., Verma A., Koch C.J. Autocrine erythropoietin signaling inhibits
hypoxia-induced apoptosis in human breast carcinoma cells. Cancer Letters. 2004; 214: 243-251.
18. Acs G., Zhang P.J., McGrath C.M., Acs P., McBroom J., Mohyeldin A., Liu S., Lu H., Verma A.
Hypoxia-inducible erythropoietin signaling in squamous dysplasia and squamous cell
carcinoma of the uterine cervix and its potential role in cervical carcinogenesis and tumor
progression. American Journal of Pathology. 2003; 162: 1789-1806.
19. Fu P., Jiang X., Arcasoy M.O. Constitutively active erythropoietin receptor expression in breast
cancer cells promotes cellular proliferation and migration through a MAP-kinase dependent pathway. Biochemical and Biophysical Research Communications. 2009; 379: 696-701.
20. Sawa S., Kamimura D., Jin G.H., Morikawa H., Kamon H., Nishihara M., Ishihara K., Murakami
M., Hirano T. Autoimmune arthritis associated with mutated interleukin (IL)-6 receptor gp130
is driven by STAT3/IL-7-dependent homeostatic proliferation of CD4+ T cells. The Journal of
Experimental Medicine. 2006; 203: 1459-1470.
21. Murray P.J. STAT3-mediated anti-inflammatory signalling. Biochemical Society Transactions.
2006; 34: 1028-1031.
22. Wincewicz A., Koda M., Sulkowska M., Kanczuga-Koda L., Wincewicz D., Sulkowski S. STAT3
and hypoxia induced proteins – HIF-1alpha, EPO and EPOR in relation with Bax and Bcl-xL
in nodal metastases of ductal breast cancers. Folia Histochemica et Cytobiologica. 2009; 47(3):
425-430.
23. Wincewicz A., Sulkowska M., Koda M., Leśniewicz T., Kanczuga-Koda L., Sulkowski S. STAT3,
HIF-1α, EPO and EPOR – signaling proteins in human primary ductal breast cancers. Folia
Histochemica et Cytobiologica.
24. Schindl M., Schoppmann S.F., Samonigg H., Hausmaninger H., Kwasny W., Gnant M., Jakesz
R., Kubista E., Birner P., Oberhuber G., Austrian Breast and Colorectal Cancer Study Group.
Overexpression of hypoxia-inducible factor 1alpha is associated with an unfavorable
prognosis in lymph node-positive breast cancer. Clinical Cancer Research. 2002; 8: 1831-1837.
25. Dales J.P., Garcia S., Meunier-Carpentier S., Andrac-Meyer L., Haddad O., Lavaut M.N., Allasia
C., Bonnier P., Charpin C. Overexpression of hypoxia-inducible factor HIF-1alpha predicts
early relapse in breast cancer: retrospective study in a series of 745 patients. International
Journal of Cancer. 2005; 116: 734-739.
26. Kronblad A., Jirstrom K., Ryden L., Nordenskjold B., Landberg G. Hypoxia inducible factor-
1alpha is a prognostic marker in premenopausal patients with intermediate to highly
Immunohistochemical expression of erythropoietin
in invasive breast carcinoma with metastasis to lymph nodes
79
differentiated breast cancer but not a predictive marker for tamoxifen response. International
Journal of Cancer. 2006; 118: 2609-2616.
27. Xu Q., Briggs J., Park S., Niu G., Kortylewski M., Zhang S., Gritsko T., Turkson J., Kay H.,
Semenza G.L., Cheng J.Q., Jove R., Yu H. Targeting Stat3 blocks both HIF-1 and VEGF
expression induced by multiple oncogenic growth signaling pathways. Oncogene. 2005; 24:
5552-5560.
28. Acs G., Acs P., Beckwith S.M., Pitts R.L., Clements E., Wong K., Verma A. Erythropoietin and
erythropoietin receptor expression in human cancer. Cancer Research. 2001; 61: 3561-3565.
29. Gritsko T., Williams A., Turkson J., Kaneko S., Bowman T., Huang M., Nam S., Eweis I., Diaz N.,
Sullivan D., Yoder S., Enkemann S., Eschrich S., Lee J.H., Beam C.A., Cheng J., Minton S., Muro-
Cacho C.A., Jove R. Persistent activation of stat3 signaling induces survivin gene expression
and confers resistance to apoptosis in human breast cancer cells. Clinical Cancer Research.
2006; 12: 11-19.
30. Reinbothe S., Larsson A.M., Vaapil M., Wigerup C., Sun J., Jögi A., Neumann D., Rönnstrand L.,
Pĺhlman S. EPO-independent functional EPO receptor in breast cancer enhances estrogen
receptor activity and promotes cell proliferation. Biochemical and Biophysical Research
Communications. 2014; 445: 163-169.
31. LaMontagne K.R., Butler J., Marshall D.J., Tullai J., Gechtman Z., Hall C., Meshaw A., Farrell F.X.
Recombinant epoetins do not stimulate tumor growth in erythropoietin receptor-positive breast carcinoma models. Molecular Cancer Therapeutics. 2006; 5: 347-355.
32. Jin W., Lin Z., Zhang X., Kong L., Yang L. Effects and mechanism of recombinant
human erythropoietin on the growth of human breast cancer MDA-MB-231 cells in nude
mice. Pathology – Research and Practice. 2015; 211(8): 570-576.
33. Liang K., Esteva F.J., Albarracin C., Stemke-Hale K., Lu Y., Bianchini G., Yang C.Y., Li Y., Li X.,
Chen C.T., Mills G.B., Hortobagyi G.N., Mendelsohn J., Hung M.C., Fan Z. Recombinant human
erythropoietin antagonizes trastuzumab treatment of breast cancer cells via Jak2-mediated
Src activation and PTEN inactivation. Cancer Cell. 2010; 18: 423-435.
34. Pelekanou V., Kampa M., Kafousi M., Dambaki K., Darivianaki K., Vrekoussis T., Sanidas E.,
Tsiftsis D.D., Stathopoulos E.N., Castanas E. Erythropoietin and its receptor in breast cancer:
correlation with steroid receptors and outcome. Cancer Epidemiology, Biomarkers &
Prevention. 2007; 16: 2016-2023.
35. Liang K., Qiu S., Lu Y., Fan Z. Autocrine/paracrine erythropoietin regulates migration and
invasion potential and the stemness of human breast cancer cells. Cancer Biology & Therapy.
2014; 15: 89-98.
36. Garber K. New drugs target hypoxia response in tumors. Journal of the National Cancer Institute.
2005; 97: 1112-1114.
37. Zhou B., Damrauer J.S., Bailey S.T., Hadzic T., Jeong Y., Clark K., Fan C., Murphy L., Lee C.Y.,
Troester M.A., Miller R., Jin J., Darr D., Perou C.M., Levine R.L., Diehn M., Kim W.Y.
Erythropoietin promotes breast tumorigenesis through tumor-initiating cell self-renewal.
Journal of Clinical Investigation. 2014; 124(2): 553-563.
38. Fillmore C.M., Kuperwasser C. Human breast cancer cell lines contain stem-like cells that self-
renew, give rise to phenotypically diverse progeny and survive chemotherapy. Breast Cancer
Research. 2008; 10(2): R25.
80
The inflammatory components in lung cancer
Bartosz Szmyd, Marcin Kaszkowiak, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-
Lasota
Department of Biomedicine and Genetics, Medical University of Lodz, Poland
The corresponding author:
Dorota Pastuszak-Lewandoska, PhD
Department of Biomedicine and Genetics Chair of Biology and Medical Parasitology,
Medical University of Lodz 251 Pomorska Street (C-5 building), 92-213 Lodz
Abstract
The immune system is one of the most complicated and least known part of the human body. According
to Virchow’s observations and the current data, its response can cause both promotion and suppression
of formation and development of the neoplasm. It is estimated that as many as 25% of all cancers are
caused by chronic inflammation. Better understanding of the delicate balance between these activities is
essential for development of proper preventive and treatment methods as well as new biological
markers. In our paper we focus on biochemical and molecular mechanisms of tumorigenic activity of
chronic inflammation in lung cancer - the most frequent malignant neoplasm characterized by high
mortality. Moreover, we describe mechanisms in which the immune system fights against tumours cells.
The last part is concerned with new strategies in lung cancer diagnosis and treatment: biomarkers and
immunotherapy.
Keywords: Lung cancer, NSCLC, inflammation, biomarkers, immunotherapy
Introduction
The immune system and its role in cancerogenesis
In 1863 Rudolf Virchow observed that:
Cells of the immune system occur in tumours
Cancerogenesis tends to appear at sites of the long-lasting inflammation
(Balkwill, Mantovani, 2001), (Schetter, Heegaard et al. 2009).
When we take into consideration all sources of inflammatory conditions (such as
viral infections, exposure to allergens, autoimmune disorders and even obesity), it
reveals that they are responsible for approximately 15 to 25% of tumours (Hussain,
Harris, 2007)(Balkwill, Mantovani, 2001). As Figure 1 shows, the response of the
immune system can result in both: anti- and pro-tumorigenic effect. The first case is
typical for efficient immune effector system (Ostrand-Rosenberg, 2009), the second
for chronic inflammation which predispose cells for oncogenic transformation
(Mantovani, Allavena et al. 2008). The range of mechanisms involved in both
responses is outlined in further chapters.
The inflammatory components in lung cancer
81
Fig. 1. The balance between both pro-tumorigenic and anti-tumorigenic effect of immune response.
Based on the diagram from Schetter, Heegaard et al. 2009.
The influence of the immune system on cancerogenesis has been described for many
organs, especially: intestines, liver, stomach and lungs. The examples of tumours
with corresponding inflammatory state are summarized in Table 1.
Table 1. The inflammatory states and their correlating tumours.
Source of inflammation Tumour References
Chronic viral hepatitis Liver cancer (Tu, Bühler et al. 2017), (Abbas,
Abbas et al. 2015), (Benali-Furet,
Chami et al. 2005)
Helicobacter pylori Gastric cancer (Sepulveda, 2013), (Wroblewski,
Peek et al. 2010)
Tumour-associated
macrophages (TAMs)
activation
Oral squamous cell
carcinoma (OSCC)
(Petruzzi, Cherubini et al. 2017),
(Fujii, Shomori et al. 2012)
Inflammatory Bowel
Disease
Colorectal cancer (Wang, Fang, 2014), (Kim, Chang,
2014)
Bartosz Szmyd, Marcin Kaszkowiak, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
82
Crohn's disease1 Colorectal cancer (Santos, Barbosa, 2017), (Freeman,
2008), (Gillen, Andrews et al. 1994)
Ulcerative colitis Colorectal cancer (Yashiro, 2014), (Isbell, Levin, 1988)
(Ransohoff, Riddell et al. 1985)
Hashimoto’s thyroiditis Papillary thyroid cancer
(PTC)
(Resende de Paiva, Grønhøj et al.
2017), (Chen, Lin et al. 2013)
Asbestos Mesothelioma, lung
cancer
(Hodgson, Darnton, 2000), (Berry,
1991), (Whitwell, Scott et al. 1977)
Tobacco smoking Lung cancer (Peto, Darby et al. 2000), (Wingo,
Ries et al. 1999), (Hecht, 1999)
Lung cancer
Lung cancer is the most common malignant neoplasm (Rafiemanesh, Mehtarpour et
al. 2016) and the leading cause of death of tumour (Islami, Torre et al. 2015). Current
research show that ASIR (standardized incidence rate) is equal 23.1 and ASMR
(standardized mortality rate) – 19.7 (per 100 000) (Rafiemanesh et al., 2016). Both
environmental and genetic factors are associated with lung cancer (Kanwal, Ding et al.
2017). Moreover, environmental factors (such as smoking and exposure to asbestos),
which results in inflammation, play an important role in the etiopathogenesis of lung
cancer (Schetter, Heegaard et al. 2009) (Pope III, Burnett et al. 2002).
Lung cancer can be classified according to histological (Zheng, 2016) and clinical
(Politi, Herbst, 2015) parameters. Focusing on the clinical classification, we can
distinguish:
Small-cell lung cancer (SCLC);
Non-small-cellular lung cancer (NSCLC) which is also divided into:
Squamous cell lung carcinoma (SCC),
Adenocarcinoma (AC),
Large-cell lung carcinoma (LCC).
The fight between cancer and the immune system
The anti-tumorigenic response is possible thanks to tumour-associated antigens
(TAAs). Dendritic cells (DCs) present these antigens by MHC class II. Other antigen
presenting cells (APCs) perform this function to a lesser extent (Szyszka-Barth,
Ramlau et al. 2014). DCs/APCs migrate to lymph nodes. It results in interaction
between TAA and naïve T-cells. Due to this interaction the first signal is conducted
by the antigen receptor – T-cell receptor (TCR). The second signal, needed to T-cells
activation, is called co-stimulation (Frauwirth, Thompson, 2002). These signals
activate effector T-cells (both CD4+ and CD8+, called helper and cytotoxic T-cells
(CTLs), respectively). Activated CD4+ T-cells secrete cytokines, which take part in
1 In memorial of Polish surgeon Antoni Leśniowski, who was first to describe ileitis terminalis, also called
Leśniowski-Crohn’s disease (Lichtarowicz, Mayberry, 1988).
The inflammatory components in lung cancer
83
e.g. CTLs’ and B-cell activation, phagocytosis promotion and stimulation of natural
killer cells (NKs).
However, the response of the immune system is often insufficient (Swann, Smyth et
al. 2007). Tumour fights with patient’s immunological system through:
the generation of antigen-loss variants;
the immunological tolerance;
the lesser expression of MHC on tumour cells’ surface;
the secretion of immunoinhibitors.
The origination of antigen-loss variants depends on selection of tumour cell
subpopulations which do not express antigens causing anti-tumorigenic immune
response. It enables protection against immune targeting (Olson, McNeel, 2012). The
tolerance mechanisms against some antigens result in the lack of response to them
with relevant immune response against others antigens (Mapara, Sykes, 2004).
Moreover, host cells fighting against tumour are recognized as autoreactive and
eliminated. The defence strategy bases on decreased expression level of MHC on
tumour cell surface limits the presenting of TAAs. More importantly, due to this
mechanism tumour cells are the target for NK-cells (Topham, Hewitt, 2009). Finally,
malignant cells can secrete immunoinhibitors. One of them is transforming growth
factor beta (TGF-β) (Szyszka-Barth, Ramlau et al. 2014). Physiologically, TGF-β
regulates many life processes, such as: cell growth, differentiation, inflammation and
angiogenesis. However, the overexpression of this cytokine is correlated with tumor
progression, metastasis and poor prognostic outcome. (Fabregat, Fernando et al.
2014).
Inflammatory biomarkers in lung cancer
In fact, all of inflammatory genes, theirs products, circulating cytokines and
microRNAs can be considered as biomarkers in lung cancer (Schetter, Heegaard et
al. 2009). For the purpose of our article, we will only focus on microRNAs affecting
inflammatory genes and circulating cytokines.
Fig. 2. The possibilities of biological material selection for assessment of expression level of different
biomarkers (microRNA, genes, proteins).
Bartosz Szmyd, Marcin Kaszkowiak, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
84
Both of them tend to be suitable biomarkers for early cancer detection because of the
relative non-invasiveness and easiness in collection of biological material (Paranjape,
Slack et al. 2009). They can be assessed in the serum samples, while the
inflammatory genes expression should be evaluated in tumour tissue obtained during
pulmonectomy/lobectomy (Mitra, Lee et al. 2011) (see Figure 2). The examples of
circulating proteins and microRNAs which can be considered as inflammatory
biomarkers in lung cancer are summarized in Table 2.
Table 2. The examples of circulating proteins and microRNAs which can be considered as inflammatory
biomarkers in lung cancer (with relevant references).
Biomarker References
circulating proteins
C-reactive protein (CRP) (Zhou, Liu et al. 2012), (Alifano, Falcoz et al. 2011)
Interleukin 6 (IL-6) (Brichory, Misek et al. 2001), (Heikkilä, Harris et al. 2009)
Interleukin 8 (IL-8) (Pine, Mechanic et al. 2011), (Zhu, Webster et al. 2004)
microRNAs
microRNA-21 (Xue, Liu et al. 2016), (Ma, Liu et al. 2012)
microRNA-17-92 cluster (Ebi, Sato et al. 2009), (Hayashita, Osada et al. 2005)
MicroRNAs are the class of short (20-25 nucleotides), evolutionarily conserved, non-
coding RNA (Williams, Cheng et al. 2017). They down-regulate genes expression by
both: translational repression and mRNA degradation (Anglicheau, Muthukumar et al.
2010). More importantly, single microRNAs can target many genes (Hashimoto,
Akiyama et al. 2013). We will focus on miRNA-21 and microRNA-17-92 cluster.
miRNA-21 regulates the expression of IL-12 (Lu, Munitz et al. 2009). This interleukin
performs many roles e.g.: it is engaged in origination of Th1 cells (Hsieh, Macatonia et
al. 1993) and is the suppressor of angiogenesis (Sgadari, Angiolillo et al. 1996). The
upregulation of microRNA-21 observed in lung cancer (Xue, Liue et al. 2016) results in
the downregulation of its targets. Thus, important functions of IL-12 are impaired.
While, microRNA-17-92 cluster is the set of seven microRNAs: microRNA-17-3p,
microRNA-17-5p, microRNA-18a, microRNA-19a, microRNA-20a, microRNA-19b-1
and microRNA-92a (Inamura, Ishikawa, 2016). Its overexpression is often observed in
lung cancer (Ebi, Sato et al., 2009). This MicroRNA class results in e.g. cell
proliferation under normoxia, inhibition of hypoxia-induced apoptosis and regulation
of angiogenesis (Osada, Takahashi, 2011).
The second group of potential biomarkers are circulating proteins, especially
cytokines. These small proteins are secreted by cells and influence these or other
cells (Zhang, An, 2007). They play key role in communication between cells
(Motaln, Turnsek, 2015). There are distinguished pro- (e.g. IL-1β, IL-8, IL-12) and
anti-inflammatory (IL-4, IL-5, IL-10) cytokines (Enewold, Mechanic et al. 2009).
The inflammatory components in lung cancer
85
IL-8 is involved in wide range of processes, such as recruitment/activation of
neutrophil, angiogenesis and metastasis in lung cancer (especially NSCLC).
Oncogenic activity of IL-8 can be result of EGFR transactivation. (Luppi, Longo et
al. 2007). IL-10 is anti-inflammatory cytokine. It is responsible for limiting host
immune response to pathogens and homeostasis maintain (Iyer, Cheng, 2012). Its
role in the tumorigenesis is controversial: its expression levels were significantly
increased in the lungs (mice model of lung cancer), higher level of IL-10 correlates
with a poor prognosis in lung cancer patients (Hsu, Wang et al. 2016). Hsu et al.
convince that IL-10 and EGFR regulate each other through positive feedback, which
leads to lung cancer formation.
The C-reactive protein (CRP) is pentameric protein, which occurs in blood in
response to inflammation. The performed meta-analysis of the role of its in lung
cancer (Zhou, Liu et al. 2012) shows:
RR of lung cancer for one unit change in natural logarithm CRP was 1.28 (95%
CI 1.17–1.41);
CRP was significantly correlated with increased risk of lung cancer among men:
RR=1.18 (95% CI 1.09–1.28);
Any statistically significant difference among women wasn’t observed.
The immunotherapy of lung cancer
The immunotherapy in treatment of neoplasm has been used for the first time in 1890
by doctor William Coley. He noticed that facial sarcoma could regress after bacterial
infection (Coley, 1893). Since that moment research in tumours immunotherapy have
developed significantly (Szyszka-Barth, Ramlau et al. 2014). Two types of
immunotherapy are distinguished: active and passive (see Table 3).
Table 3. The definitions of active and passive immunotherapy with examples. Based on (Szyszka-Barth,
Ramlau et al. 2014).
Type Definition Examples
Active Stimulation of immunological
system to fight against neoplasm
Vaccines (containing tumours antigens and
adjuvants)
Nonspecific immunomodulators
Passive The application of
immunologically active factors
synthesized/formed outside
patient’s organism
Adoptive T-cells
Monoclonal antibodies
Active immunotherapy depends on stimulation of immunological system to fight
against tumour. It comprised of vaccination and administration of nonspecific
immunomodulators. The possibility of vaccinations usage is notably exploited
especially in combination therapy of NSCLC. Many clinical trials have been
performed based on that topic (e.g. MAGE-A3, PRAME, Lucanix, CimaVax) (De
Pas, Giovannini et al. 2012). In this paragraph, we focus on vaccines targeting MUC-
Bartosz Szmyd, Marcin Kaszkowiak, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
86
1 (L-BLP25 and TG4010). It is transmembrane protein, which increased expression
is observed in NSCLC (AC: 86%, others: 74%) (Bouillez, Adeegbe et al. 2017)
(Szyszka-Barth, Ramlau et al. 2014). Genetic abnormalities of MUC-1 cause:
resistance to chemotherapy;
protection against apoptosis;
immunosuppression (Zappa, Mousa, 2016).
Since overexpression of MUC-1 and decreased level of patients’ own antibodies
against MUC-1 result in poor prognosis, this direction of research seems to be
reasonable (Lakshmanan, Ponnusamy et al. 2015).
As shown in Table 3, passive immunotherapy can be defined as application of
immunologically active factors synthesized or formed outside patient’s body.
Adoptive cell transfer (ACT) is considered as its example. Its following steps are
shown on Figure 3.
Fig. 3. Following steps of adoptive cell transfer. The last step is infusion of patient’s own T-cells, which are
characterized by anti-tumour properties.
Thanks to this procedure infused T-cells have relatively high avidity, which can be
also increased by genetic modification. Clinical trials of ACT in lung cancer focus
on: lymphokine-activated killer (LAK), cytokine-induced killer (CIK) and natural
killer (NKT) cells (Yang, Wang et al. 2016).
Summary
Better understanding of the delicate balance between opposite activities is essential
to the preparation of proper prophylactic methods, finding new biological markers
and the development of new treatment methods. Circulating microRNAs isolated
from serum exosomes or serum samples are promising biomarkers for detection and
predication of the clinical curse of lung cancer. Moreover, the immunotherapy can be
considered as the possible treatment.
The inflammatory components in lung cancer
87
References
Abbas Z., Abbas M., Abbas S., & Shazi L. Hepatitis D and hepatocellular carcinoma. World Journal
of Hepatology. 2015; 7(5):777-786.
Alifano M., Falcoz P. E., Seegers V., Roche N., Schussler O., Younes M., Antonacci F., Forgez P.,
Dechartres A., Massard G., Damotte D., & Régnard J.-F. Preresection serum C-reactive protein
measurement and survival among patients with resectable non–small cell lung cancer. The Journal of
Thoracic and Cardiovascular Surgery. 2011; 142(5):1161-1167.
Anglicheau D., Muthukumar T., & Suthanthiran M. MicroRNAs: small RNAs with big effects.
Transplantation. 2010; 90(2):105-112.
Balkwill F., & Mantovani A. Inflammation and cancer: Back to Virchow? Lancet. 2001;
357(9255):539-545.
Benali-Furet N. L., Chami M., Houel L., De Giorgi F., Vernejoul F., Lagorce D., Buscail L.,
Bartenschlager R., Ichas F., Rizzuto R., & Paterlini-Bréchot P. Hepatitis C virus core triggers
apoptosis in liver cells by inducing ER stress and ER calcium depletion. Oncogene. 2005;
24(31):4921-4933.
Berry G. Prediction of mesothelioma, lung cancer, and asbestosis in former Wittenoom asbestos
workers. British Journal of Industrial Medicine. 1991; 48(12):793-802.
Bouillez A., Adeegbe D., Jin C., Hu X., Tagde A., Alam M., Rajabi H., Wong K.-K., & Kufe D.
MUC1-C promotes the suppressive immune microenvironment in non-small cell lung cancer.
OncoImmunology. 2017; 6(9):e1338998.
Brichory F. M., Misek D. E., Yim A.-M., Krause M. C., Giordano T. J., Beer D. G., & Hanash S. M. An
immune response manifested by the common occurrence of annexins I and II autoantibodies and high circulating levels of IL-6 in lung cancer. Proceedings of the National Academy of Sciences.
2001; 98(17):9824-9829.
Chen Y.-K., Lin C.-L., Cheng F. T.-F., Sung F.-C., & Kao C.-H. Cancer risk in patients with
Hashimoto’s thyroiditis: a nationwide cohort study. British Journal of Cancer. 2013; 109(9):2496-2501.
Coley W. B. The treatment of malignant tumors by repeated inoculations of erysipelas. The
American Journal of the Medical Sciences. 1893; 105(5):487-510.
De Pas T., Giovannini M., Rescigno M., Catania C., Toffalorio F., Spitaleri G., Delmonte A., Barberis
M., Spaggiari L., Solli P., Veronesi G., & De Braud F. Vaccines in non-small cell lung cancer:
Rationale, combination strategies and update on clinical trials. Critical Reviews in
Oncology/Hematology. 2012.
Ebi H., Sato T., Sugito N., Hosono Y., Yatabe Y., Matsuyama Y., Yamaguchi T., Osada H., Suzuki M., &
Takahashi T. Counterbalance between RB inactivation and miR-17–92 overexpression in reactive
oxygen species and DNA damage induction in lung cancers. Oncogene. 2009; 28(38):3371-3379.
Enewold L., Mechanic L. E., Bowman E. D., Zheng Y.-L., Yu Z., Trivers G., Alberg A. J., & Harris C.
C. Serum concentrations of cytokines and lung cancer survival in African Americans and
Caucasians. Cancer Epidemiology, Biomarkers & Prevention : A Publication of the American
Association for Cancer Research, Cosponsored by the American Society of Preventive Oncology. 2009;
18(1):215-222.
Fabregat I., Fernando J., Mainez J., & Sancho P. TGF-beta signaling in cancer treatment. Current
Pharmaceutical Design. 2014; 20(17):2934-2947.
Frauwirth K. A., & Thompson C. B. Activation and inhibition of lymphocytes by costimulation.
Journal of Clinical Investigation. 2002; American Society for Clinical Investigation.
Freeman H.-J. Colorectal cancer risk in Crohn’s disease. World Journal of Gastroenterology. 2008;
14(12):1810-1811.
Bartosz Szmyd, Marcin Kaszkowiak, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
88
Fujii N., Shomori K., Shiomi T., Nakabayashi M., Takeda C., Ryoke K., & Ito H. Cancer-associated
fibroblasts and CD163-positive macrophages in oral squamous cell carcinoma: their
clinicopathological and prognostic significance. Journal of Oral Pathology & Medicine. 2012;
41(6):444-451.
Gillen C. D., Andrews H. A., Prior P., & Allan R. N. Crohn’s disease and colorectal cancer. Gut.
1994; 35(5):651-655.
Hashimoto Y., Akiyama Y., & Yuasa Y. Multiple-to-Multiple Relationships between MicroRNAs
and Target Genes in Gastric Cancer. PLoS ONE. 2013; 8(5):e62589.
Hayashita Y., Osada H., Tatematsu Y., Yamada H., Yanagisawa K., Tomida S., Yatabe Y., Kawahara
K., Sekido Y., & Takahashi T. A Polycistronic MicroRNA Cluster, miR-17-92, Is Overexpressed in
Human Lung Cancers and Enhances Cell Proliferation. Cancer Research. 2005; 65(21):9628-9632.
Hecht S. S. Tobacco Smoke Carcinogens and Lung Cancer. JNCI Journal of the National Cancer
Institute. 1999; 91(14):1194-1210.
