Embed Size (px)
Citation preview
COMPARATIVE APPROACHES TO CHARACTERIZATION OF
LYMPHATIC ENDOTHELIAL CELLS AS
PHENOTYPICALLY DISTINCT FROM BLOOD ENDOTHELIAL CELLS
by
Victoria (Vicky) Phương Kim Hoàng Nguyễn
A thesis submitted in conformity with the requirements
for the degree of Doctor of Philosophy
Graduate Department of Medical Biophysics
University of Toronto
© Copyright by Victoria PKH Nguyen (2010)
ii
COMPARATIVE APPROACHES TO CHARACTERIZATION OF LYMPHATIC ENDOTHELIAL CELLS
AS PHENOTYPICALLY DISTINCT FROM BLOOD ENDOTHELIAL CELLS
Victoria P. K. H. Nguyen
Doctor of Philosophy
Department of Medical Biophysics, University of Toronto
2010
ABSTRACT
The lymphatic system complements the blood circulatory system in absorption and
transport of nutrients, and in the maintenance of homeostasis. Historically, the angiogenesis
field has advanced faster and farther than the field of lymphangiogenesis. The discovery of
lymphatic markers and the emerging evidence implicating the lymphatic system as a central
player in a variety of pathological conditions has attracted research interest and driven the field
forward. Research efforts have produced the observation that regulators of the blood
endothelium are frequently members of the same protein families of regulators of the lymphatic
endothelium. More importantly, these regulators do not act discretely, restricting their
regulatory activities to one endothelial cell (EC) type. Two examples of regulators that behave
in this manner are the VEGF and the Angiopoietin families of proteins, which have cell-type-
dependent effects on EC processes such as migration, proliferation and survival. The study of
these regulators therefore requires an in vitro EC system capable of accommodating the
simultaneous characterization of the signaling pathways downstream of these shared molecular
regulators in venous, arterial and lymphatic endotheliums. To build such an in vitro system, I
isolated and validated lymphatic, venous, and arterial ECs derived from vessels of bovine
mesentery. The proteomes of the three cell types were comparatively studied using two-
iii
dimensional polyacrylamide gel electrophoresis followed by mass spectrometric identification.
The three cell types were used in a subtractive immunization scheme for the production of a
monoclonal antibody selectively reactive to a potentially novel surface protein marker of
lymphatic ECs. The studies recorded herein all share the common goal of identifying and
characterizing unique molecular signatures that distinguish lymphatic ECs from blood ECs, and
that may underline the cellular biology of the lymphatic endothelium as distinct from the blood
endothelium.
iv
ACKNOWLEDGEMENTS
I would first like to thank my mentor Dr. Daniel J. Dumont for his unwearied support,
and for his unmatched professional guidance so that I may become the best physician and
scientist I can be. Many sincere thanks to my committee members Dr. Mike Moran, for
expecting the best from me, and Dr. Dwayne Barber for always believing in me. Dr. Edwin E.
Daniel helped me to discover research as a worthy life-long pursuit. Dr. Peter F. Whyte was a
patient and kind teacher who inspired me to strive for excellence in teaching. Dr. Eric Yang
was my tireless tutor in the intricacies of proteomics. Many special thanks to Stephen, Paul,
Harold, Jamie, Maribelle, and Sue for their supportive friendship. Lani and Farhah deserve
praise for their unconditional affection. A very large thank you goes out to my very large
extended family for always being more than uncles, aunts, and cousins, but the very best of
friends. I am forever indebted to my uncle Hoàng Trọng Phú who brought my family to this
wonderful country and opened up a whole world of possibility for my future. I know that my
beloved brothers Andy and Tim, and my dear father, Nguyễn Văn Kim will always be on my
side. To Feroz Sarkari, I want to give thanks for changing my life, for always knowing what
really matters in life, and for being my compass and my anchor in rough seas. I will always
remember with the deepest gratitude my late grandfather Hoàng Trọng Phụ for his Vietnamese
nick name for me, Như ý, reminding me that I am already as he wished and that I am already
‗good enough‘, for his grand vision of what life should be for his children and grandchildren,
and for the ultimate sacrifice he paid so that they would have a shot at the life he never had.
Most importantly, I want to thank my mother, Hoàng Thị Mỹ Phương, for all her motherly
deeds, great and small, for her true unchanging, everlasting love, for her wisdom and goodness,
and determination to fulfill her father‘s dreams so that his sacrifice may not be in vain.
v
TABLE OF CONTENTS
ABSTRACT .................................................................................................................................. II
ACKNOWLEDGEMENTS ........................................................................................................IV
TABLE OF CONTENTS ............................................................................................................. V
TABLE OF ILLUSTRATIONS .................................................................................................IX
ABBREVIATIONS ...................................................................................................................... X
CHAPTER 1: BACKGROUND ....................................................................................................... 1
1.1. Rationale and Objective ..................................................................................................... 2
1.2. Introduction to Lymphatic Anatomy ................................................................................ 5
1.2.1. Lymphatic Anatomy and Function ................................................................................ 5
1.2.2. Lymphatic Endothelium Function in Immunity ............................................................ 7
1.2.3. Lymphatic Endothelium Function in Homeostasis ........................................................ 8
1.2.4. Lymphatic Endothelium Dysfunction in Lymphoedema ............................................. 10
1.2.5. Lymphatic Endothelium Dysfunction in other Human Pathologies ............................ 13
1.2.6. Protein Markers of the Lymphatic Endothelium.......................................................... 15
1.3. Data Chapter Organization ............................................................................................. 19
1.4. Attributions ....................................................................................................................... 20
CHAPTER 2: LYMPHATIC, VENOUS AND ARTERIAL ENDOTHELIAL CELLS ISOLATED FROM
BOVINE MESENTERIC VESSELS RETAIN THEIR DEFINING CHARACTERISTICS IN CULTURE
AND ARE DIFFERENTIALLY RESPONSIVE TO ANGIOPOIETIN-1 AND -2 .....................................21
2.1. Introduction ...................................................................................................................... 22
2.2. Methods ............................................................................................................................. 24
vi
2.2.1. BmEC isolation and culture ......................................................................................... 24
2.2.2. Eliminating contaminating cells .................................................................................. 25
2.2.3. Detection of transcripts by RT-PCR ............................................................................ 25
2.2.4. Cell imaging ................................................................................................................. 26
2.2.5. Stimulation of bovine ECs ........................................................................................... 26
2.2.6. Cell Lysis and Tie-2 Immunoprecipitation .................................................................. 27
2.2.7. Immunoblotting ............................................................................................................ 27
2.2.8. 3H-Thymidine uptake assay ......................................................................................... 27
2.2.9. Cell count assay ........................................................................................................... 28
2.2.10. Modified Boyden Chamber Migration Assay ............................................................ 28
2.2.11. Cell death ELISA ....................................................................................................... 29
2.3. Results ................................................................................................................................ 29
2.3.1. BmLEC isolation and culture. ...................................................................................... 29
2.3.2. Comparing bmLECs to bmVECs, and to bmAECs. .................................................... 33
2.3.3. Activation of Tie-2 in bmECs by Ang-2 ...................................................................... 39
2.3.4. BmLEC proliferation is enhanced by Ang-1 and Ang-2 ............................................. 45
2.3.5. Ang-1 promotes bmEC migration ................................................................................ 46
2.3.6. Ang-2 protects bmLEC from cell death ....................................................................... 46
2.4. Discussion .......................................................................................................................... 56
CHAPTER 3: DIFFERENTIAL PROTEOMIC ANALYSIS OF LYMPHATIC, VENOUS, AND
ARTERIAL ENDOTHELIAL CELLS EXTRACTED FROM BOVINE MESENTERIC VESSELS ............. 59
3.1. Introduction ...................................................................................................................... 60
3.2. Methods ............................................................................................................................. 63
vii
3.2.1. Cell isolation and culture ............................................................................................. 63
3.2.2. 2D-PAGE ..................................................................................................................... 63
3.2.3. Mass Spectrometric Analysis of Protein Spot Intensities ............................................ 64
3.2.4. Validation of Protein Level Differences by Immunoblotting ...................................... 64
3.3. Results ................................................................................................................................ 65
3.3.1. Quantitative Analysis of 2D-PAGE-Resolved bmEC Proteins ................................... 65
3.3.2. Mass Spectrometric Analysis of 2D-PAGE-Resolved bmEC Proteins ....................... 71
3.3.3. Validation of Identified Proteins by Immunoblotting .................................................. 83
3.3.4. Comparing Protein Levels of Different Family Members in the same ECs ................ 84
3.3.5. Comparing Protein Levels in ECs of Dissimilar Anatomical Origin .......................... 84
3.4. Discussion .......................................................................................................................... 92
CHAPTER 4: GENERATION OF A MONOCLONAL ANTIBODY SELECTIVELY REACTIVE TO
BOVINE MESENTERIC LYMPHATIC ENDOTHELIAL CELLS BUT NOT TO BOVINE MESENTERIC
VENOUS OR ARTERIAL ENDOTHELIAL CELLS ............................................................................. 98
4.1. Introduction ...................................................................................................................... 99
4.2. Methods ........................................................................................................................... 100
4.2.1 Endothelial Cell Culture.............................................................................................. 100
4.2.2. Subtractive Immunization and Hybridoma Production ............................................. 101
4.2.3. Antibody Screening and Isotyping ............................................................................. 101
4.2.4. Immunoblot Analysis ................................................................................................. 102
4.2.5. Immunohistochemistry .............................................................................................. 103
4.2.6. ULSP180 enrichment by sucrose-gradient-based cell fractionation .......................... 103
viii
4.2.7. Immunoprecipitation and Identification of ULSP180 by LC-MS/MS ...................... 104
4.3. Results .............................................................................................................................. 106
4.3.1 Hybridoma Production and Screening ........................................................................ 106
4.3.2. MAb isotyping ........................................................................................................... 112
4.3.3. MAb is Selective for bmLECs ................................................................................... 112
4.3.4. 8C8 mAb recognizes a protein of MW approximately 180kDa ................................ 117
4.3. 5. Mass Spectrometric Analysis of ULSP180 ............................................................... 123
4.4. Discussion ........................................................................................................................ 131
CHAPTER 5: CONCLUSIONS AND FUTURE DIRECTIONS ........................................................ 136
REFERENCES .......................................................................................................................... 142
ix
TABLE OF ILLUSTRATIONS
Illustration
Page
Figure 2.1. Isolation of bmLECs. 31-32
Figure 2.2. Similarities and differences between bmLECs, bmVECs and
bmAECs. 35-36
Figure 2.3. BmLECs, bmVECs and bmAECs express cell-lineage-specific
markers. 37-38
Figure 2.4. Alignment of bovine, murine, and human Tie-2 amino acid
sequences. 41-42
Figure 2.5. Stimulation of Tie-2 phosphorylation in bmECs by Ang-1 and Ang-2. 43-44
Figure 2.6. Proliferation of Ang-1 and Ang-2 stimulated bmECs. 48-49
Figure 2.7. Migration of Ang-1 and Ang-2 stimulated bmECs. 50-51
Figure 2.8. Survival of Ang-1 and Ang-2 stimulated bmECs. 52-53
Table 2.1. Summary of results. 54-55
Figure 3.1. 2D-PAGE separation of bmEC proteins. 67-68
Figure 3.2. Quantitative Analysis of 2D-PAGE-Resolved bmEC Proteins. 69-70
Table 3.1.A. Summary of proteins identified. 73-76
Table 3.1.B. Summary of biological information on protein identified. 77-80
Figure 3.3. Biological Functions of Proteins Identified in Table 4.1. 81-82
Figure 3.4. Validation of HSPA1B, HSPB1, and UBE2D3 protein levels in the
three cell types by IB. 86-87
Figure 3.5. Determination of HSPB2 protein levels in the three cell types by IB 88-89
Figure 3.6. Comparing ANAX2 Protein Levels in ECs of Dissimilar Anatomical
Origin. 90-91
Figure 4.1. Screening Assay Validation. 108-109
Figure 4.2. Serum and Hybridoma Screening. 110-111
Figure 4.3. MAb Isotyping. 113-114
Figure 4.4. Antibody Validation. 115-116
Figure 4.5. ULSP180 Characterization. 119-120
Figure 4.6. Schema of Mass Spectrometric Analysis of ULSP180. 121-122
Figure 4.7. Enrichment and Immunoprecipitation of ULSP180. 125-126
Figure 4.8. Mass Spectrometric Analysis of ULSP180. 127-128
Figure 4.9. Mass Spectrometric Analysis of ULSP180 (continued). 129-130
x
ABBREVIATIONS
Ab – Antibody
ACh - Acetylcholine
ANAX – Annexin
Ang – Angiopoietin
Anti-His – Anti-Polyhistidine Monoclonal Antibody
baAECs – Bovine Aortic Arterial Endothelial Cells
bmAECs – Bovine Mesenteric Arterial Endothelial Cells
bmECs – Bovine Mesenteric Endothelial Cells
bmBECs – Bovine Mesenteric Blood Endothelial Cells
bmLECs – Bovine Mesenteric Lymphatic Endothelial Cells
bmVECs – Bovine Mesenteric Venous Endothelial Cells
CCL21 – Cysteine-Cysteine Motif Ligand 21
CCR7 – Cysteine-Cysteine Chemokine Receptor 7
CD86 – Cluster of Differentiation 86
cDNA – Complementary Deoxyribonucleic Acid
CI – Confidence Interval
COMP-Ang-1 – Cartilage Oligomeric Matrix Protein-Angiopoietin-1
CXCL1 – Cysteine-X-Cysteine motif ligand 1
CID – Collision Induced Dissociation
CO2 – Carbon Dioxide
DAB – 3,3‘-Diaminobenzindine
DAPI – 4',6-diamidino-2-phenylindole
xi
ºC – Degrees Celcius
DMEM – Dulbecco‘s Modified Eagle‘s Medium
2D-PAGE – Two Dimensional Polyacrylamide Gel Electrophoresis
ECs – Endothelial Cells
EDTA – Ethylene-diamine-tetra-acetate
EGF – Epidermal Growth Factor
ELISA – Enzyme Linked Immunosorbant Assay
EphB4 – Ephrin B4
Erk1/2 – Extracellular Signal-Regulated Kinases 1/2
ETD – Electron Transfer Dissociation
EtOH - Ethanol
FBS – Fetal Bovine Serum
FOXC2 – Forkhead Box Protein C2
g – Gram
> – Greater than
GM1 – monosialotetrahexosylganglioside
3H - tritium
HAT – Hypoxanthine-Aminopterin-Thymidine
HPLC – High Performance Liquid Chromatography
HRP – Horse Radish Peroxidase
HSP – Heat Shock Protein
HUVEC – Human Umbilical Venous Endothelial Cells
IB – Immunoblotting
xii
ICAM – Intercellular Adhesion Molecule 1
IEF – Isoelectric Focusing
IPG – Immobilized pH Gradient
IgG - Immunoglobulin
IHC – Immunohistochemistry
IMDM – Iscove‘s Modified Dulbecco‘s Medium
i.p. – intra-peritoneal (injection)
IP - Immunoprecipitation
JNK/SAPK – c-Jun N-terminal Kinase/Stress-Activated Protein Kinase
kDa – Kilodalton
LC – Liquid Chromatography
LC-MS/MS – Liquid Chromatography Tandem Mass Spectrometry
LEC – Lymphatic Endothelial Cells
< – Less than
LMCs – Lymphatic Muscle Cells
LYVE-1 – Lymphatic Vessel Endothelial Hyaluronic Receptor 1
mAb – Monoclonal Antibody
MBS – MES Buffered Saline
MeOH – Methanol
MES – 2-(N-morpholino) ethanesulfonic acid
µCi – Microcurie
µg – Microgram
µm – Micrometer
xiii
µM – Micromoles
mg – Milligram
mL – Millilitre
MODS – Multiple Organ Dysfunction Syndrome
MS – Mass Spectrometry
MS/MS – Tandem Mass Spectrometry
MW – Molecular Weight
NB – Nota bene
NO – Nitric Oxide
nm - Nanometres
NP-40 – Nonyl phenoxylpolyethoxylethanol.
O2 – Oxygen
PBS – Phosphate Buffered Saline
PECAM-1 – Platelet/Endothelial Cell Adhesion Molecule-1
% – Percent
PCR – Polymerase Chain Reaction
PFA – Paraformaldehyde
PKB – Protein Kinase B
Prox-1 – Prospero Homeobox Protein 1
PRDX – Peroxiredoxin
PVDF – Polyvinylidene Fluoride
RIPA – Radio-immunoprecipitation Assay
RNA – Ribonucleic Acid
xiv
ROS – Reactive Oxygen Species
RT-PCR – Reverse Transcriptase Polymerase Chain Reaction
SDS – Sodium Dodecyl Sulphate
SI – Subtractive Immunization
SIRS – Acute Systemic Inflammatory State
SPI – Scored Peak Intensity
Tie-2 – Tyrosine Kinase with Immunoglobulin-Like and EGF-like Domains 2
TLR – Toll-like Receptors
TNFα or β – Tumour Necrosis Factor alpha or beta
TMB – 3, 3‘, 5, 5‘ -Tetramethylbenzidine
TritonX-100 – t-octylphenoxypolyethoxyethanol,
UBE – Ubiquitin Conjugating Enzyme
VCAM – Vascular Cell Adhesion Molecule 1
VEGF – Vascular Endothelial Growth Factor
VEGFR-3/Flt-4 – Vascular Endothelial Growth Factor Receptor/Fms-Like Tyrosine Kinase 4
1
CHAPTER 1
BACKGROUND
2
1.1. Rationale and Objective
The body‘s network of blind-ended lymphatic capillaries collects interstitial protein-rich
fluid containing clotting factors, lipids, and lymphocytes extravasated from blood vessels. The
milky white lymph then drains into collecting lymphatic vessels, which actively pump the fluid
into larger transporting lymphatic vessels. In pathology, malformed lymphatic vessels—often
resulting from aberrant lymphangiogenesis during development—can lead to lymphangiectasia,
lymphangioma, lymphatic dysplasia, and associated lymphoedema. In normal physiology,
lymph is efficiently returned to blood circulation at the thoracic and right lymphatic ducts
where the two systems meet [1, 2]. Lymph nodes and immune cells continually monitor lymph
on its way back to circulation. Immune cells migrate into and out of lymphatic vessels by direct
contact with molecules on the surface of lymphatic endothelial cells (LECs) [3, 4].
Far from being ‗just‘ a supportive layer of tissue encasing a unidirectional flow of
lymph that may occasionally harbour dangerous hitch-hiking cancerous cells, the lymphatic
endothelium has been shown to actively attract and recruit tumour cells [5-13]. Similarly,
instead of being a passive protective lining of a fluid conduit, the lymphatic endothelium has
been shown to be highly adaptive to changes in tissue microenvironment [14-18]. The
lymphatic endothelium actively participates in local inflammatory responses [19-26] and is
capable of modifying both lymph flow [27-34] and composition [35-39] in response to various
stimuli.
Studies of interactions between lymphatic vessels and putative therapeutic molecules or
drug delivery systems offer great illustrations of the specialized function of the lymphatic
endothelium as distinct from the blood endothelium. Supersaxo and colleagues observed that as
the molecular weight of a subcutaneously applied protein increased, the absorption of the
3
molecule by lymphatics increased, and that proteins larger than 16kDa were exclusively cleared
from the site of injection by lymphatics and not by blood vessels [40]. Liposome drug carriers
do not directly enter blood circulation but must first be taken up by lymphatic vessels in a
manner that is dependent on the size of the liposomes as it relates to the interstitial pressure at
the sight of subcutaneous injection [41]. Kaminskas and colleagues were able to control
whether a drug was taken up directly by the blood circulatory system, foregoing the lymphatics,
or was preferentially taken up by lymphatic vessels by varying the length of 4-benzene
sulphonate or polyethylene glycol conjugated to the nanoparticular drug carrier made of
polylysine dendrimers [42]. These examples clearly show that phenotypic differences between
the lymphatic and blood vasculature have profound impact on practical applications such as
drug absorption and drug delivery systems. Building in vitro tools to understand these
phenotypic differences would complement efforts to intelligently design therapeutics targeting
one or both of these systems.
Another clear example of why it is important to understand cell-specific regulatory
mechanisms in the design of therapeutics is the case of the angiopoietins. Induction of Ang-2,
an established agonist of lymphangiogenesis [43, 44], has been shown to associate with
progression and prognosis of a variety of human cancers; high expression of Ang-2 in
malignant tissue correlates with poor survival rate, and/or high frequency of metastasis, and/or
high microvascular density in patients with advanced colorectal carcinoma, breast cancer,
gastric carcinoma, hepatocellular carcinoma, non-small cell lung cancer, prostate cancer, and
ovarian cancer [45-54]. In contrast, expression of Ang-1, an established agonist of
angiogenesis [55, 56], in these tumours was often found to be at comparatively low to
undetectable levels. What effects does Ang-2 have on tumour LECs as compared to tumour
4
blood endothelial cells (BECs)? More importantly, what cell-specific effects would anti-cancer
therapies that target Ang-2 function have on the normal healthy lymphatic or blood
endothelium?
Although Ang-2 is a specific case, it can be expected that Ang-2 will not remain a
unique example as evidence continues to accumulate implicating direct lymphatic involvement
in pathologies beyond cancer and lymphatic insufficiency. The lymphatic system has been
shown to be potentially directly involved in chronic inflammation [57, 58], acute systemic
inflammatory state (SIRS) and multiple organ dysfunction syndrome (MODS) [59], diabetic
nephropathy [60], heart disease [61-63], hypertension [64], systemic sclerosis [65],
hyperlipidemia [66] and obesity [67]. Thus, lymphatic endothelial regulators other than the
angiopoietins will inevitably emerge as key targets in proposed therapies. Since it is reasonable
to propose that lymphangiogenic regulators would tend to also have angiogenic effects, striving
to understand the mechanisms of cell-type-specific responses elicited by regulators of lymphatic
vessel functions can therefore aid rational design of drugs with selectivity for one system or the
other.