Heikkilä K., Harris R., Lowe G., Rumley A., Yarnell J., Gallacher J., Ben-Shlomo Y., Ebrahim S., &
Lawlor D. A. Associations of circulating C-reactive protein and interleukin-6 with cancer risk:
findings from two prospective cohorts and a meta-analysis. Cancer Causes & Control. 2009;
20(1):15-26.
Hodgson J. T., & Darnton A. The quantitative risks of mesothelioma and lung cancer in relation to
asbestos exposure. The Annals of Occupational Hygiene. 2000; 44(8):565-601.
Hsieh C. S., Macatonia S. E., Tripp C. S., Wolf S. F., O’Garra A., & Murphy K. M. Development of
TH1 CD4+ T cells through IL-12 produced by Listeria-induced macrophages. Science (New York,
N.Y.). 1993; 260(5107):547-549.
Hsu T.-I., Wang Y.-C., Hung C.-Y., Yu C.-H., Su W.-C., Chang W.-C., & Hung J.-J. Positive feedback
regulation between IL10 and EGFR promotes lung cancer formation. Oncotarget. 2016;
7(15):20840-20854.
Hussain S. P., & Harris C. C. Inflammation and cancer: An ancient link with novel potentials.
International Journal of Cancer. 2007; 121(11):2373-2380.
Inamura K., & Ishikawa Y. MicroRNA In Lung Cancer: Novel Biomarkers and Potential Tools for
Treatment. Journal of Clinical Medicine. 2016; 5(3).
Isbell G., & Levin B. Ulcerative colitis and colon cancer. Gastroenterology Clinics of North America.
1988; 17(4):773-791.
Islami F., Torre L. A., & Jemal A. Global trends of lung cancer mortality and smoking prevalence.
Translational Lung Cancer Research. 2015; 4(4):327-338.
Iyer S. S., & Cheng G. Role of interleukin 10 transcriptional regulation in inflammation and
autoimmune disease. Critical Reviews in Immunology. 2012; 32(1):23-63.
Kanwal M., Ding X.-J., & Cao Y. Familial risk for lung cancer. Oncology Letters. 2017; 13(2):535-542.
Kim E. R., & Chang D. K. Colorectal cancer in inflammatory bowel disease: the risk, pathogenesis,
prevention and diagnosis. World Journal of Gastroenterology. 2014; 20(29):9872-9881.
Lakshmanan I., Ponnusamy M. P., Macha M. A., Haridas D., Majhi P. D., Kaur S., Jain M., Batra S. K.,
& Ganti A. K. Mucins in lung cancer: Diagnostic, prognostic, and therapeutic implications. Journal
of Thoracic Oncology. 2015; Elsevier.
Lichtarowicz A. M., & Mayberry J. F. Antoni Lésniowski and his contribution to regional enteritis
(Crohn’s disease). Journal of the Royal Society of Medicine. 1988; 81(8):468-470.
The inflammatory components in lung cancer
89
Lu T. X., Munitz A., & Rothenberg M. E. MicroRNA-21 is up-regulated in allergic airway
inflammation and regulates IL-12p35 expression. Journal of Immunology (Baltimore, Md. : 1950).
2009; 182(8):4994-5002.
Luppi F., Longo A. M., de Boer W. I., Rabe K. F., & Hiemstra P. S. Interleukin-8 stimulates cell
proliferation in non-small cell lung cancer through epidermal growth factor receptor transactivation. Lung Cancer. 2007; 56(1):25-33.
Ma X.-L., Liu L., Liu X.-X., Li Y., Deng L., Xiao Z.-L., Liu Y.-T., Shi H.-S., & Wei Y. Prognostic
role of microRNA-21 in non-small cell lung cancer: a meta-analysis. Asian Pacific Journal of Cancer
Prevention : APJCP. 2012; 13(5):2329-2334.
Mantovani A., Allavena P., Sica A., & Balkwill F. Cancer-related inflammation. Nature. 2008;
454(7203):436-444.
Mapara M. Y., & Sykes M. Tolerance and cancer: Mechanisms of tumor evasion and strategies for
breaking tolerance. Journal of Clinical Oncology. 2004;
Mitra R., Lee J., Jo J., Milani M., McClintick J. N., Edenberg H. J., Kesler K. A., Rieger K. M., Badve
S., Cummings O. W., Mohiuddin A., Thomas D. G., Luo X., Juliar B. E., Li L., Mesaros C., Blair I. A.,
Srirangam A., Kratzke R. A., McDonald C. J., Kim J., & Potter D. A. Prediction of postoperative
recurrence-free survival in non-small cell lung cancer by using an internationally validated gene expression model. Clinical Cancer Research : An Official Journal of the American Association for
Cancer Research. 2011; 17(9):2934-2946.
Motaln H., & Turnsek T. L. Cytokines play a key role in communication between mesenchymal
stem cells and brain cancer cells. Protein and Peptide Letters. 2015; 22(4):322-331.
Olson B. M., & McNeel D. G. Antigen loss and tumor-mediated immunosuppression facilitate
tumor recurrence. Expert Review of Vaccines. 2012; 11(11):1315-1317.
Osada H., & Takahashi T. let-7 and miR-17-92: Small-sized major players in lung cancer
development. Cancer Science. 2011; 102(1):9-17.
Ostrand-Rosenberg S. Immune Surveillance: A Balance Between Pro- and Anti-tumor Immunity.
Curr Opin Genet Dev. 2009; 18(1):11-18.
Paranjape T., Slack F. J., & Weidhaas J. B. MicroRNAs: tools for cancer diagnostics. Gut. 2009;
58(11):1546-1554.
Peto R., Darby S., Deo H., Silcocks P., Whitley E., & Doll R. Smoking, smoking cessation, and lung
cancer in the UK since 1950: combination of national statistics with two case-control studies. BMJ
(Clinical Research Ed.). 2000; 321(7257):323-329.
Petruzzi M. N. M. R., Cherubini K., Salum F. G., & de Figueiredo M. A. Z. Role of tumour-associated
macrophages in oral squamous cells carcinoma progression: an update on current knowledge. Diagnostic Pathology. 2017; 12(1):32.
Pine S. R., Mechanic L. E., Enewold L., Chaturvedi A. K., Katki H. A., Zheng Y.-L., Bowman E. D.,
Engels E. A., Caporaso N. E., & Harris C. C. Increased Levels of Circulating Interleukin 6,
Interleukin 8, C-Reactive Protein, and Risk of Lung Cancer. JNCI: Journal of the National Cancer
Institute. 2011; 103(14):1112-1122.
Politi K., & Herbst R. S. Lung cancer in the era of precision medicine. Clinical Cancer Research : An
Official Journal of the American Association for Cancer Research. 2015; 21(10):2213-2220.
Pope III C. A., Burnett R. T., Thun M. J., Calle E. E., Krewski D., Ito K., & Thurston G. D. Lung
Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate Air Pollution.
JAMA. 2002; 287(9):1132.
Bartosz Szmyd, Marcin Kaszkowiak, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
90
Rafiemanesh H., Mehtarpour M., Khani F., Hesami S. M., Shamlou R., Towhidi F., Salehiniya H., Makhsosi
B. R., & Moini A. Epidemiology, incidence and mortality of lung cancer and their relationship with the
development index in the world. Journal of Thoracic Disease. 2016; 8(6):1094-1102.
Ransohoff D. F., Riddell R. H., & Levin B. Ulcerative colitis and colonic cancer. Diseases of the
Colon & Rectum. 1985; 28(6):383-388.
Resende de Paiva C., Grønhøj C., Feldt-Rasmussen U., & von Buchwald C. Association between
Hashimoto’s Thyroiditis and Thyroid Cancer in 64,628 Patients. Frontiers in Oncology. 2017; 7:53.
Santos S. C. D. dos, & Barbosa L. E. R. Crohn’s disease: risk factor for colorectal cancer. Journal of
Coloproctology. 2017; 37(1):55-62.
Schetter A. J., Heegaard N. H. H., & Harris C. C. Inflammation and cancer: Interweaving
microRNA, free radical, cytokine and p53 pathways. Carcinogenesis. 2009; 31(1):37-49.
Sepulveda A. R. Helicobacter, Inflammation, and Gastric Cancer. Current Pathobiology Reports.
2013; 1(1):9-18.
Sgadari C., Angiolillo A. L., & Tosato G. Inhibition of angiogenesis by interleukin-12 is mediated by
the interferon-inducible protein 10. Blood. 1996; 87(9):3877-3882.
Swann J. B., Smyth M. J., Dunn G. P., Bruce A. T., Ikeda H., Old L. J., Schreiber R. D., Dunn G. P., Old L. J.,
Schreiber R. D., Shankaran V., Smyth M. J., Street S. E., Trapani J. A., MacGregor D., Smyth M. J., Street S.
E., Clementi R., Zerafa N., Davidson W. F., Giese T., Fredrickson T. N., Airoldi I., Enzler T., Coussens L. M.,
Werb Z., Engel A. M., Svane I. M., Rygaard J., Werdelin O., Svane I. M., Girardi M., Smyth M. J., Nishikawa
H., Crowe N. Y., Smyth M. J., Godfrey D. I., Broek M. E. van den, Takeda K., Kaplan D. H., Dunn G. P.,
Smyth M. J., Qin Z., Kim H. J., Hemme J., Blankenstein T., Liu J., Xiang Z., Ma X., Girardi M., Oppenheim
D. E., Langowski J. L., Tartour E., Numasaki M., Norbury K. C., Kripke M. L., Ward P. L., Koeppen H. K.,
Hurteau T., Rowley D. A., Schreiber H., Mitra-Kaushik S., Harding J., Hess J., Schreiber R., Ratner L.,
Willimsky G., Blankenstein T., Ye X., McCarrick J., Jewett L., Knowles B. B., Buell J. F., Gross T. G.,
Woodle E. S., Penn I., Guidoboni M., Buckowitz A., Naito Y., Dolcetti R., Ishikawa T., Saeterdal I., Lu B. J.,
Lai M., Cheng L., Xu J. Y., Huang Q., Nakata B., Duval A., Clark W. H., Clemente C. G., Scanlan M. J.,
Simpson A. J., Old L. J., Haanen J. B., Dhodapkar M. V., Kyle R. A., Dhodapkar M. V., Krasovsky J., Osman
K., Geller M. D., Gultekin S. H., Graus F., Darnell R. B., DeAngelis L. M., Tanaka M., Tanaka K., Tokiguchi
S., Shinozawa K., Tsuji S., Albert M. L., Mathew R. M., Cohen A. B., Galetta S. L., Alavi A., Dalmau J.,
Ferradini L., Zorn E., Hercend T., Zorn E., Hercend T., Halliday G. M., Patel A., Hunt M. J., Tefany F. J.,
Barnetson R. S., Iihara K., Inoue T., Yoneda K., Manabe M., Demitsu T., Robinson B. W., Robinson C., Lake
R. A., Kerr K. M., Miura K., Han G., Sano M., Tsutsui Y., Davis T. A., Maloney D. G., Czerwinski D. K.,
Liles T. M., Levy R., Montagna D., Stewart T. H., Hollinshead A. C., Raman S., Callaway M. P., Briggs J. C.,
Demicheli R., Abbattista A., Miceli R., Valagussa P., Bonadonna G., Kauffman H. M., McBride M. A.,
Delmonico F. L., Kauffman H. M., Oijen M. van, Kyewski B., Klein L., Schell T. D., Knowles B. B.,
Tevethia S. S., Elsas A. van, Hurwitz A. A., Allison J. P., Park J. M., Otahal P., Schell T. D., Hutchinson S.
C., Knowles B. B., Tevethia S. S., Radoja S., Rao T. D., Hillman D., Frey A. B., Kurt R. A., Kurt R. A., Park
J. A., Schluter S. F., Marchalonis J. J., Akporiaye E. T., Blohm U., Zippelius A., Guilloux Y., Schwartz R. H.,
Dong H., Tripathi P., Agrawal S., Derre L., Campoli M., Chang C. C., Ferrone S., Terabe M., Berzofsky J. A.,
Zhang Q., Chen M. L., Gorelik L., Flavell R. A., Gabrilovich D. I., Ishida T., Nadaf S., Ohm J. E., Carbone D.
P., Kawamura K., Bahar R., Natsume W., Sakiyama S., Tagawa M., McKallip R., Li R., Ladisch S., Munn D.
H., Mellor A. L., Takahashi H., Wohlfart S., Shin M. S., Kataoka T., Hinz S., Serafini P., Borrello I., Bronte
V., Kusmartsev S., Gabrilovich D. I., Kusmartsev S., Mirza N., Melani C., Chiodoni C., Forni G., Colombo
M. P., Song X., Serafini P., Mazzoni A., Bronte V., Kusmartsev S., Nefedova Y., Yoder D., Gabrilovich D. I.,
Mantovani A., Sozzani S., Locati M., Allavena P., Sica A., Smyth M. J., Dunn G. P., Schreiber R. D., Bosma
G. C., Custer R. P., Bosma M. J., Hayashi T., Faustman D. L., Engel A. M., Street S. E., Cretney E., Smyth
M. J., Cretney E., Loser K., Johnsen A. K., Yue H. H., Diehl G. E., & Winoto A. Immune surveillance of
tumors. The Journal of Clinical Investigation. 2007; 117(5):1137-1146.
Szyszka-Barth K., Ramlau K., Goździk-Spychalska J., Spychalski L., Bryl M., Gołda-Gocka I., Kopczyńska
A., Barinow-Wojewódzki A., & Ramlau R. Actual status of therapeutic vaccination in non-small cell lung
cancer. Contemporary Oncology (Poznań, Poland). 2014; 18(2):77-84.
The inflammatory components in lung cancer
91
Topham N. J., & Hewitt E. W. Natural killer cell cytotoxicity: How do they pull the trigger?
Immunology. 2009; Wiley-Blackwell.
Tu T., Bühler S., & Bartenschlager R. Chronic viral hepatitis and its association with liver cancer.
Biological Chemistry. 2017;. De Gruyter.
Wang Z.-H., & Fang J.-Y. Colorectal Cancer in Inflammatory Bowel Disease: Epidemiology,
Pathogenesis and Surveillance. Gastrointestinal Tumors. 2014; 1(3):146-154.
Whitwell F., Scott J., & Grimshaw M. Relationship between occupations and asbestos-fibre content
of the lungs in patients with pleural mesothelioma, lung cancer, and other diseases. Thorax. 1977;
32(4):377-386.
Williams M., Cheng Y. Y., Blenkiron C., & Reid G. Exploring Mechanisms of MicroRNA
Downregulation in Cancer. MicroRNA. 2017; 6(1):2-16.
Wingo P. A., Ries L. A. G., Giovino G. A., Miller D. S., Rosenberg H. M., Shopland D. R., Thun M. J.,
& Edwards B. K. Annual Report to the Nation on the Status of Cancer, 1973-1996, With a Special
Section on Lung Cancer and Tobacco Smoking. JNCI Journal of the National Cancer Institute. 1999;
91(8):675-690.
Wroblewski L. E., Peek R. M., & Wilson K. T. Helicobacter pylori and gastric cancer: Factors that
modulate disease risk. Clinical Microbiology Reviews. 2010; American Society for Microbiology (ASM).
Xue X., Liu Y., Wang Y., Meng M., Wang K., Zang X., Zhao S., Sun X., Cui L., Pan L., & Liu S. MiR-
21 and MiR-155 promote non-small cell lung cancer progression by downregulating SOCS1, SOCS6, and PTEN. Oncotarget. 2016; 7(51):84508-84519.
Yang L., Wang L., & Zhang Y. Immunotherapy for lung cancer: advances and prospects. American
Journal of Clinical and Experimental Immunology. 2016; 5(1):1-20.
Yashiro M. Ulcerative colitis-associated colorectal cancer. World Journal of Gastroenterology. 2014;
20(44):16389-16397.
Zappa C., & Mousa S. A. Non-small cell lung cancer: current treatment and future advances.
Translational Lung Cancer Research. 2016; 5(3):288-300.
Zhang J.-M., & An J. Cytokines, inflammation, and pain. International Anesthesiology Clinics. 2007;
45(2):27-37.
Zheng M. Classification and Pathology of Lung Cancer. Surgical Oncology Clinics of North
America. 2016; 25(3):447-468.
Zhou B., Liu J., Wang Z.-M., & Xi T. C-reactive protein, interleukin 6 and lung cancer risk:
a meta-analysis. PloS One. 2012; 7(8):e43075.
Zhu Y. M., Webster S. J., Flower D., & Woll P. J. Interleukin-8/CXCL8 is a growth factor for
human lung cancer cells. British Journal of Cancer. 2004; 91(11):1970.
92
Normalisation of microRNA qPCR results
in lung cancer – strategies and problems
Marcin Kaszkowiak, Bartosz Szmyd, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-
Lasota
The corresponding author: [email protected],
Dorota Pastuszak-Lewandoska, PhD
Department of Biomedicine and Genetics Chair of Biology and Medical Parasitology,
Medical University of Lodz 251 Pomorska Street (C-5 building), 92-213 Lodz
Abstract
Lung cancer (LC) is one of the most common and death-causing carcinomas in the world. There is an
urgent need for innovative diagnostic tools and treatment strategies. MicroRNAs have a great potential
for being reliable LC biomarkers and therapy targets. qPCR is one of the best methods for evaluating
their expression during research. However, it has to be performed in a very careful way. Any committed
mistakes, especially during normalization of output data may lead to misleading conclusions. Data
Normalisation (DS) is a crucial step, required for levelling different sample sizes, volumes of nucleic
acids and other varying factors, allowing us to compare results with one another. There is no universal
strategy for this process. Every known method (volume size, housekeeping genes, global mean, etc.) has
many advantages and drawbacks. This article will discuss different approaches, their pros and cons, and
suggest ways to minimize qPCR inaccuracy.
Keywords: qPCR, microRNA, normalization, quantile, global mean
Background
Lung cancer incidence and mortality has been steadily growing since 1930s (Mao,
Yang, et al., 2016). Nowadays, it has become a global problem. LC was responsible
for 13% of all cancer incidences worldwide and caused 160340 deaths in United
States (Mao et al., 2016) in 2012, what made it the most common carcinoma and
a leading cause of cancer deaths.
The prognosis of Lung cancer strongly depends on the moment of diagnosis. The
5-year survival rate ranges from 52% (local disease) to 4% (distant disease). It results
in relatively high mortality, since only about 15% of LCs are diagnosed at early
stages. (Liam, Andarini, et al., 2015) It raises an urgent need for reliable biomarkers,
that would make diagnosing process faster and more accurate.
MicroRNAs are small strands of ribonucleic acid (built by about 23 nucleotides),
regulating genes expression at the transcriptomic level. They affect mRNA by either
repressing translation or inducing degradation (Wahid, Shehzad, et al., 2010). Many
recent studies suggest their significant role in cancer etiopathogenesis (Leva,
Garofalo, et al., 2014) and therefore consider them as promising biomarkers and
therapeutic targets (Berindan-neagoe, Monroig, et al., 2015; Kong, Zhang, et al.,
2015; Yang, Tang, et al., 2014; Zhao, Li, et al., 2015). That makes an accurate
evaluation of microRNAs expression a crucial step for developing new drugs and
diagnostic procedures.
Normalisation of microRNA qPCR results in lung cancer – strategies and problems
93
Technical introduction
Real-time PCR is a commonly used and reliable technique for measuring microRNA
relative expression. Among its advantages are sensitivity, high throughput and
accuracy. However, as every method it is prone to errors. There are two types of
variation in the obtained experimental data: biological (describing real molecular
processes) and experimentally-induced (due to experiment handling), that needs to be
excluded from further analysis. It is the purpose of data normalization. Reducing the
influence of material’s quantity, integrity, purity, etc. makes it possible to compare
the results from different experiments (Vandesompele, 2012).
Classic Normalization methods
Normalization to sample size
The simplest, but also the least used method is to carefully obtain the same sample
volume (or mass, number of cells, etc.) form each patient. This approach may seem
obvious and straightforward, but is extremely prone to errors. It is impossible to
measure sample size precisely enough. Moreover, biological material often contains
a different percentage of microRNA, what makes accurate analysis impossible
(Huggett, Dheda, et al., 2005).
Normalization to total sample RNA
Formula 1. Normalization to total sample RNA. δ’ – expression after normalization of j-th microRNA in i-th
sample; δ - expression of j-th microRNA in i-th sample; r – total RNA in i-th sample
Another volume-orientated technique bases on quantification of all RNAs present in
a sample. This measurement can be easily performed prior to reverse transcription.
Unfortunately, it can be affected by poor material quality. Moreover, this method has
two serious disadvantages. Firstly, it is unable to normalize variation formed in
further steps of qPCR method. Secondly, it bases on the assumption that proportion if
different RNA types (mRNA, rRNA, miRNA, etc.) is constant within all cells, which
has been proven wrong (Huggett et al., 2005).
Normalization to artificial molecule
Formula 2. Normalization to exogenous RNA. δ’ – expression after normalization of j-th microRNA in i-th
sample; δ - expression of j-th microRNA in i-th sample; γ – exteral molecule’s expression in i-th sample.
This method uses an external molecule as a reference microRNA. Exogenous RNA
(usually C. elegans cel-miR-39) of known quantity is added to a sample before
performing Reverse Transcription. On the contrary to previously described
technique, this method allows to reduce variation caused by further qPCR steps, but
has no use in normalization of differences in samples’ volume and quality
Marcin Kaszkowiak, Bartosz Szmyd, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
94
(Schwarzenbach, Calin, et al., 2016). Among its advantages there is also insensitivity
to in-cell biological fluctuations. However, this may be considered as a drawback,
since it raises a need for result validation (Huggett et al., 2005). Generation of
suitable, not always commercially available external particles may also be a serious
problem in smaller laboratories (Huggett et al., 2005).
Normalization to endogenous reference microRNAs
Formula 3. Normalization to “housekeeping microRNA”. δ’ – expression after normalization of j-th
microRNA in i-th sample; δ - expression of j-th microRNA in i-th sample; θ – housekeeping microRNA
expression in i-th sample
Endogenous control is the most popular and most universal tool for a qPCR data
normalization (Schwarzenbach, Da Silva, et al., 2015). Similarly to gene expression
analysis, this method is based on the assumption, that some microRNAs’ expression
(the so called “housekeeping microRNAs”) is close to constant in all samples and
experimental conditions (Mar, Kimura, et al., 2009). With this hypothesis, other
expressions are adjusted using given formula. It allows to reduce variation caused by
all stages of an experiment. Unfortunately, this technique has numerous limitations.
There are no ‘100% universal’ reference molecules. Many studies report that, on
contrary to our assumption, some housekeeping microRNAs / genes expression can
significantly vary depending on tissue type and experimental conditions (Dheda,
Huggett, et al., 2004; Schmittgen & Zakrajsek, 2000; Thellin, Zorzi, et al., 1999;
Tricarico, Pinzani, et al., 2002; Ullmannová & Haskovec, 2003; Xiang, Zeng, et al.,
2014). This makes proper selection of endogenous control a very complex problem
and should be solved for each project individually in preliminary research. This step
is extremely important, since it is necessary for correct interpretation of results
(Hellemans & Vandesompele, 2014; Saviozzi, Cordero, et al., 2006).
Many researchers suggest usage of more sophisticated version of this method
– considering combination (geometric mean) of a few housekeeping microRNAs
expression as stable and reliable control (Mar et al., 2009). Candidates for mean
arguments can be selected using bioinformatics methods.
Formula 4. Normalization to multiple housekeeping microRNAs. δ’ – expression after normalization of j-th
microRNA in i-th sample; δ – expression of j-th microRNA in i-th sample; θ – k-th housekeeping microRNA
expression in i-th sample, n – number of housekeeping microRNAs
The most commonly used molecules in endogenous microRNA expression normali-
zation are small coding RNAs – RNU6A and RNU6B. Despite being considered as
universal housekeeping genes (Schwarzenbach et al., 2016), many studies report
difficulties in their application – e.g., RNU6B is not applicable in circulating
microRNA analysis (Xiang et al., 2014). RNUs are not microRNAs, which
Normalisation of microRNA qPCR results in lung cancer – strategies and problems
95
introduces bias due to their different expression profiles and biochemical character.
Usage of microRNAs and their combinations as endogenous control is advised
(Schwarzenbach et al., 2016).
Among different miRNAs studied, miR-16 is highly and invariantly expressed in
different tissues. Bioinformatic analysis has proven its high stability, especially in
combination with miR-93. They are followed by miR-191, miR-106a, miR-17-5p
and miR-25 (Schwarzenbach et al., 2016).
Data-driven Normalization methods
Normalization to global mean
Formula 5. Normalization to global mean. δ’ – expression after normalization of j-th microRNA in i-th
sample; δ - expression of j-th microRNA in i-th sample; φ – k-th microRNA expression in i-th sample,
n – number of microRNAs
Gold-standard workflow in qPCR data analysis consists of a pilot experiment in
search for molecules with stable expression, selecting a few of them and using their
expression combination for normalizing the results in further research. This approach
often presents many problems and difficulties. Suitable endogenous controls are
often either very hard, or even impossible to find (Pieter Mestdagh, Van Vlierberghe,
et al., 2009).
In response to those difficulties, relatively new method has been developed
– global mean normalization. It is based on the observation that in a large and
unbiased group of microRNAs, the average expression can be used for normalization
that would reduce variation from all steps of the experiment.(Vandesompele, 2012).
Studies suggest, that this method either outperforms or is as good as endogenous
control in terms of stability. It also allows us to observe real and significant
biological changes, that could be blurred when using other normalization techniques
(P Mestdagh, Van Vlierberghe, et al., 2009).
For usage in projects with analysis of smaller microRNA sets (i.e. 4 microRNAs),
with help of bioinformatic tools, a few microRNAs or small RNAs controls can be
selected that reassemble the mean expression value (P Mestdagh et al., 2009).
Quantile Normalization algorithm
It is simple, easy-applicable method that bases on simple statistical parameter
– quantile. This technique is mainly used in normalization of microarray data
(Hellemans & Vandesompele, 2014), but there is no obstacle for usage in qPCR
results. It bases on an assumption, that average distribution of gene expression levels
in the cell remains constant (Mar et al., 2009).
The algorithm treats sample as n data points (number of microRNAs), sorts them in
ascending/descending order and transforms i-th row in all samples with its mean
Marcin Kaszkowiak, Bartosz Szmyd, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
96
value. Those steps change expression levels while preserving rank-order (it is an
origin of method’s name – quantiles remain constant for each sample) (Mar et al.,
2009). The detailed step-by-step description is presented on Figure 1.
Create a matrix M with k rows (number of microRNAs in the study) and p
columns (number of samples in a study). Fill this array with your experimental
data (microRNA name and its expression).
Create a matrix M’ by sorting each column in ascending order. It corresponds to a
quantile distribution of each sample.
Iterate over rows and replace each column with the row’s mean value. After this
process normalized microRNAs expressions can be rearranged in the starting
order.
Figure 1. Quantile normalization
There may occur situations, where all investigated microRNAs for each sample
won’t fit in one plate. Technical variation occurring between those plates can (and
should) be reduced using same algorithm, considering plates as different samples.
However, it is very important that examined microRNAs should be distributed
randomly across plates. Afterwards, normal procedure may be performed (Mar et al.,
2009).
Rank-Invariant Set Normalization algorithm
In “housekeeping” classic expression normalization method, there has to be taken an
a priori hypothesis, that one (or more) microRNA /gene expression is constant across
all samples. This approach was often proven wrong (Dheda et al., 2004; Schmittgen
& Zakrajsek, 2000; Thellin et al., 1999; Tricarico et al., 2002; Ullmannová &
Haskovec, 2003; Xiang et al., 2014). In a dataset large enough, suitable microRNAs
can be selected from the data itself, after the experiment (Mar et al., 2009).
Bioinformatic analysis allows us to find rank-invariant (having the same rank across
all datasets) microRNAs across all samples, which are verified to be suitable for
normalization purpose (Tseng, 2001). The detailed algorithm description is presented
on Figure 2.
Normalisation of microRNA qPCR results in lung cancer – strategies and problems
97
Create a matrix M with k rows (number of microRNAs in the study) and p
columns (number of samples in a study). Fill this array with your experimental
data (microRNA name and its expression).
Create a matrix M’ by sorting each column in ascending order.
Select one microRNA set as a reference, regarding experiment’s purpose and
methodology. This can be one special patient, commercially available set, global
mean or average, etc.
Compare each of p samples with R and find microRNAs that have the same rank
in both datasets. Create a set of genes S, by intersecting the results of all pairwise
comparisons.
Let αi be the average expression of rank invariant microRNAs (stored
in S) in i-th sample. Calculate
ratio.
Multiplicate i-th column of M by βi . Dataset contains normalized
microRNA expression values.
Figure 2. Rank-Invariant Set Normalization
Marcin Kaszkowiak, Bartosz Szmyd, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
98
Conclusions
There is no universal, perfect normalization method. All techniques have many
advantages and also drawbacks. While Data-Driven methods are more accurate,
cheaper, universal, and easy to use in large, genom-wide experiments, they may be
inaccurate, or even not-applicable in smaller projects. Classic methods, especially
“Housekeeping microRNA normalization” are considered gold-standard for
normalizing qPCR results. Among its advantages are low-price and easy usage even
in single-gene research. They are, however, extremely prone to errors and may cause
a serious misinterpretation of expression data.
The perfect normalization method has yet to be discovered. Nowadays, suitable
technique should be chosen individually for each experiment, after cautious pros and
cons consideration, literature study and public databases analysis. Careful selection is
crucial for research to succeed.
Table 1. List of key features of each normalization method. Shading: Classic Normalization Methods, Data-
Driven Normalization Methods
normalization method + -
sample size Cheap
Straightforward
Extremely prone to errors
Impractical
total sample RNA Easy to apply
Reduces Reverse Transcription
Bias
Prone to mechanical errors
Based on wrong assumptions
artificial molecule Insensitive to biological
fluctuations
Precise
Expensive
Does not reduce variation created
before Reverse Transcription step
housekeeping microRNAs Reduces variation from all steps
of experiment
Popular
Easily applicable and replicable
Prone to errors
Not universal
Based on wrong assumptions
global mean Reduces variation form all steps
of experiment
Universal
Easy to use
Accurate
Based on reasonable assumptions
Accuracy dependent on study size
quantile normalization Reduces variation from all steps
of the experiment
Accurate
Universal
Straightforward
Accuracy dependent on study size
Requires random microRNA
distribution in a dataset
Normalisation of microRNA qPCR results in lung cancer – strategies and problems
99
Based on reasonable assumptions
Normalizes cross-plate variation
Rank-invariant set Combines advantages of
housekeeping microRNAs
method and data-driven
techniques
Applicable post-hoc
Allows to identify suitable
endogenous control
Accuracy dependent on study size
References
Berindan-neagoe I., Monroig P., Pasculli B., George A., Medicine T., Hatieganu P. I., Juan S., Rico P., &
Sciences P. MicroRNAome genome: a treasure for cancer diagnosis and therapy2015, 64(5):311-336.
Dheda K., Huggett J. F., Bustin S. A., Johnson M. A., Rook G., & Zumla A. Validation of
housekeeping genes for normalizing RNA expression in real-time PCR. BioTechniques 2004,
37(1):112-119.
Hellemans J., & Vandesompele J. Selection of Reliable Reference Genes for RT-qPCR Analysis. In
R. Biassoni & A. Raso (Eds.) Quantitative Real-Time PCR: Methods and Protocols 2014 (pp. 19-26).
New York, NY: Springer New York.
Huggett J., Dheda K., Bustin S., & Zumla A. Real-time RT-PCR normalisation; strategies and
considerations. Genes and Immunity 2005, 6(4):279-284.
Kong X.-M., Zhang G.-H., Huo Y.-K., Zhao X.-H., Cao D.-W., Guo S.-F., Li A.-M., & Zhang X.-R.
MicroRNA-140-3p inhibits proliferation, migration and invasion of lung cancer cells by targeting ATP6AP2. International Journal of Clinical and Experimental Pathology 2015, 8(10):12845-12852.
Leva G. Di, Garofalo M., & Croce C. M. microRNAs in cancer. Annu Rev Pathol 2014, 9:287-314.
Liam C.-K., Andarini S., Lee P., Ho J. C.-M., Chau N. Q., & Tscheikuna J. Lung cancer staging now
and in the future. Respirology 2015, 20(4):526-534.
Mao Y., Yang D., He J., & Krasna M. J. Epidemiology of Lung Cancer. Surgical Oncology Clinics of
North America 2016, 25(3):439-445.
Mar J. C., Kimura Y., Schroder K., Irvine K. M., Hayashizaki Y., Suzuki H., Hume D.,
& Quackenbush J. Data-driven normalization strategies for high-throughput quantitative RT-
PCR. BMC Bioinformatics 2009, 10.
Mestdagh P., Van Vlierberghe P., De Weer A., Muth D., Westermann F., Speleman F.,
& Vandesompele J. A novel and universal method for microRNA RT-qPCR data normalization.
Genome Biology 2009, 10(6).
Mestdagh P., Van Vlierberghe P., De Weer A., Muth D., Westermann F., Speleman F.,
& Vandesompele J. A novel and universal method for microRNA RT-qPCR data normalization.
Genome Biol 2009, 10(6):R64.
Saviozzi S., Cordero F., Lo Iacono M., Novello S., Scagliotti G. V., & Calogero A. R. Selection of
suitable reference genes for accurate normalization of gene expression profile studies in non-small
cell lung cancer. BMC Cancer 2006, 6:1-10.
Schmittgen T. D., & Zakrajsek B. A. Effect of experimental treatment on housekeeping gene
expression: validation by real-time, quantitative RT-PCR. Journal of Biochemical and Biophysical
Methods 2000, 46(1):69-81.
Marcin Kaszkowiak, Bartosz Szmyd, Dorota Pastuszak-Lewandoska, Ewa Brzeziańska-Lasota
100
Schwarzenbach H., Calin G., & Pantel K. Which is the accurate data normalization strategy for
microRNA quantification?2016, 61(11):1333-1342.
Schwarzenbach H., Da Silva A. M., Calin G., & Pantel K. Data normalization strategies for
microRNA quantification. Clinical Chemistry 2015, 61(11):1333-1342.
Thellin O., Zorzi W., Lakaye B., De Borman B., Coumans B., Hennen G., Grisar T., Igout A., & Heinen
E. Housekeeping genes as internal standards: Use and limits. Journal of Biotechnology 1999, 75(2-
3):291-295.
Tricarico C., Pinzani P., Bianchi S., Paglierani M., Distante V., Pazzagli M., Bustin S. A.,
& Orlando C. Quantitative real-time reverse transcription polymerase chain reaction:
normalization to rRNA or single housekeeping genes is inappropriate for human tissue biopsies.
Analytical Biochemistry 2002, 309(2):293-300.
Tseng G. C. Issues in cDNA microarray analysis: quality filtering, channel normalization, models
of variations and assessment of gene effects. Nucleic Acids Research 2001, 29(12):2549-2557.
Ullmannová V., & Haskovec C. The use of housekeeping genes (HKG) as an internal control for the
detection of gene expression by quantitative real-time RT-PCR. Folia Biologica 2003, 49:211-216.
Vandesompele J. From reference genes to global mean normalization Critical elements
contributing to successful qPCR results2012.
Wahid F., Shehzad A., Khan T., & Kim Y. Y. MicroRNAs: Synthesis, mechanism, function, and recent
clinical trials. Biochimica et Biophysica Acta - Molecular Cell Research 2010, 1803(11):1231-1243.
Xiang M., Zeng Y., Yang R., Xu H., Chen Z., Zhong J., Xie H., Xu Y., & Zeng X. U6 is not a suitable
endogenous control for the quantification of circulating microRNAs. Biochemical and Biophysical
Research Communications 2014, 454(1):210-214.
Yang H., Tang Y., Guo W., Du Y., Wang Y., Li P., Zang W., Yin X., Wang H., Chu H., Zhang G., &
Zhao G. Up-regulation of microRNA-138 induce radiosensitization in lung cancer cells. Tumor
Biology 2014, 35(7):6557-6565.
Zhao Q., Li P., & Yu J. M. and X. MicroRNAs in Lung Cancer and Lung Cancer Bone Metastases:
Biomarkers for Early Diagnosis and Targets for Treatment. Recent Patents on Anti-Cancer Drug
Discovery 2015.
101
Statistical comparison of deleteriousness prediction
methods for single nucleotide variants from next
generation sequencing
Dominik Wawrzuta, Students Research Association for Genetics of Medical University
of Warsaw
Contact author: Dominik Wawrzuta, Students Research Association for Genetics of
Medical University of Warsaw, Żwirki i Wigury 61, 00-001 Warszawa
Abstract
In the analysis of data from whole exome sequencing, algorithms are commonly used to calculate the
deleteriousness of single nucleotide variants. The commonly used deleteriousness prediction methods
are SIFT, Polyphen2, LRT, MutationTaster, MutationAssessor, FATHMM, MetaSVM and MetaLR.
The aim of this study is a statistical analysis of the results of deleteriousness obtained by algorithms. In
the analysis whole exome sequencing data are used, the deleteriousness of the variants in exomes is
assessed by the examined methods. Then deleteriousness is categorized as 1 for variants marked as
harmful, 0 for unmarked and -1 for tolerable ones. Analysis of interrelations between the tested
algorithms is made using Spearman correlation, as well as k-means method and Multiple
Correspondence Analysis. The applied statistical methods indicate the occurrence of high correlation
and homogeneity between the prediction results made by the MetaSVM and MetaLR algorithms, which
similarly indicate the deleteriousness of the same variants. Different results were obtained for the SIFT,
Polyphen2, LRT, MutationTaster, MutationAssessor and FATHMM algorithms. This observation
suggests that the results obtained using these methods should not be analyzed at the same time, as they
multiply the information carried.
Introduction
A non-synonymous single nuclear polymorphism (nsSNP) is a variation at a single
position in a DNA sequence, which cause change in encoded amino acid and in
corresponding protein product. Some of the nsSNPs may affect the phenotype and be
associated with disorders. Variation is called nsSNP if less than 1% of a population
carry nucleotide at a specific position in the DNA sequence (Butler 2012). Due to
potential pathogenicity of nsSNPs, there is a significant need to characterize the
effect of nsSNPs on the corresponding protein function. In human genetics, it is
important to determine which nsSNPs negatively affect the protein product and
consequently the phenotype. In the analysis of data from whole exome sequencing,
algorithms are commonly used to calculate the deleteriousness of nsSNPs. The most
used deleteriousness prediction methods are Sorting Tolerant From Intolerant (SIFT),
PolyPhen-2, Likelihood Ratio Test (LRT), MutationTaster, MutationAssessor,
Functional Analysis Through Hidden Markov Models (FATHMM), MetaSVM and
MetaLR.
SIFT (Ng, Henikoff 2003) assumes that important amino acids in protein do not
change during evolution, otherwise the substitution affects protein function. SIFT
uses PSI-BLAST algorithm (Altschul, Madden et al. 1997) to searching protein
database for functionally related proteins. Then creates a dataset of found proteins
and aligns them with query sequence. In the next step the algorithm calculates
Dominik Wawrzuta
102
conservation value and scale for each position the probability that the variant is
tolerated. (Kumar, Henikoff et al. 2009)
PolyPhen2 (Adzhubei, Schmidt et al. 2010) uses 11 features for predicting damaging
effects. Eight of them are sequence-based and three are structure-based. The most of
these features compare a properties of wild allele with a properties of analyzed allele.
The deleteriousness of the nsSNP is predicted from this individual features by Naive
Bayes classifier approach coupled with entropy-based discretization (Fayyad, Irani
1993).
LRT treats rare variants as random effects in linear mixed effects models. Likelihood
ratio test compare the goodness of fit of the models and provide probability of
deleteriousness of variant. (Zeng, Zhao et al. 2014) (Qian, Shao 2013)
MutationTaster predicts not only functional effects of non-synonymous variants but
also intronic and synonymous mutations, short insertion and deletion mutations.
Homozygous alterations found more than four times in HapMap Project or in 1000
Genomes Project are automatically predicted as neutral. Mutations described as
pathogenic in ClinVar are predicted as deleterious. (Schwarz, Cooper et al. 2014)
(Schwarz, Rödelsperger et al. 2010)
MutationAssessor calculates functional impact score (FIS) for amino acids changes.
FIS is based on evolutionary conservation patterns. To evaluate FIS score multiple
homologous sequences of analyzed sequence are clustered and aligned, then
difference of the entropy caused by mutation is estimated. (Reva, Antipin et al. 2011)
FATHMM uses Hidden Markov Models (HMM) to estimate deleteriousness of
nsSNPs. Probabilities of substituted amino acids are calculated by HMM, reduction
of probability comparing to wild type indicates negative impact upon protein
function. (Shihab, Gough et al. 2013)
MetaSVM and MetaLR use Support Vector Machine (SVM) and Logistic Regression
(LR) respectively. They are based on nine scoring methods (SIFT, PolyPhen-2
HDIV, PolyPhen-2 HVAR, GERP++, MutationTaster, Mutation Assessor,
FATHMM, LRT, SiPhy, PhyloP) and the maximum minor allele frequency (MMAF)
observed in the 1000 Genomes project population. (Dong, Wei et al. 2015)
Despite many analyzes comparing sensitivity and specificity of these algorithms
(Gnad, Baucom et al. 2013) (Wu, Kuang et al. 2016), it is not clear which method is
best to use and provides best results.
The purpose of this paper is to compare a results established by methods SIFT,
PolyPhen-2, LRT, MutationTaster, MutationAssessor, FATHMM, MetaSVM and
MetaLR using a statistical clustering techniques.
Methods
In the analysis I use dataset named dbnsfp33a containing all known variants, which
deleteriousness is assessed by SIFT, PolyPhen-2, LRT, MutationTaster,
MutationAssessor, FATHMM, MetaSVM and MetaLR simultanously. The dataset is
provided by software ANNOVAR (Wang, Li et al. 2010). Then deleteriousness is
Statistical comparison of deleteriousness prediction methods for single nucleotide variants
from next generation sequencing
103
categorized as 1 for variants marked as harmful, 0 for unmarked and -1 for tolerable
ones. As a result, for each analyzed method is created vector filled with values -1, 0,
1. Analysis of interrelations between the tested algorithms is made using Spearman
correlation, as well as k-means clustering and Multiple Correspondence Analysis. To
make calculations are used R packages factoextra (Kassambara, Mundt 2017) and
cluster (Maechler, Rousseeuw et al. 2017).
Spearman correlation The Spearman correlation calculates correlation between
ranks of vectors x and y.
Where
k-means clustering K-means clustering (MacQueen 1967) (Hartigan, Wong 1979) is
unsupervised machine learning algorithm which divides data into k groups.
Parameter k is determined by the analyst. For each cluster the centroid, which is the
mean of objects within cluster, is calculated. Finally a variation of objects within the
same cluster is as low as possible and it is calculated as a sum of squared Euclidean
distances between objects and centroid.
where
is number of clusters is object belonging to cluster is centroid of cluster
Multiple Correspondence Analysis Multiple correspondence analysis (MCA)
(Michailidis, de Leeuw 1998) is a technique for detecting and representing
underlying structures in nominal categorical data. Consider data of n objects on J =
{1, 2, …, J} categorical variables, where every category j J can take mj possible
values. Let X be a matrix containing the coordinates of the objects. Let Yj denote
matrix which contains the coordinates of every category. The data for every category
is encoded in matrix Gj , which is an n by mj matrix of category j where entry is 1 if
object belongs to category value and 0 otherwise. The scores are then obtained by
minimizing the loss-function which simultaneously minimizes for every category the
distances between the object score and the scores of that object on that category for
all objects.
The loss function is given by:
Dominik Wawrzuta
104
Results
Table 1 Results of Spearman correlation
Poly
Phen-2
Mutation
Taster
Mutation
Assessor
LRT FATHMM SIFT MetaLR MetaSVM
PolyPhen-2 1 -0.046 -0.085 -0.317 -0.471 0.200 -0.384 -0.327
Mutation
Taster
-0.046 1 0.278 -0.211 -0.404 -0.201 -0.430 -0.391
Mutation
Assessor
-0.085 0.278 1 -0.142 -0.334 -0.175 -0.412 -0.379
LRT -0.317 -0.211 -0.142 1 0.268 -0.239 0.032 -0.001
FATHMM -0.471 -0.404 -0.334 0.268 1 -0.353 0.610 0.498
SIFT 0.200 -0.201 -0.175 -0.239 -0.353 1 -0.334 -0.286
MetaLR -0.384 -0.430 -0.412 0.032 0.610 -0.334 1 0.820
MetaSVM -0.327 -0.391 -0.379 -0.001 0.498 -0.286 0.820 1
Table 1 shows the results of Spearman correlation between algorithms. The highest correlation occurs between the MetaSVM and MetaLR methods. The FATHMM algorithm also correlates with these two methods. We can distinguish two other groups, however weakly correlated, MutationTaster with MutationAssessor and SIFT with PolyPhen2.
Figure 1 graphs the results of the second applied method, which is k-means clustering. There is a clear division into three groups. One of them consists of the methods MetaSVM, MetaLR and FATHMM, which were also shown as highly correlated. In the second group the methods MutationTaster, MutationAssessor, SIFT and PolyPhen2 are clustered. The LRT algorithm creates a separate group. It should be noted that the variance explained by the axes is 25.9 and 39.2 percent.
The last method is Multiple Correspondence Analysis. The results are shown in the Figure 2. For every analyzed algorithm three binary vectors were created according to the convention “Name_Type” where Name is the name of an algorithm and Type is the letter denoting d for deleterious, u for unknown and t for tolerable values. The two directions explain almost 50% of the variation of the data. With respect to the first axis a definite split is found. We can notice division into two groups. The first group includes the MetaSVM and MetaLR algorithms. What is important, they are related in each of the statistical methods under study. The algorithms PolyPhen-2, LRT, FATHMM, SIFT, MutationAssessor and MutationTaster are in the second group. The cause of the division may be the type of data used by the algorithms. MetaSVM and MetaLR use results of other methods, while the other algorithms are based on molecular data about the tested proteins.
Statistical comparison of deleteriousness prediction methods for single nucleotide variants
from next generation sequencing
105
Figure 1 Results of k-means clustering
Figure 2 Results of Multiple Correspondence Analysis
Dominik Wawrzuta
106
Conclusions
The analyzed algorithms use various methods and data to assess the pathogenicity of
variants, but the multivariate techniques applied in the paper do not reveal
differences in the results of their calculations. The only distinctive group, shown by
clustering methods, are the MetaSVM and MetaLR methods, which differ from the
others in the way that they calculate the results. They use the results of
deleteriousness generated by other algorithms and do not calculate them on the basis
of biological parameters alone.
The results show that it is recommended to be very careful using the PolyPhen2,
SIFT, LRT, MutationAssessor, MutationTaster, FATHMM methods simultaneously
during genetic investigations. The multivariate analysis based on k-means method,
Multiple Correspondence Analysis and Spearman correlation reveals similar patterns
between different methods used to do pathogenicity assessment. It confirms that
applying multiple methods, which are based on similar assumptions implies little
gain in information and it should not be used to verify robustness of results of genetic
investigations.
Discussion
The evaluation of the impact of point mutations on protein activity is extremely
important during genetic research (Best, Wou et al. 2018) (Yang, Chen et al. 2017).
The creation of tools that would be able to make such an assessment with high
sensitivity and specificity would revolutionize the speed and cost of genetic testing.
Despite the creation of many deleteriousness assessment tools, there are no
significant differences in the results of their calculations. The MetaSVM and MetaLR
algorithms are the exception, which differ from other methods both in terms of
operation and calculated results. (Liu, Wu et al. 2016) showed also, that MetaSVM
and MetaLR have higher sensitivity and specificity than other methods. It can be
noticed that creating solutions based on existing algorithms may result in
significantly better methods than already existing.
An important source of data that can help to increase the efficiency of
deleteriousness assessment methods are the ClinVar (Landrum, Lee et al. 2018) and
HGMD (Stenson et al. 2009) databases. They contain information on pathogenic
mutations that have been confirmed by scientific research. At the moment
MutationTaster algorithm uses ClinVar data to calculate predictions of
deleteriousness.
Other data sources that carry important information about SNPs, and are rarely used
in deleteriousness assessment algorithms, are the ExAC, gnomAD (Lek, Karczewski
et al. 2016) and 1000 genomes (The 1000 Genomes Project Consortium 2015)
databases. They contain information on the incidence of SNPs in the population,
which is important information in the analysis, because the frequent occurrence of
the SNP in population probably does not negatively affects the phenotype.
In the future, the development of machine learning methods, especially based on
deep neural networks (LeCun, Bengio et al. 2015) may increase the accuracy of the
Statistical comparison of deleteriousness prediction methods for single nucleotide variants
from next generation sequencing
107
estimations. One of the first such tools is DeepVariant, which is created by the
Genomics team in Google Brain. Its results were positively evaluated by the 2016
PrecisionFDA Truth Challenge.
References
Adzhubei I. A., Schmidt S., Peshkin L., Ramensky V. E., Gerasimova A., Bork P., Kondrashov A.
S., Sunyaev S. R. A Method and Server for Predicting Damaging Missense Mutations. Nature
Methods. 2010; 7(4): 248-49.
Altschul S. F., Madden T. L., Schäffer A. A., Zhang J., Zhang Z., Miller W., Lipman D. J. Gapped
BLAST and PSI-BLAST: A New Generation of Protein Database Search Programs. Nucleic Acids
Research. 1997; 25(17): 3389-3402.
Best S., Wou K., Vora N., Van der Veyver I. B., Wapner R., Chitty L. S. Promises, Pitfalls and
Practicalities of Prenatal Whole Exome Sequencing. Prenatal Diagnosis. 2018; 38(1): 10-19.
Butler J. Single Nucleotide Polymorphisms and Applications. In: Advanced Topics in Forensic
DNA Typing: Methodology. 2012; Academic Press (San Diego). ISBN: 9780123745132.
Dong C., Wei P., Jian X., Gibbs R., Boerwinkle E., Wang K., Liu X. Comparison and Integration of
Deleteriousness Prediction Methods for Nonsynonymous SNVs in Whole Exome Sequencing Studies. Human Molecular Genetics. 2015; 24(8): 2125-37.
Fayyad U., Irani K. Multi-Interval Discretization of Continuous-Valued Attributes for
Classification Learning. In: Pollack M. (Editor) Proceedings of the 13th International Joint
Conference on Artificial Intelligence. 1993; Morgan Kaufmann Publisher; ISBN:
1-55860-300-X.
Gnad F., Baucom A., Mukhyala K., Manning G., Zhang Z. Assessment of Computational Methods for
Predicting the Effects of Missense Mutations in Human Cancers. BMC Genomics. 2013; 14(3): S7.
Hartigan J. A., Wong M. A. Algorithm AS 136: A K-Means Clustering Algorithm. Journal of the
Royal Statistical Society. Series C (Applied Statistics). 1979; 28(1): 100-108.
Kassambara A., Mundt F. Factoextra: Extract and Visualize the Results of Multivariate Data
Analyses (version 1.0.5). http://cran.us.r-project.org/web/packages/factoextra/index.html. 2017;
Kumar P., Henikoff S., Ng. P. C. Predicting the Effects of Coding Non-Synonymous Variants on
Protein Function Using the SIFT Algorithm. Nature Protocols. 2009; 4(7): 1073-81.
Landrum M. J., Lee J. M., Benson M., Brown G. R., Chao C., Chitipiralla S., Gu B., Hart J., Hoffman
D., Jang W., Karapetyan K., Katz K., Liu C., Maddipatla Z., Malheiro A., McDaniel K., Ovetsky
M., Riley G., Zhou G., Holmes J. B., Kattman B. L., Maglott D. R. ClinVar: Improving Access to
Variant Interpretations and Supporting Evidence. Nucleic Acids Research. 2018; 46: D1062-67.
LeCun Y., Bengio Y., Hinton G. Deep Learning. Nature. 2015; 521(7553): 436-44.
Lek M., Karczewski K. J., Minikel E. V., Samocha K. E., Banks E., Fennell T., O'Donnell-Luria A.
H., Ware J. S., Hill A. J., Cummings B. B., Tukiainen T., Birnbaum D. P., Kosmicki J. A., Duncan L.
E., Estrada K., Zhao F., Zou J., Pierce-Hoffman E., Berghout J., Cooper D. N., Deflaux N., DePristo
M., Do R., Flannick J., Fromer M., Gauthier L., Goldstein J., Gupta N., Howrigan D., Kiezun A., Kurki
M. I., Moonshine A. L., Natarajan P., Orozco L., Peloso G. M., Poplin R., Rivas M. A., Ruano-Rubio
V., Rose S. A., Ruderfer D. M., Shakir K., Stenson P. D., Stevens C., Thomas B. P., Tiao G., Tusie-
Luna M. T., Weisburd B., Won H. H., Yu D., Altshuler D. M., Ardissino D., Boehnke M., Danesh
J., Donnelly S., Elosua R., Florez J. C., Gabriel S. B., Getz G., Glatt S. J., Hultman C. M., Kathiresan
S., Laakso M., McCarroll S., McCarthy M. I., McGovern D., McPherson R., Neale B.M., Palotie
A., Purcell S.M., Saleheen D., Scharf J.M., Sklar P., Sullivan P. F., Tuomilehto J., Tsuang M.
T., Watkins H. C., Wilson J. G., Daly M. J., MacArthur D. G. Analysis of Protein-Coding Genetic
Variation in 60,706 Humans. Nature. 2016; 536(7616): 285-91.
Dominik Wawrzuta
108
Liu X., Wu C., Li C., Boerwinkle E. DbNSFP v3.0: A One-Stop Database of Functional Predictions
and Annotations for Human Non-Synonymous and Splice Site SNVs. Human Mutation. 2016;
37(3): 235-41.
MacQueen J. Some Methods for Classification and Analysis of Multivariate Observations. In: Le
Cam L. M., Neyman J. (Editors) Proceedings of the Fifth Berkeley Symposium on Mathematical
Statistics and Probability Volume 1: Statistics. 1967; University of California Press (Berkeley).
Maechler M., Rousseeuw P., Struyf A., Hubert M., Hornik K., Studer M., Roudier P., Gonzalez J.
Cluster: ‘Finding Groups in Data’: Cluster Analysis Extended Rousseeuw et Al. (version 2.0.6). http://cran.us.r-project.org/web/packages/cluster/index.html. 2017;
Michailidis G., de Leeuw J. The Gifi System of Descriptive Multivariate Analysis. Statistical
Science. 1998; 13(4): 307-36.
Ng P. C., Henikoff S. SIFT: Predicting Amino Acid Changes That Affect Protein Function. Nucleic
Acids Research. 2003; 31(13): 3812-14.
Qian M., Shao Y. A Likelihood Ratio Test for Genomewide Association under Genetic
Heterogeneity. Annals of Human Genetics. 2013; 77(2): 174-82.
Reva B., Antipin Y., Sander C. Predicting the Functional Impact of Protein Mutations: Application
to Cancer Genomics. Nucleic Acids Research. 2011; 39(17): e118.
Schwarz J. M., Cooper D. N., Schuelke M., Seelow D. MutationTaster2: Mutation Prediction for the
Deep-Sequencing Age. Nature Methods. 2014; 11(4): 361-62.