The studies recorded in these chapters all share the common goal of answering the
research question: What molecular signatures give rise to the phenotypic character of the LEC
lineage distinguishable from BEC lineages? The establishment of an in vitro system of LECs,
venous, and arterial ECs (VECs and AECs) derived from vessels of bovine mesentery (chapter
2) made possible the application of mass spectrometry for the comparative study of 2D-PAGE-
resolved protein profiles of these three endothelial cell (EC) types (chapter 3), and the use of
subtractive immunization (SI) for the production of a monoclonal antibody (mAB) selectively
reactive to a surface protein of LECs (chapter 4). These studies share the hypothesis that an
5
unknown number of unidentified proteins is responsible for maintaining lymphatic
endothelium distinctiveness and for executing unique lymphatic system functions.
Understanding the mechanisms of cell-type-specific responses elicited by regulators of
lymphatic vessel functions as distinct from blood vessel functions, can aid the rational design of
drugs specifically targeting one system and not the other.
1.2. Introduction to Lymphatic Anatomy
1.2.1. Lymphatic Anatomy and Function
The lymphatic system is a series of vessels that form a one-way active drainage system
for protein-rich fluid containing clotting factors, lipids and lymphocytes from peripheral tissues.
After this fluid is collected from tissue spaces by initial (or terminal) lymphatic capillaries, is it
milky white in appearance and is then called lymph. From initial lymphatics, lymph is pumped
into larger pre-collecting and collecting lymphatic vessels, which propel the fluid by rhythmic
contractions of lymphangions separated by valves. From collecting lymphatics, lymph is
sequentially pumped into larger prenodal lymphatic vessels, lymph nodes, and postnodal
lymphatic vessel trunks. Lymph from legs, abdomen, the left arm, and the left side of the head
and chest, return to venous circulation at the left base of the neck via the junction between the
thoracic lymphatic duct and the left subclavian vein. Lymph from the right arm, thorax, the
right side of the head and chest, return to venous circulation via the junction between right
lymphatic duct at the right base of the neck [1, 2, 68, 69].
Salient features of the lymphatic system exist that are absent in the blood circulatory
system. Firstly, unlike the blood circulatory system, the lymphatic system does not circulate
lymph. As afore mentioned, lymph formed at initial lymphatics does not return to the point of
6
origin in tissues through lymphatics. Lymph flow ends when lymph empties into venous blood
at the base of the neck. Secondly, initial lymphatics consisting of a single layer of ECs and
devoid of support cells are blind-ended tubes that are more flat, larger (10-60 µm as compared
to blood capillary diameters of about 8 µm) [68, 69] and are much more selectively retentive of
large molecules (20-100 nm or larger than 12kDa) than blood capillaries, which is more
permeable to particles of a few nanometres [38, 40, 69, 70]. Thirdly, ―anchoring filaments‖ are
hypothesized to anchor the outer flap of over-lapping initial LECs to surrounding tissues,
maintaining vessel patency, and allowing the inner flap to form a one-flap valve, thereby
stopping fluid from gushing freely back into the interstitium [69, 71]. Fourthly, the opening
and closing of discontinuous lymphatic endothelial cell-cell junctions is an active process
highly responsive to pressure changes in the interstitium [15, 16]. Fifthly, lymphangion
contractility in larger lymphatic vessels (responsible for the uni-directionality of lymph flow) is
tonal, phasic, and highly coordinated as lymph is actively pumped from one lymphangion
compartment to the next [72]. Finally, lymphatic muscle cells surrounding the endothelium in
larger lymphatic vessels are highly specialized contractile cells called lymphatic muscle cells
(not smooth muscle cells) innervated by surrounding non-myelinated nerve fibres [69, 73].
These lymphatic characteristics in combination produce a highly adaptive system
responsive to physical changes in tissue microenvironment. For example, Poggi and colleagues
observed differences between the lymphatic vessels of the mucous membrane/sub-mucous layer
of the bladder regularly subjected to changes in volume and pressure, and the intramuscular
lymphatic vessels located at trigone regions where changes in volume and pressure are virtually
absent [17]. The most notable difference was the presence of elastic and collagen fibres joining
the lymphatic vessel wall to surrounding tissues in the former and the absence of those fibres in
7
the latter [17]. This comparative observational study of lymphatics in the urinary bladder has
probably produced some of the clearest evidence for the adaptability of lymphatic vessels to
tissue microenvironment. This adaptability is especially important considering the crucial role
the lymphatic endothelium plays in immunity. A discussion of the lymphatic endothelium as a
vital participant in the body‘s immune system follows in the next section.
1.2.2. Lymphatic Endothelium Function in Immunity
More instantaneous adaptive responses exhibited by the lymphatic endothelium have
been demonstrated in inflammation when interstitial fluid pressure and lymphatic drainage are
both generally thought to increase [14]. Marchetti and colleagues showed that the lymphatic
endothelium in inflamed dental pulp exhibited distended endothelial walls accompanying
opening of cell-cell junctions, and disappearance of micro-pinocytotic vesicles, which are
suggestive signs of an endothelium optimized for bulk lymph drainage [26]. Miteva and
colleagues found expression level of the dendritic-homing chemokine (Cysteine-Cysteine
motif) ligand 21 (CCL21) in LECs to correlate positively with transmural flow rate [14].
These observations suggest that the lymphatic endothelium dynamically responds to
inflammatory events by preparing appropriately for increased intramural flow and leukocyte
infiltration.
In addition to increased lymph drainage, other inflammatory changes to the tissue
microenvironment have been shown to activate the lymphatic endothelium. Johnson and
colleagues showed inflammatory cytokines TNF-α and TNF-β (tumour necrotic factors) to
stimulate LEC expression of other proteins important for leukocyte transmigration and homing
to the lymphatic endothelium, such as intercellular adhesion molecule 1 (ICAM-1), vascular
8
cell adhesion molecule 1 (VCAM-1), E-selectin, CCL5, CCL2, and CCL20 [19]. More direct
involvement of the lymphatic endothelium with inflammatory events were shown by Sawa and
colleagues, who showed that these same leukocyte adhesion and homing proteins along with the
chemokines (Cysteine-X-Cysteine motif) ligands CXCL1, 3, 5, 6, and 8 were induced in LECs
with lipoteichoic acid [20], a major cell-wall component of Gram-positive bacteria. In fact, the
lymphatic endothelium serves the function of immune sentinel by detecting the presence of
pathogen-associated molecules such as lipoteichoic acid through LEC expression of receptors
central to innate immunity responses: Toll-like receptors TLR2 and 4 [24] and TLR6 and 9
[25]. These observations suggest that LECs are highly sensitive to inflammatory events and
play an integral role in the body‘s response to infection.
In the absence of high interstitial fluid pressure and therefore high lymph load during
infection, the lymphatic endothelium has been shown to be suppressive of immune cell activity.
Podgrabinska and colleagues demonstrated that in the absence of pathogen-derived signals,
high dendritic cell expression of CD86 correlated with adhesive interactions with LECs and
attenuation of dendritic cell maturation and its ability to stimulate T-cells [22]. These results
imply that in addition to providing support during active inflammation, the resting lymphatic
endothelium contributes to prevention of undesired immune reactions and thereby provides an
important service to bodily health. A discussion of other functions of the lymphatic
endothelium in homeostasis follows below.
1.2.3. Lymphatic Endothelium Function in Homeostasis
In health, cardiovascular homeostasis is maintained by the ability of blood vessel walls
to constrict or relax when appropriate. The dependency of blood vessel vasoconstriction and
vasodilatation on intact functional endothelium has been well established since 1980 [74-79].
9
By contrast, the role the lymphatic endothelium plays in regulating lymphatic vessel tone,
relaxation, and pumping action has only been explored with significant depth since 1990 [27-
34, 72, 73]. Like the arterial endothelium and unlike the venous endothelium, which is
relatively unresponsive to the neurotransmitter acetylcholine (ACh) [74-76, 79], the lymphatic
endothelium was shown by Ohhashi and Takahashi to release a vessel muscle ―relaxing factor‖
upon stimulation with ACh [27]. The observation by these researchers that mechanical ablation
of the lymphatic endothelium suppressed responsiveness of the vessel ring to ACh was
corroborated by Ferguson [33], and Hashimoto and colleagues [34]. One of the relaxing factors
released by the lymphatic endothelium upon stimulation with fluid flow or ACh was later
shown to be very likely nitric oxide (NO) [28, 80]. These studies establish the dependency of
lymphatic vessel tone regulation on LEC-derived factors.
Whether ACh and LEC-derived factors also play a role in the regulation of phasic
contractions of lymphangions in larger lymphatic vessels was examined by Yokohama and
Ohhashi [29]. The researchers studied spontaneous phasic contractions in bovine mesenteric
lymphatic vessels and found ACh to dose-dependently reduce both the rate of phasic
contractions (negative chronotropic effect) and the strength of contractions (negative inotropic
effect) [29]. The vessels studied had a rhythmic contraction rate of approximately 3 beats per
minute that was lowered to approximately 2 beats per minute in the presence of ACh. The
strength of contraction was lowered by approximately 10% in the presence of ACh. More
importantly, physical ablation of the endothelium abolished these effects suggesting that LEC-
derived factors such as NO or other related compounds mediate not only lymphatic vessel tone
but also vessel pumping action [29]. The dependency of lymphatic vessel phasic contractility
regulation on an intact and functional lymphatic endothelium was confirmed in other studies by
10
Elias and colleagues [31], von der Weid and colleagues [32], Ferguson [33], and Hashimoto and
colleagues [34].
Augmenting these studies are findings by Koller and colleagues, whose findings suggest
that LEC-dependent chronotropic regulation can be separated from inotropic regulation of
lymphatic vessel phasic contractions [30]. The researchers used lymphatic microvessels from
rat iliac to demonstrate increases in intramural flow reduced the amplitude of phasic
contractions by approximately 40%, but increased frequency of rhythmic contraction by
approximately 20% [30]. Interestingly, increased intramural flow did not change lymphatic
vessel diameter suggesting that LEC-regulated changes in lymphatic vessel tone (or resistance),
phasic contraction, and contractile strength can be modulated independently of one another to
respond sensitively to fluid pressure dynamics in local interstitial environments. Studies of the
kind described here, thus far, are few in number. However they all add insightful explanatory
value to clinically observed severity and complexities of lymphoedema and related lymphatic
insufficiency disorders. As basic knowledge about lymphatic system anatomy and function
continues to accumulate, the list of human pathologies which involve the lymphatic system
grows longer. Established and emerging lymphatic-associated disorders are discussed in detail
in the next two sections.
1.2.4. Lymphatic Endothelium Dysfunction in Lymphoedema
The most well studied lymphatic-associated disease is called lymphoedema, which is an
umbrella term for lymph transport insufficiency resulting from congenital lymphatic
vessel/valve malformation (primary lymphoedema), or from interference with normally formed
functional lymphatic vessels/valves by injury, infection, or iatrogenic procedures (secondary
lymphoedema). Congenital lymphoedema is represented by Milroy‘s and Meige‘s syndromes,
11
both of which show an autosomal-dominant pattern of inheritance and are caused by mutations
of the vascular endothelial growth factor receptor 3 (VEGFR-3/Flt-4) tyrosine kinase receptor
and the forkhead box protein C2 (FOXC2) transcription factor, respectively [81]. Primary
lymphoedemas like these syndromes are usually characterized by swelling of both limbs
whereas secondary lymphoedema is more localized to limbs normally served by the obstructed
lymphatic vessel. Incidences of secondary lymphoedema in North America are predominantly
related to complications following surgical treatment of breast cancer, melanoma, gynecologic
cancers, lymphoma and urologic cancers [82]. Secondary lymphoedema in the developing
world is predominantly related to infection with the nematode worm Wuscheria bancrofti
(lymphatic filariasis), which matures in lymphatic vessels and causes disruption to lymph
transport [82].
In general, failure of lymph transport functions results in a build-up of protein-rich fluid
in tissues drained by lymphatic vessels at the point of dysfunction. Over time, mass swelling of
affected area worsens due to deposition of fat, collagen and other fibrous tissues, abnormal
localized tissue growth and skin thickening, and inappropriate accumulation of immune cells
[82-86]. Since lymphatic vessels play a vital role in immunity and inflammatory responses,
lymphoedema often accompanies local immune suppression, and associated complications such
as chronic bacterial infections of the skin (erysipelas), chronic inflammation of the skin
(dermatitis) of lymphatic vessels (lymphangitis) and of subcutaneous tissues (cellulitis), and
consequently delayed wound healing [87-93]. In severe chronic lymphoedema, outcomes can
range in severity from disfigurement and resultant psychological distress, loss of sensory nerve
function, and loss of limb mobility due to sepsis, gangrene, and amputation [94-97].
12
Venous capillaries play a similar role to lymphatic capillaries in the maintenance of
homeostasis. However, important differences have significant implications for the differential
diagnosis of lymphatic disorders and the management of treatment strategies. This can be
highlighted by noting that the terms ‗oedema‘ and ‗lymphoedema‘ are not interchangeable and
that the unique biology of the lymphatic system allows it to play a unique role in maintaining
protein-fluid balance.
In the resting adult human, approximately 15 mL of plasma-protein-rich fluid filters out
at the arterial end of capillary beds each minute. Approximately 80-200 g of plasma protein is
extravasated into interstitial spaces with this fluid. At the venous end of capillary beds, 12-13.5
mL of extravasated fluid is reabsorbed each minute with the exclusion of extravasated plasma
macromolecules. The remaining concentrated filtrate is effectively removed from interstitial
spaces by a functional lymphatic system [68, 71]. Congestive heart failure, venous
hypertension, and chronic venous insufficiency are causes of disturbance to hydrostatic forces
governing fluid reabsorption at the venous end of capillary beds [98]. Failure to reabsorb
extravasated fluid at the venous end of capillary beds can result in fluid oedema and tissue
swelling without the immune dysfunction frequently associated with pure lymphoedema [99].
Typical treatment for the ensuing oedema is diuretics. By contrast, swelling associated with
lymphoedema and other related disorders cannot be resolved by diuretic therapy or any other
existing molecular or pharmaceutical intervention [98]. Standard treatment strategies of
lymphoedema revolve around multilayer compression bandaging, mechanical massage,
physical therapy involving skin hygiene, limb compression, regimented exercise, and as a last
resort, surgical resection [82].
13
Unfortunately, these treatment strategies for lymphoedema, in development since the
early 1900s, have not improved dramatically since the 1960s [100]. It is hoped that initial
characterizations of lymphatic involvement in human pathologies other than lymphoedema over
the past 10 years will continue to renew research interest in the unique biology and function of
lymphatic vessels, and thereby lead to more speedy progress toward effective lymphatic
endothelium-targeted therapies.
1.2.5. Lymphatic Endothelium Dysfunction in other Human Pathologies
The variety of human pathologies in which lymphatic vessels as implicated reflects the
lymphatic endothelium‘s diverse functions as discussed in sections 1.2.2. and 1.2.3. Firstly, in
diseases with presentation of chronic inflammation, aberrant inflammation-driven
lymphangiogenesis is thought to produce lymphatic endotheliums that exacerbate the
inflammation [57] rather than resolves or suppresses it as a healthy lymphatic endothelium
normally would [22]. This phenomenon has been observed in skin psoriasis [101], Crohn‘s
disease [102], Kaposi‘s sarcoma [58], tubulointerstitial fibrosis (nephropathy) [60], and
systemic sclerosis [65]. Secondly, in familial combined hyperlipidemia, low-grade
inflammation is thought to promote growth of new lymphatic vessels which facilitate fat
deposition thereby contributing to obesity [66, 67]. Thirdly, proper lymph transport is thought
to be a requirement of healthy chronotropic and inotropic regulation of heart contractions since
cardiac lymph flow impairment is concomitantly observed with arrhythmia, cardiac injury,
terminal heart failure, and even transplant rejection [61-63]. Fourthly, the endothelium as a
manufacturer of secreted factors has a potentially impactful role in SIRS/MODS [59] and
hypertension induced by high-salt diet [64].
14
From results of work with animal models of major trauma and shock, Deitch and
colleagues proposed the hypothesis that yet-to-be-identified factors derived from lymphatics in
the gut are responsible for systemic vascular damage and subsequent organ failure in SIRS and
MODS [59]. From their work with rats fed a high salt diet, Machnik and colleagues proposed
that clearance of excess sodium from interstitial spaces and thereby maintenance of volume and
blood pressure homeostasis depend on communication between lymphoreticular cells and the
lymphatic endothelium [64]. Inhibition of the resultant increase in the expression of endothelial
NO synthase in LECs led to elevated blood pressure [64].
Finally, the functions of the lymphatic endothelium in cell trafficking in immunity,
lymphangiogenesis (growth of new lymphatic vessels), lymph formation and transport, are co-
opted by metastasizing cells to disseminate in the body. For example, constitutive LEC
expression of lymphocyte attractants CCL19 and CCL21 promotes CCR7-positive melanoma
chemotaxis toward lymphatics and migration to regional lymph nodes (CCR7 is the Cysteine-
Cysteine motif chemokine receptor of CCL19 and 21) [5, 12, 13]. For another example, tumour
cells have been shown to produce the LEC proliferation inducer VEGF-C (vascular endothelial
growth factor-C), and both the increase in tumour-associated lymphatic vessel number and
lymph node metastasis could be blocked by inhibiting the VEGF-C receptor VEGFR-3/Flt-4
[9]. VEGF-A dependent tumour-associated lymphangiogenesis in lymph nodes also potentially
contributes to tumour cell dissemination via lymphatic vessels [6-8]. These few studies suggest
that the lymphatic endothelium plays an active role in tumour cell recruitment, dissemination to
lymph nodes and perhaps even metastatic spread to distant sites in the body [6, 7, 11].
Wong and Hynes took and step further and proposed the hypothesis that tumour cells
may ―prefer‖ to disseminate via the lymphatics over blood vessels. The review authors
15
proposed that gaps in LEC intercellular junctions (unlike the tight inter-endothelial junctions
characteristic of BECs), lack of support cells surrounding LECs in smaller lymphatic vessels
(that more frequently surround BECs in comparable blood vessels), and the low flow/shear
stress fluid environment within lymphatic vessel lumens make lymphatics the ideal conduit for
tumour cell metastasis [10]. This intriguing hypothesis deserves further scrutiny. Research
focus on understanding the ways that LECs are similar to and different from BECs in function
in contribution to cancer metastasis may help to shed light on the proposition. Such research
focus may also help to expand the range of potential therapeutic targets and designs possibly
conceived to treat cancer metastasis.
Research interest in the unique biology of the lymphatic endothelium and in its role in a
large variety of human pathologies, presented in this section, have produced some of the most
impactful publications in the field within the past ten years. This was made possible by the
discovery and characterization of LEC protein markers starting with VEGFR-3/Flt-4 in 1995
[103]. A brief summary of known lymphatic endothelium markers follows below.
1.2.6. Protein Markers of the Lymphatic Endothelium
Research in the past two decades has contributed most to what we now know about the
unique biology of the lymphatic system, despite over two centuries of effort [100]. This is
largely due to the discovery of lymphangiogenic factors VEGF-C and VEGF-D [104] and
specific LEC protein markers VEGFR-3/Flt-4 [103], lymphatic vessel endothelial hyaluronic
acid receptor 1 (LYVE-1) [105] , and prospero homeobox protein 1 (Prox-1) [106] in 1995,
1999, and 2002, respectively. Specific LEC protein markers made possible visualization and
characterization of LECs, the lymphatic endothelium, and lymphangiogenesis in studies of
various human diseases. Equally importantly, specific LEC protein markers make possible the
16
isolation and validation of cultures of LECs for in vitro experiments. While modern progress in
lymphatics research is palpable, a balance of caution and optimism is warranted considering the
mixed success of outcomes from recent endeavours to discover novel markers of LECs.
The advent of microarray technology created opportunities to profile entire
transcriptomes in isolated LECs and BECs in comparative studies. All 30 novel candidate
markers of LECs with expression of at least three fold higher in LECs than in BECs listed in the
review by Farnsworth, Achen and Stacker [11] were identified in two microarray studies both
utilizing the Human GeneChips-U95Av2; Podgrabinska and colleagues contributed 40% of the
list [36], and Hirakawa and colleagues contributed 60% of the list [107]. Only four out of the
30 novel candidate markers were validated by RT-PCR, two by other means, and the rest of the
list is yet to be validated. Despite the fact that both studies used LECs sorted from
disaggregated human neonatal foreskins, only one novel candidate marker was found in
common between the two microarray studies: RANTES or CCL5, a member of the 8-kDa
family of pro-inflammatory cytokines with chemotactic activity for monocytes and T-cells
[108, 109]. The high expression of CCL5 in LECs is consistent with their purported role in
immunity and inflammation.
Furthermore, consistent with the active role lymphatics are thought to play in cellular
trafficking in immunity and inflammatory responses, Hirakawa and colleagues found LECs to
express a broader range of cytokines than BECs [107]. Consistent with the purported role
LECs have in modifying lymph flow and composition, Podgrabinska and colleagues found
enrichment in genes involved in intracellular protein transport, vesicle formation and fusion,
and vesicle associated membrane proteins [36]. Consistencies notwithstanding, many of the 30
17
genes on the list would need to be extensively validated and characterized before they can be
considered bona fide protein markers of LECs.
Several good reasons exist for follow-up experiments. One reason concerns
reproducibility of microarray results. Other reasons have to do with the imperfect fidelity of
identified protein markers of LECs and BECs. Podgrabinska and colleagues used LYVE-1-
coated magnetic beads to purify LECs from disaggregated mixed cultures of dermal cells [36].
Cause for concern with this approach is the discovery of LYVE-1 expression in pleural
mesothelial cells [110] and macrophages [111]. The LEC cultures in this study may not be
contaminated with pleural mesothelial cells absent in skin, but would highly likely to be with
macrophages present in skin. Similarly, Hirakawa chose to define LECs as Sialomucin-
negative and platelet/endothelial cell adhesion molecule-1 (PECAM-1)-positive cells, and
BECs as Sialomucin-positive cells in similar cell sorting procedures with magnetic beads [107].