Schwarz J. M., Rödelsperger C., Schuelke M., Seelow D. MutationTaster Evaluates Disease-Causing
Potential of Sequence Alterations. Nature Methods. 2010; 7(8): 575-76.
Shihab H. A., Gough J., Cooper D. N., Stenson P. D., Barker G. L., Edwards K. J., Day I. N., Gaunt T.
R. Predicting the Functional, Molecular, and Phenotypic Consequences of Amino Acid
Substitutions Using Hidden Markov Models. Human Mutation. 2013; 34(1): 57-65.
Stenson P. D., Mort M., Ball E. V., Howells K., Phillips A. D., Thomas N. S., Cooper D.N. The
Human Gene Mutation Database: 2008 Update. Genome Medicine. 2009; 1(1): 13.
The 1000 Genomes Project Consortium. A Global Reference for Human Genetic Variation. Nature.
2015; 526(7571): 68-74.
Wang K., Li M., Hakonarson H. ANNOVAR: Functional Annotation of Genetic Variants from
High-Throughput Sequencing Data. Nucleic Acids Research. 2010; 38(16): e164.
Wu Y., Kuang Q., Dong Y., Huang Z., Li Y., Li Y., Li M. Predicting Pathogenic Single Nucleotide
Variants through a Comprehensive Analysis on Multiple Level Features. Chemometrics and
Intelligent Laboratory Systems. 2016; 156(15): 224-30.
Yang L., Chen J., Shen B. Newborn Screening in the Era of Precision Medicine. Advances in
Experimental Medicine and Biology. 2017; 1005: 47-61.
Zeng P., Zhao Y., Liu J., Liu L., Zhang L., Wang T., Huang S., Chen F. Likelihood Ratio Tests in Rare
Variant Detection for Continuous Phenotypes. Annals of Human Genetics. 2014; 78(5): 320-32.
109
Amnion as a potential source of neural
and glial cell precursors – from cell culture
to a practical application
Mateusz Król1, Łukasz Limanówka
2, Patrycja Szmytkowska
3, Piotr Czekaj
4
[email protected], Students Scientific Society, School of Medicine in Katowice,
Medical University of Silesia, Medyków 18, 40-758 Katowice, Poland [email protected], Students Scientific Society, School of Medicine in Katowice,
Medical University of Silesia, Medyków 18, 40-758 Katowice, Poland [email protected], Department of Cytophysiology, Chair of Histology and
Embryology, School of Medicine in Katowice, Medical University of Silesia, Medyków
18, 40-758 Katowice, Poland [email protected], Department of Cytophysiology, Chair of Histology and Embryology,
School of Medicine in Katowice, Medical University of Silesia, Medyków 18, 40-758
Katowice, Poland
Abstract
Chronic disease and acute lesions of the central and peripheral nervous system are the difficulty for
contemporary medicine due to limited ability of the nervous tissue to regenerate. A promising
alternative to pharmacological treatments would be therapy with the use of neural and glial cell
precursors obtained by differentiation of amniotic cells. Human amniotic cells (hAC) are viable, do not
form cancerous foci, and display features of pluripotency. They can differentiate in vitro into cells of all
three germ layers, e.g. endodermal cells of the liver and pancreas, mesenchymal muscle and bone cells,
as well as ectodermal neurons, astrocytes and oligodendrocytes.
Cells with pluripotency characteristics, are usually isolated from the amnion by enzymatic methods.
Factors stimulating ex vivo differentiation of amniotic cells towards the cells of nervous tissue are:
ATRA (all-trans retinoic acid), FGF (fibroblast growth factor) and EGF (epidermal growth factor). To
provide stem cell differentiation into specific types of nerve cells, additional factors are used, like: NGF
(nerve growth factor), SHH (sonic hedgehog homolog), FGF-8 (fibroblast growth factor 8), melatonin
and ligands of Wnt-4 receptors.
Preclinical studies with the use of amnion cells and neuron-like cells derived from amnion culture have
revealed considerable effectiveness in recovery of neurological functions after TBI (traumatic brain
injury) or ischemic stroke. The increase of cell survival and migration rate, have been also reported.
However, only a few percent of grafted cells survive longer than one month, and the neuroprotective
effect is achieved by secretion of neurotrophic factors like BDNF, NGF, CNTF, GDNF and NT-3 at the
site of injury.
Therapies involving the transplantation of neural cell precursors may contribute to the reduction of
pathological changes resulting from traumatic brain injury and ischemic stroke, slowing down the
course of neurodegenerative disease and could be used in regenerative medicine.
Introduction
Chronic and acute nervous system diseases always result in progressive inflammation, massive loss of nervous tissue cells, diffuse areas of impairment and changes in concentrations of specific neurotransmitters. Classical therapeutic methods, like pharmacological treatment, are not able to effectively prevent progression of injured area in conditions like TBI (traumatic brain injury) and stroke. Moreover, there is not
Mateusz Król, Łukasz Limanówka, Patrycja Szmytkowska, Piotr Czekaj
110
known any therapy that could effectively suppress development of neurodegenerative and demyelinating disease. Therefore, a reasonable approach to tackle this issue could be intensification of regenerative processes, suppression of the inflammation and replacement of impaired cells at the site of injury. Such the strategy of treatment could contribute to partial or even full regression of lost tissue functions.
In recent years, there has been an increasing interest in various sources of stem cells which could be used in cell transplantation (Mamede, Carvalho et al. 2012). One promising source of stem cells is the human placenta. The innermost layer of the placenta is amniotic membrane containing cells characterized by low immuno-genicity, no teratoma formation and features of pluripotency. Since the amniotic membrane together with placenta is ordinarily discarded after childbirth, collecting placentae do not pose moral or ethical constraints. Furthermore, amniotic membrane-derived cells can differentiate into cells of all three germ layers, including neuron and glial cell precursors. These above-mentioned features make the amniotic membrane as an alternative source of stem cells for bone-derived mesenchymal stromal cells or embryonic-like stem cells (Wolbank, Peterbauer et al. 2007).
In preclinical studies, transplantation of undifferentiated amnion cells inhibited inflammation, promoted angiogenesis and improved cells viability. However, the major problem with this kind of application was the ability to differentiate these cells into mature neuron-like cells. Previous research has demonstrated that only few of grafted, undifferentiated cells possess neural morphology (Okawa, Okuda et al. 2001). The appropriate method for obtaining higher percentage of neural-like cells would be in vitro stem cells differentiation, before transplantation. On the other hand, in vivo regeneration can be achieved with the use of neuroprotective substances secreted by the grafted cells (Kim KS, Kim HS et al. 2013). Moreover, there are indications that differentiated amniotic cells would have greater clinical output due to higher production of neurotransmitters and neurotrophins, and ability to survive in host tissue (Yan, Zhang et al. 2013).
The purpose of this mini-review is to summarize recent studies concerning the potential of amniotic cells to differentiate into neural and glial cell precursors, as well as the effects of application both, differentiated and undifferentiated cells in preclinical trials.
Human amnion as a source of stem cells and mesenchymal tissue
Human amniotic membrane contains cells (hAC; human amnion cells) of different embryological origin (Fig. 1). hAEC (human amnion epithelial cells) originate from epiblast and form continuous monolayer of cells situated at the border between the amniotic fluid and the basement membrane (Miki, Marongiu et al. 2010). Sparsely distributed hAM-MSC (human amnion membrane-mesenchymal stromal cells) are located in mesenchymal tissue underlying the basement membrane (Soncini, Vertua et al. 2007). Both types of amnion cells can be obtained by means of enzymatic digestion. The received cell populations consist of cells with variable ability to self-renewal and differentiation (Parolini, Alviano et al. 2008). Important feature of hAC is low expression level of human leukocyte antigen class I (HLA-A, B, C) and lack of class II antigens (McDonald, Payne et al. 2015).
Amnion as a potential source of neural and glial cell precursors
– from cell culture to a practical application
111
Figure 1. Human amniotic membrane. Semithin section, stained with toluidine blue (Dept. of Cytophysiology,
MUS Katowice). Magnification 1000x.
Not only the cells from amniotic membrane, but also amniotic membrane itself has
been used in regenerative medicine. Several publications have documented the use of
amniotic membrane in the management of burns, creation of surgical dressings, as
well as reconstruction of ocular surface or even peripheral nerve regeneration
(Niknejad, Peirovi et al. 2008; Diaz-Prado, Muinos-López et al. 2011; Mohammad,
Shenaq et al. 2000). Especially, mechanical properties of amniotic membrane such as
permeability, flexibility and plasticity allow its application to tissue replacement with
suitable strength necessary to resist stress during the growth of tissue. Amniotic
membrane also have the ability to prevent inflammatory cells and fibroblasts
migration into nerve tissue. (Sadraie, Parivar et al. 2016). Furthermore, the presence
of ECM (extracellular matrix) components, including collagen, laminin, fibronectin
and vitronectin, serving as ligands for surface transmembrane receptors called integrins,
is crucial for intracellular processes like cell transport (exocytosis and endocytosis),
proliferation, migration, differentiation and apoptosis. ECM components also
contribute to maintaining appropriate concentrations of growth factors and anti-
inflammatory cytokines, partly by binding them directly, as well as indirectly
– by absorption of water, in which these factors and cytokines are dissolved
(Niknejad, Peirovi et al. 2008). Thereby, ECM provides microenvironment that is
conducive to viability, proliferation and migration of grafted tissue.
Mateusz Król, Łukasz Limanówka, Patrycja Szmytkowska, Piotr Czekaj
112
Extraction of human amnion cells
After obtaining written informed consent from the patient, a placenta is collected following term cesarean section. It must be shipped into laboratory immediately, so that the exposure time for hypoxia should be as short as possible. Subsequently, the amniotic membrane is mechanically peeled from the underlying chorion, and then it should be washed in PBS (phosphate-buffered saline) carefully and thoroughly to nearly ensure complete removal of blood, thus providing appropriate conditions for enzyme activities (Burton, Sebire et al. 2013).
Once the amniotic membrane is separated and washed, the isolation of amnion cells can be carried out. To release hAC cells from the membrane, the amnion tissue ought to be subjected to incubation with various enzymes. For example, hAEC can be obtained by incubation with dispase II or trypsin-EDTA solutions. The above described method allows to achieve from 12 to 14 million epithelial cells per gram of tissue (Gramignoli, Srinivasan et al. 2016).
Likewise, the isolation of hAM-MSC can be performed by enzymatic digestion, though with the use of various enzymes and slightly variant time periods of incubation. Collagenase and DNase (deoxyribonuclease) are also used. As a result, from 14 to 34 million of mesenchymal cells per one placenta can be obtained (Soncini, Vertua et al. 2007).
Characteristics of main cell populations derived from amnion
For the determination of amnion cell populations obtained after isolation, expression of specific cell markers should be assessed. In general, some of hAC cells show the expression of pluripotency markers specific for embryonic stem cells, e.g. OCT-4 (Octamer-binding protein 4), SOX-2 (SRY-related HMG-box gene 2), SSEA-3 (Stage Specific Embryonic Antigen-3), SSEA-4 (Stage Specific Embryonic Antigen-4), TRA (Tumor Rejection Antigen)-1-60 and TRA-1-81. These markers are typically present in cells having the capacity to self-renewal and differentiate into cells of three germ layers. However, hAC population contains both, pluripotent and multipotent stem cells. Moreover, hAC exhibit limited proliferation capacity in vitro (Diaz-Prado, Muinos-López et al. 2011), and the number of cells initially obtained from one amnion membrane is crucial for preparing appropriately dense cell suspension suitable for transplantation.
Presence of specific markers allows prompt distinction between hAEC and hAM-MSC, which is important due to variant differentiation capacity towards neuro-ectodermal lineages between both populations. Classical antigens of amniotic epithelial cells are cytokeratins 1-8, 10, 13-16, 19, integrins CD9 and CD24, CD324 (E-cadherin). On the other hand, cell markers such as SSEA-1, CD34, CD133 and CD117 (c-kit) are not present on hAEC or their expression was found to be very low (Miki, Lehman et al. 2005). In turn, hAM-MSC express classical mesenchymal markers, such as: CD29, CD44, CD49e, CD90, CD105 and CD166, though these cells are negative for hematopoietic and monocyte markers (CD34, CD45 and CD14) (Antoniadou, David et al. 2015). Unique characteristic feature, which merits attention, is absence of telomerase in amnion cells, which distinguishes these cells from embryonic stem cells and decreases risk of mutations (Tahan C., Tahan V. et al. 2014 and Kwon, Kim et al. 2016).
Amnion as a potential source of neural and glial cell precursors
– from cell culture to a practical application
113
Human amniotic cells culture and differentiation
Usually, hAEC and hAM-MSC isolated enzymatically with the use of enzymes such
as dispase, collagenase or trypsin, are resuspended in DMEM (Dulbecco’s modified
Eagle medium) mainly with addition of FBS (fetal bovine serum) and antibiotics and
seeded into plastic flasks. After achieving an appropriate confluence, the cells are
passaged and cultured (Sanluis-Verdes A., Sanluis-Verdes N. et al. 2017). Cell
culture can be supplemented with EGF (epidermal growth factor) to maintain
proliferation (Miki, Marongiu et al. 2010). After several passages, hAEC and hAM-
MSC cell cultures can be directly transplanted, to ameliorate regenerative processes
of nervous tissue. The grafted cells may produce growth factors and cytokines at the
site of injury whereby it can enhance the survival of injured neurons and glial cells.
Transplanted cells may also differentiate into appropriate neurons and glial cells at
the site of application and thus substitute injured cells. It was shown that only
a fraction of grafted cells undergoes differentiation into functional neurons and glial
cells in vivo. In order to transplant more differentiated cells, it would be important to
achieve the cells that are committed to a neuronal fate by induction of neural
differentiation in vitro (Wennersten, Meijer et al. 2004). In general, hAEC and hAM-
MSC can be subjected to differentiation after several passages. Then the medium
should be changed to induce neurospheres (Amendola, Nardella et al. 2014) and/or
terminal neural differentiation directly (Fig. 2).
Figure 2. In vitro methods for obtaining neural and glial cell precursors from amniotic cells. Terminal
differentiation can be achieved directly by placing hAEC or hAM-MSC in the specific differentiation medium
(A) or indirectly by placing the cells in the inductory medium (B) followed by differentiation medium (C).
Mateusz Król, Łukasz Limanówka, Patrycja Szmytkowska, Piotr Czekaj
114
Neurospheres are complex structures containing cells of different phenotypes and
expressing various neural lineage-specific markers. Cytoarchitecture of these
structures is most likely caused by cell adaptation to the conditions prevailing in in
vitro culture. The inner part of neurosphere is composed of poorly proliferating cells
with low level of gene transcription. It is caused by a greater difficulty in nutrient
exchanging. On the other hand, the outer layer is rich in proliferating cells, and
actively produces proteins, such as neurotrophins and cytokines that can be
accumulated in the inner part and/or released into the culture medium (Bez, Corsini
et al. 2003 and Amendola, Nardella et al. 2014).
One of the standard medium that has been used to make the cells committed to
a neuronal fate is DMEM supplemented with, among others, FBS, RA (retinoic acid),
bFGF (basic fibroblast growth factor), EGF, L-glutamine and non-essential amino-
acids, 2-mercaptoethanol, sodium pyruvate, insulin-transferrin-selenium and ascorbic
acid. In this medium, hAC acquired some morphological neuron-like features, e.g.
presence of forming extensions, after 4 days of culture. At the end of culture in the
differentiating medium, these features were much more distinct and synapses or Nissl
bodies, were seen. In addition, the presence of NF (neurofilaments) and markers such
as GFAP (glial fibrillary acidic protein), vimentin, NGFR (nerve growth factor
receptor), and S100β (S-100β protein) has been reported (Sanluis-Verdes A., Sanluis-
Verdes N. et al. 2017). Different number of cells positive for individual markers
suggest the presence of different cell subpopulations in culture. Moreover, higher
expression of glial marker S100β and neural markers: NF and NGFR, has been
detected in hAEC in comparison with hAM-MSC (Sanluis-Verdes A., Sanluis-
Verdes N. et al. 2017). hAEC synthesize some neurotransmitters (dopamine and
other catecholamines) and enzymes, like tyrosine hydroxylase (TH), necessary for
the formation of these neurotransmitters (Kakishita, Elwan et al. 2000). Furthermore,
hAEC secrete neurotrophic factors, such as BDNF (brain-derived neurotrophic
factor), NT-3 (neurotrophin-3), insulin-like growth factors, EGF, prostaglandin E2
etc., which may be conducive to dopaminergic neurons survival after amniotic cells
transplantation in neurodegenerative disorders like PD (Parkinson’s disease) (Yang,
Song et al. 2010; Allen,Watson et al. 2013; Bothwell, 2014; Bennett, Lagopoulos,
2014; Ziegler, Levison et al. 2015; Oyagi, Hara, 2012). All these observations
indicate that potentially, hAEC can easier differentiate towards neural precursors.
Preclinical application of amnion tissue
Both, chronic and acute pathological conditions lead to massive neural cells degene-
ration resulting in progressive idiosyncratic clinical picture with characteristic signs
and symptoms. Nonstationary course of these pathological conditions inevitably
leads to deterioration of mental and neurological functions, and finally to dementia.
Unfortunately, none of currently used therapies is not able to effectively impede this
escalating neurodegenerative process. Since the cell therapy has proved its clinical
effectiveness of treatment of above-mentioned conditions, there are great
expectations of using these cells on a broad scale.
There are several studies concerning positive effects of undifferentiated hAC
transplantation for treating various diseases (Li, Guo et al. 2014; Kim KS, Kim HS et
Amnion as a potential source of neural and glial cell precursors
– from cell culture to a practical application
115
al. 2013; Xue, Chen et al. 2012; Sheng, McShane et al. 1993; Yang, Song et al. 2010;
Liu, Vaghjiani et al. 2012; Liu, Chan et al. 2014; McDonald, Payne et al. 2015;
Chen, Lu et al. 2011; Liu, Wu et al. 2008). For example, in TBI, external forces act
on nervous tissue, resulting in short and long-term undesirable consequences, like
decreased level of consciousness, amnesia, neurologic deficits and ischemic
complications (Mckee, Daneshvar 2015). In the case of TBI, it has been shown that
for 2 weeks after ipsilateral intracerebroventricular hAC transplantation in rat model
of PBBI (penetrating ballistic-like brain injury), alleviation of axonal degeneration in
CC (corpus callosum), ipsilateral thalamus and around lesioned area were observed.
Transplanted cells accumulated in subventricular zone and corpus callosum where
they could survive up to 4 weeks. Moreover, hAC had ability to intensively migrate
toward lesioned area. Neuroprotective effect of amnion cells transplantation,
probably results from well-known ability to secrete soluble factors, which support
neural cells viability and induce endogenous NSC (neural stem cells) population to
differentiation (Chen, Tortella et al. 2009).
Another example of potential hAC application is ischemic stroke. After ischemic
incident it is enormously important to immediately initiate neuroprotective therapy,
usually by means of neuroprotective drugs like piracetam or tPA (tissue plasminogen
activator), albeit limited by short time window of efficacy (Yu, Soncini et al. 2009).
In such condition, with the aid of transplanted hAC, a reduction of ischemic area and
improvement of neurological function could be obtained. In the rat model of
ischemic stroke, cells after injection to CC survived up to 3 weeks and had the
capability to migrate toward the areas of injury. Moreover, the infarct area volume
decreased by 10% in comparison to control group, and applied cells expressed neural
progenitor cell markers like Nestin and MAP2 (Li, Miao et al. 2012).
Advanced depletion of dopaminergic neurons in PD could also be ameliorate by hAC
administration. After hAC injection into striatum of rats with PD, animals achieved
better scores in behavioral tests. Immunohistochemical staining for human nuclear
antibody showed extensive migration of transplanted cells. After cells’ administration
number of TH-positive dopaminergic neurons increased, what suggested activation
of endogenous NSC responsible for physiologic neurogenesis (Kong, Cai et al. 2008).
The above examples of hAC treatment show that grafted cells contribute to restriction of injured areas and improvement of neurological function. It is very interesting because it was also reported that only a few percent of applied cells differentiate into mature nervous cells (Okawa, Okuda et al. 2001). It is important to consider the differences of effectiveness between differentiated and undifferentiated cells, and what stems from it, whether hAC should be differentiated in vitro, before transplantation. There is little study about preclinical application of neural precursors derived from hAC. Detailed examination of differences between hAM-MSC and hAM-NP (hAM-MSC differentiated into neural precursors) by Yan et al. (2013) suggests that cells subjected to differentiation in vitro have greater therapeutic potential. Despite the absence of significant differences in cell survival between hAM-MSC and hAM-NP in the rat model of TBI, there was observed more reduced area of damage in case of hAM-NP, in comparison to hAM-MSC. Moreover, rats treated with hAM-NP exhibited better scores in tests evaluating neurological
Mateusz Król, Łukasz Limanówka, Patrycja Szmytkowska, Piotr Czekaj
116
recovery, compared with rats treated with hAM-MSC. Quantitative real-time PCR showed significantly stronger expression of neurotrophins like NGF (nerve growth factor), CNTF (ciliary neurotrophic factor), GDNF (glial cell line-derived neurotrophic factor), BDNF and NT-3 genes by hAM-NP in comparison to hAM-MSC (Yan, Zhang et al. 2013).
In another study, Gao et al. (2016) found that application of previously differentiated bovine amnion epithelial cells into spinal cord in the animal model of injury, gave distinctly better rate of transplant survive in comparison to application of undifferentiated AEC. In this study, it was also reported that the recovery of motor functions was significantly better in the group treated with neural-like cells derived from amnion epithelial cells. These findings confirm the thesis that amnion cells with acquired neural fate commitment have superior potential in treatment of nervous system diseases in comparison to undifferentiated cells (Gao, Bai et al. 2016).
Summary
It can be concluded that human amnion membrane can be an attractive source of cells for therapy of nervous system diseases. Obtained amnion cells can effectively differentiate toward neural and glial cell precursors. Application of the cells with previously committed neural fate in vitro, yields better neural function recovery than the use of undifferentiated cells. Higher rate of survival and substantial secretion of neurotrophins in vivo are also features of neural precursors which attract attention in view of clinical therapy. However, more studies are required to fully understand processes of both, hAC differentiation and regeneration before their clinical application in humans.
References Allen SJ., Watson JJ., Shoemark DK., Barua NU., Patel NK. GDNF, NGF and BDNF as therapeutic options for neurodegeneration. Pharmacology & Therapeutics. 2013; 138(2):155-75.
Amendola D., Nardella M., Guglielmi L., Cerquetti L., Carico E., Alesi V., Porru M., Leonetti C., Bearzi C., Rizzi R., D'Agnano I., Stigliano A., Novelli G., Bucci B. Human placenta-derived neurospheres are susceptible to transformation after extensive in vitro expansion. Stem Cell Research and Therapy. 2014; 5(2): 55.
Antoniadou E., David A. L., Placental stem cells. Best Practice & Research: Clinical Obstetrics & Gynaecology. 2015.
Bennett MR., Lagopoulos J. Stress and trauma: BDNF control of dendritic-spine formation and regression. Progress in Neurobiology. 2014; 112:80-99.
Bez A., Corsini E., Curti D., Biggiogera M., Colombo A., Nicosia R. F., Pagano S. F., Parati E. A. Neurosphere and neurosphere-forming cells: morphological and ultrastructural characterization. Brain Research. 2003; 993(1-2):18-29.
Bothwell M. NGF, BDNF, NT3, and NT4. Neurotrophic Factors. 2014; 220:3-15.
Burton G. J., Sebire N. J., Myatt L., Tannetta D., Wang Y. L., Sadovsky Y., Staff A. C., Redman C. W. Optimising sample collection for placental research. Placenta. 2014;35(1):9-22.
Chen Z., Lu XC., Shear DA., Dave JR., Davis AR., Evangelista CA., Duffy D., Tortella FC. Synergism of human amnion-derived multipotent progenitor (AMP) cells and a collagen scaffold in promoting brain wound recovery: pre-clinical studies in an experimental model of penetrating ballistic-like brain injury. Brain Research. 2011; 1368:71-81.
Chen Z., Tortella FC., Dave JR., Marshall VS., Clarke DL., Sing G., Du F., Lu XC. Human amnion-derived multipotent progenitor cell treatment alleviates traumatic brain injury-induced axonal degeneration. Journal of Neurotrauma. 2009; 26(11):1987-97.
Amnion as a potential source of neural and glial cell precursors
– from cell culture to a practical application
117
Díaz-Prado S., Muiños-López E., Hermida-Gómez T., Cicione C., Rendal-Vázquez M.E, Fuentes-Boquete I., Toro F. J,. Blanco F. J. Human amniotic membrane as an alternative source of stem cells for regenerative medicine. Differentiation. 2011; 81(3).
Gramignoli R., Srinivasan RC., Kannisto K., Strom SC. Isolation of Human Amnion Epithelial Cells According to Current Good Manufacturing Procedures. Current Protocols in Stem Cell Biology. 2016;37:1E.10.1-1E.10.13.
Gao Y., Bai C., Zheng D., Li C., Zhang W., Li M., Guan W., Ma Y. Combination of melatonin and Wnt-4 promotes neural cell differentiation in bovine amniotic epithelial cells and recovery from spinal cord injury. Journal of Pineal Research; 2016; 60(3):303-12.
Kakishita K., Elwan M. A., Nakao N., Itakura T., Sakuragawa N. Human Amniotic Epithelial Cells Produce Dopamine and Survive after Implantation into the Striatum of a Rat Model of Parkinson's Disease: A Potential Source of Donor for Transplantation Therapy. Experimental Neurology. 2000; 165(1):27-34.
Kim KS., Kim HS., Park JM., Kim HW., Park MK., Lee HS., Lim DS., Lee TH., Chopp M., Moon J. Long-term immunomodulatory effect of amniotic stem cells in an Alzheimer's disease model. Neurobiology of Aging. 2013; 34(10):2408-20.
Kong XY., Cai Z., Pan L., Zhang L., Shu J., Dong YL., Yang N., Li Q., Huang XJ., Zuo PP. Transplantation of human amniotic cells exerts neuroprotection in MPTP-induced Parkinson disease mice. Brain Research. 2008; 1205:108-15.
Li F., Miao ZN., Xu YY., Zheng SY., Qin MD., Gu YZ., Zhang XG. Transplantation of human amniotic mesenchymal stem cells in the treatment of focal cerebral ischemia. Molecular Medicine Reports. 2012; 6(3):625-30.
Li Y., Guo L., Ahn HS., Kim MH., Kim SW. Amniotic mesenchymal stem cells display neurovascular tropism and aid in the recovery of injured peripheral nerves. Journal of Cellural and Molecular Medicine. 2014; 18(6):1028-34.
Liu T., Wu J., Huang Q., Hou Y., Jiang Z., Zang S., Guo L. Human amniotic epithelial cells ameliorate behavioral dysfunction and reduce infarct size in the rat middle cerebral artery occlusion model. Shock. 2008; 29(5):603-11.
Liu YH., Chan K., Vaghjiani V., Murthi P., Manuelpillai U., Toh BH. Human amniotic epithelial cells suppress relapse of corticosteroid-remitted experimental autoimmune disease. Cytotherapy. 2014; 16(4):535-44.
Liu YH., Vaghjiani V., Tee JY., To K., Cui P., Oh DY., Manuelpillai U., Toh BH., Chan J. Amniotic Epithelial Cells from the Human Placenta Potently Suppress a Mouse Model of Multiple Sclerosis. PloS One. 2012; 7(4):e35758.