Cause for concern with this approach is the expression of Sialomucin in eosinophils and mast
cells, both likely to be present in skin [112]. More disconcertingly, subtypes of macrophages
are also known to be Sialomucin-negative and PECAM-1-positive [113].
The concern of imperfect fidelity of identified protein markers of LECs also applies to
Prox-1 and the vascular endothelial growth factor receptor VEGFR-3/Flt-4, which, along with
LYVE-1, are also expressed by pleural mesothelial cells [110]. Another well known marker of
LECs, podoplanin, was a podocyte protein before its expression in LECs and not in BECs was
discovered [114]. More recently, podoplanin was found to be expressed by peritoneal
mesothelial cells, osteocytes, glandular myoepithelial cells, ependymal cells, and other cells of
the lymphoreticular system [115]. A good brief summary of the limitations of these and other
LEC markers is the review by Baluk and McDonald [116].
18
One way to circumvent marker infidelity is to incorporate into best practices the use of
multiple markers to define cells of interest as belonging to the lymphatic endothelial lineage
prior to experimentation. Another way to circumvent marker infidelity is to functionally
characterize cells of interest prior to experimentation. If cells of interest behave in culture as
LECs are expected to behave, for example, in terms of responsiveness to known
lymphangiogenic factors such as VEGF-C and Angiopoietin 2 (Ang-2), then they are more
likely to belong to the lymphatic endothelial lineage.
The VEGF and the angiopoietin family of proteins have vessel-type specific effects.
VEGF-A signalling through its cognate receptor VEGFR-2 induces sprouting angiogenesis
(growth of new blood vessels) in vivo, promotes BEC proliferation, survival, migration, and
permeability in vitro (reviewed in [117]). VEGF-C signalling through its cognate receptor
VEGFR-3/Flt-4 induces sprouting lymphangiogenesis in vivo, and promotes LEC proliferation,
survival, and migration in vitro (reviewed in [118]). As reviewed by Thurston [119], knocking
out Ang-1 or its cognate receptor tyrosine kinase Tie-2 in mouse results in embryonic lethality
due to defects in the heart and to under-development of the blood vasculature. Knocking out
Ang-2 results in lymphatic vasculature under-development and blood vessel defects in the
retina [119]. Unlike VEGF-A and -C which have selective affinity for VEGFR-2 and VEGR-3,
respectively, both Ang-1 and Ang-2 can bind to and activate Tie-2 in vitro [120]. Mechanisms
by which these endothelial cell-type and vessel-type specific effects take place are currently
unknown. These observed effects provide rationale for further investigation of the lymphatic
endothelium as anatomically and functionally distinct from the blood endothelium.
19
1.3. Data Chapter Organization
Chapter 2
This chapter is a description of the isolation and validation an in vitro culture system of
lymphatic, venous, and arterial endothelial cells derived from vessels of bovine mesentery.
Chapter 3
This chapter is a description of the comparative study of 2D-PAGE-resolved protein
profiles of these three EC types described in chapter 2. This comparative analysis examines the
reproducibility and feasibility of mass spectral analysis of 2D-PAGE in the attempt to identify
novel cell-type specific proteins expressed by the three endothelial cells.
Chapter 4
This chapter is a description of the use of the three EC cultures for the production of a
monoclonal antibody selectively reactive to a potential surface protein marker of lymphatic
ECs. The attempted procedure examines the feasibility of subtractive immunization as a tool
for the discovery of novel surface markers unique to the lymphatic endothelium.
20
1.4. Attributions
Chapter 2
Jason Trinh carried out reverse transcriptase PCR experiments from RNA prepared by
me. Dr. Harold Kim provided the designed primers for the PCR experiments. Dr. Stephen
Chen assisted in counting cells in the migration assay. Dr. Brenda Coomber provided
laboratory space for the isolation of the endothelial cells. All contributed to the editing of the
manuscript. Raj Kukreja obtained bovine mesentery from the abattoir. This chapter was
published on March 6th
2007 in BMC Cell Biology 8:10
Chapter 3
Dr. Natalie Rodrigues performed two-dimensional polyacrylamide gel electrophoresis
(2D-PAGE) using my cell preparations and helped me with the silver-staining of the gels.
George Hanna and Dr. Eric Yang assisted with the quantitative analysis of the gel spots.
George Hanna excised gel spots of interest. Dr. Eric Yang provided technical support in
identification of proteins by mass spectrometry. Katie Pizzuto provided assistance with some
of the immunoblots (IBs). Dr. Paul Van Slyke isolated the bovine aortic arterial endothelial
cells (baAECs). This chapter was published on Feb 22, 2010 in Proteomics 10:1658
Chapter 4
Maribelle Cruz and the Animal Health Technicians of the Department of Comparative
Research cared for the mice and performed injections on mice as needed. Meena Bali prepared
the ECs I provided her for injections, produced the hybridomas, and screened about a quarter of
the hybridoma colonies. Dr. Eric Yang provided technical support for mass spectrometry work.
21
CHAPTER 2
LYMPHATIC, VENOUS AND ARTERIAL ENDOTHELIAL CELLS
ISOLATED FROM BOVINE MESENTERIC VESSELS
RETAIN THEIR DEFINING CHARACTERISTICS IN CULTURE
AND ARE DIFFERENTIALLY RESPONSIVE TO ANGIOPOIETIN-1 AND -2
22
2.1. Introduction
Tie-2 (tyrosine kinase with immunoglobulin-like and EGF-like domains 2) is the
receptor tyrosine kinase for the angiopoietin family of ligands (Ang-1, 2, 3, and 4). The role of
Tie-2 in endothelial cells has been extensively studied over the years, and the discovery of
impaired lymphatic vessel patterning and function in Ang-2 knockout mice has since added
extra complexity to this growing field [43]. Tie-2 has also been shown to be expressed in
lymphatic endothelial cells [121]. In vivo studies using slow-release pellets of an
engineered form of Ang-1, cartilage oligomeric matrix protein-Angiopoietin-1 (COMP-Ang-
1), have further showed induction of angiogenesis and ectopic lymphangiogenesis in mouse
cornea [121]. In addition, over-expression of Ang-1 in the skin of mouse ears via recombinant
adeno-associated virus gene delivery induced lymphatic endothelial cell proliferation,
lymphatic vessel enlargement, sprouting, and branching [55].
Prior to the discovery of the involvement of Tie-2 and its ligands in lymphangiogenesis,
the role of Tie-2 and it ligands in angiogenesis was the focus of mouse genetic studies. Ang-1
deficient mice exhibited phenotypes similar to those of Tie-2 knockout mice. These phenotypes
included impaired myocardial trabeculation and endocardial development as well as lack of
perivascular cell recruitment to endothelial cells undergoing angiogenesis [56]. Ang-2 was
initially characterized as an Ang-1 competitive antagonist since transgenic over-expression of
Ang-2 produced angiogenic defects resembling those of Ang-1 or Tie-2 knockout mice [44].
In recent studies, evidence suggesting condition-dependent agonistic roles for Ang-2
brought into question the initial characterization of Ang-2 as simply a competitive antagonist of
Ang-1. Ang-2 has been shown to activate Tie-2 phosphorylation at high concentrations leading
to cell survival via the PKB signalling pathway [122]. Following a 24-hour pre-treatment,
23
Ang-2 can lead to vessel growth in a fibrin matrix model [123]. Ang-2 can also induce tubule
formation in murine brain capillaries and promote endothelial cell migration [124]. In addition
to inhibiting JNK/SAPK phosphorylation, Ang-2 has been shown to induce phosphorylation of
Tie-2, PKB, Erk1/2, and p38 members of the mitogen activated protein kinases in the human
umbilical vein endothelial cells (HUVEC) [125]. Other functions of Ang-2 in vasculogenesis
include promotion of EC adhesion independently of Tie-2 [126], regulation of differentiation of
cells surrounding the cortical peritubular capillaries of the kidneys [127], and modulation of
retinal and hyaloid blood vessel remodelling [128].
Complementing these studies are mouse knockout experiments that further elucidate the
importance of Ang-2 in endothelial cell function. Ang-2-null mice (strain 129/J) exhibited
defects in the hyaloid blood vasculature and gross defects in lymphatic remodelling [43]. Skin
oedema in Ang-2-null mice correlated with improper recruitment of support cells such as
lymphatic muscle cells (LMCs) responsible for the contractile function of lymphangions in
peripheral or dermal lymphatic vessels [43]. The Ang-2-null mice also developed lethal
chylous ascites and oedema in the peritoneal cavity around day 14, correlating with improperly
remodelled lacteals in villi, disorganization and hypoplasia of intestinal lymphatic capillaries,
and deficient smooth muscle association in mesenteric-collecting lymphatic vessels [43].
Therefore, the function of Ang-2 is not limited to antagonistic regulation of Ang-1 in Tie-2
signalling.
In this chapter, I demonstrate that bovine mesenteric venous, arterial, and lymphatic
endothelial cells (bmVEC, bmAEC, and bmLEC) respond differentially to Ang-1 and Ang-2.
Firstly, both Ang-1 and Ang-2 activated Tie-2 and downstream Erk1/2 in bmLEC. However,
whereas Ang-1 promoted migration in bmLECs, Ang-2 slightly more effectively promoted
24
proliferation and survival in these cells. Secondly, Ang-1 more effectively activated Tie-2 and
downstream Erk1/2 in bmVECs and bmAECs than did Ang-2. However, whereas Ang-1
promoted survival, and migration in bmVECs and bmAECs, Ang-2 did not stimulate the same
responses in these cells. Whereas Ang-1 produced a small proliferative response in bmAECs,
neither Ang-1 nor Ang-2 produced a proliferative response in bmVECs. Taken together, my
results suggest cellular responses to Ang-1 and Ang-2 stimulation vary depending on the origin
of the ECs. These results lend support to the use of EC culture systems such as the one I have
built to compare and contrast signaling events downstream of vascular modulators such as the
angiopoietins simultaneously in ECs of different vessel origins.
2.2. Methods
2.2.1. BmEC isolation and culture
Methods for isolation and culture of bovine ECs are adapted from previously established
protocols [13, 14]. Sections of the gut mesentery were taken from freshly slaughtered cattle at a
local abattoir (Better Beef, Guelph, Ontario). The mesenteric sections were brought back to the
laboratory in warm phosphate buffered saline (PBS). Connective and fatty tissues, which were
superficially rinsed with 75% ethanol, were removed to expose lymph nodes and vessels.
Sterile 0.1% Evan‘s Blue dye was injected into the lymph nodes to highlight the lymphatic
vessels. Dye was flushed from excised vessels with warm PBS supplemented with penicillin
(100 Units/mL, Life Technologies), streptomycin (100 µg/mL, Life Technologies), and
fungizone (2.5 µg/mL Gibco, Invitrogen). Distal ends of excised vessels were occluded with
surgical suture in order to infuse a solution of dispase I (2.6 Units/mL, Roche) and collagenase
A (1 mg/mL, Roche). Vessels occluded at both ends to trap the enzyme solution were
25
incubated in PBS at 37ºC for 10 minutes. Cells dislodged by the treatment were grown on
plates coated with 1% bovine gelatin (Sigma-Aldrich), in DMEM (Dulbecco‘s modified eagle‘s
medium, Sigma-Aldrich), and standard conditions.
Isolation of cells from venous and arterial vessels from bovine mesentery followed a
similar strategy as described for bmLECs. However, the use of dye was unnecessary.
2.2.2. Eliminating contaminating cells
In order to establish high-purity cultures of bovine mesenteric endothelial cells
(bmECs), a combination of strategies was used. Firstly, cells that were uninhibited by contact
and grew atop the monolayer of ECs were removed by several washes with trypsin such that the
monolayer underneath was not disturbed. Secondly, visible regions of the monolayer that did
not appear cobble stoned at confluency were removed by direct aspiration. Finally, limiting
dilution into 96-well plates was used to isolate groups of homogenous ECs. Groups of cells
comprising of more than 90% ECs were used in experiments. 90% estimates of purity were
based on monolayer cobblestone appearance.
2.2.3. Detection of transcripts by RT-PCR
Total RNA was prepared from cells isolated from bovine gut mesentery using Tri-
Reagent (Sigma-Aldrich) according to manufacturer‘s instructions. Synthesis of cDNA from
1µg of total RNA was done with Thermoscript reverse transcriptase (Invitrogen) according to
manufacturer‘s instructions. PCR was performed with Taq polymerase (Qiagen) with primer
sequences: Tie-1 (forward primer 5‘-TGA CTT TGC GGG AGA ACT GG-3‘, reverse primer
5‘-CTC CGA CCA GCA CGT TTC GG-3‘), Tie-2 (forward primer 5‘-GAT TTT GGA TTG
TCC CGA GGT CAA G-3‘, reverse primer 5‘-CAC CAA TAT CTG GGC AAA TGA TGG-
3‘), Neuropilin-1 (forward primer 5‘-CAG AAC GCT GCC CAC TGC AT-3‘, reverse primer
26
5‘-CTT TCT GGG TCC TTT TTA TC3‘), VEGFR3 (forward primer 5‘-CGG TGC CCA GTG
CGT GGG ACG-3‘, reverse primer 5‘-TTG ACT AGC CAT CGT AGG ACA-3‘), Prox-1
(forward primer 5‘-TTG TCA CCC AAT CAC TTG AAA-3‘, reverse primer 5‘-CTT CCA
GGA AGG ATC AAC ATC-3‘) and GAPDH (forward 5‘-ACC ACA GTC CAT GCC ATC
AC-3‘, reverse primer 5‘-TCC ACC ACC CTG TTG CTG TA-3‘)
2.2.4. Cell imaging
BmECs were grown to confluency on gelatin-coated glass coverslips, fixed with 4%
paraformaldehyde in PBS for 10 minutes at room temperature, and permeabilized with 0.1%
TritonX-100 (Sigma-Aldrich) for 1 minute at room temperature. Fixed and permeabilized cells
were stained with haematoxylin and eosin (H&E) or with immunofluorescence using
antibodies: Prox-1(RDI), LYVE-1 (RDI), and podoplanin. The following protocol was used for
Rhodamine-Phalloidin staining: Fixed and permeabilized cells were incubated in Rhodamine-
phalloidin (Molecular Probes, 1:40 dilution) in PBS for 15 minutes, and washed. Cells were
mounted on microscope slides with aqua-polymount (Polysciences Inc.) or DAPI-containing
Vectashield mounting medium (Vector Laboratories) where appropriate. Bright-field
microscope images were produced with the Zeiss Axioplan 2 light microscope (Carl Zeiss).
Confocal images were produced using Carl Zeiss LSM 510 laser scanning confocal microscope.
Fluorescent images were produced using the Zeiss Axiovert 200M.
2.2.5. Stimulation of bovine ECs
BmECs grown on gelatinized plates were moved to bare plates 24 hours prior to
stimulation. Cells were stimulated for 15 minutes with human recombinant Ang-1 or Ang-2
(800 ng/mL, R&D Systems) in 10% FBS. For stimulations with recombinant Ang-1, which
contains a polyhistidine tag, anti-polyhistidine monoclonal antibody (anti-His, R&D Systems)
27
was used to cluster the ligand according to manufacturer‘s instructions (ligand to antibody ratio
1:20). Where indicated, mock treatment refers to cells incubated with anti-His antibody alone.
2.2.6. Cell Lysis and Tie-2 Immunoprecipitation
Bovine ECs were washed twice with ice-cold PBS supplemented with 2mM activated
sodium orthovanadate (Sigma-Aldrich, see [129] for activation protocol). Cells were lysed on
ice for 30 minutes with RIPA lysis buffer (150mM NaCl, 1% Igepal, 0.5% Sodium
Deoxycholate, 0.1% SDS, 50mM Tris pH 7.4, 1mM EDTA) supplemented with 2 mM sodium
orthovanadate, and complete protease inhibitors (Roche). Tie-2 was immunoprecipitated from
equal protein amounts of cleared whole cell lysates with 2 µg of anti-Tie-2 antibody C-20
(Santa Cruz Biotechnology) pre-coupled to 25 µL protein A-sepharose (Amersham
Biosciences) for 1 hour.
2.2.7. Immunoblotting
Proteins were resolved on 10% PAGE gels and transferred to PVDF (Perkin Elmer)
membranes. Antibodies used in immunoblots were: anti- phosphotyrosine 4G10 antibody (1
µg/ml, Upstate Biotechnologies), phosphoTie-2-specific anti-pTyr992 antibody (1:1000, Cell
Signaling), anti-Tie-2 antibody 33.1 (0.5 µg/ml, BD Biosciences Pharmingen), and phospho and
pan Erk1/2 (1:1000, Cell Signaling).
2.2.8. 3H-Thymidine uptake assay
Bovine ECs were seeded in 96-well plates at a density of 4500 cells/well in 100 µL 10%
FBS DMEM. After 24 hours, 10% FBS DMEM was replaced with 1% FBS DMEM + anti-His
antibody (16 ug/mL, ―Mock‖), with 1% FBS DMEM + Ang-1 (800 ng/mL) clustered with anti-
His antibody (1:20 ligand to antibody ratio), or with 1% FBS DMEM + Ang-2 (800 ng/mL).
28
Cells treated thus were pulsed for 6 hours with 2 µCi of 3H-thymidine (Amersham). After 6
hours, the 96-well plates were placed in -80ºC for cell lysis before 3H-thymidine incorporation
was measured with the TopCount NXT Microplate Scintillation and Luminescence Counter
(Packard).
2.2.9. Cell count assay
400 000 cells were plated in each well of 6-well plates at the start of the assay in
10%FBS media or in mock media supplemented with 800 ng/mL super-clustered Ang-1 or 800
ng/mL Ang-2 and cell numbers were monitored every 24 hours for 96 hours via cell counting
with the hemocytometer. Mock media of bmLECs consisted of 5% FBS, of bmVECs consisted
of 2.5% FBS, and of bmAECs consisted of 1% FBS. These percentages of FBS were required
to maintain the respective cells over 96 hours while maintaining a decent level of cellular
proliferation and avoiding cell death. All mock media contained 16 ug/mL clustering anti-His
antibody.
2.2.10. Modified Boyden Chamber Migration Assay
Bovine ECs were grown for 24 hours before they were serum starved for 2 hours in
0.1% FBS DMEM. Serum starved bovine ECs were then trypsinized and seeded in 500 µL of
0.1% FBS DMEM at a density of 84 000 cells/well in the upper chamber of 6 well plates
containing 8 µm pore-size inserts (Falcon). 1.5 mL of the following were placed in the bottom
wells: 10% FBS DMEM (positive control); 0.1% FBS DMEM + anti-His antibody (16 µg/mL,
―Mock‖); 0.1% FBS DMEM + Ang-1 (800 ng/mL) clustered with anti-His antibody (1:20
ligand to antibody ratio); 0.1% FBS DMEM + Ang-2 (800 ng/mL). BmECs migration occurred
over a 4-hour time period. Membrane of inserts were fixed and stained with filtered Harris‘
hematoxilin prior to mounting on microscope slides with aqua-polymount (Polysciences Inc.).
29
2.2.11. Cell death ELISA
Bovine ECs were seeded in 6-well plates at a density of 100 000 cells/plate and grown
for 24 hours in standard tissue culture conditions and DMEM supplemented with 10% FBS
before the start of the assay. To examine cell death due to serum starvation, 10% FBS DMEM
was replaced with 0.1% FBS DMEM + anti-His (16 µg/mL), with 0.1% FBS DMEM + Ang-1
(800 ng/mL, clustered with anti-His antibody), or with 0.1% FBS DMEM + Ang-2 (800
ng/mL). In order to achieve detectable levels of cell death, bmLECs were harvested after 24
hour serum starvation, bmVECs after 48 hours, and bmAECs after 72 hours. Cell death was
determined by quantifying cytoplasmic histone-associated-DNA-fragments (mono- and
oligonucleosomes) via the ELISA (enzyme-linked immunosorbent assay, Cell Death ELISAplus
,
Roche) according to manufacturer‘s specifications.
2.3. Results
2.3.1. BmLEC isolation and culture.
In order to distinguish translucent lymphatic vessels from surrounding fatty tissue,
Evan‘s blue dye was injected into exposed mesenteric lymph nodes. The blue staining of post-
nodal lymphatic vessels facilitated excision and processing of the vessels to extract ECs. Blood
vessels remained red after dye injection (Figure 2.1.A) and therefore could not be mistaken for
lymphatic vessels. Initial cultures of cells extracted from dispase- and collagenase-treated
excised bovine mesentery lymphatic vessels were at least 50% endothelial based on cellular
morphology. Cells cultured from treated vessels were a mixture of LMCs, fibroblasts, and
cobble stoned bmLECs.
Unlike high-purity endothelial cell populations, a mixed cell population did not form a
30
cobblestone monolayer when grown to confluency (Figure 2.1.B). The morphologies of the
different cells within the mixed population were visually distinguishable. A mixed population
of predominantly LMCs, fibroblasts, and a few bmLECs formed circular ―swirly‖ patterns
(Figure 2.1.C). This morphology has previously been described by Leak and Jones, and
Johnston and Walker [130, 131]. Populations of cells that predominantly consisted of
fibroblasts (Figure 2.1.D) or LMCs (Figure 2.1.E) had distinct morphologies unlike the cobble
stoned appearance of bmLEC at confluency (Figure 2.1.F).
To obtain purity greater than 90%, contaminating LMCs and fibroblasts were eliminated
from the culture by direct aspiration, differential trypsinization, and by limited dilution.
Resulting cultures of the primary cells bmLECs had doubling times of approximately 24 hours
(Figure 2.2.B), consistent with doubling times reported by Leak and Jones [130]. These
bmLECs were robust and could be frozen and thawed past 20 passages.