Mamede AC., Carvalho MJ., Abrantes AM., Laranjo M., Maia CJ., Botelho MF. Amniotic membrane: from structure and functions to clinical applications. Cell and Tissue Research. 2012; 349(2):447-58.
McDonald CA., Payne NL., Sun G., Moussa L., Siatskas C., Lim R., Wallace EM., Jenkin G., Bernard CC. Immunosuppressive potential of human amnion epithelial cells in the treatment of experimental autoimmune encephalomyelitis. Journal of Neuroinflamation. 2015; 12:112.
Mckee AC., Daneshvar DH. The neuropathology of traumatic brain injury. Handbook of Clinical Neurology. 2015; 127:45-66.
Miki T., Lehmann T., Cai H.,.Stolz D.B, Strom S.C. Stem cell characteristics of amniotic epithelial cells. Stem Cells. 2005. 1549-59.
Miki T., Marongiu F., Dorko K., Ellis E. C. S., Strom S. C. Isolation of Amniotic Epithelial Stem Cells. Current Protocols in Stem Cell Biology. 2010; Chapter 1:Unit 1E.3.
Mohammad J., Shenaq J., Rabinovsky E., Shenaq S. Modulation of peripheral nerve regeneration: a tissue-engineering approach. The role of amnion tube nerve conduit across a 1-centimeter nerve gap. Plastic & Reconstructive Surgery. 2000; 105(2):660-6.
Mateusz Król, Łukasz Limanówka, Patrycja Szmytkowska, Piotr Czekaj
118
Niknejad H., Peirovi H., Jorjani M., Ahmadiani A., Ghanavi J., Seifalian AM. Properties of the amniotic membrane for potential use in tissue engineering. European cells & materials. 2008; 15: 88-99.
Okawa H., Okuda O., Arai H., Sakuragawa N., Sato K. Amniotic epithelial cells transform into neuron-like cells in the ischemic brain. NeuroReport. 2001; 18: 4003-4007.
Oyagi A., Hara H. Essential roles of heparin-binding epidermal growth factor-like growth factor in the brain. CNS Neuroscience & Therapeutics. 2012; 18(10):803-10.
Parolini O., Alviano F., Bagnara G. P., Bilic G., Bühring H. J., Evangelista M., Hennerbichler S., Liu B., Magatti M., Mao N., Miki T., Marongiu F., Nakajima H., Nikaido T., Portmann-Lanz C. B., Sankar V., Soncini M., Stadler G., Surbek D., Takahashi T. A., Redl H., Sakuragawa N., Wolbank S., Zeisberger S., Zisch A., Strom S. C. Concise review: isolation and characterization of cells from human term placenta: outcome of the first international Workshop on Placenta Derived Stem Cells. Stem Cells (Dayton, Ohio). 2008; 26(2):300-11.
Sadraie SH., Parivar K., Arabi F., Moattari M., Kaka G., Mansouri K. Study of Transected Sciatic Nerve Repair by Amniotic Membrane with Betamethasone in Adult Albino Wistar Rats. Archives of Iranian Medicine. 2016; 19(9):612-7.
Sanluis-Verdes A., Sanluis-Verdes., Manso-Revilla M. J., Castro-Castro A. M., Pombo-Otero J., Fraga-Mariño M., Sanchez-Ibañez J., Doménech N., Rendal-Vázquez M. E. Tissue engineering for neurodegenerative diseases using human amniotic membrane and umbilical cord. Cell and Tissue Banking. 2017; 18(1):1-15.
Sheng JG., McShane LM., Plunkett RJ., Cummins AC., Oldfield EH., Kopin IJ., Palmatier MA. Dopaminergic neuronal sprouting and behavioral recovery in hemi-parkinsonian rats after implantation of amnion cells. Experimental Naurology. 1993; 123(2):192-203.
Soncini M., Vertua E., Gibelli L., Denegri M., Albertini A., Wengler G.S., Parolini O., Isolation and characterization of mesenchymal cells from human fetal membranes. Journal of Tissue Engineering and Regenerative Medicine. 2007. 1, 296-305.
Tahan A. C., Tahan V., Placental amniotic epithelial cells and their therapeutic potential in liver diseases. Frontiers in Medicine. 2014. 1. 1-4.
Wennersten A., Meijer X., Holmin S., Wahlberg L., Mathiesen T. Proliferation, migration, and differentiation of human neural stem/progenitor cells after transplantation into a rat model of traumatic brain injury. Journal of Neurosurgery. 2004; 100(1):88-96.
Wolbank S., Peterbauer A., Fahrner M., Hennerbichler S., van Griensven M., Stadler G., Redl H., Gabriel C. Dose-dependent immunomodulatory effect of human stem cells from amniotic membrane: a comparison with human mesenchymal stem cells from adipose tissue. Tissue Engineering. 2007; 13(6):1173-83.
Yan ZJ., Zhang P., Hu YQ., Zhang HT., Hong SQ., Zhou HL., Zhang MY., Xu RX. Neural Stem-Like Cells Derived from Human Amnion Tissue are Effective in Treating Traumatic Brain Injury in Rat. Neurochemical Research. 2013; 38(5):1022-33.
Yang X., Song L., Wu N., Liu Z., Xue S., Hui G. An experimental study on intracerebroventricular transplantation of human amniotic epithelial cells in a rat model of Parkinson's disease. Neurological Research. 2010; 32(10):1054-9.
Yu SJ., Soncini M., Kaneko Y., Hess DC., Parolini O., Borlongan CV. Amnion: a potent graft source for cell therapy in stroke. Cell Transplantation. 2009; 18(2):111-8.
Xue S., Chen C., Dong W., Hui G., Liu T., Guo L. Therapeutic effects of human amniotic epithelial cell transplantation on double-transgenic mice co-expressing APPswe and PS1ΔE9-deleted genes. Science China Life Sciences. 2012; 55(2):132-40.
Ziegler AN., Levison SW., Wood TL. Insulin and IGF receptor signalling in neural-stem-cell homeostasis. Nature Reviews Endocrinology. 2015; 11(3):161-70.
119
What makes the bacterial vaginosis difficult
to diagnose: our experience in Silesia, Poland
Patryk Madeja1,3
, Olga Smolarek1, Agnieszka Mosler
1, Małgorzata Romanik
2
1Student Research Group at the Department of Medical Microbiology, School
of Medicine in Katowice, Medical University of Silesia, Katowice, Poland 2Department of Medical Microbiology, ul. Medyków 18, 40-752 Katowice, School
of Medicine in Katowice, Medical University of Silesia, Katowice, Poland 3Corresponding author: Department of Medical Microbiology, ul. Medyków 18,
40-752 Katowice, School of Medicine in Katowice, Medical University of Silesia,
Katowice, Poland; [email protected]
Abstract
INTRODUCTION: Bacterial vaginosis (BV) is manifested mainly by vaginal discharge and is defined
as an imbalance of the bacteria that are normally present in the vagina.
During our last study we find out that some smears are hard to classify according to Nugent’s Criteria.
The aim of this study was to characterize patients with problematical clinical diagnosis.
MATERIALS AND METHODS: Vaginal smears were collected from 100 women of childbearing age
with vaginal discharge without cervicitis and antibiotic therapy. Classification of the vaginal microbiota
by Nugent Criteria was performed by laboratory staff and students. In 44 cases there was diagnostic
discrepancies between students and laboratory staff.
RESULTS: Among the selected 44 women vaginal pH ≥ 4.5 had been noted in 38.63%, while only
16.07% of the other group patients had incorrect vaginal pH. Statistically 31.82% of the first group was
in age range 36-40 and only 14.29% of the properly diagnosed was in the same age range. Vaginal
douching had been used by 45.45% of incorrectly diagnosed group while in the other group only by
33.93%. Results of chi-square test revealed statistically significance in pH (χ2 = 13.2665; p = 0.00027 at
p < 0.05), patient’s age (χ2 = 12.411; p = 0.014543 at p < 0.05).
CONCLUSIONS: Our study showed that in case of diagnostic difficulties the vaginal pH and patient’s
age must be considered for the diagnosis of bacterial vaginosis.
Introduction
At present, over 19.8 million women live in Poland (CSO, 2016). About 7.6 million
are in childbearing age (CSO, 2013). It is obvious that the most common cause of
vaginal symptoms among women in this age group should be considered as serious
health problem (CDC, 2018).
Bacterial vaginosis (BV) is defined as chronic imbalance of the morphotypes
normally present in the vagina. The mixed microflora consisting of Gardnerella
vaginalis, Bacteroides spp., Mobiluncus spp., Mycoplasma hominis and multiple
other bacteria replaces population of Lactobacillus spp. – the predominant species in
healthy vaginal microbiota (Antonio, Hawes et al. 1999). Because the aetiology of
BV is still vague and associated with multiple genital microorganisms there are some
clinical signs (adherent, homogenous, white vaginal discharge with a characteristic
fishy odour and vaginal pH ≥ 4.5) that has been defined for diagnosis of BV and
called the Amsel criteria (Amsel, Totten et al. 1983). On the other hand, there are
several microbiological criteria, which may be used to confirm BV established by
Patryk Madeja, Olga Smolarek, Agnieszka Mosler, Małgorzata Romanik
120
three other groups of authors (Ison, Hay, 2002, Spiegel, Amsel et al. 1983, Nugent,
Krohn et al. 1991).
Correct diagnosis of BV may be problematic, because there are couple of factors
interfering the assessment: multitude of forms Lactobacillus morphotypes take,
problematic counting of some Lactobacillus morphotypes forming a leptotricha or
assessment of smears lacking any morphotypes (because of imperfections in
diagnostic criteria). Many factors including smoking, vaginal douching, taking
antibiotics, and high-risk sexual behaviour are associated with diversity of the
vaginal microbiota and lack of Lactobacillus spp. (Hellberg, Nilsson et al. 2000,
Cotrell, Hansen, 2010, Olson, Boohaker, et. al. 2018, Wessels, Lajoie, et al. 2017).
Recent data shows that using of copper intrauterine device may increase colonization
by BV-associated microbiota (Achilles, Austin et al. 2018).
BV increases risk of acquisition and transmission of sexually transmitted infections
(STI) including Human Immunodeficiency Virus (HIV), Herpes Simplex Virus Type
2 (HSV-2), Chlamydia trachomatis (C. trachomatis), and Neisseria gonorrhoeae (N.
gonorrhoeae) (Bautista, Wurapa, et al. 2016).
Many studies indicate clinical importance of BV, showing that this condition is a risk
factor for adverse pregnancy outcomes, including preterm deliveries, low birth weight
infants, and enhances risk of miscarriage in the first trimester and postpartum
endometritis (Nelson, Hanlon et al. 2009, Chawla, Bhalla et al. 2013, Ralph, Rutherford
et al. 1999). It may also lead to infertility (Mania-Pramanik, Kerkar et al. 2009).
According to Centers for Disease Control and Prevention (CDC) treatment guidelines
women with BV symptoms should be treated orally or intravaginally with
metronidazole or clindamycin (CDC, 2015; access: 14.03.2018). Therapy is calculated
to relieve vaginal symptoms and signs of infection, and potentially to reduce the risk
for acquiring C. trachomatis, N. gonorrhoeae, Trichomonas vaginalis, HIV, and
HSV-2 (Brotman, Klebanoff et al. 2010, Cherpes, Wiesenfeld, et al. 2006,
Schwebke, Desmond 2007). There are also presumptions that oral and vaginal
administration of probiotics may result in the cure of BV or cause reduction relapses
of symptoms (Kim, Park 2017). Despite antimicrobial treatment, BV has tendency to
relapse: recurrence was noted in 80% of cases within a year after treatment and in
50% of the cases within 5-9 years after treatment (Romanik, Ekiel et al. 2007).
The aim
The aim of this study was to characterize patients with problematical clinical diagnosis
and expose potential factors that may make BV difficult to diagnose, considering the
reasons mentioned in the introduction, we believe that early diagnosis and treatment of
BV is very important especially for women of childbearing age.
Materials and methods
Vaginal smears were collected by clinician from 100 women of childbearing age with vaginal discharge and without cervicitis (in clinical and cytological examination). None of the patients were undergoing antibiotic therapy at the time of the collection. Classification of the vaginal microbiota by laboratory staff (two laboratory
What makes the bacterial vaginosis difficult to diagnose: our experience in Silesia, Poland
121
diagnosticians) and students was performed using three scales based on microscopic analysis of Gram-stained vaginal smears: Nugent scoring system, Hay and Ison score and Spiegel score. Smears were microscopically evaluated under 1000x magnification with oil immersion. Then clinical data of the patients on anonymous written form were analysed.
Table 1. Hay and Ison Criteria and Spiegel Criteria
Hay and Ison Criteria Spiegel Criteria Diagnosis
Grade 1 Predominance of Lactobacillus morphotypes
Domination of morphotypes of Lactobacillus spp.
Vaginal physiological
microbiota
Grade 2 Mixed vaginal flora with approximately equal amounts of Lactobacillus spp. and Gardnerella vaginalis morphotypes
Equal amount of Lactobacillus spp. morphotypes and Gardnerella-like bacteria
Intermediate condition
Grade 3 Predominance of Gardnerella vaginalis morphotypes and Mobiluncus morphotypes and a few Lactobacillus morphotypes
A few or no Lactobacillus spp. and many Gram-negative and/or Gram-variable morphotypes (typical for G. vaginalis)
BV
There are three grades in scale established by Hay and Ison’s (Table 1).
Another scale, called Spiegel score, is very similar to Hay and Ison’s method (Table 1). According to Spiegel’s system bacteria are categorized into groups of morphotypes in terms of its length. Lactobacillus spp. Gram-positive morphotypes are classified as long bacteria, but on the contrary Gram-variable morphotypes typical for Gardnerella vaginalis, Bacteroides spp., Prevotella spp., or Mobiluncus spp. are classified as short bacteria.
However, the current ‘gold standard’ in BV diagnostics is Nugent scoring system. (Hoffman, Bellad, et al. 2017) (Table 2).
Table 2. Nugent scoring system
Lactobacillus
morphotypes
Gardnerella or Bacteroides
morphotypes
Curved variable rods
Score 0 for >30 Score 0 for 0 Score 0 for 0
Score 1 for 5-30 Score 1 for <1 Score 1 for <5
Score 2 for 1-4 Score 2 for 1-4 Score 2 for 5+
Score 3 for <1 Score 3 for 5-30
Score 4 for 0 Score 4 for >30
Patryk Madeja, Olga Smolarek, Agnieszka Mosler, Małgorzata Romanik
122
The average of all 20 visual fields is taken as final Nugent score. Nugent score 0-3 is
considered physiological microbiota, 4-6 is considered ‘intermediate’ state, while ≥7
confirms BV.
Statistical analysis was performed using Microsoft Excel (Office 2007) and Chi-
Square Test Calculator available on www.socscistatistics.com/tests/chisquare2/
Default2.aspx (access: 14.12.2017) and Mann-Whitney U Test Calculator available
on www.socscistatistics.com/tests/mannwhitney/ (access: 14.12.2017), p<0.05 was
considered statistically significant.
During this study, patients were divided into two subgroups according to inter-
observer (students’ vs laboratory staff) agreement. The first subgroup, consisted of
patients misdiagnosed by students and counted 44 women. For simplicity we called
them ‘incoherent’. Another subgroup of patients, correctly diagnosed by students,
consisted of 56 women was analogically called ‘coherent’. We analysed these two
groups in terms of different parameters: vaginal pH, age, parity, contraceptives,
vaginal douching, cigarette smoking and recurrence of the symptoms after treatment
with metronidazole according to CDC guidelines (CDC, 2015; access: 14.03.2018).
Results
The age of women in both groups (coherent and incoherent) was comparable. Median
age of incoherent group was 35.5 years and coherent was 29 years. In incoherent group
age average was 33.6 years and in the coherent group was 30.6 (Table 3).
Table 3. Age characteristics (in years) in women from incoherent and coherent group
Group Age U-MW* p-value**
Min-max Median Average
-1.90266 0.05744 Incoherent 22-45 35.5 33.6
Coherent 20-45 29 30.6
*Mann–Whitney U test; **result is statistically significant p <0.05
The most numerous group with the results of the BV assessment were different
between the students and the diagnosticians was group of patients with age ranged
36-40. In coherent group 8 of 56 (14.29%) patients were in this age range, while in
incoherent group 14 of 44 (31.82%) patients were in mentioned age range (Table 4).
What makes the bacterial vaginosis difficult to diagnose: our experience in Silesia, Poland
123
Table 4. Percentage of age range distribution in incoherent and coherent groups
Age range Incoherent Coherent
22-25 18.18% 23.21%
26-30 22.73% 35.71%
31-35 9.09% 14.29%
36-40 31.82% 14.29%
41-45 18.18% 12.5%
Out of 44 patients of incoherent group, the proper diagnosis of physiological microbiota was noted in 52.27% of cases, intermediate state in 29.55% and BV in 18.18% of cases (Table 5).
Table 5. Distribution of different diagnoses among women from incoherent and coherent group
Groups Nugent score χ2 p-value*
0-3 Normal vaginal microbiota
4-6 Intermediate state
7-10 BV
14.2489 0.000805 Incoherent Number (%)
23/44 (52.27%)
13/44 (29.55%)
8/44 (18.18%)
Coherent Number (%)
43/56 (76.79%)
7/56 (12.5%)
6/56 (10.71%)
*result is statistically significant p <0.05
Out of 56 patients of coherent group, majority (76.79%) possessed physiological microbiota, 12.5% – intermediate state, and in 10.71% of women BV was diagnosed. This analysis shows that out of 66 women, who in fact did not suffer from BV, 43 was diagnosed by students properly and 23 was diagnosed improperly. There were 20 patients with intermediate state and only 7 of them were diagnosed by students properly, while 13 was misdiagnosed. Among 14 women suffering from BV, 6 was diagnosed correctly, while 8 was diagnosed incorrectly.
Analysing subsequent factors, such as parity (Table 6) and contraception and its type (Table 7), there were no statistically significant differences between groups, this would indicate a high homogeneity of the groups covered by this study and a lack of association of parity and contraception type with diagnostic difficulties using Nugent score.
Among the incoherent group, 52.27% of women had never been pregnant while 25% (11 women) were multiparous. Primiparous were 10/44 (22.73%) of incoherent and 19/56 (33.93%) of coherent group.
Patryk Madeja, Olga Smolarek, Agnieszka Mosler, Małgorzata Romanik
124
Table 6. Distribution of parity in group incoherent and coherent in the study
Parity Incoherent Coherent χ2 p-value*
Nulliparas
Number (%)
23/44
(52.27%)
24/56
(42.86%)
3.0588 0.216669 Primiparas
Number (%)
10/44
(22.73%)
19/56
(33.93%)
Multiparas
Number (%)
11/44
(25%)
13/56
(23.21%)
*result is not statistically significant at p <0.05
Table 7. Distribution of contraception type used by patients of evaluated groups and performing vaginal
douching
Contraception Incoherent Coherent χ2 p-value***
No contraception
Number (%)
31/44
(70.45%)
38/56
(67.86%)
0.0935 0.759772
OC *
Number (%)
10/44
(22.73%)
11/56
(19.64%)
IUD **
Number (%)
0/44
(0%)
1/56
(1.79%)
Contraceptive transdermal
patch
Number (%)
2/44
(4.55%)
3/56
(5.36%)
Contraceptive vaginal ring
Number (%)
1/44
(2.27%)
3/56
(5.36%)
Vaginal douching
Number (%)
20/44
(45.45%)
19/56
(33.93%)
*OC – oral contraception; **IUD – intrauterine device; ***result is not statistically significant at p <0.05
What makes the bacterial vaginosis difficult to diagnose: our experience in Silesia, Poland
125
Vaginal douching had been used by 20/44 (45.45%) of incorrectly diagnosed women,
while in another group it has been performed only by 19/56 (33.93%) of patients.
Majority of women from both groups claimed not to smoke tobacco. Only one
woman (1.79%) of coherent and three women (6.82%) of incoherent group have been
smoking cigarettes. Relapses of BV after treatment occurred only in 6.82% women
(3/44) of incoherent and 1.79% woman (1/56) of coherent group.
Abnormal vaginal pH noted in physical examination was strongly associated with
results obtaining in incoherent group (table 8).
Table 8. Distribution of vaginal pH in incoherent and coherent group in the study
Vaginal pH range Incoherent
N (%)
Coherent
N (%)
χ2 p-value*
≥4.5 17/44
(38.63%)
47/56
(16.07%)
13.2665 0.00027
<4.5 27/44
(61.36%)
9/56
(83.93%)
*result is statistically significant p <0.05
Furthermore, the microbiological assessment according to Nugent score shows
higher percentage of intermediate state in incoherent group (Table 9).
Table 9. Distribution of correct and incorrect diagnoses among women with different state of microbiota
Groups Nugent Score
0-3 (Normal vaginal
microbiota)
4-6 (Intermediate state) 7-10 (BV)
Incoherent Number (%) 23/66
(34.85%)
13/20
(65%)
8/14
(57.14%)
Coherent Number (%) 43/66
(65.15%)
7/20
(35%)
6/14
(42.86%)
x2 19.3488.
p-value 0.000063.
*result is statistically significant p <0.05
Patryk Madeja, Olga Smolarek, Agnieszka Mosler, Małgorzata Romanik
126
Discussion
There are many factors altering the result of microscopic evaluation of Gram-stained
smears using Nugent scoring system. Current study is the first to associate clinical
features with difficulty of making correct BV diagnosis. From all characteristics, we
took into consideration, all can alter vaginal ecosystem (Achilles, Austin, et al. 2018,
Sabour, Arzanlou, et al. 2018, Aslan, Bechelaghem 2018), and can complicate
microscopic evaluation.
Our study showed that among many factors as: using contraception and especially
type of contraception, age, vaginal douching, vaginal pH and parity, incorrect vaginal
pH value (≥ 4.5) demonstrated a statistically significant difference between two
studied groups. Percentage of misdiagnoses was the highest in the group of patients
diagnosed with intermediate state, especially in women between 36-40 years old.
Correct diagnosis of BV still remains problematic as its aetiology is vague. Nugent
scoring system is a method considered the ‘gold standard’ as it is the most objective
laboratory tool for diagnosing BV, but its usage may be challenging for
inexperienced diagnosticians.
First problem we met during this study was wide range of forms that Lactobacillus
morphotypes take, like large and intermediate rods, spherical or very long forms, so
called leptotrichas. Often it is possible to mistake intermediate rods with Gardnerella
spp. morphotypes or spherical forms with Gram-positive cocci. When mistaken with
another bacteria, morphotype is not taken into account and it may change the result
of the assessment. Gram variable rods – Gardnerella vaginalis has a very thin cell
wall with a characteristic Gram-negative staining pattern but cell walls are
unequivocally Gram-positive in their ultrastructural characteristics and chemical
composition (Sadhu, Domingue et al. 1989).
Another problem is connected with counting of Lactobacillus morphotypes, forming
a leptotricha, as long as it is hard to point a place where one morphotype ends and
another one begins. Assessment of smears without any morphotypes at all is also
challenging: such smears get 4 points for absence of Lactobacillus spp. morphotypes
and 0 points, for absence of Gardnerella spp. and Mobiluncus spp. morphotypes,
what gives in the end 4 points and include the smear in intermediate state group
– this not have anything to do with reality.
The evaluation of intermediate state is the most frequent diagnostic problem (Taylor-
Robinson, Morgan, et al. 2003, Bradshaw, Morton et al. 2005). Some researchers
suggest that there is no need for distinguishing intermediate state and a new cut-off
should be established (Rodrigues, Peixoto et al. 2015). Although, this might be
difficult, as intermediate state shares both characteristics of normal microbiota
(Santos-Greatti, da Silva, et al. 2016) and BV (Guédou, Van Damme et al. 2016,
Guédou, Van Damme et al. 2014, Waqqar, Aziz et al. 2017).
Using Nugent scoring system factors such as microscopic image area size (Larsson,
Carlsson et al. 2004), experience of slide reader (Chawla, Bhalla, et al. 2013),
What makes the bacterial vaginosis difficult to diagnose: our experience in Silesia, Poland
127
thickness of sample or subjectivity of assessment of different morphotypes can alter
the result (Forsum, Larsson et al. 2008).
We did not calculate the prevalence of BV, providing that the slides we evaluated
were not randomly selected, because we needed a representative number of smears
with different diagnoses. BV is the most common cause of abnormal vaginal
discharge in women of reproductive age, however, frequency of this condition is
variable between different ethnic and BV risk groups and depends on diagnostic
method used.
At present there are many molecular methods used to detect BV-associated bacteria
(Lannon, Waldorf et al. 2018). In our opinion for BV diagnosis more appropriate is
using clinical (Amsel) or microscopic criteria (Nugent, Hay’s), than PCR based
methods. Not only because of its expensiveness, but also because of too high
sensitivity of PCR methods, especially in cases when physiological microbiota is
present. Further analysis among women with different BV risk is needed to
determine their potential impact on proper BV diagnosis.
CONCLUSION
The borderline diagnoses of vaginal abnormal microbiota (intermediate state) should
be confronted with clinical symptoms and vaginal pH before making final clinical
diagnosis using Nugent scoring system.
REFERENCES
Achilles S. L., Austin M. N., Meyn L. A., Mhlanga F., Chirenje Z. M., Hillier S. L. Impact of
contraceptive initiation on vaginal microbiota. American Journal of Obstetrics and Gynecology.
2018; 18: 30176-5.
Amsel R., Totten P. A., Spiegel C. A., Chen K. C., Eschenbach D., Holmes K. K. Nonspecific
vaginitis. The American Journal of Medicine. 1983; 74(1): 14-22.
Antonio M. A., Hawes S. E., Hillier S. L. The identification of vaginal Lactobacillus species and the
demographic and microbiologic characteristics of women colonized by these species. The Journal
of Infectious Diseases. 1999; 180(6): 1950-1956.
Aslan E., Bechelaghem N. To 'douche' or not to 'douche': hygiene habits may have detrimental
effects on vaginal microbiota. Journal of Obstetrics and Gynaecology. 2018; 13: 1-4
Bautista C. T., Wurapa E., Sateren W. B., Morris S., Hollingsworth B., Sanchez J. L. Bacterial vaginosis:
a synthesis of the literature on etiology, prevalence, risk factors, and relationship with chlamydia and
gonorrhea infections. Military Medical Research. 2016; 3: 4
Bradshaw C. S., Morton A. N., Garland S. M., Horvath L. B., Kuzevska I., Fairley C. K. Evaluation of
a Point-of-Care Test, BVBlue, and Clinical and Laboratory Criteria for Diagnosis of Bacterial Vaginosis. Journal of Clinical Microbiology. 2005; 43(3): 1304-1308.
Brotman R. M., Klebanoff M. A., Nansel T. R., Yu K. F., Andrews W. W., Zhang J., Schwebke J. R.
Bacterial vaginosis assessed by gram stain and diminished colonization resistance to incident
gonococcal, chlamydial, and trichomonal genital infection. The Journal of Infectious Diseases. 2010;
202(12): 1907-1915.
Centers for Disease Control and Prevention, 2015 Sexually Transmitted Diseases Treatment
Guidelines, Bacterial Vaginosis, https://www.cdc.gov/std/tg2015/bv.htm, date of access: 14.03.2018
Patryk Madeja, Olga Smolarek, Agnieszka Mosler, Małgorzata Romanik
128
Chawla R., Bhalla P., Chadha S., Grover S., Garg S. Comparison of Hay’s Criteria with Nugent’s
Scoring System for Diagnosis of Bacterial Vaginosis. BioMed Research International. 2013; 2013:
365194.
Cherpes T. L., Wiesenfeld H. C., Melan M. A., Kant J. A., Cosentino L. A., Meyn L. A., Hillier S. L.
The associations between pelvic inflammatory disease, Trichomonas vaginalis infection, and positive herpes simplex virus type 2 serology. Sexually Transmitted Diseases 2006; 33(12): 747-752.
Cottrell B. H. An Updated Review of of Evidence to Discourage Douching. MCN, The American
Journal of Maternal/Child Nursing. 2010; 35(2): 102-107.
Data from Central Statistical Office of Poland, 2013, Population projection 2014-2050
Data from Central Statistical Office of Poland, 31.12.2016
Forsum U., Larsson P. G., Spiegel C. Scoring vaginal fluid smears for diagnosis of bacterial
vaginosis: need for quality specifications. APMIS: acta pathologica, microbiologica, et immunologica
Scandinavica. 2008; 116(2): 156-159.
Guedou F. A., Van Damme L., Deese J., Crucitti T., Mirembe F., Solomon S., Becker M., Alary M.
Intermediate vaginal flora and bacterial vaginosis are associated with the same factors: findings from an exploratory analysis among female sex workers in Africa and India. Sexually Transmitted
Infections. 2014; 90(2): 161-164.
Guédou F. A., Van Damme L., Mirembe F., Solomon S., Becker M., Deese J., Crucitti T., Alary M.
Intermediate vaginal flora is associated with HIV prevalence as strongly as bacterial vaginosis in a cross-sectional study of participants screened for a randomised controlled trial. Sexually
Transmitted Infections. 2012; 88(7): 545-551.
Hellberg D., Nilsson S., Mårdh P-A. Bacterial vaginosis and smoking. International Journal of STD
and AIDS. 2000; 11(9): 603-606.
Hoffman M. K., Bellad M. B., Charantimath U. S., Kavi A., Nagmoti J. M., Nagmoti M. B., Mallapur
A. A., Katageri G. M., Ramadurg U. Y., Bannale S. G., Revankar A. P., Ganachari M. S., Derman R. J.,
Goudar S. S. A Comparison of Colorimetric Assessment of Vaginal pH with Nugent Score for the
Detection of Bacterial Vaginosis. Infectious Diseases in Obstetrics and Gynecology. 2017; 2017:
1040984.
Centers for Disease Control and Prevention, Sexually Transmitted Diseases (STDs), Bacterial
Vaginosis (BV) Statistics, https://www.cdc.gov/std/bv/stats.htm, date of access: 15.03.2018.
Ison C. A., Hay P. E. Validation of a simplified grading of Gram stained vaginal smears for use in
genitourinary medicine clinics. Sexually Transmitted Infections. 2002; 78(6): 413-415.
Kenyon C., Colebunders R., Crucitti T. The global epidemiology of bacterial vaginosis: a systematic
review. American Journal of Obstetrics and Gynecology. 2013; 209(6): 505-523.
Kim J. M., Park J. P. Probiotics in the Prevention and Treatment of Postmenopausal Vaginal
Infections: Review Article. Journal of Menopausal Medicine. 2017; 23(3): 139-145.
Lannon S. M. R., Adams Waldorf K. M., Fiedler T., Kapur R. P., Agnew K., Rajagopal L., Gravett M.
G., Fredricks D. N. Parallel detection of lactobacillus and bacterial vaginosis-associated bacterial
DNA in the chorioamnion and vagina of pregnant women at term. The Journal of Maternal-Fetal &
Neonatal Medicine. 2018; 12: 1-9
Larsson P., Carlsson B., Fahraeus L., Jakobsson T., Forsum U. Diagnosis of bacterial vaginosis: need
for validation of microscopic image area used for scoring bacterial morphotypes. Sexually
Transmitted Infections. 2004; 80(1): 63-67.
Mania-Pramanik J., Kerkar S. C., Salvi V. S. Bacterial vaginosis: a cause of infertility? International
Journal of STD and AIDS. 2009; 20(11): 778-781.
What makes the bacterial vaginosis difficult to diagnose: our experience in Silesia, Poland
129
Modak T., Arora P., Agnes C., Ray R., Goswami S., Ghosh P., Das N.K. Diagnosis of bacterial
vaginosis in cases of abnormal vaginal discharge: comparison of clinical and microbiological criteria. JIDC The Journal Of Infection In Developing Countries. 2011; 5(5): 353-60.
Nelson D. B., Hanlon A., Hassan S., Britto J., Geifman-Holtzman O., Haggerty C., Fredricks D. N. Preterm labor and bacterial vaginosis-associated bacteria among urban women. Journal of Perinatal Medicine. 2009; 37(2): 130-134.
Nugent R. P., Krohn M. A., Hillier S. L. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. Journal of Clinical Microbiology. 1991; 29(2): 297-301.
Olson K. M., Boohaker L. J., Schwebke J. R., Aslibekyan S., Muzny C. A. Comparisons of vaginal
flora patterns among sexual behaviour groups of women: implications for the pathogenesis of bacterial vaginosis. Sexual Health. 2018; 15(1): 61-67.
Ralph S. G., Rutherford A. J., Wilson J. D. Influence of bacterial vaginosis on conception and miscarriage in the first trimester: cohort study. BMJ: British Medical Journal. 1999; 319(7204): 220-223.
Rodrigues F. S., Peixoto S., Adami F., Alves B. C., Gehrke F. S., Azzalis L. A., Junqueira V. B., Fonseca F. L. Proposal of a new cutoff for Nugent criteria in the diagnosis of bacterial vaginosis. Journal of Microbiological Methods. 2015; 115: 144-146.
Romanik M., Ekiel A., Friedek D., Martirosian G. Can chlamydial cervicitis influence diagnosis of
bacterial vaginosis? Journal of Clinical Microbiology. 2005;43(9): 4914-4915.
Romanik M. K., Ekiel A. M., Tomana L., Martirosian G. The bacterial vaginosis --treatment problems. Wiadomości Lekarskie. 2007; 60(1-2): 64-7
Romanik M., Martirosian G. Frequency, diagnostic criteria and consequences of bacterial vaginosis in pregnant women. Przegląd Epidemiologiczny. 2004; 58(3): 547-553.
Sabour S., Arzanlou M., Vaez H., Rahimi G., Sahebkar A., Khademi F., Prevalence of bacterial vaginosis in pregnant and non-pregnant Iranian women: a systematic review and meta-analysis. Archives of Gynecology and Obstetrics. 2018.
Santos-Greatti M. M. V., da Silva M. G., Ferreira C. S. T., Marconi C. Cervicovaginal cytokines,
sialidase activity and bacterial load in reproductive-aged women with intermediate vaginal flora. Journal of Reproductive Immunology. 2016; 118: 36-41.
Schwebke J. R., Desmond R. A randomized trial of metronidazole in asymptomatic bacterial vaginosis to prevent the acquisition of sexually transmitted diseases. American Journal of Obstetrics and Gynecology. 2007; 196(6): 517.e1-e6.
Spiegel C. A., Amsel R., Holmes K. K. Diagnosis of bacterial vaginosis by direct gram stain of
vaginal fluid. Journal of Clinical Microbiology. 1983; 18(1): 170-177.
Taylor-Robinson D., Morgan D. J., Sheehan M., Rosenstein I. J., Lamont R. F. Relation between
Gram-stain and clinical criteria for diagnosing bacterial vaginosis with special reference to Gram grade II evaluation. International Journal of STD and AIDS. 2003; 14(1): 6-10.
Waqqar S., Aziz S., Shakeel R., Waheed A., Ahmed S.S., Ali A., Khan A.A., Shoaib M. Redox
imbalance correlates with high Nugent score in bacterial vaginosis. The Journal of Obstetrics and Gynaecology Research. 2018; 44(3): 509-517.
Wessels J. M., Lajoie J., Vitali D., Omollo K., Kimani J., Oyugi J., Cheruiyot J., Kimani M., Mungai J. N., Akolo M., Stearns J. C., Surette M. G., Fowke K. R., Kaushic C. Association of high-risk sexual
behaviour with diversity of the vaginal microbiota and abundance of Lactobacillus. PLoS One. 2017; 12(11): e0187612.
Wolrath H., Forsum U., Larsson P. G., Borén H. Analysis of Bacterial Vaginosis-Related Amines in Vaginal Fluid by Gas Chromatography and Mass Spectrometry. Journal of Clinical Microbiology. 2001; 39(11): 4026-4031.
130
Detection of sleepiness in narcolepsy
Renata Plucińska1, Mirosław Łątka
2, Aleksandra Wierzbicka-Wichniak
3, Wojciech
Jernajczyk3, Grzegorz Domański
4
1Warsaw University of Technology, Faculty of Electronics and Information Technology,
Institute of Electronic Systems, Nowowiejska 15/19, 00-665 Warsaw, e-mail:
[email protected]. 2Wrocław University of Science and Technology, Faculty of Fundamental Problems of
Technology, Department of Biomedical Engineering, Wybrzeże Stanisława
Wyspiańskiego 27, 50-370 Wrocław. 3Institute of Psychiatry and Neurology, Department of Neurophysiology, Jana
Sobieskiego 9, 02-957 Warsaw. 4Warsaw University of Technology, Faculty of Electronics and Information Technology,
Institute of Radioelectronics and Multimedia Technology, Nowowiejska 15/19, 00-665
Warsaw.
Abstract
Narcolepsy is a neurological disorder that involves impairment of sleep-wake cycle regulation.
Intermittent sleepiness which may occur at any time is a cardinal symptom of this disease.
In this preliminary study we attempt to differentiate group of narcoleptic patients and healthy subjects,
using the spectral parameters of EEG signal.
Using routine 19-channel EEG recordings (10-20 standard) of 25 narcoleptic patients and 18 healthy
volunteers we found the statistically significant differences in the normalized FFT power of alpha, beta
and theta rhythms. We exploited these differences to correctly identify all patients using neural network
classifier.
Keywords: Digital Signal Processing; EEG; Narcolepsy; MSLT; FFT
Introduction
Sleep disorders with predominant excessive daytime sleepiness are classified
according to the International Classification of Sleep Disorders, 3rd edition (ICSD-3)
as Central disorders of hypersomnolence. One of the best described and known
hypersomnia is narcolepsy (American Academy of Sleep Medicine, 2014).
Narcolepsy is a chronic, neurological disease, caused by hypocretin deficiency in the
brain. Hypocretin (also known as orexin) is a neurotransmitter produced by neurons
localized in the hypothalamus. It is believed that loss of the hypocretin neurons is an
effect of autoimmune process (Savvidou, Knudsen et al. 2013) Narcolepsy is a rare
disease, affecting 0,03-0,16% of population, depending on the ethnic origin. Most of
the cases of narcolepsy are sporadic, only about 5% are familial (Guilleminault,
Mignot et al 1989). The onset is usually within the second decade of life but can be at
any age. During the first years after the onset of first symptoms, there may be
progression of the disease.
Narcolepsy is characterized by the presence of excessive daytime sleepiness with
sleep attacks and clinical manifestation of dissociated rapid eye movement sleep
including cataplexy, hypnagogic hallucinations and sleep paralysis. Other sleep
disturbances, as REM sleep behaviour disorder (RBD), sleep apnea and periodic limb
Detection of sleepiness in narcolepsy
131
movements in sleep are also frequent in narcolepsy (Wierzbicka, Wichniak et al
2006). Cataplexy is one of the most characteristic symptom in narcolepsy,
manifested by sudden decrease of muscle tone, leading to drop of the jaw, head,
bending knees, even falling down and is triggered by emotions, such as laughing,
excitation, amazement. Sleep paralysis is described by the patients as impossibility to
move the limbs during falling asleep or awakening. At the same time dream-like
sensations known as hypnagogic hallucinations may occur.
According to diagnostic criteria of ICSD-3 the diagnosis of narcolepsy is based on
the clinical features and diagnostic examinations and tests. The disease is divided in
two subtypes: narcolepsy type 1 and type 2. Both types require the presence of
3 months of excessive daytime sleepiness (American Academy of Sleep Medicine,
2014). The standard diagnosis recommends performing neurophysiological
procedures, such as polysomnography (PSG) and Multiple Sleep Latency Test
(MSLT), consisting of five scheduled naps during the day. The mean sleep latency
shorter than 8 min. and presence at least two sleep onset REM periods (SOREMPs)
in MSLT meet criteria for narcolepsy (Carskadon, Dement et al 1986). SOREMP
recorded during the PSG is also included into diagnostic criteria. In type 1 there is
clear-cut cataplexy or a proven undetectable amount of hypocretin-1 in the
cerebrospinal fluid (below 110pg/ml). In type 2, there is no cataplexy and no proven
hypocretin deficiency, but the diagnosis is confirmed in a sleep laboratory with
polysomnography and MSLT.
Beside deficiency of hypocretin in cerebral spinal fluid, another biological marker is
frequent in narcolepsy. It is associated with specific alleles of immunological system,
HLA DQB1*0602. These both biological markers are strongly associated with the
occurrence of cataplexy (Krahn, Pankratz et al 2002).
Narcolepsy significantly affects educational skills, work and social life of the
patients, lowers their quality of life, leads to emotional problems and depression
(Wierzbicka, Wichniak et al 2007).
The treatment consists of behavioural methods and pharmacological compounds,
especially stimulating medications.
Routine electroencephalography (EEG) is a short-lasting examination of brain
activity and usually helpful to estimate the vigilance state of the patient. However,
sleepiness is frequently missed by only visual evaluation of the EEG recordings.
Therefore, a novel method of analysing the EEG signal may enable to detect
sleepiness more precisely.
Some studies of alpha wave power in EEG signal and its change related to the status
of the subject were conducted (Benca, Obermeyer et al 1999; Cantero, Atienza et al,
2002). Variety of methods (including classical and model based spectral methods) to
describe state of alertness level from full spectrum EEG recordings were tested
(Subasi, 2005). There are already some automatic methods of detecting drowsiness,
sleepiness and alertness using power spectral density of discrete wavelet transform
followed by neural network (Kiymik, Akin et al 2004) and using multimodal analysis
followed by neural network as well (Garcés Correa, Orosco, et al. 2014). Some
Renata Plucińska, Mirosław Łątka, Aleksandra Wierzbicka-Wichniak,
Wojciech Jernajczyk, Grzegorz Domański
132
studies with detection drowsy driving using Discrete Wavelet Transform and
unsupervised learning through K-means clustering also were carried (Gurudath,
Riley, 2014). Detection of differences between patients responding and not
responding to the treatment for depression using wavelets with a power spectrum
covering the alpha wave range has been studied (Jernajczyk, Gosek et al. 2017).
Recent studies show that spectral analysis of EEG signal in nocturnal poly-
somnography of sleeping patients and control subjects shows statistically significant
differences in theta and delta waves activity as well (Yub, Choi et al 2017).
Aim of the study was to attempt to distinguish narcolepsy patients with clinically
significant sleepiness during the day from the healthy volunteers, using power
analysis of 3 EEG waves (alpha, beta, theta) and then neural network to recognise the
sleepiness.
Subjects and Methods
Subjects
We performed retrospective analysis of EEG recordings of 25 patients with
narcolepsy (20 men and 5 women, mean age 30 ± 8 years). 16 of them were
diagnosed with type 1 narcolepsy the others with type 2. The control group consisted
of 18 healthy volunteers (4 men and 14 women, mean age 26 ± 9 years). The analysis
was approved by the Ethics Committee of Institute of Psychiatry and Neurology in
Warsaw (IPIN).
All EEG recordings were performed at the Department of Neurophysiology of IPIN.
Data were collected using the Braintronics EEG ISO-1032CE electroencephalograph
and Elmiko's EEGDigiTrack software. 19 electrodes were positioned according to 10-
20 standard with the ground electrode at Fpz. The sampling frequency was 250 Hz.
EEG measurements were performed between 2009 and 2016 (patients) and in 2017
(control group).
The MATLAB R2016a programming environment was used for all analyses.
EEG analysis
The block diagram of the EEG data processing algorithm used in this study is shown
in Fig. 1.
The artifacts-free parts were extracted from the 3 sections of the EEG recording: eyes
closed (which preceded hyperventilation and photostimulation tests), hyperventilation
(which preceded photostimulation test) and right after hyperventilation (but before
photostimulation or eyes open). All sections were analysed independently, resulting
in three different approaches. The results were presented in Tab. 1-9.
Every part of every section was divided into a set of EEG segments of length equal to
512 samples (approximately 2 s) for which was calculated the FFT (Lyons, 2010)
power of alpha (8-13 Hz), beta (13-30 Hz) and theta (4-8 Hz) rhythms in each of 19
channels (Augustyniak, 2001; Jus, Jus, 1967). For each subject the number of
extracted segments depends on the length of the artefact-free parts.
Detection of sleepiness in narcolepsy
133
Fig. 1. The EEG data processing algorithm block diagram
For a given rhythm, the power was normalized in a single channel by the sum of FFT
power in all 19 channels. Data filtration was not used, because only Fourier power
spectrum was extracted from the signal.
Afterwards the neural network was used for obtained segments. The task of the
neural network was to determine to which of the two groups the subject belongs.
Depending on the length of artifact-free parts, during the examination of one patient,
several dozens to several hundred segments were obtained. For each segment three
normalized power values were obtained. Calculation of the ratio of the number of
segments classified to one group and segments classified to the second group,
determinates the probability that the subject belongs to a given group.
For each segment 57 values of normalized FFT power (3 values of rhythms X 19
channels) were used for classification of subjects as patients or healthy controls. The
neural network with 57 input neurons, one hidden layer and the output layer with two
neurons was used. The number of neurons in a hidden layer was chosen by trials and
errors (Basheer, Hajmeer, 2000) and for closed eyes it consisted of 11 neurons, for
hyperventilation – 15 and after hyperventilation – 9. The backpropagation algorithm
was used to train a network. The 70% of input vectors were randomly assigned to the
training set. The others were used for testing. Main parameters of neural network
were calculated like in (Kiymik, Akin et al 2004). The task of the neural network was
to detect excessive sleepiness in a given segment. After completion of the training,
the network was used to classify all of testing input vectors. Each subject was
classified as healthy depending on whether the percentage of “awake” segments was
greater than 50%. If this percentage was smaller than 50% the subject was classified
as narcoleptic.
Renata Plucińska, Mirosław Łątka, Aleksandra Wierzbicka-Wichniak,
Wojciech Jernajczyk, Grzegorz Domański
134
Results
The presented tables (Plucińska, 2017) show the effectiveness of identifying two-second segments of EEG recordings (during eyes closed, hyperventilation and right after hyperventilation). For all sections data were averaged after obtaining results for 10 attempts for teaching neural networks.
Segments with closed eyes
Table. 1. Average parameters obtained from teaching neural networks 10 times for segments with closed eyes.
Name of the parameter
Sensitivity False positive rate
Selectivity Accuracy Specificity
Average value 0.95 0.08 0.95 0.93 0.92
Standard deviation 0.012 0.016 0.012 0.003 0.016
For each subject, the average errors of classifying the test data was calculated.
Table. 2. Average errors of the test data for neural networks. Indications of control subjects during closed eyes
Subject number 1 2 3 4 5 6 7 8 9
Average error [%] 2.41 3.39 7.89 19.36 3.62 2.37 32.34 3.81 3.65
Minimum error [%] 1.85 0.00 5.26 12.77 2.13 1.69 27.66 1.59 1.92
Maximum error [%] 3.70 11.86 10.53 31.91 6.38 5.08 40.43 9.52 5.77
Subject number 10 11 12 13 14 15 16 17 18
Average error [%] 0.00 1.67 1.32 21.40 4.57 8.42 6.67 5.20 17.37
Minimum error [%] 0.00 0.00 0.00 16.00 2.86 5.26 2.38 2.00 15.79
Maximum error [%] 0.00 16.67 3.95 26.00 5.71 13.16 9.52 10.00 26.32
Table. 3. Average errors of test data for neural networks. Indications of narcoleptic patients during closed eyed
Patient number 1 2 3 4 5 6 7 8 9
Average error [%] 4.48 2.38 6.43 22.86 10.56 1.62 5.79 3.59 13.93
Minimum error [%] 3.45 0.00 3.57 17.86 5.56 0.00 5.26 2.56 7.14
Maximum error [%] 6.90 7.50 7.14 32.14 19.44 5.41 10.53 7.69 17.86
Patient number 10 11 12 13 14 15 16 17 18
Average error [%] 25.11 7.78 10.36 0.00 0.00 0.00 1.18 1.30 6.50
Minimum error [%] 19.15 3.70 7.14 0.00 0.00 0.00 0.00 0.00 5.00
Maximum error [%] 34.04 14.81 17.86 0.00 0.00 0.00 8.82 1.85 10.00
Patient number 19 20 21 22 23 24 25
Average error [%] 23.96 5.71 7.13 8.98 3.14 1.37 10.98
Minimum error [%] 18.75 3.57 3.96 6.12 1.96 0.00 7.32
Maximum error [%] 27.08 7.14 10.89 16.33 3.92 3.92 14.63
Detection of sleepiness in narcolepsy
135
Tab. 2-3 show that with the 50% threshold controls group and narcoleptic patients
were correctly assigned to their group. The average error for control subject was
8.08% and for narcoleptics 7.40%.
Segments during hyperventilation
The same calculations were made for hyperventilation (Tab. 4-6).
Table 4. Average parameters obtained from teaching neural networks 10 times for segments with
hyperventilation
Name of the
parameter Sensitivity
False positive
rate Selectivity Accuracy Specificity
Average value 0.94 0.09 0.94 0.92 0.91
Standard deviation 0.009 0.009 0.009 0.003 0.009
Table. 5. Average errors of the test data for neural networks. Indications of control subjects during
hyperventilation
Subject number 1 2 3 4 5 6 7 8 9
Average error [%] 10.59 0.00 6.30 4.00 0.00 0.00 48.46 2.19 0.00
Minimum error [%] 5.88 0.00 3.70 4.00 0.00 0.00 42.31 0.00 0.00
Maximum error [%] 17.65 0.00 11.11 4.00 0.00 0.00 53.85 6.25 0.00
Subject number 10 11 12 13 14 15 16 17 18
Average error [%] 0.00 0.00 11.92 36.25 8.08 2.92 0.00 7.69 1.18
Minimum error [%] 0.00 0.00 7.69 29.17 7.69 0.00 0.00 7.69 0.00
Maximum error [%] 0.00 0.00 19.23 41.67 11.54 4.17 0.00 7.69 5.88
Table. 6. Average errors of test data for neural networks. Indications of narcoleptic patients during
hyperventilation
Patient number 1 2 3 4 5 6 7 8
Average error [%] 14.17 0.45 5.56 8.00 0.00 6.25 0.00 0.00
Minimum error [%] 8.33 0.00 0.00 6.67 0.00 6.25 0.00 0.00
Maximum error [%] 16.67 4.55 11.11 13.33 0.00 6.25 0.00 0.00
Patient number 9 10 11 12 13 14 15 16
Average error [%] 0.00 16.00 50.53 0.45 0.00 6.00 0.00 2.86
Minimum error [%] 0.00 16.00 42.11 0.00 0.00 4.00 0.00 0.00
Maximum error [%] 0.00 16.00 57.89 4.55 0.00 8.00 0.00 14.29
Patient number 17 18 19 20 21 22 23 24
Average error [%] 11.33 33.53 14.40 3.50 6.27 0.63 4.55 5.88
Minimum error [%] 6.67 23.53 12.00 0.00 5.88 0.00 4.55 5.88
Maximum error [%] 13.33 41.18 16.00 5.00 7.84 6.25 4.55 5.88
Renata Plucińska, Mirosław Łątka, Aleksandra Wierzbicka-Wichniak,
Wojciech Jernajczyk, Grzegorz Domański
136
The average error for control subjects was 7.75%, and for narcoleptics 7.93%.
Looking only on average error, one narcoleptic patient had more “awake” segments,
which would lead to incorrect classification. Looking only on maximum error, also
one healthy volunteer wouldn’t be assigned to a proper group.
Segments after hyperventilation
Tables (Tab. 7-9) shows that after hyperventilation worse results were obtained.
Table 7. Average parameters obtained from teaching neural networks 10 times for segments after
hyperventilation
Name of the
parameter Sensitivity
False
positive rate Selectivity Accuracy Specificity
Average value 0.92 0.13 0.92 0.89 0.87
Standard deviation 0.016 0.017 0.016 0.004 0.017
Table. 8. Average errors of the test data for neural networks. Indications of control subjects after
hyperventilation
Subject number 1 2 3 4 5 6 7 8 9
Average error [%] 18.52 0.00 3.75 6.00 6.30 3.23 51.07 6.25 16.55
Minimum error [%] 18.52 0.00 3.13 5.00 3.70 3.23 46.43 0.00 10.34
Maximum error [%] 18.52 0.00 6.25 10.00 11.11 3.23 53.57 18.75 24.14
Subject number 10 11 12 13 14 15 16 17 18
Average error [%] 0.00 0.00 23.60 9.09 3.75 0.77 7.22 16.80 16.80
Minimum error [%] 0.00 0.00 16.00 9.09 3.13 0.00 0.00 12.00 12.00
Maximum error [%] 0.00 0.00 28.00 9.09 6.25 7.69 11.11 24.00 20.00
Table. 9. Average errors of test data for neural networks. Indications of narcoleptic patients after
hyperventilation
Patient number 1 2 3 4 5 6 7 8
Average error [%] 15.63 1.05 0.00 6.32 15.00 3.33 0.00 5.56
Minimum error [%] 15.63 0.00 0.00 5.26 14.29 0.00 0.00 5.56
Maximum error [%] 15.63 5.26 0.00 10.53 21.43 5.56 0.00 5.56
Patient number 9 10 11 12 13 14 15 16
Average error [%] 0.00 51.90 20.00 0.00 7.50 14.00 13.64 2.78
Minimum error [%] 0.00 47.62 13.64 0.00 0.00 0.00 9.09 0.00
Maximum error [%] 0.00 57.14 27.27 0.00 8.33 30.00 27.27 5.56
Patient number 17 18 19 20 21 22 23 24
Average error [%] 19.03 20.00 6.67 23.08 0.87 0.00 1.43 16.50
Minimum error [%] 9.68 20.00 6.67 23.08 0.00 0.00 0.00 10.00
Maximum error [%] 22.58 20.00 6.67 23.08 4.35 0.00 14.29 20.00
Detection of sleepiness in narcolepsy
137
The average error for control subjects was 10,54% and for narcoleptics 10,18%. For
both groups one narcoleptic patient and one control subject has the average error
greater than 50%.
Summary of results
The comparison between different sections analysed above was shown in the summary
table (Tab. 10). Better results were obtained during analysis of the eyes closed section.
Table. 10. Summary of results. Comparison between different sections of EEG recordings based on the
maximum error of classification. Positive means patients with narcolepsy
Eyes Closed Hyperventilation After Hyperventilation
True positive 25 23 23
False positive 0 1 1
True negative 18 17 17
False negative 0 1 1
Discussion
The best discrimination between patients and controls was obtained for closed eyes.
In the other conditions there were at least one false positive and one false negative.
The presented results demonstrate that the EEG Fourier power spectrum may be used
to detect pathological sleepiness in narcoleptic patients. Further studies are needed to
validate this approach. We plan to increase the size of both cohorts and explore the
possibility of reducing the number of electrodes used in classification.
References
American Academy of Sleep Medicine. The International Classification of Sleep Disorders,
Diagnostic and coding manual., 3nd ed. IL. 2014; American Academy of Sleep Medicine
(Westchester).
Augustyniak P., Elektroencefalografia. In: Wichur A. (Editor) Przetwarzanie sygnałów
elektrodiagnostycznych. 2001; Uczelniane Wydawnictwa Naukowo-Dydaktyczne (Cracow/Poland);
ISBN: 83-88408-37-2.