Previously, Leak and Jones [130] described the spontaneous formation of lymphatic
tube structures in vitro from confluent monolayer of LECs. I also observed the formation of
these tubes in 96-well plates of bmLEC cell populations left at confluency for at least a few
weeks (Figures 2.1.G and 2.1.H). Cells at the ends of the tubes attached to and migrated up the
sidewalls of the tissue culture wells (Figure 2.1.G). The formation of lymphatic tubes in culture
indicated that extracted bmLECs maintain lymphatic phenotypic character in spite of having
been removed from the in vivo environment.
31
Figure 2.1. Isolation of bmLECs. A/ Evan‘s blue dye, when injected into the lymph node of
bovine mesentery (***), outlined lymphatic vessels (**), but not blood vessels (*). B/ A mixed
cell population did not form a monolayer in culture. C/ A mixed cell population of
predominantly LMCs and fibroblasts formed circular patterns in culture. D/ Typical
morphology of a cell population predominantly comprising of fibroblasts. E/ Morphology of a
cell population predominantly comprising of LMCs. F/ Morphology of a cell population
predominantly comprising of LECs. G/ Mixed cultures of cells extracted from lymphatic
vessels spontaneously form lymphatic tube-like structures with ends attached to the walls of the
tissue culture dish. H/ Enlargement of the lymphatic tube-like structure seen in G/.
32
33
2.3.2. Comparing bmLECs to bmVECs, and to bmAECs.
In essence, the strategy used to isolate bmLECs was also used to isolate bmAECs and
bmVECs. Dye injection was unnecessary since large mesenteric arteries and veins were
discernable from surrounding fatty tissue, unlike milky white lymphatic vessels. Arteries were
thick-walled vessels that were easily distinguished from slightly larger and thin-walled veins.
Cultured bmLECs, bmVECs, and bmAECs exhibited the same cobble stoned monolayer
morphology at confluency (Figure 2.2.A), although the packing of cells together was different.
After the formation of the confluent monolayer, bmAECs, bmVECs, and bmLECs continued to
grow at different rates to form tighter-packed monolayer (Figure 2.2.B). BmAECs grew fastest
at this stage and formed the most dense monolayer. BmAECs were followed by bmVECs,
which formed the second most dense monolayer. BmLECs grew at the slowest rate at this stage
and formed the least dense monolayer (Figure 2.2.B).
Staining the cells with fluorescent phalloidin revealed that the arrangement of actin
microfilaments in the three cell types appeared similar, typically with dense peripheral bands of
actin filaments, and prominent stress fibers throughout the cytoplasm (Figure 2.2.A). However
there did seem to be some qualitative differences in the intracellular distribution of these actin
filaments. There were extensive dense peripheral bands of actin, with relatively few stress
fibers in bmVECs, whereas stress fibers were more prominent in bmAECs. In contrast,
bmLECs seemed to have an intermediate combination of both actin filament arrangements
(Figure 2.2.A).
Upon achieving confluency, apart from differences in cell packing, bmLECs were
indistinguishable from bmVECs and bmAECs by phase contrast microscopy alone. However,
reverse transcriptase-PCR analysis, and immunofluorescence with antibodies specific for Prox-
34
1, LYVE-1, or Podoplanin and immunofluorescence confirmed that the EC preparations
retained their lineage-specific properties in culture. All three cell preparations expressed Tie-1
and Tie-2 whereas bmLECs but not blood ECs expressed VEGFR-3/Flt-4 (Flt-4) and Prox-1
(Figure 2.3.A). Immunofluorescence also showed bmLECs to express LYVE-1 and Podoplanin
(Figure 2.3.B), both of which have been shown to be highly expressed in LECs but not in BECs
[132]. Furthermore, bmAECs and not bmVECs expressed neuropilin-1, which has been shown
to be highly expressed in AECs [133] (Figure 2.3.A). Although each of these protein markers
for LECs may not exclusively identify LECs per se, however, the co-detection of these markers
together reliably identifies the cultured cells as of lymphatic endothelium origin.
35
Figure 2.2. Similarities and differences between bmLECs, bmVECs and bmAECs. A/
BmLECs cannot be distinguished from bmVECs or bmAECs by cellular morphology alone.
Actin staining of the three cells using Rhodamine-phalloidin showed actin filaments arranged in
dense bands around the cell periphery and actin stress fibers extending throughout the
cytoplasm. B/ ECs of each type (bmLEC, bmVEC, and bmAEC) were seeded in equal
numbers (approximately 30-40% confluency) and counted every 24 hours. Trypan blue
exclusion was used to determine cell viability. Numbers from 3 independent counts were used
to compile the figure.
36
37
Figure 2.3. BmLECs, bmVECs and bmAECs express cell-lineage-specific markers. A/
Despite being similar in morphology, RT-PCR analysis showed bmLECs and not bmVECs or
bmAECs expressed Prox-1 and VEGR-3/Flt-4; bmAECs and not bmVECs expressed
neuropilin-1. All cells expressed Tie-2 and Tie-1. B/ As shown by immunofluorescence,
bmLECs stained positive for podoplanin, Prox-1, and LYVE-1.
38
39
2.3.3. Activation of Tie-2 in bmECs by Ang-2
In order to ensure Tie-2 in bovine cells is sufficiently similar to human Tie-2 to be
stimulated with human angiopoietin ligands, amino acid sequences of bovine, mouse, and
human Tie-2 were aligned with CLUSTALW (Figure 2.4). Protein sequence alignment of
human, mouse and bovine Tie-2 indicated 95% identity between bovine and human sequences
and 92% identity between mouse and human, or mouse and bovine sequences. More
importantly, the residues found by Barton and colleagues [134] to be important for the
interaction of Tie-2 with Ang-2 were conserved across the three species (Figure 2.4, blue
boxes), and all tyrosine residues in the kinase domain were perfectly conserved (Figure 2.4).
Based on this observation, bovine ECs should be highly suitable for stimulation with human
ligands.
The Tie-2 receptor protein levels were greatest in bmAECs, whereas the levels in
bmVECs and bmLECs were lower (Figure 2.5.A, anti-Tie-2 (33.1)). Once the differences in
expression levels of Tie-2 in each cell were accounted for, activation of Tie-2 as judged by its
autophosphorylation by Ang-2, relative to basal autophosphorylation level (mock treatment),
was highest in bmLECs, and was considerably lower in bmAECs and bmVECs (Figure 2.5.A
and 2.5.B). Clustered Ang-1, which was previously demonstrated in our laboratory to be the
most potent form of Ang-1 [135], stimulated the activation of Tie-2 in all three cell types above
basal levels (Figure 2.5.A and 2.5.B). Of note, the level of phosphorylation of Tie-2 due to
Ang-1 stimulation as compared to the level due to Ang-2 stimulation was the same in bmLECs
but different in bmAECs and bmVECs (Figure 2.5.A and 2.5.B)
Studies of HUVEC and capillary endothelial cell systems have shown that Ang-1 and
Ang-2 stimulation resulted in the activation of extracellular signal-regulated kinase or Erk1/2
40
[125, 136-138]. In my endothelial cell system, whereas the endogenous protein levels of Erk1/2
were not different in bmLECs, bmVECs and bmAECs, active phosphorylated Erk 1/2
(phospho-Erk) levels were quite different. Mock-treated bmVECs contained high amounts of
phospho-Erk, whose levels increased slightly upon Ang-1 stimulation (Figure 2.5.A and 2.5.C).
Ang-2 stimulation of bmVECs did not activate Erk1/2 according to phosphorylation levels of
the protein (Figure 2.5.A and 2.5.C). Although bmAECs had relative low basal levels of
phospho-Erk in the mock-treated sample, these cells responded in a similar fashion as bmVECs
to Ang-1 stimulation. Phospho-Erk levels were increased above basal levels upon Ang-1
stimulation, but not upon Ang-2 stimulation in bmAECs (Figure 2.5.A and 2.5.C). In contrast
to bmAECs and bmVECs, stimulation of bmLECs with either Ang-1 or Ang-2 resulted in an
increase in phospho-Erk levels, although Ang-1 seemed to provide a more potent signal (Figure
2.5.A and 2.5.C).
In order to ascertain whether the autophosphorylation of Tie-2 detected in these
immunoblots reflected ligand-dependent activation of Tie-2 and not just phosphorylation by
another kinase, I immunoblotted membranes from bmLECs stimulated with Ang-2 with an
activation-specific antibody directed to tyrosine 992. Phosphorylation of tyrosine 992 has been
shown to reflect activation of kinase activity in Tie-2 [139]. Tie-2 was phosphorylated at this
residue when bmLECs were stimulated with Ang-2 (Figure 2.5.D), indicating that the kinase
activity of the tyrosine receptor was activated.
41
Figure 2.4. Alignment of bovine, murine, and human Tie-2 amino acid sequences. Amino
acid sequences were obtained from Uniprot/Swiss-prot and aligned with CLUSTALW [140]:
Human Q02763, Bovine Q06807, Mouse Q02858. Boxed in blue are residues found to be
important for Ang-2 interaction with Tie-2 [18], all of which are conserved between the three
species. The transmembrane domain is indicated by the black bar. All tyrosine residues in the
cytoplasmic domain of Tie-2 are well conserved between the three species.
42
43
Figure 2.5. Stimulation of Tie-2 phosphorylation in bmECs by Ang-1 and Ang-2. A/
Stimulation of bmLECs, bmVECs, and bmAECs with 800 ng/mL Ang-1 (clustered with anti-
His antibody), or 800 ng/mL Ang-2, or with clustering antibody alone (Mock). Ang-1
stimulation resulted in activation of Tie-2 in bmLECs and to a lesser extent bmAECs and
bmVECs. B/ Phosphorylated Tie-2 signals from blots of 3 independent experiments were
converted to pixel density, quantified using ImageQuant version 5.0, and normalized to
corresponding total Tie-2 levels. Paired Ang-1- and Ang-2- stimulated, normalized Tie-2
phosphorylation levels in each experiment were compared by t-test and p-values for 95%
confidence interval (CI) are shown. Both Ang-1 and Ang-2 stimulated Tie-2 to comparable
levels in bmLECs, whereas Ang-1 was more effective in stimulating Tie-2 in bmVECs and
bmAECs. C/ Phosphorylated Erk1/2 signals from blots of the same 3 independent experiments
from B/ were converted to pixel density, quantified using ImageQuant version 5.0, and
normalized to corresponding total Erk1/2 levels. Paired Ang-1- and Ang-2-stimulated,
normalized Erk1/2 phosphorylation levels in each experiment were compared by t-test and p-
values for 95% CI are shown. Ang-1 was more effective in stimulating Erk1/2 phosphorylation
in all three cell types than Ang-2. D/ Stimulation of bmLECs with 800 ng/mL Ang-2 resulted
in phosphorylation of tyrosine residue 992 on Tie-2.
44
45
2.3.4. BmLEC proliferation is enhanced by Ang-1 and Ang-2
A number of studies have shown that Ang-1 stimulation results in the activation of
Erk1/2 [125, 136-138]. More importantly, Ang-1 has been shown to stimulate proliferation of
ECs [137]. Endothelial proliferation in response to growth factors is considered a key
angiogenic and lymphangiogenic response. Therefore, I used tritiated (3H) thymidine uptake
and viable cell counting to determine whether Ang-1 and Ang-2 had a proliferative effect on
bmLECs, and to compare the response of all three cell types to these factors (Figure 2.6).
BmVECs did not respond to either Ang-1 or Ang-2, whereas bmAECs responded best to Ang-1
stimulation (Figure 2.6.A). In contrast, bmLECs responded to both Ang-1 and Ang-2 but best
to Ang-2-stimulation (Figure 2.6.A). These results suggest that the endothelial cell type
influences proliferative responses to the either Ang-1 or Ang-2.
In order to ensure that Ang-1 and Ang-2 stimulation did not simply increase thymidine
uptake in bmECs independently of cell cycle progression and cellular proliferation, I performed
cell count experiments (Figure 2.6.B, 2.6.C, 2.6.D). Consistent with the trends shown by
thymidine uptake assays, Ang-1 and Ang-2 both provided proliferative signals to bmLECs,
increasing bmLEC numbers over the 96-hour period of the assay (Figure 2.6.B). Furthermore,
consistent with thymidine uptake results, Ang-2 was marginally more effective than Ang-1 in
increasing cell counts of bmLECs, although the result was not statistically significant (Figure
2.6.B). However, Ang-1 and Ang-2 increased bmLEC cell counts significantly above mock
levels indicating that both Ang-1 and Ang-2 were proliferation stimulators of bmLECs (Figure
2.6.B).
Consistent with thymidine uptake results, neither Ang-1 nor Ang-2 provided
proliferative signals to bmVECs (Figure 2.6.C). Neither ligand increased cell counts of
46
bmVECs above mock levels over the 96-hour period (Figure 2.6.C). In contrast and consistent
with the trends shown by the thymidine uptake results, Ang-1 provided a small proliferative
signal to bmAECs, whereas Ang-2 did not increase bmAEC cell counts above mock levels over
the 96-hour period (Figure 2.6.D). Taken together, these cell count results again indicate that
whether Ang-1 and Ang-2 are mitogenic depends on the endothelial cell type.
2.3.5. Ang-1 promotes bmEC migration
EC cell migration is another well-accepted criterion for the angiogenic response. Ang-1
has previously been shown to induce cell migration, whereas Ang-2 has previously been shown
not to drive the migration of HUVECs [141]. Using a modified Boyden chamber assay I tested
the ability of Ang-1 and Ang-2 to drive migration of these three cell types. I found that Ang-2
did not promote migration of these cells whereas Ang-1 was able to drive bmEC migration
(Figure 2.7). These results suggest that although bmLECs, bmVECs, and bmAECs responded
differentially to Ang-1 and Ang-2 in proliferative assays, they responded similarly to the
chemotactic effects of Ang-1.
2.3.6. Ang-2 protects bmLEC from cell death
Many angiogenic factors are known to protect ECs from serum-deprivation-induced
cellular apoptosis [142]. Ang-1 has previously been shown by numerous research groups
including the Dumont group to protect HUVEC from apoptosis [143-150]. However, a role for
Ang-2-mediated EC survival remains unclear. Results of some studies suggested that Ang-2
counteracts the anti-apoptotic effects of Ang-1 and leads to cell death [44, 151-153]. Results of
other studies suggested that Ang-2 mediates a PKB-dependent cell survival signal [122, 125,
154]. To investigate the role of Ang-2 in serum-deprivation-induced cell death, I tested the
ability of Ang-1 or Ang-2 to promote cell survival on the three cell preparations. Ang-1
47
protected bmVECs and bmAECs from cell death, whereas it had minimal effect on bmLECs
(Figure 2.8). In contrast Ang-2 had virtually no effect on bmVECs and a marginal effect on
bmAECs (Figure 2.8). Interestingly, Ang-2 produced a survival response in bmLECs (Figure
2.8). Together with results from survival and proliferation assays (Summarized in Table 2.1),
these results further suggest that the origin of ECs dramatically influences their response to
either Ang-1 or Ang-2.
48
Figure 2.6. Proliferation of Ang-1 and Ang-2 stimulated bmECs. A/ BmLECs, bmAECs and
bmVECs were tested for proliferative response in the presence of Ang-1 and Ang-2 via tritiated
thymidine uptake. Clustered Ang-1 (800ng/mL) and Ang-2 (800ng/mL) did not increase
thymidine uptake in bmVECs. Ang-1 induced increased thymidine uptake in bmAECs and
bmLECs above mock levels (clustering anti-His antibody alone). Ang-2-treated bmLECs
showed higher levels of thymidine uptake than did Ang-1-treated cells. Results from 2
independent experiments were compiled for the figure. ANOVA p-values for 95% CI are
indicated with the colours of the bars corresponding to the colours used to represent cell type.
BmLECs (B), bmVEC (C), and bmAEC (D) were tested for proliferative response in the
presense of clustered Ang-1 (800 ng/mL) and unclustered Ang-2 (800 ng/mL) by cell counting.
400 000 cells were seeded (hour 0, approximately 10-15% confluency) and monitored every 24
hours. Trypan blue exclusion was used to determine cell viability. Results are compiled from
one representative experiment of 2 independent experiments, each done in tripplicate. ANOVA
p-values for 95% CI for cell counts at the 96 hour time points are as follows: B/ BmLEC p=9E-
10 (*); C/ BmVEC p=0.22 (**); D/ BmAEC p=6E-7 (***).
49
50
Figure 2.7. Migration of Ang-1 and Ang-2 stimulated bmECs. BmECs were tested for
migration in a modified Boyden chamber assay in the presence of Ang-1 and Ang-2. Migration
of bmLECs, bmVECs, and bmAECs towards clustered Ang-1-containing media was
significantly increased compared to towards control media (Mock) containing only 1% FBS and
clustering anti-His antibody. In contrast, Ang-2 did not significantly stimulate migration in any
of the cell types, compared to control media. Migration was measured using the Boyden
Chamber. Results from 2 independent experiments were compiled for the figure. ANOVA p-
values for 95% CI are indicated.
51
52
Figure 2.8. Survival of Ang-1 and Ang-2 stimulated bmECs. Death in bmECs was induced
by serum starvation and relative amounts of cell death compared to 0.1%FBS control were
measured by quantification of cytoplasmic histone-associated-DNA-fragments (mono- and
oligonucleosomes) via ELISA (Roche). Ang-2 provided marginally better protection from
serum-deprivation-induced cell death than did Ang-1 for bmLECs. Ang-2 provided marginally
worse protection from serum-deprivation-induced cell death than did Ang-1 for bmVECs and
bmAECs. Results of 3 independent experiments were compiled for this figure. P-values of
unpaired, two-tailed t-test for 95% CI are indicated.
53
54
Table 2.1. Summary of results. Table depicts relative levels of Tie-2 and Erk1/2 activation
and relative levels of proliferation, survival in serum-deprivation conditions, and migration of
Ang-1 and Ang-2 stimulated bmECs.
55
56
2.4. Discussion
In this study, I set out to examine whether the type of EC could affect its response to
Ang-1 or Ang-2. I prepared ECs from lymphatic, venous and arterial vessels taken from bovine
mesentery and subjected them to angiogenesis-related in vitro assays. My results demonstrate
that the activation of Tie-2 by Ang-2 stimulation was as effective as by Ang-1 stimulation in
bmLECs. In bmVECs and bmAECs, Ang-1 stimulation was more effective than Ang-2 in
activating Tie-2. The activation of Tie-2 in all three cell types by Ang-1 or in bmLECs by Ang-
2 correlated with the activation of Erk1/2. Ang-1 activated Erk1/2 in all three cell types,
whereas Ang-2 only activated Erk1/2 in bmLECs. Of note, Ang-1 seemed to be a more potent
activator of Erk1/2 than Ang-2 in bmLECs. These results suggest Erk1/2 is a downstream
signaling pathway of Tie2 in ECs. Although Ang-1 stimulation correlated with Erk1/2
activation in all three cells, Ang-1 did not promote proliferation in all three cell types but only
weakly in bmAECs and bmLECs (Table 2.1). Unlike bmVECs, both bmAECs and bmLECs
had low levels of basal phosphorylated Erk1/2 levels in the mock samples. This finding
suggests that the proliferative response of bmLECs to Ang-2 stimulation is likely not solely
under the regulation of Erk1/2 downstream of Tie-2. Complex cell-type specific mechanisms
are likely present in the ECs and are responsible for suppressing the proliferative signal from
Ang-1 in bmVECs, while promoting the weak proliferative responses in bmLEC and bmAECs.
Likewise, additional cell-specific mechanisms may be present in BECs that suppress the
proliferative and survival signals from Ang-2, but are present in LECs and are responsible for
promoting the proliferative and survival signals from Ang-2. This was implied by my findings
that Ang-2 was not a proliferative signal for bmAECs or bmVECs but only for bmLECs.
Similarly, Ang-2 did not support serum-free survival of bmVECs and bmAECs but promoted
57
the survival of bmLECs (Table 2.1). Of note, the level of Erk1/2 activation stimulated by Ang-
2 in bmLECs was lower than the level stimulated by Ang-1, which did not provide as potent a
proliferative or survival signal as Ang-2 in bmLECs.
In summary, my data shows that ECs of different origin have different sets of cellular
responses to Ang-1 and Ang-2 stimulation, suggesting that Ang-1 and Ang-2 activate distinct
sets of signaling molecules in the signaling cascades of ECs of different origin. Thus, Ang-2
may be a full agonist in ECs of lymphatic origin, but a partial agonist in ECs of blood origin.
Since Ang-1 and Ang-2 also elicited slightly different responses from bmAECs and bmVECs in
my study (Table 2.1), the subtype of ECs of blood origin need to be taken into consideration in
future studies of Ang-1 and Ang-2 signaling and function.
Until recently, Ang-2 was thought to be an Ang-1 competitive antagonist. This
conclusion was made based on the finding that transgenic over-expression of Ang-2 produced
blood vasculature defects resembling those of Ang-1 or Tie-2 knockout mutants [43]. Ang-2
competitively blocked activation of Tie-2 by Ang-1 but Ang-2 was equally as effective as Ang-
1 in activating Tie-2 phosphorylation in NIH 3T3 fibroblasts ectopically expressing Tie-2 [43].
This finding suggested that ECs have additional components that allow functional
discrimination between the two angiopoietin [43]. My findings clearly indicates that the
additional components suggested by these initial studies of Ang-2 are endothelial cell-type
specific and may be responsible for the agonistic activities of Ang-2 in LECs. Even though
there are some functional redundancies, each angiopoietin may be driving a distinct set of
cellular events within a specific type of endothelial cell.
An important implication of the results described in this chapter is that the design of
therapies aimed at enhancing or mitigating angiogenesis and lymphangiogenesis need to include
58
considerations of biological differences between vascular systems that share molecular
regulators. The angiopoietins may be only one example of many other molecular regulators
that behave differently in the two vascular systems. Cell systems such as the one I have built
and describe in this chapter make it possible for simultaneous characterization of the signaling
pathways downstream of these shared molecular regulators in each of the three EC types:
venous, arterial and lymphatic. Such a cell system as this combined with high throughput mass
spectrometry-based methods currently in refinement produces a potentially very powerful tool
to indentify and characterize new lymphatic and blood vascular markers and phenotypic
regulators.
59
CHAPTER 3
DIFFERENTIAL PROTEOMIC ANALYSIS OF LYMPHATIC,
VENOUS, AND ARTERIAL ENDOTHELIAL CELLS
EXTRACTED FROM BOVINE MESENTERIC VESSELS
60
3.1. Introduction
Proteins perform functions essential for most cellular processes unique to lymphatic and
blood endothelial cell types. Protein profiling is therefore an important tool in characterizing in
vitro EC models used to understand cell-type specific cellular processes at the molecular level.
As early as 1984, when research endeavours to understand EC molecular biology behind
angiogenesis and lymphangiogenesis was still in its infancy, three separate groups were the first
to independently establish in vitro cultures of LECs isolated from bovine mesentery [131],
collecting ducts [155], and cavernous lymphangiomas [156] (as reviewed in [157]). Beside
similarities to BECs cultures such as monolayer cobblestone morphology and presence of
Weibel-Palade bodies, the researchers noted unique features displayed by LECs in culture
including overlapping intercellular junctions and anchoring filaments or microfibrils connecting
abluminal cell membranes of capillary LECs to the surrounding connective tissue [157, 158]
The capacity to isolate and establish stable lines of primary, immortalized, or
transformed LECs in culture allowed lymphatic researchers to compare and contrast LECs with
BECs in terms of morphology, physiology, and function. More recently, the advent of
technologies to study entire cell transcriptomes and proteomes have allowed lymphatic
researchers to characterize in vitro LEC cultures at the molecular level and perform
comparative studies with BECs similarly isolated [36, 107, 159-166]. The collective effort of
these and other researchers has yielded a reliable list of observed differences and similarities
between LECs and BECs.
I described in chapter 2 some of these differences in primary cultures of LECs, VECs,
and AECs isolated and cultured from vessels of bovine mesentery [167] by the method of
Johnston and Walker [131]. In culture, early passaged cells retained distinct molecular
61
signatures; bmLECs expressed well-known lymphatic markers such as Prox-1, Podoplanin, and
VEGF-R3, which were absent in bmVEC and bmAECs [167]. Other differences between the
three cell types that I observed include lower packing density of cells and relatively higher
responsiveness to stimulation by Angiopoietin-2 (Ang-2) in bmLECS as compared to the blood
ECs with bmVECs being the least responsive [167]. (Ang-2 has previously been shown to be
required for lymphangiogenesis during embryonic development [43, 168].)
In the study described in this chapter, I set out to characterize primary cultures of
bmLECs, bmVECs, and bmAECs described above by comparative analysis of protein profiles
resolved by 2-dimensional SDS polyacrylamide gel electrophoresis (2D-PAGE). I
hypothesized that potential molecular players that may be responsible for the observable
differences between the three ECs described in my previous work are identifiable by rigorous
quantification of silver-stained 2D-PAGE followed by mass spectrometric identification of
proteins showing statistically significant differences in levels between the three cell types. The
main rationale for attempting to identify proteins potentially responsible for phenotypic
differences I observed in these three cells, is to establish that my three-cell model is potentially
useful for studies aimed at developing drugs targeting disease states of the lymphatic
vasculature, without adversely affecting blood vessels. A highly relevant example of such a
disease state is intestinal lymphangiectasia. This condition has complex etiology generally
involving obstructed lymph flow in lymphatic vessels of the gastrointestinal tract. A three-cell
system such as the one I have built from vessels of the mesentery would be useful for in vitro
cell-based assays to understand similarities and differences between these three cells types and
how they respond to stimuli.
62
To the best of my knowledge, the comparative study reported in this chapter is one of a
very small number of studies that compare ECs from all three representative vessel types,
lymphatic, venous, and arterial, since the early 1990s. Weber and colleagues [169] assayed for
presence of 5‘-nucleotidase, alkaline phosphatase, adenylate and guanylate cyclase in cultured
ECs from lymphatic, venous, and arterial vessels from bovine mesentery isolated by the method
of Johnston and Walker [131]. Although I used the same cell system as Weber and colleagues,
I aimed to identify new differentially expressed proteins in the three cell types. Veikkola and
colleagues [170] compared responses of lymphatic, venous, and arterial ECs isolated from
human umbilical cord, aorta, skin, and saphenous vein to VEGFs. Ando and colleagues [171]
compared levels of cell adherence molecules in mouse mesenteric LECs to aortic AECs, and
VECs from vena cava. Bianchi and colleagues [161] compared 2D-PAGE-resolved protein
profiles of bovine thoracic aortic AECs, inferior vena cava VECs, and thoracic duct LECs.
Unlike these three studies, I used ECs cultured from smaller vessels of bovine mesentery.
Three 2D-PAGE electrophoretograms for each of the three mesenteric cell types were
produced and quantitatively analysed. Protein identification by liquid chromatography tandem
mass spectrometry (LC-MS/MS) was performed to identify 39 proteins excised from 2D gels
found to be present at statistically significantly different levels (>1.5 fold) in the three cell types
(p<0.05). Additionally, I was able to validate by immunoblotting (IB) three proteins, HSPA1B
(a member of the HSP70 family), HSPB1 (a member of the HSP27 family), and UBE2D3 (a
member of E2 ubiquitin-conjugating enzymes) found to be at highest levels in bmAECs,
bmVECs, and bmLECs. The lack of substantial overlap between my results, those of Bianchi
and colleagues [161], and those of other groups‘ comparative studies are fully discussed.
63
Functional implications of differences in levels of various proteins identified in the three cell
types are also discussed.
3.2. Methods
3.2.1. Cell isolation and culture
Bovine ECs from veins, arteries, and lymphatic vessels of the mesentery were isolated
and cultured as previously described [167] by the method of Johnston and Walker [131].
Bovine aortic arterial endothelial cells (baAECs) were isolated by a similar method. Cells were
briefly dislodged by dispase and collagenase from bovine aortic vessels. All EC types were
cultured in 10% fetal bovine serum in standard tissue culture conditions: 5% CO2, atmospheric
O2, and 37°C. Cells were at passage 20 at the time of harvest. Cells were harvested at
confluency and lysed in solubilisation buffer containing 8M urea, detergents, 2% immobilized
3-10 pH gradient buffer (IPG buffer, Amersham Pharmacia Biotech, Uppsala, Sweden).
Protein concentrations were quantified and equal protein amounts were subjected to 2D-PAGE
separation.
3.2.2. 2D-PAGE
Isoelectric focusing was performed on ReadyStrip IPG 11 cm strips (Biorad
Laboratories Hercules, California) with a linear range of pH 3-10 (using the same method as
described in http://proteome.tmig.or.jp/2D/2DE_method.shtml). IPG strips were first
rehydrated under passive conditions for 12 hours on the PROTEAN IEF cell (Biorad
Laboratories) followed by focusing at step-wise increases of voltage for a total of 46 700
64
voltage-hour with a current limit of 50 μA. IPG strips were equilibrated as directed by
manufacturer. The second dimension separation was carried out on 12% acrylamide gels.
2D-PAGE gels were silver stained and digitally recorded using the ProExpress gel
imaging system (PerkinElmer, Waltham, Massachusetts). Quantitative comparisons of spot
intensities were analysed using Progenesis Samespots (Nonlinear Dynamics, Durham, North
Carolina). Background subtraction and spot intensity normalization were automatically
performed by Progenesis Samespots. Relative spot intensity values were calculated on a
logarithmic scale and manually converted to relative fold-difference levels. Statistical analysis
was also automatically performed by Progenesis Samespots on intensity values of protein spots
from three separate gels for each of the cell types (total of nine gels).
3.2.3. Mass Spectrometric Analysis of Protein Spot Intensities
Excised protein spots of interest were trypsin digested, extracted, and analysed by LC-
MS/MS (1100 nanoflow LC system, XCT plus ion trap, Agilent Technologies, Santa Clara, CA,
USA). Spectrum Mill proteomics workbench (Agilent) was used to match MS/MS spectra to
non-redundant bovine protein sequences NCBInr (released in 2005). Peptide scores >10,
protein scores >11, SPI values >70%, were deemed acceptable since Spectrum Mill workbench
is not based on statistical methods.
3.2.4. Validation of Protein Level Differences by Immunoblotting
Cells were harvested at 90% confluency and lysed in RIPA buffer (150mM NaCl, 1%
Igepal, 0.5% Sodium Deoxycholate, 0.1% SDS, 50mM Tris pH 7.4, 1mM EDTA, protease
inhibitor tablets (Roche)). Lysates were cleared and quantified before proteins are separated by
SDS-PAGE. To detect the 17kDa protein UBE2D3 by immunoblotting, 4%-20% gradient gel
(Thermo) was used along with 0.2 μM PVDF transfer membrane (Sigma-Aldrich). All other
65
proteins were separated by 10% gel and 0.45 μM PVDF transfer membrane. Mouse
monoclonal antibodies against UBE2D3 (M01, clone 4C1), ANXA2, polyclonal mouse
antibodies against HSPA1B (HSP70-2, HSPA1A, whole antisera), and polyclonal rabbit
antibodies against HSPB1 and HSPB2 were all from Abnova. Anti-beta-actin mouse
monoclonal antibodies were from Sigma-Aldrich (Clone AC-15, mouse ascites fluid). Anti-
Tie-2 mouse monoclonal antibodies were from BD Biosciences Pharmingen. Autoradiograms
of chemiluminescent signal from immunoblots were quantified using ImageQuant 5.0. Pixel
densities were converted to relative values with lymphatic signal intensities taken as 1.
Statistical analysis reported for figures 3.4 and 3.5 was performed on averaged relative values
from at least 4 immunoblots for each of the three proteins validated.
3.3. Results
3.3.1. Quantitative Analysis of 2D-PAGE-Resolved bmEC Proteins
Three representative 2D-PAGE electrophoretograms for three EC types are shown in
figure 3.1.A. Spots chosen for MS/MS identification are indicated in the electrophoretogram
for bmLECs (Figure 3.1.A). Protein spots from all nine electrophoretograms (three per cell
type) were automatically matched by vector functions performed by Progenesis Samespots
(Figure 3.2.B). Matches were then manually checked to ensure correct assignment of vectors.
Quantitative analysis of protein level differences between bmLECs, bmVECs, and
bmAECs is demonstrated in Figure 3.2 with representative spot 32 (Figure 3.2.A). Progenesis
Samespots generates a three dimensional rendition of the imaged spot for better visualization
(Figure 3.2.B). Each data point in Figure 3.2.C represents each of the same spot number 32 in
each of the nine gels. Each set of three normalized relative volumes shown belonged to one of
66
the three indicated cell types (Figure 3.2.C). Averaged log normalized volumes automatically
calculated by Progenesis Samespots for representative spot 32 for each of the three cell types
are shown in Figure 3.2.D. P-values from ANOVA and power values associated with averaged
log normalized volumes are listed along with spot numbers in Table 3.1.A.
67
Figure 3.1. 2D-PAGE separation of bmEC proteins. A) Representative 2D silver stained
electrophoretograms of bmLECs, bmVECs, and bmAECs, generated using IPG strips pH 3-10.
Spot numbers correspond to protein identities listed in Table 4.1. Approximate molecular
weights and isoelectric points are indicated. B) Demonstration of spot matching function in
Progenesis Samespots. Automatic matching was manually checked to ensure accuracy.
68
69
Figure 3.2. Quantitative Analysis of 2D-PAGE-Resolved bmEC Proteins. Quantitative
analysis of protein level differences between bmLECs, bmVECs, and bmAECs. A) Enlarged
silver stained image of one representative spot numbered 32 is shown. B) Three dimensional
rendition of imaged spot by Progenesis Samespots. C) Normalized relative volumes of the
same spot from nine gels (three per cell type). D) Averaged log normalized volumes
determined for representative spot 32 for each of the three cell types.
70
71
3.3.2. Mass Spectrometric Analysis of 2D-PAGE-Resolved bmEC Proteins
Also summarized in Table 3.1.A are protein identities determined by LC-MS/MS and
database search with Spectrum Mill and Bovine NCBI non redundant database [172] along with
distinct summed MS/MS search scores, number of unique peptides that were matched, and
percent amino acid sequence coverage by the matched unique peptides. Proteins identified are
relisted in Table 3.1.B and are organized according to their subcellular localization, Gene
Ontology (GO) biological process and molecular function categories [173] as summarized by
GeneCards [174]. The distribution of these proteins in biological processes listed in Table
3.1.B are shown graphically in Figure 3.3. An overwhelming majority of listed proteins have
protein processing, chaperone, turnover, synthesis, and protein binding in skeletal structural
support functions.
Proteins found to be present at statistically significantly different levels in each of three
cell types included two peroxiredoxin proteins: PRDX2 and PRDX6, two ubiquitin conjugating
enzymes: UBE2N and UBE2D3, and two heat shock proteins: HSPB1 (HSP27) and HSPA1A
(HSP70) (Table 1A). PRDX6 protein level was higher in both bmLECs (close to 2-fold above
bmAECs) and bmVECs (1.7-fold above bmAECs). PRDX2 protein level was markedly higher
in bmLECs than in the blood ECs, 1.8-fold above the level in bmVECs and 2.5-fold above that
in bmAECs. UBE2N and UBE2D3 protein levels were higher in bmLECs than in the blood
ECs, 1.7- and 2.2-fold above levels in bmVECs, 2.5- and 3.8-fold above those in bmAECs,
respectively. HSPB1 protein level was highest in bmVECs, over 3.5- and over 5.7-fold above
the level in bmLECs and in bmAECs, respectively. HSPA1A protein level was highest in
bmAECs, over 30- and 12-fold over bmLECs and bmVECs, respectively. Like other proteins
on the list, I generally found a gradient of differences in protein levels between the three cells.
72
Limits on sensitivity of the method used probably prevented the detection of known and new
protein markers that distinguish a particular EC type.
73
Table 3.1.A. Summary of proteins identified. Proteins separated by 2D-PAGE found to be
statistically different in endothelial cells isolated from bovine mesenteric lymphatic, venous,
and arterial vessels (bmLEC, bmVEC, and bmAEC) identified by mass spectrometry. Proteins
are organized by relative fold differences in the three bmECs. At the top of the table are
proteins most up-regulated in bmAECs, proteins most up-regulated in bmVECs are in middle of
the table, and at the bottom of the table are proteins most up-regulated in bmLECs. Distinct
summed MS/MS search score, number of distinct peptides identified, and percent of amino acid
covered by distinct peptides provided by Spectrum Mill workflow are listed. Theoretical and
experimentally observed molecular weight (MW, kilodaltons, kDa) listed differed by at most 12
kDA (HMGCS1). Theoretical and experimentally observed isoelectric points (pI) differed by at
most 0.9 (GOT2). Accession numbers are from NCBI non-redundant database of bovine
sequences [172]. The table has been split into three parts and displayed on pages 74-76.
74
75
76
77
Table 3.1.B. Summary of biological information on protein identified. Proteins identified in
Table 4.1.A are relisted in this table and organized by subcellular localization, biological
process and GO molecular function (Gene Ontology [173]) as summarized by GeneCards [174].
The table has been split into three parts and displayed on pages 78-80
78
79
80
81
Figure 3.3. Biological Functions of Proteins Identified in Table 4.1. Percentage
breakdown—according to biological functions—of proteins found to be present at statistically
significantly different levels in the three endothelial cell types.
82
83
3.3.3. Validation of Identified Proteins by Immunoblotting
For proof of principle, I selected three proteins, each found to be at highest levels in
each of the three cell types, to validate by immunoblotting followed by pixel density
quantification with ImageQuant. The composite of resulting blots is shown Figure 3.4.A and
relative averaged signal levels in the three cell types from at least four blots for each protein are
shown in Figures 3.4.B, 3.4.C, and 3.4.D. In contrast to results from comparative analysis of
the 2D-PAGE electrophoretograms, HSPA1B was only about 2-fold higher by IB (as compared
to 30-fold) in bmAECs than in bmLECs (Figure 3.4.B). Furthermore, the level of HSPA1B in
bmVECs was not found by IB to be double that in bmLECs as recorded in Table 1A from
analysis of silver-stained 2D-PAGE electrophoretograms (Figure 3.4.B). Likewise, while the
level of HSPB1 was certainly highest in bmVECs, consistent with results from 2D-PAGE
summarized in Table 3.1.A, the magnitude of the difference was not reproduced by IB (over 5-
fold versus 3-fold by IB, Figure 3.4.C). UBE2D3 protein level was almost 4-fold higher in
bmLECs than in bmAECs by 2D-PAGE analysis but only approximately 2.5-fold over
bmAECs by IB (Figure 3.4.D). In all three cases, the pattern of differences was reproduced by
validation experiments but the magnitude of differences in protein levels was not reproduced.
In figures 3.4.B and 3.4.F, I showed the same method of quantification done on Tie-2
and beta-actin IBs. Tie-2 is a well-known receptor tyrosine kinase whose expression is thought
to be restricted to ECs almost exclusively. Tie-2 levels appeared to be slightly higher (1.5-fold)
in bmVECs and bmAECs than bmLECs (Figure 3.4.E). Beta-actin was not significantly
different between the three cell types as expected (Figure 3.4.F). The slightly higher level of
actin in bmLECs and bmVECs (1.2 fold) indicated that bmLEC and bmVEC lysates were made
up consistently slightly more concentrated than bmAEC lysates (Figure 3.4.F). However, this
84
small difference was not enough to account for all the relative differences in HSPA1B, HSPB1,
and UBE2D3 protein levels in the three cell types (Figures 3.4.A-D).
3.3.4. Comparing Protein Levels of Different Family Members in the same ECs
Figure 3.5, when I used antibodies specific to HSPB2, another member of the HSP27
family of proteins, I found protein levels to be highest in bmAECs instead of bmVECs as was
the case for HSPB1. This result may indicate that family members of the HSP27 family may be
differentially expressed in the three EC types. Different family members may play a unique
role in the biology of each of the three EC types. The same may also apply to the HSP70
family, to which HSPA1B belongs, and the E2 ubiquitin conjugation enzyme family, to which
UBE2D3 belongs.
3.3.5. Comparing Protein Levels in ECs of Dissimilar Anatomical Origin
Comparing my results to those of Bianchi and colleagues [161] I found very little
overlap. Similar to my results, Bianchi and colleagues found ANXA2 (Annexin A2) levels to
be highest in bovine aortic (baAECs), roughly 3 times the level in thoracic duct LECs [161]. I
found ANXA2 levels in bmAECs to be roughly 1.5 times that in bmLECs (Table 3.1.A). In
both my results and those of Bianchi and colleagues [161], the trend was the same in that the
silver stained signal corresponding to ANXA2 identification by mass spectrometry: bmVECs
had the lowest level of the three cell types. Other than ANXA2 similarities, there was little
overlap in both the identities of proteins found differentially up or down regulated and in the
trends of relative differences between the three cell types (Table 3.1.A).
I hypothesized that discrepancies between the two sets of data may be partly due to
differences in anatomical source of bovine lymphatic, venous, and arterial ECs. To test this
hypothesis, I quantified immunoblotted protein signals of ANXA2 in my bmECs as well as in
85
baAECs similarly isolated and cultured (Figure 3.6). Surprisingly, the trend of ANXA2 being
lowest in endothelial cells of venous origin was not validated by the IBs. ANXA2 was actually
highest in both bmAECs and bmVECs (Figure 3.6). I clearly saw statistically significantly
higher levels of ANXA2 in bmAECs as compared to baAECs (p = 0.0002, 12 blots). This
result indicates that protein level differences are highly dependent on anatomical context.
86
Figure 3.4. Validation of HSPA1B, HSPB1, and UBE2D3 protein levels in the three cell
types by IB. A) Protein from the three cell types were detected using isoform-specific
antibodies to HSPA1B (70 kDa), HSPB1 (27 kDa), and UBE2D3 (17 kDa). For loading
controls, Tie-2, an endothelial cell maker, and beta-actin were used. Autoradiograms from
chemiluminescent immunoblots were quantified using ImageQuant and averaged relative fold
differences in pixel densities are shown in B-F. B) Results of 5 blots are shown with HSPA1B
protein levels in bmLEC taken to be 1. T-test comparing relative levels of HSPA1B in
bmLECs to those in bmAECs resulted in p = 0.003 C) Results of 5 blots are shown with
HSPB1 protein levels in bmLEC taken to be 1. T-test comparing relative levels of HSPB1 in
bmLECs to those in bmVECs resulted in p = 0.008 D) Results of 5 blots are shown with
UBE2D3 protein levels in bmLEC taken to be 1. T-test comparing relative levels of UBE2D3
in bmLECs to those in bmVECs resulted in p = 0.009 E) and F) results from 4 blots are shown,
t-tests done as above resulted in p values of 0.013 for Tie-2 and 0.2 for beta-actin. (All t-tests
used α = 0.05)
87
88
Figure 3.5. Determination of HSPB2 protein levels in the three cell types by IB. A) Protein
from the three cell types were detected using isoform-specific antibodies to HSPB2 (27 kDa).
For loading controls, anti-beta-actin antibodies were used. B) Autoradiograms from
chemiluminescent immunoblots were quantified using Image Quant and averaged relative fold
differences in pixel densities are shown. Results of 5 blots are shown with UBE2D3 protein
levels in bmLEC taken to be 1. T-test (α = 0.05) comparing relative levels of UBE2D3 in
bmLECs to those in bmAECs resulted in p = 0.003.
89
90
Figure 3.6. Comparing ANAX2 Protein Levels in ECs of Dissimilar Anatomical Origin.
Determination of Annexin A2 (ANAX2) protein levels in bmLECs, bmVECs, bmAECs, and
bovine aortic endothelial cells (baAECs) by immunoblotting. A) Protein from four cell types
were detected using antibodies to ANAX2 (35 kDa). For loading controls, anti-beta-actin
antibodies were used. B) Autoradiograms from chemiluminescent immunoblots were
quantified using ImageQuant and averaged relative fold differences in pixel densities are
shown. Results of 12 blots are shown with Annexin A2 protein levels in bmLEC taken to be 1.