Basheer I.A., Hajmeer M. Artifical neural network: fundamentals, computing, design and
application. Journal of Microbiological Methods. 2000; 43, 3-31.
Benca R. M., Obermeyer W. H., Larson C. L., Yun B., Dolski I., Kleist K. D., Weber S. M., Davidson
R. J. EEG alpha power and alpha power asymmetry in sleep and wakefulness. Psychophysiology.
1999; 36(4), 430-436.
Cantero J. L., Atienza M., Salas R. M. Human oscillations in wakefulness, drowsiness period, and
REM sleep: different electroencephalographic phenomena within the alpha band.
Neurophysiologie Clinique-Clinical Neurophysiology. 2002; 32(1), 54-71.
Carskadon M.A., Dement W.C., Mitler M.M., Roth T., Westbrook P. R. Keenan S. Guidelines for the
multiple sleep latency test (MSLT): a standard measure of sleepiness. Sleep. 1986; 9(4): 519-524.
Garcés Correa A., Orosco L., Laciar E. Automatic detection of drowsiness in EEG records based on
multimodal analysis. Medical Engineering and Physics. 2014; 36(2): 244-249.
Guilleminault C., Mignot E., Grumet F.C. Familial patterns of narcolepsy. Lancet. 1989; 2(8676):
1376-1379.
Renata Plucińska, Mirosław Łątka, Aleksandra Wierzbicka-Wichniak,
Wojciech Jernajczyk, Grzegorz Domański
138
Gurudath N., Riley H. B. Drowsy Driving Detection by EEG Analysis Using Wavelet Transform
and K-means Clustering. Procedia Computer Science. 2014; 34, 400-409.
Jernajczyk W., Gosek P., Łątka M., Kozłowska K., Święcicki Ł., West B. J. Alpha Wavelet Power as
Biomarker of Antidepressant Treatment Response in Bipolar Depression. In: Pokorski M. (Editor.),
Advances in experimental medicine and biology, Neuroscience and respiration, Influenza and respiratory care vol. 968. 2017; Springer; ISBN: 978-3-319-51712-4.
Jus K., Jus A., Strukturalne elementy graficzne elektroencefalogramu. In: Melanowska H. D.,
Lisiak Cz. (Editors), Elektroencefalografia kliniczna. 1967; Państwowy Zakład Wydawnictw
Lekarskich (Warsaw); 011-064453-00-0.
Kiymik M. K., Akin M., Subasi A. Automatic recognition of alertness level by using wavelet
transform and artificial neural network. Journal of Neuroscience Methods. 2004; 139(2), 231-240.
Krahn L.E., Pankratz V.S., Oliver L., Boeve B., Silber M. Hypocretin (orexin) levels in cerebrospinal
fluid of patients with narcolepsy: relationship to cataplexy and HLA DQB1*0602 status. Sleep.
2002; 25, 733-736.
Lyons R. G. Dyskretne przekształcenie Fouriera. In: Wprowadzenie do cyfrowego przetwarzania
sygnałów. 2010; Wydawnictwa Komunikacji i Łączności (Warszawa); ISBN: 978-83-206-1764-1.
Lyons R. G. Szybkie przekształcenie Fouriera. In: Wprowadzenie do cyfrowego przetwarzania
sygnałów. 2010; Wydawnictwa Komunikacji i Łączności (Warszawa); ISBN: 978-83-206-1764-1.
Plucińska R. M., Analiza metod cyfrowego przetwarzania sygnałów do efektywnego
diagnnozowania narkolepsji w rutynowym badaniu EEG. 2017, Politechnika Warszawska
(Warszawa).
Savvidou A., Knudsen S., Olsson-Engman M., Gammeltoft S., Jennum P., Palm L. Hypocretin
deficiency develops during onset of human narcolepsy with cataplexy. Sleep. 2013; 36(1): 147-8.
Subasi A. Application of Classical and Model-Based Spectral Methods to Describe the state of
Alertness in EEG. Journal of Medical Systems. 2005; 29(5), 473-486.
Wierzbicka A., Wichniak A., Szatkowska E, Musińska I., Czasak K., Waliniowska E., Jakubczyk T.,
Jernajczyk W. Różnicowanie narkolepsji z innymi hipersomniami na podstawie obrazu klinicznego
i badań neurofizjologicznych. Sen. 2006; 2, 47-53.
Wierzbicka A., Wichniak A., Waliniowska E., Szatkowska E., Czasak K., Musińska I., Jernajczyk W.
Ocena nasilenia objawów klinicznych, stopnia niepełnosprawności oraz jakości życia w
narkolepsji i innych hipersomniach. Sen. 2009; 7(2), 88-93.
Yub S. H., Choi H. D., Seo W. S. Spectral analysis of Polysomnography in Narcolepsy. Psychiatry
Investigation. 2017; 14(2), 193-197.
139
Increased incidence of CNS tumours
in Parkinson Disease
Marcin Cackowski, Aleksandra Rosiek, Jarosław Jóźwiak*
*Center for Biostructure Research, Department of Histology and Embryology,
Medical University of Warsaw, Chałubińskiego Street 5, 02-004 Warsaw, Poland
Abstract
Purpose: Parkin, DJ-1, PINK1, LRRK2 and UCHL1 are neuroprotective proteins which play a key role
in Parkinson’s disease (PD) etiology. Central nervous system (CNS) tumors are more frequent in
patients with PD, hence these proteins may also be involved in oncogenesis. There are several papers
indicating that Parkinson’s-related proteins are dysfunctional in different types of CNS tumors.
Materials and methods: In this paper we are investigating activity of Parkin, DJ-1, PINK1, LRRK2,
UCHL1 in CNS tumors. We review recent publications in the PubMed database using the following
phrases: “Parkin/DJ-1/PINK1/LRRK2/UCHL1 and CNS tumor” and “Parkinson’s disease and CNS
tumor”. We also study the role of these proteins in physiology, PD and oncogenesis, and their relation to
common cell signaling pathways.
Results: Parkin is a tumor suppressor gene (TSG) and its expression is lowered in various CNS tumors.
PINK-1 is induced by another TSG called PTEN, which levels are also decreased in brain lesions. DJ-1 is
an oncogene and its hyperactivity is associated with higher proliferation and lower differentiation of these
tumors. LRRK-2 gene encodes MAPKKK protein, and UCHL-1 is overexpressed in juvenile gliomas.
Conclusion: Parkin, DJ-1, PINK1, LRRK2, UCHL1 may be used as markers of PD, as well as
prognostic factors for CNS tumors. Patients with diagnosed PD should undergo CNS tumor screening
tests more often. Drugs targeting these proteins may be used against both CNS tumors and PD, what
might be beneficial pharmacoeconomically.
Keywords: Parkin, DJ-1, PINK1, LRRK2, UCHL1, CNS tumor, Parkinson disease
Introduction
Parkinson’s disease (PD) is the second most frequent neurodegenerative disorder
(Elbaz, Carcaillon et al. 2016). Major symptoms, such as rigidity, tremor and
akinesia, are caused by the dysfunction of dopaminergic neurons of the substantia
nigra (Lang, Lozano 1998). Levy’s bodies, which are intracellular aggregates of
accumulated α-synuclein (α-SNCA) and ubiquitin, are the most common
pathological findings in PD neurons (Polymeropoulos, Lavedan et al. 1997;
Wakabayashi, Tanji et al. 2007). Overall worldwide prevalence of PD is 315 (95%
CI 113-873) per 100,000 and is correlated with increasing patient age (Pringsheim,
Jette et al. 2014). Central Nervous System tumours are relatively rare lesions – they
represent 1.8% of cancer cases and were responsible for 2.3% of cancer-related
deaths in 2012 worldwide (Ferlay, Soerjomataram et al. 2015). The correlation
between PD and various types of cancer seems to be complex. PD increases the
incidence of several malignancies (e.g. melanoma, thyroid and breast carcinoma),
however it also decreases the frequency of many other cancers (e.g. lung and
bladder) (Inzelberg, Jankovic 2007). Furthermore, CNS tumours are more frequent in
patients with the history of PD, as reported in the metanalysis by Rong Ye and
colleagues. Overall, OR in this study was 1.51 (95%CI 1.21-1.89, p<0.001,
n=329,276). The incidence of tumours detected prior to the diagnosis of Parkinson’s
Marcin Cackowski, Aleksandra Rosiek, Jarosław Jóźwiak
140
disease was not significant, but it was significantly higher after the diagnosis (Ye,
Shen et al. 2016). Neuroprotective proteins (PINK1, Parkin, DJ-1, UCHL1 and
LRRK2) are dysfunctional in both familial and sporadic types of PD and their
mutations are considered to be risk factors (Klein, Schlossmacher 2007).
Additionally, they can act as oncogenes, tumour suppressor genes(TSGs) or can be
involved in multiple cell signalling pathways (Devine, Plun-Favreau et al. 2011)
(Table 1.). This is one of the leading hypotheses explaining why malignancies are
significantly more frequent in PD patients.
Table 1. Summary of PD-related proteins and their contribution to PD and oncogenesis
Gene Physiology Type of PD Role in oncogenesis
Parkin Mitophagy Familial AR TSG
PINK1 Mitophagy Familial AR Induced by PTEN (TSG)
DJ-1 Chaperone Familial AR Oncogene, PTEN inhibitor
LRRK2 Cytoplasmic kinase Familial AD MAPKKK protein
UCHL1 Deubiquitinase Familial AD?a Oncogene/TSG degradationb
TSG – Tumour suppressor gene, AR – Autosomal recessive, AD – Autosomal dominant,
PD – Parkinson disease a function unclear – limited amount of cases
b depending on cancer type
Parkin
Physiologically, parkin (PARK2) acts as an E3 ubiquitin-ligase (Shimura, Hattori et
al. 2000) and promotes autophagy of mitochondria (Chan, Salazar et al. 2011).
So-called mitophagy protects the cell from oxidative stress by depleting damaged
mitochondria (Lemasters, 2005). Ubiquitination regulates proteasomal and lysosomal
protein degradation (Komander, Rape 2012), thus reduced activity of parkin may
result in higher level of reactive oxygen species (ROS) (Xiao, Goh et al. 2017) and
protein accumulation (Shimura, Schlossmacher et al 2001) (Figure 1.). Certain
mutations of parkin gene cause familial autosomal recessive PD with early onset
(Kitada, Asakawa et al. 1998). Furthermore, parkin is a TSG (Cesari, Martin et al.
2003) and its levels are lowered in various malignancies (Inzelberg, Jankovic 2007).
In glioblastoma multiforme cell lines, parkin concentration decreases during hypoxia
and it regulates the expression pattern of hypoxia inducible factors (HIF-1α and
HIF-3α) (Maugeri, D’Amico et al. 2016). HIFs are proteins inducing angiogenesis in
normal tissues and cancers (Semenza 2003). Yeo and colleagues deduce that the
dysfunction of parkin corelates with higher mortality; therefore it can be considered
as a prognostic marker for patients with gliomas (Yeo, Ng et al. 2012). It seems that
p53 cooperates with parkin and their disturbances might result in both CNS
tumorigenesis and development of PD (Checler, Alves da Costa 2014).
Increased incidence of CNS tumours in Parkinson Disease
141
PINK1
PINK1 (PTEN-induced putative kinase 1, also known as PARK6) is a kinase that
functions as a mitochondrial sensor. In normal mitochondria, PINK1 levels are
reduced by PARL protease. However, in damaged cells its concentration is elevated
and PINK1 promotes parkin activation and therefore mitophagy (Youle, van der
Bilek 2012). Nonsense or missense mutations of PINK1 are responsible for
autosomal recessive hereditary PD with early onset (Valente, Abou-Sleiman et al.
2004). Due to its function as parkin activator, the role of PINK1 in oncogenesis and
PD etiology seems to be similar to that of parkin. Nevertheless, PINK1’s contribution
is more complex. It is activated by PTEN, which is one of the most frequently
mutated TSGs (Di Crostofano, Pandolfi 2000). Thus, PTEN abnormalities may also
lead to oxidative stress (Figure 1.). In CNS malignancies, PINK1 seems to act as
a TSG. It decreases growth of glioblastoma and inhibits aerobic glycolysis
(Warburg’s Effect). Loss of PINK1 was associated with low patient survival and was
observed in numerous brain lesions (Agnihothri, Golbourn et al. 2016). Notch in
noncanonical signalling interacts with PINK1, which alters mitochondrial function
and activates mTORC2/Akt pathway in Drosophila and human brain cancer stem
cells (Lee, Wu et al. 2013).
DJ-1
DJ-1 (PARK7) is a mitochondrial chaperone protein which stops α-synuclein
accumulation caused by oxidative stress (Zhou, Zhu et al. 2006). Hence, its
disturbances may lead to PD. Some DJ-1 mutations may result in autosomal
recessive parkinsonism with early-onset (Bonifati, Rizzu et al. 2003). It also protects
cells from heavy-metal induced cytotoxicity (Björkblom, Adilbayeva et al. 2013).
Unlike above-mentioned PD-related proteins, DJ-1 is an oncogene and it is involved
in Ras-related signalling (Nagakubo, Taira et al. 1997). It functions as a negative
regulator of the previously mentioned PTEN protein (Kim, Peters et al., 2005). Both
Ras pathway’s hyperactivity and TSG suppression might result in tumorigenesis
(Figure 1.). DJ-1 expression may be considered a prognostic factor in
medulloblastoma (Lin, Pan et al. 2014), as well as survival marker in astrocytoma
and glioblastoma (Abd El Atti, Abou Gabal et al. 2013). Its nuclear levels correlate
with WHO grading of astrocytomas (Miyajima, Sato et al. 2010).
LRRK2
LRRK2 gene (leucin-rich repeat kinase2 or PARK8) encodes a protein called
dardarin. It is a cytoplasmic kinase which has an important function in vesicular
transport system, protein homeostasis (Bae, Lee 2015) and inflammation (Gardet,
Benita et al. 2010). α-synuclein accumulation and apoptosis are observed in mice
with mutated LRRK2 gene (Tong, Yamaguchi et al. 2010). Mutations of LRRK2 are
responsible for significant number of cases of autosomal dominant type of PD as
well as some sporadic cases. Primarily, dardarin disturbances cause benign late-onset
PD (Zimprich, Biskup et al. 2004). LRRK2 also functions as a MAPKKK protein,
which can explain its role in tumorigenesis (Gloeckner, Schumacher et al. 2009)
(Figure 1.). LRRK2 mutations (especially G2019S) are present in various
Marcin Cackowski, Aleksandra Rosiek, Jarosław Jóźwiak
142
malignancies and they boost the risk of developing these lesions (Agalliu, San
Luciano et al. 2015). LRRK2 is a well-identified target for brain-blood barrier
penetrating inhibitors in innovative concepts of PD treatment (Gunosewoyo, Yu, et
al. 2017). The same substances might be viable candidates for use in brain tumour
therapy. The number of papers concerning CNS tumours and LRRK2 disturbances is
very limited and their role in brain malignancies needs to be investigated.
UCHL1
UCHL1 (ubiquitin carboxy-terminal hydrolase L1 or PARK5) functions as
a deubiquitinase (Larsen, Krantz et al. 1998) and is a brain-specific enzyme (Doran,
Jackson et al. 1983). In particular, it is responsible for α-synuclein degradation (Liu,
Fallon et al. 2002) (Figure 1.). mRNA expression studies show higher concentration
of UCHL-1 in substantia nigra than in other structures of the brain (Solano, Miller et
al. 2000). In one case, a missense UCHL1 mutation was found in a German family
with history of familial PD (Leroy, Boyer et al. 1998). On the other hand, Healy and
colleagues’ metanalysis argues that UCHL-1 mutations don’t cause PD in any
hereditary model (Healy, Abou-Sleiman et al. 2006). UCHL-1 contribution to
familial PD remains unclear and further studies are required. The role of UCHL-1 in
tumorigenesis seems to be complex, since deubiquitinases can be both oncogenes or
TSGs, depending on tissue type (Shi, Grossman 2010). In juvenile high-grade
gliomas UCHL-1 is overexpressed and associated with cancer stem cells population
(Sanchez-Diaz, Chang et al. 2017). The number of studies regarding the role of
UCHL1 in CNS tumours is also very limited and further research is required.
1 depending on cell type
Figure 1. PD-related proteins’ signalling and their role in neuroprotection and tumorigenesis. DJ-1, PINK1
and parkin are related with mitochondrial compartment and LRRK2 and UCHL1 are cytoplasmic proteins
Increased incidence of CNS tumours in Parkinson Disease
143
Conclusion
PD-related proteins may be considered as potential markers in brain lesions, as well
as in sporadic or familial types of PD. Maugeri and colleagues propose parkin
isoforms as candidates for glioma markers (Maugeri, D’Amico et al. 2015). Drug
development targeting parkin, PINK-1, DJ-1, LRRK2 and UCHL1 might be possibly
beneficial for both PD and brain cancer patients, which could be encouraging for
pharmacological companies. Drapalo and colleagues suggest that Parkin, DJ-1 and
PINK1 should be considered as new targets in ganglioglioma treatment (Drapalo,
Jozwiak 2018). Tumour screening tests might be considered for PD (especially
hereditary types) patients more frequently.
Discussion
In this paper we focused on one of several hypotheses explaining why tumours are
more frequent in PD. While PD-related proteins might be dysfunctional in these
lesions, other explanations should be considered. PD patients undergo neurological
examination or radiological imaging scans more often than regular population, which
may explain the positive correlation between the occurrence of brain malignancies
and PD (detection bias). According to some studies, L-DOPA (first line treatment in
PD) may cause oxidative stress (Vatessery, Smith et al. 2006), hence cells utilizing
this molecule (e.g. melanocytes or glia) might be exposed to ROS, commonly
associated with carcinogenesis.
The number of studies concerning PD proteins and CNS tumours is limited and many
more would have to be performed to confirm our suggestions and conclusions.
Research concerning incidence of brain malignancies in families with history of
hereditary types of PD is especially required.
Acknowledgement
Authors declare no conflict of interest.
References
Abd El Atti R. M., Abou Gabal H. H., Osman W. M., Saad A. S. Insights into the prognostic value of
DJ-1 and MIB-1 in astrocytic tumors. Diagnostic Pathology. 2013; 8:126.
Agalliu I., San Luciano M., Mirelman A., Giladi N., Waro B., Aasly J., Inzelberg R., Hassin-Baer S.,
Friedman E., Ruiz-Martinez J., Marti-Masso JF., Orr-Urtreger A., Bressman S., Saunders-Pullman R.
Higher frequency of certain cancers in LRRK2 G2019S mutation carriers with Parkinson disease:
a pooled analysis. JAMA Neurology. 2015; 72(1):58-65.
Agnihotri S., Golbourn B., Huang X., Remke M., Younger S., Cairns R. A., Chalil A., Smith C. A.,
Krumholtz S. L., Mackenzie D., Rakopoulos P., Ramaswamy V., Taccone M. S., Mischel P. S., Fuller
G. N., Hawkins C., Stanford W. L., Taylor M. D., Zadeh G., Rutka J. T. PINK1 Is a Negative
Regulator of Growth and the Warburg Effect in Glioblastoma. Cancer Research. 2016;
76(16):4708-4719
Bae J. R., Lee B. D. Function and dysfunction of leucine-rich repeat kinase 2 (LRRK2):
Parkinson's disease and beyond. BMB Reports. 2015; 48(5):243-248.
Björkblom B., Adilbayeva A., Maple-Grødem J., Piston D., Ökvist M., Xu X. M., Brede C., Larsen J.
P., Møller S. G. Parkinson disease protein DJ-1 binds metals and protects against metal-induced
cytotoxicity. Journal of Biological Chemistry. 2013; 288(31):22809-22820.
Marcin Cackowski, Aleksandra Rosiek, Jarosław Jóźwiak
144
Bonifati V., Rizzu P., van Baren M. J., Schaap O., Breedveld G. J., Krieger E., Dekker M. C., Squitieri
F., Ibanez P., Joosse M., van Dongen J. W., Vanacore N., van Swieten J. C., Brice A., Meco G., van
Duijn C. M., Oostra B. A., Heutink P. Mutations in the DJ-1 gene associated with autosomal
recessive early-onset parkinsonism. Science. 2003; 299(5604):256-259.
Cesari R., Martin E. S., Calin G. A., Pentimalli F., Bichi R., McAdams H., Trapasso F., Drusco A.,
Shimizu M., Masciullo V., d'Andrilli G., Scambia G., Picchio M. C., Alder H., Godwin A. K., Croce C.
M. Parkin, a gene implicated in autosomal recessive juvenile parkinsonism, is a candidate tumor
suppressor gene on chromosome 6q25–q27. Proceedings of the National Academy of Sciences of the
United States of America. 2003; 100(10):5956-5961.
Chan N. C., Salazar A. M., Pham A. H., Sweredoski M. J., Kolawa N. J., Graham R. L. J., Hess S.,
Chan D. C. Broad activation of the ubiquitin-proteasome system by Parkin is critical mitophagy.
Human Molecular Genetics. 2011; 20(9):1726-1737.
Checler F., Alves da Costa C. Interplay between Parkin and p53 Governs a Physiological
Homeostasis That Is Disrupted in Parkinson's Disease and Cerebral Cancer. Neurodegenerative
Diseases. 2014; 13:118-121.
Devine M. J., Plun-Favreau H., Wood N. W., Parkinson’s Disease and cancer: two wars, one front.
Nature Reviews Cancer. 2011; 11:812-823.
Di Crostofano A., Pandolfi P. P. The multiple roles of PTEN in Tumor Supression. Cell. 2000;
100(4):387-390.
Doran J. F., Jackson P., Kynoch P. A. M., Thompson R. J. Isolation of PGP 9.5, a New Human
Neurone-Specific Protein Detected by High-Resolution Two-Dimensional Electrophoresis. Journal
of Neurochemistry. 1983; 40(6):1542-1547.
Drapalo K., Jozwiak J. Parkin, PINK1 and DJ1 as possible modulators of mTOR pathway in
ganglioglioma. The International Journal of Neuroscience. 2018; 128(2):167-174.
Elbaz A., Carcaillon L., Kab S., Moisan F. Epidemiology of Parkinson’s disease. Revue neurologique.
2016; 172(1): 14-26.
Ferlay J., Soerjomataram I., Dikshit R., Eser S., Mathers C., Rebelo M., Parkin D. M., Forman D., Bray
F. Cancer incidence and mortality worldwide: Sources, methods and major patterns in
GLOBOCAN 2012. International Journal of Cancer. 2015; 136(5):E359-E386.
Gardet A., Benita Y., Li C., Sands B. E., Ballester I., Stevens C., Korzenik J. R., Rioux J. D., Daly M.
J., Xavier R. J., Podolsky D. K. LRRK2 is involved in the IFN-gamma response and host response
to pathogens. Journal of Immunology. 2010; 185(9):5577-5585.
Gloeckner C. J., Schumacher A., Boldt K., Ueffing M. The Parkinson disease-associated protein
kinase LRRK2 exhibits MAPKKK activity and phosphorylates MKK3/6 and MKK4/7, in vitro.
Jounral of Neurochemistry. 2009; 109(4):959-968.
Gunosewoyo H., Yu L., Munoz L., Kassiou M. Kinase targets in CNS drug discovery. Future
Medicinal Chemistry. 2017; 9(3):303-314.
Healy D. G., Abou-Sleiman P. M., Casas J. P., Ahmadi K. R., Lynch T., Gandhi S., Muqit M. M.,
Foltynie T., Barker R., Bhatia K. P., Quinn N. P., Lees A. J., Gibson J. M., Holton J. L., Revesz T.,
Goldstein D. B., Wood N. W. UCHL-1 is not a Parkinson's disease susceptibility gene. Annals of
neurology. 2006; 59(4):627-633.
Inzelberg R., Jankovic J. Are Parkinson disease patients protected from some but not all cancers?
Neurology. 2007; 69(15):1542-1550.
Kim R. H., Peters M., Jang Y., Shi W., Pintilie M., Fletcher G. C., DeLuca C., Liepa J., Zhou L., Snow
B., Binari R. C., Manoukian A. S., Bray M. R., Liu F. F., Tsao M. S., Mak T. W. DJ-1, a novel
regulator of the tumor suppressor PTEN. Cancer cell. 2005; 7(3):263-273.
Increased incidence of CNS tumours in Parkinson Disease
145
Kitada T., Asakawa S., Hattori N., Matsumine H., Yamamura Y., Minoshima S., Yokochi M., Mizuno
Y., Shimizu N. Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism.
Nature. 1998; 392:605-608.
Klein C., Schlossmacher M. G. Parkinson disease, 10 years after its genetic revolution: Multiple
clues to complex disorder. Neurology. 2007; 69(22):2093-2104.
Komander D., Rape M. The Ubiquitin Code. Annual Review of Biochemistry. 2012; 81:203-229.
doi:10.1146/annurev-biochem-060310-170328
Lang A. E., Lozano A. M. Parkinson’s Disease first of two parts. The New England Journal of
Medicine. 1998; 339:1044-1053
Larsen C. N., Krantz B. A., Wilkinson K. D. Substrate specifity of deubiquitinating enzymes:
ubiquitin C-terminal hydrolases. Biochemistry. 1998; 37(10):3358-3368.
Lee K. S., Wu Z., Song Y., Mitra S. S., Feroze A. H., Cheshier S. H., Lu B. Roles of PINK1,
mTORC2, and mitochondria in preserving brain tumor-forming stem cells in a noncanonical Notch signaling pathway. Genes & Development. 2013; 27:2642-2647.
Lemasters J. L. Selective Mitochondrial Autophagy, or Mitophagy, as a Targeted Defense Against
Oxidative Stress, Mitochondrial Dysfunction, and Aging. Rejuvenation Research. 2005; 8(1):3-5.
Leroy E., Boyer R., Auburger G., Leube B., Ulm G., Mezey E., Harta G., Brownstein M. J., Jonnalagada
S., Chernova T., Dehejia A., Lavedan C., Gasser T., Steinbach P. J., Wilkinson K. D., Polymeropoulos M.
H. The ubiquitin pathway in Parkinson's disease. Nature. 1998; 395(6701):451-452.
Lin J. P., Pan B. C., Li B., Li Y., Tian X. Y., Li Z. DJ-1 is activated in medulloblastoma and is
associated with cell proliferation and differentiation. World Journal of Surgical Oncology. 2014;
12:373.
Liu Y., Fallon L., Lashuel H. A., Liu Z., Lansbury P. T. Jr. The UCH-L1 gene encodes two opposing
enzymatic activities that affect alpha-synuclein degradation and Parkinson's disease susceptibility.
Cell. 2002; 111(2):209-218.
Maugeri G., D’Amico A. G., Reitano R., Saccone S., Federico C., Cavallaro S., D’Agata V. Parkin
modulates expression of HIF-1α and HIF-3α during hypoxia in gliobastoma-derived cell lines in vitro. Cell and Tissue Research. 2016; 364(3):465-474.
Maugeri G., D'Amico A. G., Magro G., Salvatorelli L., Barbagallo G. M., Saccone S., Drago F.,
Cavallaro S., D'Agata V. Expression profile of parkin isoforms in human gliomas. International
Journal of Oncology. 2015; 47(4):1282-1292..
Miyajima Y., Sato Y., Oka H., Utsuki S., Kondo K., Tanizaki Y., Nagashio R., Tsuchiya B., Okayasu I.,
Fujii K. Prognostic significance of nuclear DJ-1 expression in astrocytoma. Anticancer Research.
2010; 30(1):265-269.
Nagakubo D., Taira T., Kitaura H., Ikeda M., Tamai K., Iguchi-Ariga S. M., Ariga H. DJ-1, a novel
oncogene which transforms mouse NIH3T3 cells in cooperation with ras. Biochemical and
biophysical research communications. 1997; 231(2):509-513.