T-test (α = 0.05) comparing relative levels of Annexin A2 in bmAECs to those in baAECs
resulted in p = 0.0002.
91
92
3.4. Discussion
I have generated and analysed silver-stained electrophoretograms of 2D-PAGE-resolved
proteins from bovine mesenteric ECs extracted from venous, arterial, and lymphatic vessels. I
identified 39 proteins that showed variations in the three cell types with ANOVA p-values
<0.05. Most proteins were found at high levels in both bmAECs and bmVECs (49%).
Similarly, most proteins were found to be present at high levels in both bmLECs and bmVECs
(41%). Of the three cell types, protein profiles of bmLECs and bmAECs were the most
divergent; only 10% were found to be high in both of these cell types. This observation is
consistent with findings by Bianchi and colleagues [161] and with theories on lymphatic lineage
specification from venous lymphatic progenitors during embryonic development [175].
My study design is very similar to that of Bianchi and colleagues [161] but I found little
overlap in the resulting list of proteins. For example, Bianchi and colleagues [161] found
PRDX2 levels to be highest in venous ECs and lowest in arterial ECs. In contrast, I found
PRDX2 levels to be highest in bmLECs but were at similar levels in the blood ECs. Similar to
my results (Table 3.1.A), Bianchi and colleagues found silver stained signal of the spot
corresponding to ANXA2 protein identification to be strongest in baAECs compared to LECs
from bovine thoracic duct and VECs from bovine vena cava [161]. In both my results and
those of Bianchi and colleagues, the trend was the same in that the silver stained signal
corresponding to ANXA2 identification by mass spectrometry in VECs was the lowest of the
three cell types. When I attempted to validate ANXA2 levels in ECs by immunoblotting, the
trend shown by silver stained 2D-PAGE was not reproduced. Perhaps contaminants in spot 94
confounded silver staining signals and/or antibodies commercially available may be cross-
reacting with other members of the Annexin family of proteins that are also differentially
93
regulated between the three cells. Results shown in Figure 5 certainly prove that different
members of the same family of proteins can be differentially regulated in the three cell types.
Overall, the lack of substantial overlap between my results and those of Bianchi and
colleagues [161] may simply be due to differences in anatomical sources of the cultured bovine
ECs used. Bianchi and colleagues [161] isolated VECs from bovine inferior vena cava, AECs
from thoracic aorta, and LECs from thoracic duct. My results shown in Figure 6 indicate that
this may be the case. ANXA2 protein level in bmAECs was statistically significantly higher
than in baAECs. Evidence that same EC types from different anatomical sites have different
gene expression programmes has been reported by Chi and colleagues [159]. Other reasons for
the discrepancy may be at the proteomics and informatics level where it is generally accepted
that search results are somewhat dependent on the instruments used for mass spectrometry, and
highly dependent on the algorithms used for the search [176]. Until consensus is reached on
statistical methods to validate database searching strategies to analyse MS/MS spectra,
independent but similar protein profiling studies may be considered complementary to one
another.
None of the resulting proteins from my analysis of 2D-PAGE resolved profiles can be
considered novel markers of lymphatic, venous, and arterial EC lineages. However, some
trends found in my results are still noteworthy. Consistent with previously reported
observations that ECs in arterial vessels, and not ECs from venous vessels (unless placed under
stress), express HSP70 [177], I found HSP70 to be much higher in abundance in bmAECs as
compared to bmLECs and bmVECs (Table 3.1.A). Also consistent with previously reported
observations that venous ECs express HSPB1 but not HSPA1B under non-stress conditions
[178], I found HSPB1 to be highly abundant in bmVECs as compared to bmAECs and bmLECs
94
(Table 3.1.A). Both HSPB1 and HSPA1B belong to large families of heat shock proteins
(HSP27 and HSP70) that function as molecular chaperones. HSP27 and HSP70 aid proper
protein folding during normal physiology as well as during cellular crises such as heat and
oxidative stress. HSP27 and HSP70 have both been shown to be cytoprotective and potentially
play protective roles in mouse and rat models of brain and heart ischemia [179].
Although the magnitudes of differences in 2D-PAGE silver-stained signals of the three
proteins HSPA1B, HSPB1, and UBE2D3 in the three cell types were not reproduced in my
validation with isoform-specific antibodies, the pattern of differences was mostly reproduced.
HSPA1B was consistently highest in bmAECs, HSPB1 was consistently highest in bmVECs,
and UBE2D3 was consistently highest in bmLECs. Limits of quantification techniques used
may account for the lack of agreement in magnitude of differences in protein levels between the
two methods for HSPA1B and HSPB1. Another possible explanation for the discrepancy may
be the co-migration of proteins that are similarly differentially present in the three cell types.
Differences in the silver-stained protein spots may be amplified while validation of one protein
out of all the proteins present in the spot shows smaller degrees of differences.
Actual changes in the proteomes of the cells may also be a cause; cells used for
validation have been frozen and thawed and passed for a greater number of times than cells
used in the initial 2D-PAGE analysis. Whatever the cause, what is important is that I was able
to reproduce the larger patterns in protein differences as mentioned above.
To the best of my knowledge, the VEGFR family of proteins whose members, for
example: VEGFR-1 versus VEGFR-3/Flt-4, are differentially expressed in different EC types, I
know of only a few other protein families whose members serve as EC-type markers. I report
here a new family of proteins, HSP27, whose members HSPB1 and HSPB2 are expressed at
95
relatively highest levels in VECs and in AECs, respectively. This finding further corroborates
the idea that different members of protein families may play unique roles in the biology of each
of the three EC types. This same idea may also apply to the HSP70 family, to which HSPA1B
belongs, and the E2 ubiquitin conjugation enzyme family, to which UBE2D3 belongs. Further
experimentation is required to confirm whether other HSP70 proteins and other E2 ubiquitin
enzymes are differentially expressed in the three EC types. Understanding the potentially
unique roles of these proteins in EC biology, for example the response of ECs to various
cellular stressors, would be an important avenue of research.
One important stressor relevant to this discussion is chronic hyperoxia. Excess
molecular oxygen or hyperoxia is toxic to cells due to ROS (reactive oxygen species)
generation. At the other end of the oxygen pressure gradient, oxygen and glucose depletion
during an ischemic attack can lead to a chain of molecular events also resulting in ROS
generation and further cellular damage [179]. HSP27 has been shown to directly prevent
protein mis-folding by acting as a sort of antioxidant [180]. Similarly, evidence exists for the
role of HSP70 in protecting cells against ROS-induced cellular damage [181]. Furthermore,
both HSP70 and HSP27 have been shown to be highly inducible in brain cells undergoing
ischemia, epileptic seizure and hyperthermia and associate with cellular resistance to such
insults [182].
Partial pressure of O2 available to healthy cells in the body has been estimated to be 80-
100 mm Hg or 10-12.5% [183]. In a relatively more O2-rich environment such as our tissue
culture incubators at atmospheric O2 partial pressure of 20%, the cultured bovine ECs may be
exposed to chronic hyperoxic conditions, and thereby may be exposed to higher than
physiologically typical levels of intracellular ROS. In fact, lowering oxygen pressure has been
96
shown to result in reduced intracellular ROS levels and ROS-induced damage to in vitro
cultured cells [184]. The elevated HSPB1 and HSPA1B in bmVECs and bmAECs,
respectively, may be evidence of selective pressure of populations of vascular ECs in culture
that can best respond to relatively higher intracellular ROS at atmospheric O2 pressures.
Interestingly, HSP27 and HSP70 are not the only proteins in my list of identified protein
(Table 1) to have an association with cellular responses to ROS. Peroxiredoxins (such as
PRDX2 and PRDX6) are a family of proteins that protect cells from oxidative stress by
catalyzing reduction of peroxides [185]. Ubiquitin conjugating enzymes (such as UBE2N and
UBE2D3) are a family of proteins that play essential roles in the unfolded protein response and
in the removal by proteolysis of oxidized proteins irreversibly damaged [179]. Treating cells
with hydrogen peroxide to generate ROS-induced oxidative stress on cells has been shown to
increase ubiquitin conjugating activity 3.5-9.5 fold and to correlate with increased intracellular
proteolysis [186]. Perhaps the presence of oxidative stress-related proteins such as HSP27 and
HSP70 in bmVECs and bmAECs, PRDX2 in bmLECs, PRDX6 in bmLECs and bmVECs,
UBE2N and UBE2D3 in bmLECs, are evidence of the unique responses of each of the three
cultured bovine mesenteric ECs to elevated internal ROS levels in an artificial environment
with atmospheric O2 levels much higher than physiological levels. The impact of atmospheric
O2 on cultured cells has previously been reported in studies of cultured lymphocytes [187, 188].
The impact of culturing ECs on cellular genetic and phenotypic character has also been
observed by Wick and colleagues [162], Amatschek and colleagues [163], and Veikkola and
colleagues [170]. Wick and colleagues [162] compared the transcriptomes of ex vivo lymphatic
and blood ECs isolated directly from human skin to cultured LECs and BECs from the same
source and found only about 10% of transcripts examined did not alter upon culture.
97
Amatschek and colleagues [163] found a phenotypic drift from BEC phenotype toward an LEC
phenotype after BECs were separated from their in vivo environment and cultured. Veikkola
and colleagues [170] observed the suppression of LEC marker expression in BECs when these
cells were co-cultured with vascular muscle cells. These findings suggest that a genetic and
phenotypic drift occurs during the shift from physiological cellular environments to the
artificial culturing systems in the laboratory setting. Whether there is a predictable pattern or
trend to this genetic and phenotypic drift or whether the drift occurs with a certain levels of
variation contingent upon conditions in particular lab settings is a question that should be
examined carefully. In addition to laboratory variations in methods of EC isolation and
instrumentation used in analysis, the significant lack of overlap between lists of proteins and
transcripts published in comparative studies of various EC types and lineages would suggest
that much of the genetic and phenotypic drift observed may be contingent upon cell type as well
as on unique conditions in particular lab settings.
98
CHAPTER 4
GENERATION OF A MONOCLONAL ANTIBODY SELECTIVELY REACTIVE
TO BOVINE MESENTERIC LYMPHATIC ENDOTHELIAL CELLS
BUT NOT TO BOVINE MESENTERIC VENOUS OR ARTERIAL ENDOTHELIAL CELLS
99
4.1. Introduction
Basic understanding of the dynamic interaction between the lymphatic endothelium with
macromolecules and cellular components of lymph is necessary for the development of
strategies to treat lymphatic-related diseases [2, 189-191]. The deep dearth of this impactful
knowledge lends support to the continuing effort to discover and characterize surface protein
markers of LECs.
One of the methods applied to the differential identification of novel surface markers of
cells is subtractive immunization (SI). SI has been demonstrated to be successful in producing
highly selective antibodies to surface markers of subsets of normal pancreatic cells [192], of
subsets of normal kidney cells [193], of human plasma proteins [194], of epididymis-specific
sperm proteins [195], and of various types of metastatic human tumour cells [196-202]. In
chapter 2, I described a system of three EC types isolated from vessels of bovine mesentery,
bmLEC, bmAEC, bmVEC. In chapter 3, I further characterized the three cell types using a
proteomics approach in the hopes of discovering novel unique markers of bmLECs. Described
here in chapter 4, mice tolerized to bmAECs and bmVECs prior to being immunized with live
bmLECs were expected to have greater potential to produce selective antibodies to lymphatic-
specific surface antigens, thereby help to identify novel surface markers of LECs.
Standard immunization protocols routinely used to generate monoclonal antibodies
(mAb) for laboratory and clinical applications often do not result in antibodies with the desired
selectivity. SI by pre-tolerizing host animals to undesired antigens using cyclophosphamide,
has been successful where standard immunization has not been in producing mAbs targeting
highly specific epitopes within the proteome [203]. For example, SI has proven useful in
generating discriminating mAbs to proteins that are over 90% identical [204, 205]. SI is also
100
thought to be advantageous for use in generating mAbs reactive to low abundant epitopes or
epitopes that are highly similar to ―self‖ antigens and thereby too weakly immunogenic in
standard immunization protocols [206]. I have taken advantage of the promises of SI to
generate a mAb selectively reactive to bmLECs and not to bmVECs or bmAECs using
immunosuppression of BALB/c mice with cyclophosphamide treatment. Initial
characterization efforts showed that the mAb recognizes a protein approximately 180 kDa in
size predominantly expressed by bmLECs and not by bmVECs or bmAECs by immunoblotting
(IB) and immunohistochemistry (IHC). The protein co-fractionates with proteins in GM1-
enriched membrane domains such as caveolin-1and the receptor tyrosine kinases Tie-2, and
VEGRF-3/Flt-4 in bmLECs. For the sake of convenience, I shall temporarily call the protein
ULSP180 for Unidentified Lymphatic Specific 180kDa Protein. Attempts at identifying
ULSP180 by immunoprecipitation followed by gel electrophoresis and tandem mass
spectrometry (MS/MS) were unsuccessful due to the possibility that the protein sequence may
not yet be part of any mammalian protein databases. Further extensive characterization is
remains needed to establish the protein target of this LEC-selective mAb to be a novel marker
of ECs of lymphatic origin.
4.2. Methods
4.2.1 Endothelial Cell Culture
Bovine mesenteric ECs of venous, arterial, and lymphatic origin were isolated and
cultured from vessels of the mesentery as previously described (Chapter 2 [167]) by the method
of Johnston and Walker [131]. Briefly, cells were dislodged by incubation with dispase and
collagenase. Dislodged cells were briefly cultured then colonies of ECs were subjected to
101
limited dilution cloning to ensure EC purity. Cell populations tested positive for Prox-1,
LYVE-1, podoplanin, and VEGFR-3/Flt-4 were cultured in DMEM (Hyclone) supplemented
with 10% FBS (Fisher Scientific) in standard tissue culture conditions: 5% CO2, atmospheric
O2, and 37ºC. Cells were harvested and pooled during late growth phase when confluency had
not been reached. Approximately 5-10 x 107 cells per 500mL PBS (pH 7.4) were i.p. injected
into each mouse for SI. BmAECs and bmVECs served as the immune-telerogen while bmLECs
served as the immunogen.
4.2.2. Subtractive Immunization and Hybridoma Production
SI procedure outlined here has been previously established and described by Sleister and
Rao [205]. Briefly, Freund‘s adjuvant primed BALB/c mice at 4-6 weeks of age were injected
with a 1:1 mix of bmVECs and bmAECs, followed by injections with 2.5 mg of
cyclophosphamide prepared in 0.2mL total volume 2 and 48 hours later. This procedure was
repeated twice more at 2 and 4 weeks before tolerized mice were immunized with bmLECs.
Three booster injections with bmLECs at least 1 week apart was subsequently given. Serum
from blood samples taken by saphenous vein bleeds were screened by the same method
described below (―Antibody Screening‖) for selective reactivity to bmLECs but not to the
bmBECs. The third day after the last booster injection, hybridomas were generated by fusion of
disaggregated mouse spleen cells with the cell line Sp2/0 as previously described [207]. HAT
supplement (Gibco, Invitrogen) was added to the media (IMDM, Gibco, Invitrogen) the day
after fusion. The procedure is summarized in Figure 4.1.
4.2.3. Antibody Screening and Isotyping
Mouse serum from bleeds and tissue culture media supernatant from hybridoma
colonies were screened using an ELISA-like assay developed for the three cell types. Briefly,
102
bmLECs, bmVECs, and bmAECs grown to confluency on U-bottomed 96-well plates (Falcon,
Becton Dickinson) were incubated with serum or supernatant on ice for 30 minutes. Fluids
were decanted and cells were washed at least five times with PBS before incubation with rabbit-
anti-mouse antibodies conjugated to horse radish peroxidase (HRP, Bio-Rad Laboratories) on
ice for 30 minutes. After at least five more washes with PBS, HRP was exposed to the
substrate 3, 3', 5, 5' tetramethylbenzidine (TMB, Sigma-Aldrich) until coloration was visible.
The reaction was stopped by acidification with nitric acid. Optical absorbance at 450 nm was
measured using the EL800 plate reader (BioTek). Validation of the screen was performed using
rabbit-anti-human EphB4 polyclonal antibody (Santa Cruz Biotechnology). Isotype-specific
antibodies used in the same screening procedure to determine the isotype of mAbs were
obtained from Invitrogen and used according to instructions provided. Isotype of mAbs were
confirmed with IsoStrips from Roche Applied Science according to instructions provided.
4.2.4. Immunoblot Analysis
Cells were harvested at confluency and lysed in RIPA buffer (150mM NaCl, 1% Igepal,
0.5% sodium deoxycholate, 0.1% SDS, 50mM Tris, pH 7.4, 1mM EDTA, protease inhibitor
tablets (Roche)). Lysates were cleared and quantified before proteins are separated by SDS-
PAGE. Proteins were resolved on 10% PAGE gels and transferred to PVDF (Perkin Elmer)
membranes. Antibodies used in immunoblots (IB) were: anti-Tie-2 (33.1, BD Biosciences
Pharmingen), anti-VEGFR-3/Flt-4 (C-20, Santa Cruz Biotechnology), and goat-anti-
mouse/rabbit-HRP (Bio-Rad Laboratories). Enhanced chemiluminescence (ECL) substrate was
provided by SuperSignal West Pico Chemiluminescent Substrate or by West Femto
Chemiluminescent (Thermo Scientific) whenever the ECL signal at the Pico detection level was
103
inadequate. IB signals on X-ray films were converted to pixel intensity using Image Quant
version 5.0.
4.2.5. Immunohistochemistry
Confluent layers of ECs were gently lifted off the tissue culture plate and spun down
into a pellet. The entire EC pellet was fixed overnight in 4% paraformaldehyde (PFA) and
dehydrated in ethanol (EtOH) for at least 24 hours before paraffin embedding and microtome
sectioning. Paraffin sections of cells were then stained using the immunohistochemistry (IHC)
protocol available from most antibody manufacturers. Sections of cells were deparaffinised in
55ºC for 10 minutes followed by 2 xylene soakings. Decreasing concentrations of EtOH were
used to rehydrate the sections before permeabilization with 0.1% TritonX-100 in PBS.
Endogenous peroxidase activity was quenched with 3% H2O2 in methanol (MeOH) prior to
blocking with normal donkey serum. Sections were incubated with hybridoma supernatant
containing antibody raised against bmLECs for 30 minutes at room temperature. Secondary
antibody used was biotinylated donkey-anti-mouse immunoglobulin (Jackson ImmunoResearch
Laboratories Inc.). Avidin-peroxidase conjugate (Vectastain ABC kit) and the substrate 3,3‘-
diaminobenzindine (DAB) were applied as instructed by the manufacturer (Vector
Laboratories). Sections were counterstained with methyl green (Vector Laboratories) prior to
dehydration and mounting in Cytoseal-SYL (Richard-Allan Scientific).
4.2.6. ULSP180 enrichment by sucrose-gradient-based cell fractionation
Detergent-free purification of caveolin-rich membrane fractions was performed as
previously described by Song and colleagues [208] with modifications. Cells grown to
confluency in five 15-cm dishes were lifted into PBS, pelleted, and resuspended in ice-cold 500
mM sodium carbonate (pH 11), containing a cocktail of protease inhibitors (Roche Applied
104
Science) . Cells were homogenized in a tight-fitting Dounce homogenizer for 20 strokes, with
an Ultra-Turrax tissue grinder (IKA Works Inc.) for three 10-second bursts, and with a
Sonifwere 250 (Branson) for three 20-second bursts. The homogenate was mixed 1:1 with 90%
sucrose prepared in MES buffered saline (MBS, 25mM MES, pH 6.5, 150mM NaCl). The
sucrose gradient was formed in an ultracentrifuge tube (Ultra-Clear™ 14 x 89mm, Beckman) as
follows from the bottom up: 4mL homogenate (45% sucrose), 6mL 35% sucrose, 2mL 5%
sucrose (both made up in MBS, with 250mM sodium carbonate). The gradient was centrifuged
at 39,000 rpm for 24 hours in an SW41-TI rotor (Beckman XL-70 Ultracentrifuge) at 4ºC.
500uL fractions were collected from the top down and analysed by dot blotting with HRP-
conjugated cholera toxin subunit B (Sigma-Aldrich). Fractions collected around the interface
between 5% sucrose and 35% sucrose contained GM1 gangliosides reactive to cholera toxin.
4.2.7. Immunoprecipitation and Identification of ULSP180 by LC-MS/MS
MAbs selectively reactive to bmLECs were pre-coupled to Dynabeads M-270 Epoxy
(Invitrogen) pre-coated with rabbit-anti-mouse polyclonal antibody (Abcam Inc.) as indicated
by the manufacturer‘s instructions. 107 magnetic beads (Dynabeads M-270 Epoxy) were coated
with at least 3µg of the rabbit cross-linking antibody in 1mL of 1M ammonium sulphate in
PBS. 106 magnetic beads coupled to mAbs selective for bmLECs were then used to
immunoprecipitate ULSP180 from dialysed sucrose fractions described above.
Cells grown to confluency in fifteen 15-cm dishes were subject to the sucrose gradient
described above with the following changes: the sucrose gradient was formed in a larger
ultracentrifuge tube (Ultra-Clear™ 25 x 89mm, Beckman) as follows from the bottom up:
16mL homogenate (45% sucrose), 16mL 35% sucrose, 8mL 5% sucrose. The gradient was
centrifuged at 28,000 rpm for 24 hours in an SW28 rotor (Beckman XL-70 Ultracentrifuge) at
105
4ºC. One mL fractions were collected from the top down. Fractions containing ULSP180 were
pooled from four such gradients and dialysed into PBS prior to immunoprecipitation (IP) with
the bmLEC-selective mAb as described above. IP reactions contained approximately 2mg in
total protein. During the IP procedure, magnetic bead precipitation was performed using a
magnetic particle concentrator (DynaMag-Spin). Immunoprecipitates were eluted off the beads
with 100µL 10% SDS. 2x Laemmli sample buffer was added to a total volume of 200µL
before separation by SDS-PAGE (large gel format) and visualized by silver staining (Silver
Stain Plus, Bio-Rad). Gel regions of interest were excised and destained with 1% H2O2 prior to
in-gel digestion with sequencing-grade trypsin (Promega).