Polymeropoulos M. H., Lavedan C., Leroy E., Ide S. E., Dehejia A., Dutra A., Pike B., Root H.,
Rubenstein J., Boyer R., Stenroos E. S., Chandrasekharappa S., Athanassiadou A., Papapetropoulos T.,
Johnson W. G., Lazzarini A. M., Duvoisin R. C., Di Iorio G., Golbe L. I., Nussbaum R. L. Mutation in
the alpha-synuclein gene identified in families with Parkinson's disease. Science. 1997;
276(5321):2045-2047
Pringsheim T., Jette N., Frolkis A., Steeves T. D. L. The prevalence of Parkinson’s disease: A
systematic review and meta-analysis. Movement Disorders. 2014; 29(13):1583-1590.
Marcin Cackowski, Aleksandra Rosiek, Jarosław Jóźwiak
146
Sanchez-Diaz P. C., Chang J. C., Moses E. S., Dao T., Chen Y., Hung J. Y. Ubiquitin carboxyl-
terminal esterase L1 (UCHL1) is associated with stem-like cancer cell functions in pediatric high-grade glioma. PLoS ONE. 2017; 12(5):e0176879.
Semenza G. L. Targeting HIF-1 for cancer therapy. Nature Reviews Cancer. 2003; 3(10): 721-732.
Shi D., Grossman S. R. Ubiquitin becomes ubiquitous in cancer: emerging roles of ubiquitin ligases and deubiquitinases in tumorigenesis and as therapeutic targets. Cancer biology & therapy. 2010; 10(8):737-747.
Shimura H., Hattori N., Kubo S., Mizuno Y., Asakawa S., Minoshima S., Shimizu N., Iwai K., Chiba T., Tanaka K., Suzuki T. Familial Parkinson disease gene product, parkin, is a ubiquitin-protein ligase. Nature Genetics. 2000; 25:302-305
Shimura H., Schlossmacher M. G., Hattori N., Frosch M. P., Trockenbacher A., Schneider R., Mizuno Y., Kosik K. S., Selkoe D. J. Ubiquitination of a new form of alpha-synuclein by parkin from human brain: implications for Parkinson’s disease. Science. 2001; 293(5528):263-269.
Solano S. M., Miller D. W., Augood S. J., Young A. B., Penney J. B. Jr. Expression of alpha-
synuclein, parkin, and ubiquitin carboxy-terminal hydrolase L1 mRNA in human brain: genes associated with familial Parkinson's disease. Annals of neurology. 2000; 47(2):201-210.
Tong Y., Yamaguchi H., Giaime E., Boyle S., Kopan R., Kelleher R. J. 3rd, Shen J. Loss of leucine-
rich repeat kinase 2 causes impairment of protein degradation pathways, accumulation of alpha-synuclein, and apoptotic cell death in aged mice. Proceedings of the National Academy of Sciences of United States of America. 2010; 107(21):9879-9884.
Valente E. M., Abou-Sleiman P. M., Caputo V., Muqit M. M. K., Harvey K., Gispert S., Ali Z., Del Turco D., Bentivoglio A. R., Healy D. G., Albanese A., Nussbaum R., Gonzalez-Maldonado R., Deller T., Salvi S., Cortelli P., Gliks W. P., Latchman D. S., Harvey R. J., Dallapiccola B., Auburger G., Wood N. W. Hereditary Early-Onset Parkinson’s Disease Caused by Mutations in PINK1. Science. 2004; 304:1158-1160.
Vatessery G. T., Smith W. E., Quach H. T. Effect of oxidative stress induced by L-dopa on endogenous antioxidants in PC-12 cells. Annals of the New York Academy of Sciences. 2006; 1074:330-336.
Wakabayashi K., Tanji K., Mori F., Takahashi H. The Lewy body in Parkinson's disease: molecules implicated in the formation and degradation of alpha-synuclein aggregates. Neuropathology. 2007; 27(5):494-506.
Xiao B., Goh J. Y., Xiao L., Xian H., Lim K. L., Liou Y. C. Reactive oxygen species trigger
Parkin/PINK1 pathway dependent mitophagy by inducing mitochondrial recruitment of Parkin. Journal of Biological Chemistry. 2017; 292(40):16697-16708.
Ye R., Shen T., Jiang Y., Xu L., Si X., Zhang B. The relationship between Parkinson Disease and Brain Tumor: A Meta-Analysis. PLoS ONE. 2016 11(10): e0164388.
Yeo C. W. S., Ng F. S. L., Chai C., Tan J. M. M., Koh G. R. H., Chong Y. K., Koh L. W. H., Foong C. S. F., Sandanaraj E., Holbrook J. D., Ang B., Takahashi R., Tang C., Lim K. Parkin pathway activation mitigates glioma cell proliferation and predicts patient survival. Cancer Research. 2012; 72(10): 2543-2553.
Youle R. J., van der Bilek A. M., Mitochondrial Fission, Fusion and Stress. Science. 2012; 337(6098):1062-1065.
Zhou W., Zhu M., Wilson M. A., Petsko G. A., Fink L. A. The oxidation state of DJ-1 regulates its chaperone activity towards α-Synuclein. Journal of Molecular Biology. 2006; 356(4):1036-1048.
Zimprich A., Biskup S., Leitner P., Lichtner P., Farrer M., Lincoln S., Kachergus J., Hulihan M., Uitti R. J., Calne D. B., Stoessl A. J., Pfeiffer R. F., Patenge N., Carbajal I. C., Vieregge P., Asmus F., Müller-Myhsok B., Dickson D. W., Meitinger T., Strom T. M., Wszolek Z. K., Gasser T. Mutations in LRRK2 cause autosomal-dominant parkinsonism with pleomorphic pathology. Neuron. 2004; 44(4):601-607.
147
Role of vitamin D in etiology of mood disorders
– literature review
Konrad Majewski, Jarosław Jóźwiak
Department of Histology and Embryology, Medical University of Warsaw
Chałubińskiego Street 5, 02-004, Warsaw
E-mail: [email protected]
Phone number: +48 692 588 107
Abstract
Vitamin D deficiency is widely described as one of the crucial factors impacting the development of
mood disorders, especially depressive disorders. This literature review shows association between
vitamin 25(OH)D concentration in human serum and mood disorders from different points of view,
clustering on several molecular mechanisms.
In order to be included in the review, articles from the PubMed archive had to fit the following criteria:
original articles associated with vitamin D and mood disorder/depression/depressive
disorder/bipolar disorder (in accordance with DSM IV and V)
written in English
published between 1 Nov 2007 and 1 Nov 2017
Several molecular mechanisms of vitamin D effect on mental health were identified. Vitamin D
deficiency during gestation could manifest itself in disproportion in the amount of neurotransmitters in
different parts of cerebrum, compared with control group - offspring of female rats living without UVB
radiation restriction. In an analysis of transcriptome from samples harvested post mortem from patients
suffering from bipolar disorder, a dissimilar molecular structure of the receptor for calcitriol was
observed. Moreover, human orthologue of Caenorhabditis elegans gene ANK3, which mutated copy
could be corelated with a nematodes’ lifespan extension, was reported to be correlated with mood
disorders, while its overexpression is inhibited by vitamin D.
Introduction
Vitamin D3 is fat-soluble isoprenoid which play several roles in regulation of
molecular pathways in human physiology. Receptors for calcitriol, active form of
vitamin D, are found in cell nucleus; complex of receptor and molecule of calcitriol
acts as a transcription factor (Rodwell, Bender et al., 2014). Synthesis of calcitriol
(scheme 1.) starts with degradation of its first provitamin, 7-dehydrocholesterol,
stored in skin, by UVB radiation (Berg, Tymoczko et al., 2014). Lack or restriction
in sunlight connected with four seasonal cycle is tantamount to less intensive UVB
radiation and lower rate of provitamin D photolysis. Next, cholecalciferol procured
during photolysis is transported to the liver and metabolized to 25-hydroxycalciferol
– a chemical compound commonly detected during haemoanalysis. Vitamin
25(OH)D is then transported to kidneys, where metabolic cascade ends up in
production of 1,25-dihydrocalciferol - calcitriol. Hydroxylation catalyzed by
1-hydroxylase is upregulated by parathyroid hormone and downregulated by
elevated level of Ca2+
. Calcitriol, an active form of vitamin D, interconnects with
vitamin D-binding protein in serum (Langer-Gould, Lucas et al., 2018).
Konrad Majewski, Jarosław Jóźwiak
148
Scheme 1. A biosynthesis of calcitriol
Vitamin D3 triggers production of calbindin, a calcium binding protein associated
with absorption of calcium ions from small intestine and connected with enteric
neuron activity (Zetzmann, Strehl et al., 2018). Calbindin is also commonly used as
a marker of mature GABA-ergic neurons in central nervous system (Radic, Frieß et
al., 2017), especially Purkinje cells (Shawn, Palliser et al., 2017), deep layers of the
superior colliculus (Najdzion, 2017) and neurons of oculomotor nucleus (de la Cruz,
Pastor et al., 2017). Due to this fact, prolonged lack of vitamin D has an impact on
pathogenesis of osteoporosis and rachitis – low concentration of calcium ions
triggers osteomalacia. The first reports correlating depression with osteoporosis were
published as early as in mid-sixties.
The minimum optimal concentration of vitamin 25(OH)D in human serum is 30
ng/ml (Bischoff-Ferrari, 2008), while concentration bellow 20 ng/ml is known
as critical for correct bone tissue physiology, especially bone growth and fracture
prevention. The groups crucially susceptible to low level of vitamin D are neonates
(a diet based on mother’s milk), pregnant women (common unbalanced diet) and
elderly people, hospitalized or staying in their homes (lack of UVB radiation).
As supplementation, 1000 UI of calcitriol per day elevates 25-hydroxyvitamin D in
serum above 30 ng/ml in up to 50% of the population.
The association between the level of 25-hydroxyvitamin D and etiology of mood
disorders was extensively studied during the several recent years. In the current paper
we wanted to show a few interesting concepts on molecular effect of vitamin D3
concentration in human serum on emergence of psychiatric manifestations, not only
statistical correlations reported in other studies. According to WHO statistics, mood
disorders, specifically depressive disorders, in the following decades will become the
Role of vitamin D in etiology of mood disorders – literature review
149
one of the most common medical conditions in the world (Kessler, Birnbaum et. al.,
2010; Alonso, Angermeyer et. al., 2004). A fact worth mentioning is that young
people (between 18 and 25 years) are critically impacted – for this group suicide as
a complication of mood disorder is the second most common cause of death (Reddy,
2010).
New insight
In PubMed database, there are few papers showing anything else than only statistical
correlation between lower concentration of 25-hydroxyvitamin D in serum and
elevated risk of depressive disorder or bipolar disorder. During the last years, several
advanced studies based on disturbances of calcitriol-depended molecular pathways
appeared. Lack of vitamin D3 in mother’s organism during gestation could lead to
alteration in the amount of neurotransmitters in different parts of cerebrum (Kesby,
Turner et al., 2017). In later life, those disproportions could hypothetically manifest
themselves in mental illnesses, including mood disorders (Shih-Hsien, Lan-Ting et
al., 2014). Another paper reports correlation of bipolar disorder with different
structure and/or distribution of receptors for vitamin D3 (Chu, Liu et al., 2009).
Furthermore, studies based on genome-wide association (GWAS) between
nematodes’ and human transcriptome lead to the discovery of gene thought to be
strongly correlated with mood disorders. Last but not least, we show the connection
between calcitriol and development of GABA-ergic neurons in several regions of
central nervous system and its role in psychopathology (Oh, Son et al., 2012).
Review and discussion
The first of chosen extensive studies is based on four-years cohort observation on
elderly Chinese men including 629 patients (Chan R., Chan D. et al., 2011). The
importance of this study manifests itself in prolonged time of patients monitoring and
regular repetition of psychological testing. Findings show reversed correlation
between the level of 25-hydroxyvitamin D and depression – the higher level of
vitamin D was present in the patients’ serum, the lover risk of onset of depressive
disorder was observed.
Another study (Jaddou, Batieha et al., 2012), based on nearly ten times greater,
diversified population (n=5640) of Jordanian adults, connected measurement of level
of 25-hydroxyvitamin D with mood evaluation based on Depression Anxiety Stress
Scale 21. The standardized scale stands for greater repeatability of this experiment.
Results allow to approximate hypothetical effect of concentration of vitamin
25(OH)D for onset of depressive disorder – 42.3 ng/ml shows statistically significant
positive correlation with decreasing risk of depression morbidity.
Relatively new hypothesis debates impact of level of vitamin D3 on pathogenesis of
mental illnesses according to central nervous system developmental disorders. Based
on studies on rat population, deficiency in vitamin D concentration in motherly
serum during growth of embryos alters multiple neurotransmitter systems in the
neonatal cerebrum (Kesby, Turner et al., 2017). In this experiment, before
insemination, for six weeks prior to gestation female rats were cut off from UVB
radiation (inducing photolysis of provitamin D) and food rich in vitamin D, in order
Konrad Majewski, Jarosław Jóźwiak
150
to eliminate reserves of calcitriol in mother’s body. After confinement, younglings
were sacrificed and their brain tissue was analyzed with high performance liquid
chromatography to evaluate disruption in the level of neurotransmitters in different
parts of cerebrum. The most significant changes are shown in table 1.
Table 1 Changes of the greatest magnitude presented in findings of Keby, Turner et al. studies.
Neurotransmitter Brain region Character of change
in level of neurotransmitter
3-methoxytyramine (3-MT) Basal ganglia Increase
Noradrenaline Hippocampus Increase
Serotonin Caudate putamen Decrease
5-hydroxyindoleacetic acid Medial prefrontal cortex Decrease
Glutamine Thalamus Decrease
Serine Medial prefrontal cortex, basal
ganglia, midbrain, cerebellum Increase
As we can see, deficiency of calcitriol during gestation can manifest in disruption of
several neurotransmitters. The most crucial neurotransmitters for mood disorders are
serotonin (5-hydroxytriptamine) and noradrenaline; one of the most popular
conception stands for major role of serotonin in progression of major depressive
disorder (Cowen, 2008). Furthermore, pathological concentration of noradrenaline in
hippocampus could lead to disruption in physiology of limbic system, commonly
known to be responsible for emotional reactions. This hypothesis is associated with
GABA-ergic calbindin-depended disorders described later in this article.
Another concept of molecular impact of calcitriol on mood disorder is connected
with different structure of nuclear receptor for vitamin D3 (Chu, Liu et al., 2009). By
analyzing transcriptome samples harvested post mortem from brain tissue of suicidal
patients suffering from bipolar disorder, unique properties of those receptors and
calcium signaling pathway enzymes have been discovered. Unfortunately, there is no
information on direct impact of mentioned unusual structure of receptors on
psychopathology of mood disorders; comparatively small sample population (only 10
samples from bipolar patients) increases requirement for additional studies.
Uncommon research based on an experiment on Caenorhabditis elegans
transcriptome (Rangaraju, Levey et al., 2016) suggests that there is at least one gene
strongly associated with mood disorders. Differential gene expression was analyzed
for nematodes treated with mianserin, antagonist of 2-adrenergic receptor,
commonly used in the treatment of depression as atypical antidepressant (Carman,
Ahdieh et al., 2008). Next, genes ranked based on p-value distribution were classified
via Bonferroni correction to find the gene most affected by mianserin. In parallel,
samples of human transcriptome were taken from cohort of elderly patients who
suffer from depressive symptoms, including suicidal patients (45 of 737 patients).
Genome-wide association study allowed to find an orthologue of nematode gene in
Role of vitamin D in etiology of mood disorders – literature review
151
human genome based on single nucleotide polymorphism. ANK3, gene most strongly
affected by the use of mianserin, has different role in the human and nematode
organism. C. elegans with mutation of ank3 show no response to mianserin and their
lifespan is prolonged. In humans, cellular expression of ANK3 fluctuates during life
– concentration of ANK3 transcript shows negative correlation with aging.
Overexpression of ANK3 was found in samples from patients suffering from mood
disorders, usually major depressive disorder. Rangaraju et al. suggested that vitamin
D3 could be one of the factors that negatively affect expression of ANK3, which
could be correlated with decreased risk of occurrence of signs and symptoms
characteristic for mood disorders. Moreover, scientists predict that the product of
ANK3 could be a blood marker for mood disorders in the future.
In another paper, in dorsolateral prefrontal cortex harvested post mortem from
patients suffering from major depressive disorder, lower density of glial cells and
GABA-ergic neurons were observed (Oh, Son et al., 2012). One of the markers of
GABA-ergic neurons, along with calretinin and parvalbumin, is calbindin. In brains
of mouse embryos, the level of calbindin-positive neurons usually increases before
birth (Dávila, Real et al., 2005). Due to direct influence of vitamin D3 on expression
of gene for calbindin it is possible that calcitriol could affect central nervous system
by the development of GABA-ergic neurons; disturbed proportion of excitatory and
inhibitory nervous cells could lead to many psychiatric manifestations, including
mood disorders (Luscher, Shen et al., 2011).
It is interesting that differences between physiology of male and female amygdalae,
a part of limbic system responsible among others for emotional reactions, memory
correlated with emotions and sexual behavior, could also be associated with
differences in expression of calcium-binding proteins, including calbindin (Równiak,
2017), which could explain the hypothetical connection of calbindin (and indirectly
also vitamin D) with mood disorder. However, some studies show that expression of
calbindin in brain is vitamin D3-indepdendent (Hall, Norman, 1991).
At this point, we are able to distinguish three planes of hypothetical molecular impact
of vitamin D3 on the development of mood disorder:
I Depletion of vitamin D3 directly before or during gestation could disturb
proportion of neurotransmitters or neurons performing different roles in central
nervous system. Only additional data could show if those changes are
maintained during the adult life.
II For an unknown reason, calcitriol nucleus receptors in brain tissue of patients
suffering from mood disorders are not able to perform their role correctly or
result from disturbed proportion of calcitriol in the maternal organism
– additional studies could confirm one of these hypotheses in the future.
III One or many genes could be directly engaged in etiology of mood disorders.
Additional data have to confirm character of this molecular impact; at this
moment the role of connection between the newly discovered ANK3 gene with
psychiatric manifestations is based on statistically significant correlation.
Konrad Majewski, Jarosław Jóźwiak
152
An additional problem of studies on molecular character of calcitriol on human
central nervous system was reported recently. It is obvious that psychiatric
disturbances could be observed only in living patients; molecular pathway
disturbance analyzed only with samples taken post mortem exclude the possibility for
running an additional psychological examination, which is important from the point
of view of standardization.
Disturbed molecular pathways during gestation do not resolve the problem of
increased psychiatric disturbances characteristic for seasonal mood disorders – it is
not clear if there is any impact of quarterly disturbance of photolysis of
7-dehydrocholesterol rather than psychological effect of lack of light. Studies on
correlation of vitamin D3 with another nosological entity could be also helpful.
To sum up, all above hypotheses on molecular character of impact of calcitriol on
mood disorders necessitate further studies. Only additional data could answer the
question if calcitriol has indeed impact on several molecular mechanisms whose
malfunction leads to mood disorder, or maybe lower level of 25-hydroxyvitamin D
has only statistical correlation with increased risk of depressive disorder morbidity.
References Alonso J., Angermeyer M.C., Bernert S., Bruffaerts R., Brugha T.S., Bryson H., de Girolamo G., Graaf R., Demyttenaere K., Gasquet I., Haro J.M., Katz S.J., Kessler R.C., Kovess V., Lépine J.P., Ormel J., Polidori G., Russo L.J., Vilagut G., Almansa J., Arbabzadeh-Bouchez S., Autonell J., Bernal M., Buist-Bouwman M.A., Codony M., Domingo-Salvany A., Ferrer M., Joo S.S., Martínez-Alonso M., Matschinger H., Mazzi F., Morgan Z., Morosini P., Palacín C., Romera B., Taub N., Vollebergh W.A.; ESEMeD/MHEDEA 2000 Investigators, European Study of the Epidemiology of Mental Disorders (ESEMeD) Project. Disability and quality of life impact of mental disorders in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta Psychiatrica Scandinavica Supplementum. 2004;(420):38-46.
The Biosynthesis of Membrane Lipids and Steroids in Berg J. M., Tymoczko J. L., Gatto J. G. jr, Stryer L. Biochemistry. 2015, WH Freeman Publisher; ISBN: 9781137563453
Bischoff-Ferrari H.A. Optimal serum 25-hydroxyvitamin D levels for multiple health outcomes. Advances in Experimental Medicine and Bioliology. 2008; 624:55-71.
Botham K. M., Mayes P. A. Lipids of Physiologic Significance. In Rodwell V.W., Bender D., Botham K. M., Kennelly P. J., Weil P. A. Harpers Illustrated Biochemistry 30th Edition 2014, Lange Publisher; ISBN: 9780071825344
Carman J.S., Ahdieh H., Wyatt-Knowles E., Warga E., Panagides J. A controlled studyof mianserin in moderately to severely depressed outpatients. Psychopharmacol Bulletin. 1991; 27(2):135-9.
Chan R., Chan D., Woo J., Ohlsson C., Mellström D., Kwok T., Leung P. Association between serum 25-hydroxyvitamin D and psychological health in older Chinese men in a cohort study. Journal of Affective Disorders. 2011; 130, 251-259.
Chu T.T., Liu Y., Kemether E. Thalamic transcriptome screening in three psychiatric states. Journal of Human Genetics. 2009; 54: 665-675.
Cowen PJ. Serotonin and depression: pathophysiological mechanism or marketing myth? Trends in Pharmacological Science 2008 Sep;29(9):433-6.
de la Cruz R. R., Pastor A. M., Martińez-Guijarro F. J., López-García C., Delgado-García J. M. Localization of parvalbumin, calretinin, and calbindin D-28k in identified extraocular motoneurons and internuclear neurons of the cat. Journal of Comparative Neurology. 1998; 390(3):377-91.
Role of vitamin D in etiology of mood disorders – literature review
153
Dávila J.C., Real M.A., Olmos L., Legaz I., Medina L., Guirado S. Embryonic and postnatal
development of GABA, calbindin, calretinin, and parvalbumin in the mouse claustral complex. Journal of Comparative Neurology. 2005; 481(1):42-57.
Hall AK, Norman AW. Vitamin D-independent expression of chick brain calbindin-D28K. Molecular Brain Research. 1991 Jan;9(1-2):9-14.
Jaddou H. Y., Batieha A. M., Khader Y. S., Kanaan S. H., El-Khateeb M. S., Ajlouni K. M. Depression
is associated with low levels of 25-hydroxyvitamin D among Jordanian adults: result from a national population survey. European Archive of Psychiatry and Clinical Neurosciences. 2012; 262: 321-327.
Kesby J. T., Turner K. M., Alexander S., Eyles D. W., McGrath J. J., Burne T. H. J. Developmental
vitamin D deficiency alters multiple neurotransmitter system in the neonatal rat brain. International Journal of Developmental Neuroscience. 2017; 62: 1-7.
Kessler R.C., Birnbaum H.G., Shahly V., Bromet E., Hwang I., McLaughlin K.A., Sampson N., Andrade L.H., de Girolamo G., Demyttenaere K., Haro J.M., Karam A.N., Kostyuchenko S., Kovess V., Lara C., Levinson D., Matschinger H., Nakane Y., Browne M.O., Ormel J., Posada-Villa J., Sagar R., Stein D.J. Age differences in the prevalence and co-morbidity of DSM-IV major depressive episodes: results from the WHO World Mental Health Survey Initiative. Depression and Anxiety. 2010; 27(4):351-64.
Langer-Gould A., Lucas R. M., Xiang A. H., Wu J., Chen L. H., Gonzales E., Haraszti S., Smith J.B., Quach H., Barcellos L. F. Vitamin D-Binding Protein Polymorphisms, 25-Hydroxyvitamin D, Sunshine and Multiple Sclerosis. Nutrients. 2018; 7;10(2).
Luscher B., Shen Q., Sahir N. The GABAergic Deficit Hypothesis of Major Depressive Disorder Molecular Psychiatry. 2011; 16(4): 383-406.
Najdzion J. Cocaine- and amphetamine-regulated transcript peptide and calcium binding proteins
immunoreactivity in the deep layers of the superior colliculus of the guinea pig: Implications for multisensory and visuomotor processing. Journal of Chemical Neuroanatomy. 2017;88:55-69.
Oh DH, Son H, Hwang S, Kim SH. Neuropathological abnormalities of astrocytes, GABAergic
neurons, and pyramidal neurons in the dorsolateral prefrontal cortices of patients with major depressive disorder. European Neuropsychopharmacology. 2012; (5):330-8.
Radic T., Frieß L., Vijikumar A., Jungenitz T., Deller T., Schwarzacher S.W. Differential Postnatal Expression of Neuronal Maturation Markers in the Dentate Gyrus of Mice and Rats. Frontiers in Neuroanatomy 2017; 11: 104.
Rangajaru S., Levey D. F., Nho K., Jain N., Andrews K. D., Le-Niculescu H., Slomon D. R., Saykin A. J., Petrascheck M., Niculescu A. B. Mood stress and longevity: convergence on ANK3. Molecular Psychiatry. 2016; 1-13.
Reddy M. S. Depression: The Disorder and the Burden Indian Journal of Psychological Medicine. 2010; (1):1-2
Równiak M. The neurons expressing calcium-binding proteins in the amygdala of the guinea pig: precisely designed interface for sex hormones. Brain Structure and Function. 2017; 222(8): 3775-3793.
Shaw JC, Palliser HK, Dyson RM, Berry MJ, Hirst JJ. Disruptions to the cerebellar GABAergic system in juvenile guinea pigs following preterm birth. International Journal of Developmental Neuroscience. 2017; 65: 1-10.
Shih-Hsien L., Lang-Ting L., Yen Kuang Y. Serotonin and Mental Disorders: A Concise Review on Molecular Neuroimaging Evidence Clinical Psychopharmacology and Neuroscience. 2014; 12(3): 196-202
Zetzmann K., Strehl J., Geppert C., Kuerten S., Jabari S., Brehmer A. Calbindin D28k-Immunoreactivity in Human Enteric Neurons. International Journal of Molecular Science. 2018 Jan 8;19(1).
154
Index of authors
Badowska-Kozakiewicz A. M. .......... 62
Biały Ł. P. ............................... 22, 36, 48
Brzeziańska-Lasota E. ........................ 92
Cackowski M. .................................. 139
Czarnocki Z. ....................................... 48
Czekaj P. ........................................... 109
Domański G. .................................... 130
Jernajczyk W. ................................... 130
Jóźwiak J. ................................. 139, 147
Kaszkowiak M. .................................. 92
Kobiela T. ............................................. 9
Król M. ............................................. 109
Limanówka Ł. .................................. 109
Lisiecki K. .......................................... 48
Liszcz A. ............................................. 62
Łątka M. ........................................... 130
Madeja P. .......................................... 119
Majewski K. ..................................... 147
Maksimiuk M. .................................... 62
Młynarczuk-Biały I. ................ 22, 36, 48
Mosler A. .......................................... 119
Pastuszak-Lewandoska D. .................. 92
Patera J. ............................................... 62
Plucińska R. ...................................... 130
Romanik M. ...................................... 119
Rosiek A. ........................................... 139
Smolarek O. ...................................... 119
Sobiborowicz A. ................................. 62
Sobiepanek A. ....................................... 9
Sobieraj M. .......................................... 62
Sobol M. .............................................. 62
Strus P. ................................................ 48
Szmyd B. ............................................. 92
Szmytkowska P. ................................ 109
Szymański P. ....................................... 36
Wawrzuta D. ..................................... 101
Wierzbicka-Wichniak A. .................. 130
Zaręba Ł. ............................................. 22