Peptides were extracted, and analyzed by LC-MS/MS (1100 nanoflow LC system, XCT
plus ion trap, Agilent Technologies). Spectrum Mill proteomics workbench (Agilent) was used
to match MS/MS spectra to non-redundant bovine protein sequences in the NCBInr database
(NCBI non-redundant downloaded in 2010, 29624 sequence entries). Unmatched spectra were
returned to another round of MS/MS search but this time against mammalian protein sequences
in SwissProt database (downloaded in 2010, 55449 sequence entries). Unmatched spectra after
the second round of MS/MS search were subjected to Spectrum Mill de novo sequencing
powered by the Sherenga algorithm [209]. In general, default settings were used. Peptide
scores >10, protein scores >11, and Scored Peak Intensity (SPI) values >70% and Sherenga
scores >150 were deemed acceptable for further automated and/or manual validation. Peptides
from in-gel-digestion of bmVEC whole cell lysate proteins and from mock IP with bmVEC
sucrose fractions, separated on the same gel helped to discriminate between potential MS/MS
match of ULSP180 and contaminating co-IP proteins.
106
4.3. Results
4.3.1 Hybridoma Production and Screening
In order to develop monoclonal antibodies (mAbs) selectively reactive to bmLECs and
not bmVECs or bmAECs, the method of SI by tolerization with cyclophosphamide treatment as
outlined in figure 4.1A was employed. The method of mAb generation and screening strategy
described here were devised by adapting procedures previously described by Sleister and Rao
[205]. Instead of purified immunogen absorbed on to 96-well plates, cultures of the three
bmECs were grown in separate replicate wells in a 96-well U-bottom plate. mAbs selective for
antigen on the surface of bmLECs would be expected to bind the ECs and be retained on the
bottom of the plate similar to what occurs in an ELISA assay. This screening strategy was
intended to detect presence of bmLEC-selective mAbs in mouse test-bleed serum and in
hybridoma tissue culture supernatant. In figure 4.1.B, the screening method described was
validated using antibodies that recognize the extracellular portion of the receptor tyrosine
kinase Ephrin B4 (Eph B4) known to be highly expressed specifically by bmVECs [210]. The
screening strategy proved to be robust since the optical absorbance level at 450nm produced by
TMB substrate exposed to HRP was most intense in wells holding bmVECs (Figure 4.1.B).
The biggest concern with this strategy was that extensive washing of live cells with even
small amounts of detergent would destroy the cells. Without extensive washing, it was
expected that the high background 450nm absorbance level would mask the signal of interest.
Furthermore, any signal of interest would be expected to be less robust than that produced by
purified antigens in the case of classical ELISAs—given the number of antigen per well in the
strategy is limited by the number of cells grown in each well. The resulting signal to noise ratio
was not significantly low in the screen validation (Figure 4.1.B). With anti-EphB4 antibodies,
107
the absorbance was about 0.14 units, only slightly higher than 0.1 units of the secondary
antibody alone controls. Nevertheless, a small but statistically significant difference (p<0.05)
was detectable. The advantages of this screening strategy are considered in the discussion.
The strategy validated in Figure 4.1.B was used to screen serum from test-bleeds of
mice immunized with bmLECs and/or pre-tolerized with a 1:1 mix of bmVECs and bmAECs
(Figure 4.2.A and 4.2.B). In figure 4.2.A, serum from a mouse that was not pre-tolerized to the
bmBECs failed to show selective reactivity to bmLECs even after the third boost. In figure
4.2.B, serum from a mouse that was pre-tolerized to the bmBECs showed selective reactivity to
bmLECs after the third boost. Spleen from the latter mouse was subsequently used to generate
hybridomas. Interestingly, serum reactivity to bmVECs was consistently higher than to
bmAECs correlating with the greater lineage similarity between VECs and LECs than between
AECs and LECs [211].
Figure 4.2.C shows the results of tissue culture supernatant from hybridoma colonies
screened for bmLEC-selective reactivity. Out of over 850 colonies screened, only a handful
(highlighted in colour) showed selective reactivity to bmLECs and not to the bmBECs (Figure
2C). Unfortunately, not all of these colonies survived further clonal expansion. Hybridoma
8C8 and 5D5 produced the most robust signals in the screen but because 5D5 did not survive,
8C8 was chosen for further analysis.
108
Figure 4.1. Screening Assay Validation. A/ The subtractive immunization procedure used to
generate mouse monoclonal antibodies selectively reactive to bmLECs and not to bmVECs or
bmAECs is summarized. B/ The hybridoma ELISA-like screen was validated using a
polyclonal anti-EphB4 antibody that targets the extracellular portion of EphB4. The antibody
selectively binds to bmVECs grown on U-bottom 96-well plates. Absorbance of light at
wavelength of 450nm is produced by TMB substrate in the presence of HRP provided by
conjugated goat-anti-rabbit antibodies. Error bars represent triplicate wells in the screen.
109
110
Figure 4.2. Serum and Hybridoma Screening. A/ A control mouse was immunized with
bmLECs but not pre-tolerized to bmVECs and bmAECs. Mouse serum was collected after the
first immunization and the third boost. This control mouse failed to develop and release
selective antibodies for bmLECs into blood serum. B/ Pre-tolerized mouse showed evidence of
presence of antibodies selective for bmLECs in its serum. NB: Samples in panels A and B were
analysed the same day under the same conditions. Error bars in A and B represent triplicate
wells in the screen. C/ 100 representative hybridoma colonies are shown out of over 850
colonies screened. Many hybridoma colonies depicted in grey showed no selective reactivity to
bmLECs. The modest few showing selectivity only bmLECs is highlighted in colour.
Secondary antibody only control is shown in fuchsia.
111
112
4.3.2. MAb isotyping
The mAb from hybridoma sub-clone of colony 8C8 was found to be of isotype IgG1-
kappa by two different methods. In figure 3A, HRP-conjugated secondary antibodies raised
against various types of mAbs were used in the same 96-well screening method described in
figures 4.1.B and 4.2. The advantage of this method was that the isotype of the mAb that
recognizes bmLECs would be assayed and any unwanted Abs present in the hybridoma culture
supernatant would be washed away. The disadvantage was that any kind of contamination due
to failed establishment of clonal colonies may not be detectable. In figure 4.3.B, direct
absorbance of hybridoma culture supernatant to commercially available isotyping strips was
performed. It was frequently observed that before clonal expansion, mixed colonies of
hybridomas showed traces of IgM detectable by the isotyping strips but not by the 96-well
screening method. Single cell cloning was sufficient to eliminate the IgM contaminating
antibodies.
4.3.3. MAb is Selective for bmLECs
To determine whether the 8C8 mAb recognizes formalin fixed antigen on bmLECs,
pellets of bmECs were fixed, embedded in paraffin and sectioned. An antibody of the same
isotype as 8C8 but with no reactivity toward any bmECs (4H4 mAb) provided the negative
control mAb (Figure 4.4.A). Above-background staining of bmVECs by the 8C8 mAb (Figure
4.4.C), round peri-nuclear structures appeared darkly stained in bmLECs (Figure 4.4.B). In the
left panel showing one sectioning orientation of bmLECs, the peri-nuclear structure reactive to
8C8 mAb seemed to be present only once per lymphatic cell (Figure 4.4.B). 8C8 mAb also
stained the extracellular space on the apical side of the bmLEC monolayer as visible in the
sectioning orientation shown on the right panel of figure 4.4.B.
113
Figure 4.3. MAb Isotyping. The mAb from hybridoma sub-clone of colony 8C8 was isotyped
using two methods: A/ Using isotype-specific secondary antibodies in the same screen
described in Figure 4.1.B, and B/ Direct absorbance of the antibody on commercially available
isotyping strips. Both methods show 8C8 mAb to be of isotype IgG1-kappa.
114
115
Figure 4.4. Antibody Validation. Sections of bmEC pellets embedded in paraffin were used
to determine whether 8C8 mAb is reactive to formalin-fixed antigen and therefore amenable for
use in IHC. A/ 4H4 is a mAb from a hybridoma colony found in the initial screen to be non-
selective for bmLECs. 4H4 is used here as a negative control to help determine the level of
background staining. B/ Two different sectioning orientations of bmLECs are shown. The dark
round peri-nuclear staining that appears once per cell is visible in the left sectioning orientation
(arrow head). The darkly stained extracellular space on the apical side of a bmLEC monolayer
is visible in the right sectioning orientation (arrow head). C/ 8C8 mAb is not reactive to
bmVECs. Microscope magnification of 40x was used.
116
117
4.3.4. 8C8 mAb recognizes a protein of MW approximately 180kDa
In order to ascertain whether 8C8 mAb recognizes denatured antigen in membrane
fractions of bmLECs but not in those of bmVECs, proteins from sucrose gradient fractions of
bmLEC and bmVEC lysates were separated by SDS-PAGE and subjected to immunoblotting
with 8C8 mAb. EC lysates were fractionated to enrich for membranous proteins prior to SDS-
PAGE and IB. A step sucrose gradient was prepared with cell lysates in 45% sucrose at the
bottom of the gradient. Ultracentrifugation resulted in floatation of low density membrane
fractions to the interface of 35% and 5% sucrose steps in the gradient. This method was
developed to isolate caveolin-enriched plasma membrane domains [208, 212], however, the
alkalinity of the sodium carbonate buffer used during cell lysis (pH 11) results in a high
percentage of non-raft membrane proteins to accumulate at the 35%-5% sucrose interface [213]
where GM1 gangliosides also migrate (Figure 4.5.A). Thus, the method developed by Song and
colleagues [208, 212] is a convenient method of fractionating cells to enrich for membranous
proteins, wherever in the cell those membrane fractions may have originated [213].
In figure 4.5.B, 8C8 mAb used in the IB did recognize a protein of approximately
180kDa that co-migrated with GM1 gangliosides. The protein will henceforth be provisionally
named ULSP180 for Unidentified Lymphatic Specific Protein of 180 kDa. While this
experiment does not allow for the conclusion that ULSP180 is a raft associated protein, it does
provide evidence that ULSP180 is likely to be a membranous protein. The only known
lymphatic-specific membranous protein marker that resolves around the same molecular weight
(MW) region on an SDS-PAGE gel is VEGFR-3/Flt-4. However, in figure 4.5.C, anti-VEGFR-
3/Flt-4 antibodies failed to produce a signal of the same MW as ULSP180. The loading control
was provided by Tie-2, which is expressed in both bmLECs and bmVECs (Figure 4.5.D).
118
These results indicated that fractionation by step sucrose gradient is a viable method of
enriching for ULSP180 for further analysis and possible identification or sequencing. The work
flow from lab bench to in silico protein sequence analysis is detailed in figure 4.6.
119
Figure 4.5. ULSP180 Characterization. Proteins from sucrose gradient fractions of bmLEC
and bmVEC lysates were separated by SDS-PAGE and immunoblotted using antibodies raised
against proteins as shown. A/ Fractions 4 and 5 (500 µL collected from top down) were
enriched for GM1 gangliosides, which binds to HRP-conjugated cholera toxin sub-unit B in this
dot blot. B/ 8C8 mAb recognizes a protein with MW of approximately 180 kDa (temporarily
called ULSP180) co-migrating with GM1-positive membrane domains enriched in fractions 4
and 5. C/ The lymphatic marker VEGFR-3/Flt-4 is present only in bmLECs but not bmVECs.
VEGFR-3/Flt-4 does not co-migrate as tightly as ULSP180 does with GM1-positive membrane
domains in the gradient. D/ Tie-2 is present in both bmLECs and bmVECs. Tie-2 also does not
co-migrate as tightly as ULSP180 does with GM1-positive membrane domains in the gradient.
120
121
Figure 4.6. Schema of Mass Spectrometric Analysis of ULSP180. Depiction of the work
flow from It lab bench to in silico analysis of MS/MS spectra by Spectrum Mill software and
return to It lab bench for protein validation. Not shown here are the multiple rounds of MS/MS
searching possible to account for modifications such as phosphorylation that may result in
initial uninterpreted spectra. ‗Hopelessly‘ uninterpretable spectra can be subjected to Sherenga-
powered de novo sequencing before the return to classic biochemistry methods.
122
123
4.3. 5. Mass Spectrometric Analysis of ULSP180
Tandem mass spectrometry (MS/MS) is a powerful method of obtaining protein
sequence information from an unidentified protein from a sample, provided that the mixture of
protein in the sample has been sufficiently simplified [214]. The ability to fractionate bmLECs
and produce fractions enriched in ULSP180 (Figure 4.5) was a large step toward the potential
identification of ULSP180. Furthermore, collected gradient fractions containing ULSP180 can
be located by a simple dot blot with cholera toxin subunit B conjugated to HRP (Figure 4.5).
Quantification of IB signals of ULSP180 from blots of sucrose fractions showed virtually 100%
of the signal was distributed around the 35%-5% sucrose interface where approximately 8% of
the total protein in the gradient migrated (Figure 4.7.A). Fractions containing ULSP180 were
pooled and dialysed into PBS (MW cutoff 10kDa) before IP with 8C8 mAb bound to magnetic
beads coated with rabbit-anti-mouse cross-linking antibodies (Figure 4.7.B). Undialysed
sucrose fractions were not amenable to IP (Figure 4.7.B, lane 10). Direct IP with 8C8 mAb and
Protein G Sepharose was also likewise unsuccessful (Figure 4.7.B, lane 6).
In figure 4.7.C, ULSP180-enriched fractions from a scaled up sucrose gradient were
pooled, dialysed, and subject to IP with 8C8 mAb. The same sucrose gradient was produced
from bmVECs processed identically to bmLECs. Total protein in each IP was approximately
2mg. Eluted immunoprecipitates separated by SDS-PAGE and visualized by silver stain. A
faint band present only in the bmLEC IP appeared instead of the anticipated strong band where
ULSP180 (Figure 4.7.C). Notwithstanding, thin sections of the gel corresponding to where
ULSP180 resolved were excised and subjected to MS/MS analysis. Corresponding gel regions
of SDS-PAGE-resolved bmVEC proteins were likewise processed and served as a negative
control.
124
Two high scoring MS/MS mass spectra derived from tryptic peptides in bmLEC IP
samples but not detectable in bmVECs are shown in figure 4.8. One spectrum matched to the
theoretical MS/MS spectra of bovine Proteolipid 2 (PLP2, Figure 4.8.A). Surprisingly, PLP2 is
a 17kDa protein. A mismatch is highly unlikely considering the decoy database resulted in no
match and that 89.9% of spectral peak-detected ion current explained by the search
interpretation (SPI%, Figure 4.8.A). The other spectrum in figure 4.8.B was not adequately
interpreted in the first round of MS/MS search using the bovine database. The spectrum
matched to the theoretical MS/MS spectra of mouse Complement protein C3 during the second
round of searching using the mammalian database (Figure 4.8.B). C3 has a MW of 186kDa,
expected of a protein resolving in the region of ULSP180 on an SDS-PAGE gel. A mismatch is
highly unlikely considering the decoy database resulted in a poor match (forward-minus-reverse
score of 11.74) and that SPI% was 96.5% (Figure 4.8.B). Whether these matches are protein-
interacting partners of ULSP180 or whether they provide clues as to the true sequence of the
protein requires further validation with PLP2 and C3-specific antibodies. Until further
validation, ULSP180 remains ‗unidentified.‘
A large number of spectra were not interpretable by MS/MS searching and were subject
to de novo sequence analysis. Figure 4.9.A shows an example of a MS/MS spectrum
unsuccessfully matched during the initial MS/MS searches. Figure 4.9.B shows the same
MS/MS spectrum successfully sequenced by Sherenga (Sherenga score 178). However, the
amino sequences derived this way was not successfully matched to any protein in databases
using MS-Blast [215] by the Suyaev lab.
125
Figure 4.7. Enrichment and Immunoprecipitation of ULSP180. A/ One mL fractions were
collected from the top of the gradient and protein quantities were determined for each fraction.
The 8C8 mAb was used to detect ULSP180 in the different fractions by IB. IB signals of
ULSP180 were converted to pixel intensity using ImageQuant® version 5.0. Virtually 100 %
of the ULSP180 signal appeared in sucrose gradient fractions 2, 3, and 4, which—when
combined—contained approximately 8% (±1%) of total protein mass in the gradient. Error bars
represent 4 independent gradients. B/ IP for ULSP180 by 8C8 mAb from pooled fractions 2, 3,
and 4, was optimal with the use of magnetic beads covalently linked to a rabbit-anti-mouse
cross-linking antibody (Lane 4). Protein G sepharose failed to immunoprecipitate ULSP180
(Lane 6). Sucrose gradient fractions had to be dialysed overnight into PBS (pH 7.4) using
dialysis tubing with MW cut-off of 10kDa (Lane 10). Samples of dialysed bmLEC sucrose
fractionate before and after IP were loaded in lanes 2 and 8. An 8% SDS-PAGE gel was used
for lanes 1-8, 10% was used for lanes 9-11. BmVECs were identically treated and sucrose
fractionate of bmVECs does not show presence of ULSP180 (Lanes 1, 3, 5, and7). In both A/
and B/, ECL substrate was provided by Ist Femto Chemiluminescent Substrate (Pierce). C/
Scaled up sucrose fractionate of bmVECs and bmLECs were subject to IP with the 8C8 mAb.
Immunoprecipitates were eluted and separated by an 8% large format SDS-PAGE gel.
ULSP180 did not appear as a strong band by silver staining and was only as visible as the
myoferlin band indicated.
126
127
Figure 4.8. Mass Spectrometric Analysis of ULSP180. Sections of the gel in Figure 7C that
corresponded to where ULSP180 resolved were excised and subjected to in-gel digestion with
trypsin. Extracted tryptic peptides were separated by HPLC and analysed by tandem mass
spectrometry (MS/MS). Tryptic peptides from corresponding gel regions of SDS-PAGE
resolved bmVEC proteins served as a negative control. MS/MS mass spectra of two tryptic
peptides in bmLEC IP samples but not detectable in bmVECs are shown. A/ The MS/MS
spectrum of one of the two peptides matched to the theoretical MS/MS spectrum of a segment
of a 17kDa transmembrane protein called Proteolipid 2 in the bovine NCBI database. MS/MS
search of the forward database did not result in a second ranked match (Rank 2 score of 0).
MS/MS search of the decoy database did not result in any match (Reverse match score of 0). B/
The MS/MS spectrum of the second peptide matched to the theoretical MS/MS spectrum of a
segment of a 186 kDa protein excreted into the extracellular space called Complement C3 in the
mouse SwissProt database. MS/MS search of the forward database did not result in a second
ranked match (Rank 2 score of 0). MS/MS search of the decoy database resulted in a poor
match (Reverse match score of 6.09). NB: In both A and B, black peaks are unmatched ion
peaks whereas labelled and coloured peaks are successfully matched ion peaks.
128
129
Figure 4.9. Mass Spectrometric Analysis of ULSP180 (continued). A/ Another MS/MS
mass spectrum derived from tryptic peptides in bmLEC IP samples but not detectable in
bmVECs is shown. This particular spectrum is one example of a spectrum that did not match
well to any theoretical MS/MS spectra in the databases. Black peaks indicate unmatched ion
peaks. B/ The protein sequence was procured from the MS/MS spectrum in A/ using the
Spectrum Mill de novo peptide sequencing function. Coloured peaks are successfully explained
ion peaks whereas black peaks are not explained by the proposed sequence shown at top. NB:
Isoleucine (I) has the same mass as leucine (L) and is indicated as isoleucine even though
Sherenga cannot distinguish between the two amino acids. The same thing can be said of lysine
(K) and glutamine (Q), indicated as glutamine.
130
131
4.4. Discussion
I have attempted to discover a protein marker of LEC that is absent from ECs of venous
and arterial origin using subtractive immunization (SI). SI yielded a small number of
hybridoma clones producing mouse monoclonal antibodies (mAb) selectively reactive to an
antigen on the surface of bmLECs. Initial tests of the mAb from hybridoma colony 8C8
showed it to be non-reactive to bmVECs. 8C8 mAb recognizes a protein of approximately
180kDa in apparent size on an immunoblotted SDS-PAGE gel. The protein, temporarily named
ULSP180, localizes to a peri-nuclear structure as well as to the plasma membrane on the apical
side of a bmLEC monolayer. ULSP180 is highly soluble and can be immunoprecipitated from
detergent-free sucrose fractions dialysed into PBS. The low silver signal of IP products at the
expected size on the SDS-PAGE gel may be due to low IP protein yield or the possibility that
ULSP180 may not efficiently reduce silver (Figure 4.7.C). Initial attempts to perform mass
spectral analysis on ULSP180 resulted in two potential database matches that require further
validation. Taken together, SI showed tremendous potential as a viable procedure for
discovering novel candidate surface markers of LECs.
Theoretically, an ideal protein marker of LECs would be (i) absent from all sub-types of
ECs of venous and arterial lineages, (ii) absent from lymphatic vessel pericytes, LMCs,
fibroblasts, surrounding connective tissues, and other unrelated cell types (iii) present in most
anatomical subtypes of LECs in most in vivo biological and in vitro cell culture contexts, (iv)
useful for marking LECs in a variety of species, (v) useful for marking LECs at most
developmental stages. Realistically, even the best known LEC markers have been found to be
expressed in other cell types (carefully reviewed by Baluk and McDonald [116]). For
examples, FOXC2 and podoplanin are also found in podocytes [114, 216]; Prox-1 and
132
podoplanin are present in certain types of nerve cells [217, 218]; LYVE-1, Prox-1, and
VEGRF-3/Flt-4 have all been detected in mesothelial cells [110].
Beside marker infidelity, another cause for the difficulty in trying to discover a genuine
LEC protein marker using in vitro EC cultures may be the gene expression shift between LEC
and BEC cell types encapsulated by the term ―cellular plasticity.‖ For example, BECs have
been observed to shift to a more LEC phenotype upon isolation and culture in vitro [163].
When BECs were isolated and co-cultured with vascular muscle cells, LECs markers up-
regulated in the cultured BECs were repressed [170]. The impact that isolating and culturing
cells have on the genotype and phenotype of those cells was observed previously by Wick and
colleagues [219]. They noted that only about 10% of transcripts studied retained the same
differential profile between LECs and BECs before and after culture [219], suggesting that the
shift from physiological cellular environments to the artificial culturing systems in the
laboratory setting leads to measurable changes in gene expression. However, the highly stable
expression of VEGFR-3/Flt-4 by isolated bmLECs and its absence in bmVECs in my studies
suggest that if a protein were to be discovered and validated to have selective expression in
LECs and not in the BECs then its expression would also likely to be highly stable regardless of
in vitro culture conditions.
Another cause of the difficulty of discovering a LEC marker that meets all criteria (i)-
(v) is the phenotypic variability of LECs contingent upon anatomical location. Chi and
colleagues reported evidence that the same EC type from different anatomical sites have
different gene expression programs [159]. Add to this complexity the fact that the same LEC
subtype from the identical anatomical site may be responding to different physiological or
pathological stimuli such as inflammation and injury, and the task of defining a consistent LEC
133
phenotype becomes even more difficult. In addition to striving to arrive at an amino acid
sequence for ULSP180, extensive IHC profiling of the presence of the protein in a variety of
adult and embryonic tissues is also required in order to establish ULSP180 as a bona fide novel
surface marker of ECs of the lymphatic lineage. The challenges of performing mass spectral
analysis on ULSP180 add to the above difficulties in trying to discover a truly novel protein
marker of LECs.
If subsequent validation supports the conclusion that ULSP180 is a very small protein
with a theoretical MW of 17kD, the observed 180kDa may be a result of heavy post
translational modifications like glycosylation. Indeed, proteins that are comprised of 78%
carbohydrate have been observed and studied in the past, for example, mucins and salivary
glycoproteins [220, 221]. However, the much larger difference in observed and theoretical MW
would suggest that the protein mass is at least 90% due to covalently linked groups on the
amino acid chain. To the best of my knowledge, ULSP180 may be the first of this kind of
protein if validation experiments support such outcome. If validation supports the outcome that
ULSP180 is a complement protein, then the sequence of this complement protein—unmatched
with the approximately 600 sequences of complement proteins already in the database—may
become a new addition. Uninterpretable spectra unsuccessfully matched even after being
subjected to Sherenga de novo sequencing points to the third possibility that ULSP180 may
simply be an ―unknown‖ protein yet to be part of any protein database. If subsequent validation
proves this to be the case, then the outcome would become an example of a drawback of
database-driven methods to discover novel proteins.
These limitations are altogether not insurmountable given the availability and reliability
of classical biochemical methods to characterize ULSP180. Moreover, SI still stands as a
134
viable method in endeavours to discover novel markers of LECs. Future repeat of the
procedure described in this study would be improved by considering ways of increasing the
sensitivity of the screening method. I have mentioned the limit placed on the number of
antigens by assaying live cells. Disaggregated plasma membrane proteins adsorbed directly to
the bottom of 96-well plates may help increase the sensitivity of the assay. However, the
advantage of using live cells—as was done in this study—was that antibodies would also be
screened for reactivity to unaltered antigen and thereby potentiates the use of the antibody for
sorting live cells prior to tissue culture. Another way to improve the SI procedure would be to
increase the antigenicity of LECs. Some examples of refinement include using mAbs generated
from the first round of subtractive immunization to block identified antigenic sites and/or
employing neuraminidase treatment as previously described [222] to expose hidden antigens on
the surfaces of LECs. The screening procedure would need to be adjusted accordingly. Despite
these limitations, SI should still be considered a useful tool for LEC marker discovery.
There exists a strong consensus that identifying phenotypic regulators of individual cells
of LEC and BEC lineages is an important avenue of research [116]. Despite many drawbacks
to in vitro cultures and available methods, EC biology researchers continue to search for new
bona fide markers and meticulously profile expression patterns of the few LEC makers known
to date. Yet another way to mitigate the problem of limited-exclusivity in selectivity is the
reliance on detection of multiple markers in the cells of interest before they are declared as
being of the lymphatic endothelial cell lineage. Once properly validated, LECs can be used to
determine the biological functions of identified candidate protein markers. The continued effort
by our laboratory and others to identify and functionally characterize new markers of LEC is
135
important for deepening understanding of the LEC phenotype and molecular determinants of
the cell lineage.
136
CHAPTER 5
CONCLUSIONS AND FUTURE DIRECTIONS
137
To address the question of what molecular signatures give rise to the genotypic and
phenotypic character of the lymphatic endothelial cell lineage distinguishable from blood
endothelial cell lineages, I began with the isolation and culturing of ECs from lymphatic,
venous, and blood vessels from bovine mesentery. The resulting cells bmLEC, bmVEC and
bmAEC appeared to behave as expected of cells of their respective lineages. This was
evidenced by the stable expression of endothelial receptor tyrosine kinase Tie-2 and of
appropriate cellular markers Prox-1, VEGFR3, and Neuropilin-1 that define the particular
origin of each cell preparation. Furthermore, I showed in chapter 2 that while bmLECs
responded slightly more readily to Ang-2 stimulation, bmVECs and bmAECs were more
sensitive to Ang-1 stimulation. Exposure of bmLECs to Ang-2 induced marginally higher levels
of proliferation and survival than did exposure to Ang-1, whereas exposure to Ang-1 resulted in
higher levels of migration in bmLECs than did to Ang-2. These results suggest that each EC
type possesses a unique repertoire of expressed proteins that allows that cell type to
differentially respond to the angiopoietins.
To answer the question of what molecular signatures make lymphatic ECs different
from arterial or venous ECs and perhaps explain the observed behaviour of bmLECs distinct
from the bmBECs as described in chapter 2, I compared 2D-PAGE-resolved protein profiles of
bmLECs to those of bmVECs and bmAECs. Chapter 3 is a record of the results of quantitative
comparison of 2D-PAGE electrophoretograms produced in triplicate for each of the three cell
types. Protein identification by LC-MS/MS was performed to identify 39 proteins found to be
present at statistically significantly different levels in the three cell types (p<0.05). Most of the
39 proteins have not been previously reported in EC proteomic studies of 2D-PAGE
electrophoretograms. Three proteins, HSPA1B (HSP70 family member), HSPB1 (HSP27
138
family member), and UBE2D3 (a member of E2 ubiquitin-conjugating enzymes) were validated
and found to be at highest levels in bmAECs, bmVECs, and bmLECs, respectively.
The lack of substantial overlap between my results and those of other groups‘
comparative studies implies that a genetic and phenotypic drift occurs during the shift from
physiological cellular environments to the artificial culturing systems in the laboratory setting.
It is possible for this shift to further confound experimental results obtained from in vitro EC
cultures that may already be incredibly phenotypically divergent according to anatomical
origin. However disheartening this finding may be, it still should not be reason for complete
dismissal of the cell system I built since with regards to the angiopoietins, the cells seemed to
behave as expected from in vivo studies [43, 44, 55, 56]. Another important fact to note is that
despite evidence of artificially-induced changes to the EC proteome, expression of VEGR-3, a
more or less reliable lymphatic protein marker [103], remained high in bmLECs and was not
detectable in bmVECs or bmAECs despite long-term culture of the cells. Perhaps this suggests
that a similarly reliable marker that contributes to the maintenance of the lymphatic phenotype
is still discoverable with a cell system like one I have built. However, extensive post-discovery
validation necessarily follows any such discovery.
In chapter 4, I described an adapted protocol of standard immunization used to generate
monoclonal antibodies selectively reactive to bmLECs and not to bmVECs or bmAECs. The
strategy was to immunosuppress mice with cyclophosphamide treatment after injections of
bmVECs and bmAECs so that the mice would fail to develop antibodies that recognize shared
antigens on the surface of all three cell types. Initial characterization efforts showed that the
mAb recognizes a protein approximately 180 kDa in size predominantly expressed by bmLECs
and not by bmVECs or bmAECs by immunoblotting and immunohistochemistry. The protein,
139
tentatively named ULSP180, co-fractionates with GM1-enriched membranes and the receptor
tyrosine kinases Tie-2, and VEGRF-3/Flt-4 in a step sucrose gradient. Further characterization
would be required to establish the protein target of this LEC-selective mAb to be a true marker
of endothelial cells of lymphatic origin. Nevertheless, these initial results do tentatively support
the three-cell system set up and SI as promising tools for the identification of novel markers and
phenotypic regulators of LECs.
The next steps should involve answering the following questions about ULSP180:
Is the protein absent from other sub-types of ECs of venous and arterial lineages, and present in
other sub-types of LECs? For example, what is the expression of ULSP180 in microvascular
ECs from blood capillaries of the two systems? Is the protein absent from lymphatic vessel
pericytes, lymphatic muscle cells, smooth muscle cells, fibroblasts, and connective tissues
surrounding vessels? Is ULSP180 present in which anatomical subtypes of LECs in which in
vivo biological and in which in vitro cell culture contexts? Is ULSP180 present in LECs in a
variety of species including human? Is ULSP180 present in LECs during early developmental
stages? Questions listed here provide basis for validation experiments before ULSP180 can be
declared a new marker of ECs of the lymphatic lineage. Further characterization should be
aimed at elucidating its potential role as a molecular regulator of the LEC phenotype.
To answer some of these questions about ULSP180, the ability of 8C8 mAb to
recognize a 180kDa protein from immunohistochemical sections of adult mouse, rat, bovine,
sheep, and human tissues from a variety of organ sites that are co-stained with known LEC-
specific antibodies need to be verified. Alternatively, freshly isolated and uncultured ECs from
vessels belonging to different anatomical sites can be examined directly by flow cytometry
using 8C8 mAb conjugated to fluorescent tags. Developmental arrays of mouse tissues can be
140
used to establish whether ULSP180 is present at earlier developmental stages and therefore may
play regulatory roles in the development and maturation of the lymphatic system. These
experiments should aim to generate an anatomical map of where ULSP180 can be found to be
expressed.
None of the experiments proposed thus far requires the identity of ULSP180 to be
known. Alternatively, an amino acid sequence of ULSP180 should be arrived at first either by
Edman degradation or by refinement of my initial attempts to perform mass spectral analysis on
the protein. Optimization efforts should be focused on increasing the yield of
immunoprecipitated ULSP180 visualized as a silver-stained protein band on an SDS-PAGE gel.
Beside the possibility that ULSP180 is yet to be part of any protein database, the difficulty of
obtaining an unambiguous match with a candidate from the bovine protein sequence database
may also have been due to low IP yield combined with the challenges of performing mass
spectral analysis on an integral membrane protein using reversed-phase HPLC. If ULSP180 is
in fact an integral membrane protein, refinements to HPLC such as previously proposed by Ball
and colleagues [223] may help to improve recovery of separated trypsin cleavage fragments,
thereby increasing the chance of obtaining sequence information on the protein.
The advantage of knowing the protein sequence firsthand is the availability of PCR-
based methods to generate data from transcripts amplified from small amounts of tissue, tissues
even as small as laser-capture micro-dissected tissues. Another advantage of working with a
known protein amino acid sequence is the availability of bioinformatics tools to analyse or
predict homologous protein domains, conserved functional domains, and various types of
consensus or signal sequences indicative of protein localization and function.
141
In future endeavours to continue the identification and characterization of LEC-specific
molecules along similar lines, there is no reason to limit the source of ECs to lymphatic and
blood vessels of the mesentery. In fact, there is value in understanding the LEC phenotype in
specific anatomical contexts. The development of drug delivery strategies utilizing lymphatic
vessels and lymph nodes to increase bioavailability of protein-based drug compounds and
immunoactive biomolecules would greatly benefit from an understanding of how the lymphatic
endothelium at different anatomical sites absorb, process, and interact with luminal molecules
during lymph transport and formation [224]. Understanding the lymphatic phenotype during
inflammation and during wound healing would aid the development of therapies for chronic
inflammatory and chronic lymphatic insufficiency, respectively [2]. Design of pharmaceuticals
to enhance or mitigate lymphatic functions for various pathological conditions, would also
greatly benefit from development of in vitro systems that helps to predict the response of
different endothelia to therapeutic molecules. I have described such as system in these pages.
As the spectrum of human disorders involving the lymphatic system continues to expand day by
day, the rewards of understanding the biology of this unique system are bountiful. I hope to
have contributed even in the smallest way to this collective understanding.
142
REFERENCES
1. Skandalakis JE, Skandalakis LJ, Skandalakis PN: Anatomy of the lymphatics. Surg
Oncol Clin N Am 2007, 16(1):1-16.
2. Tammela T, Alitalo K: Lymphangiogenesis: Molecular mechanisms and future
promise. Cell, 140(4):460-476.
3. Johnson LA, Jackson DG: Cell traffic and the lymphatic endothelium. Ann N Y Acad
Sci 2008, 1131:119-133.
4. Vestweber D: Lymphocyte trafficking through blood and lymphatic vessels: more
than just selectins, chemokines and integrins. Eur J Immunol 2003, 33(5):1361-1364.
5. Shields JD, Emmett MS, Dunn DB, Joory KD, Sage LM, Rigby H, Mortimer PS,
Orlando A, Levick JR, Bates DO: Chemokine-mediated migration of melanoma cells
towards lymphatics--a mechanism contributing to metastasis. Oncogene 2007,
26(21):2997-3005.
6. Das S, Skobe M: Lymphatic vessel activation in cancer. Ann N Y Acad Sci 2008,
1131:235-241.
7. Achen MG, Stacker SA: Molecular control of lymphatic metastasis. Ann N Y Acad
Sci 2008, 1131:225-234.
8. Hirakawa S: From tumor lymphangiogenesis to lymphvascular niche. Cancer Sci
2009, 100(6):983-989.
9. He Y, Rajantie I, Pajusola K, Jeltsch M, Holopainen T, Yla-Herttuala S, Harding T,
Jooss K, Takahashi T, Alitalo K: Vascular endothelial cell growth factor receptor 3-
mediated activation of lymphatic endothelium is crucial for tumor cell entry and
spread via lymphatic vessels. Cancer Res 2005, 65(11):4739-4746.
143
10. Wong SY, Hynes RO: Lymphatic or hematogenous dissemination: how does a
metastatic tumor cell decide? Cell Cycle 2006, 5(8):812-817.
11. Farnsworth RH, Achen MG, Stacker SA: Lymphatic endothelium: an important
interactive surface for malignant cells. Pulm Pharmacol Ther 2006, 19(1):51-60.
12. Shields JD, Fleury ME, Yong C, Tomei AA, Randolph GJ, Swartz MA: Autologous
chemotaxis as a mechanism of tumor cell homing to lymphatics via interstitial flow
and autocrine CCR7 signaling. Cancer Cell 2007, 11(6):526-538.
13. Wiley HE, Gonzalez EB, Maki W, Wu MT, Hwang ST: Expression of CC chemokine
receptor-7 and regional lymph node metastasis of B16 murine melanoma. J Natl
Cancer Inst 2001, 93(21):1638-1643.
14. Miteva DO, Rutkowski JM, Dixon JB, Kilarski W, Shields JD, Swartz MA:
Transmural flow modulates cell and fluid transport functions of lymphatic
endothelium. Circ Res, 106(5):920-931.
15. Murfee WL, Rappleye JW, Ceballos M, Schmid-Schonbein GW: Discontinuous
expression of endothelial cell adhesion molecules along initial lymphatic vessels in
mesentery: the primary valve structure. Lymphat Res Biol 2007, 5(2):81-89.
16. Zoltzer H: Initial lymphatics--morphology and function of the endothelial cells.
Lymphology 2003, 36(1):7-25.
17. Poggi P, Marchetti C, Tazzi A, Scelsi R: The lymphatic vessels and their relationship
to lymph formation in the human urinary bladder. Lymphology 1995, 28(1):35-40.
18. Navas V, O'Morchoe PJ, O'Morchoe CC: Lymphatic system of the rat pancreas.
Lymphology 1995, 28(1):4-20.
144
19. Johnson LA, Clasper S, Holt AP, Lalor PF, Baban D, Jackson DG: An inflammation-
induced mechanism for leukocyte transmigration across lymphatic vessel
endothelium. J Exp Med 2006, 203(12):2763-2777.
20. Sawa Y, Tsuruga E, Iwasawa K, Ishikawa H, Yoshida S: Leukocyte adhesion molecule
and chemokine production through lipoteichoic acid recognition by toll-like
receptor 2 in cultured human lymphatic endothelium. Cell Tissue Res 2008,
333(2):237-252.
21. Sawa Y, Shibata K, Braithwaite MW, Suzuki M, Yoshida S: Expression of
immunoglobulin superfamily members on the lymphatic endothelium of inflamed
human small intestine. Microvasc Res 1999, 57(2):100-106.
22. Podgrabinska S, Kamalu O, Mayer L, Shimaoka M, Snoeck H, Randolph GJ, Skobe M:
Inflamed lymphatic endothelium suppresses dendritic cell maturation and function
via Mac-1/ICAM-1-dependent mechanism. J Immunol 2009, 183(3):1767-1779.
23. Teoh D, Johnson LA, Hanke T, McMichael AJ, Jackson DG: Blocking development of
a CD8+ T cell response by targeting lymphatic recruitment of APC. J Immunol
2009, 182(4):2425-2431.
24. Kuroshima S, Sawa Y, Kawamoto T, Yamaoka Y, Notani K, Yoshida S, Inoue N:
Expression of Toll-like receptors 2 and 4 on human intestinal lymphatic vessels.
Microvasc Res 2004, 67(1):90-95.
25. Pegu A, Qin S, Fallert Junecko BA, Nisato RE, Pepper MS, Reinhart TA: Human
lymphatic endothelial cells express multiple functional TLRs. J Immunol 2008,
180(5):3399-3405.
145
26. Marchetti C, Poggi P, Calligaro A, Casasco A: Lymphatic vessels of the human dental
pulp in different conditions. Anat Rec 1992, 234(1):27-33.
27. Ohhashi T, Takahashi N: Acetylcholine-induced release of endothelium-derived
relaxing factor from lymphatic endothelial cells. Am J Physiol 1991, 260(4 Pt
2):H1172-1178.
28. Tsunemoto H, Ikomi F, Ohhashi T: Flow-mediated release of nitric oxide from
lymphatic endothelial cells of pressurized canine thoracic duct. Jpn J Physiol 2003,
53(3):157-163.
29. Yokoyama S, Ohhashi T: Effects of acetylcholine on spontaneous contractions in
isolated bovine mesenteric lymphatics. Am J Physiol 1993, 264(5 Pt 2):H1460-1464.
30. Koller A, Mizuno R, Kaley G: Flow reduces the amplitude and increases the
frequency of lymphatic vasomotion: role of endothelial prostanoids. Am J Physiol
1999, 277(6 Pt 2):R1683-1689.
31. Elias RM, Eisenhoffer J, Johnston MG: Role of endothelial cells in regulating
hemoglobin-induced changes in lymphatic pumping. Am J Physiol 1992, 263(6 Pt
2):H1880-1887.
32. von der Weid PY, Crowe MJ, Van Helden DF: Endothelium-dependent modulation
of pacemaking in lymphatic vessels of the guinea-pig mesentery. J Physiol 1996, 493
( Pt 2):563-575.
33. Ferguson MK: Modulation of lymphatic smooth muscle contractile responses by the
endothelium. J Surg Res 1992, 52(4):359-363.
34. Hashimoto S, Kawai Y, Ohhashi T: Effects of vasoactive substances on the pig
isolated hepatic lymph vessels. J Pharmacol Exp Ther 1994, 269(2):482-488.
146
35. Pepper MS, Skobe M: Lymphatic endothelium: morphological, molecular and
functional properties. J Cell Biol 2003, 163(2):209-213.
36. Podgrabinska S, Braun P, Velasco P, Kloos B, Pepper MS, Skobe M: Molecular
characterization of lymphatic endothelial cells. Proc Natl Acad Sci U S A 2002,
99(25):16069-16074.
37. Cornford ME, Oldendorf WH: Terminal endothelial cells of lymph capillaries as
active transport structures involved in the formation of lymph in rat skin.
Lymphology 1993, 26(2):67-78.
38. Ono N, Mizuno R, Ohhashi T: Effective permeability of hydrophilic substances
through walls of lymph vessels: roles of endothelial barrier. Am J Physiol Heart Circ
Physiol 2005, 289(4):H1676-1682.
39. Takahashi T, Shibata M, Kamiya A: Mechanism of macromolecule concentration in
collecting lymphatics in rat mesentery. Microvasc Res 1997, 54(3):193-205.
40. Supersaxo A, Hein WR, Steffen H: Effect of molecular weight on the lymphatic
absorption of water-soluble compounds following subcutaneous administration.
Pharm Res 1990, 7(2):167-169.
41. Oussoren C, Storm G: Liposomes to target the lymphatics by subcutaneous
administration. Adv Drug Deliv Rev 2001, 50(1-2):143-156.
42. Kaminskas LM, Kota J, McLeod VM, Kelly BD, Karellas P, Porter CJ: PEGylation of
polylysine dendrimers improves absorption and lymphatic targeting following SC
administration in rats. J Control Release 2009, 140(2):108-116.
43. Gale NW, Thurston G, Hackett SF, Renard R, Wang Q, McClain J, Martin C, Witte C,
Witte MH, Jackson D et al: Angiopoietin-2 is required for postnatal angiogenesis
147
and lymphatic patterning, and only the latter role is rescued by Angiopoietin-1.
Dev Cell 2002, 3(3):411-423.
44. Maisonpierre PC, Suri C, Jones PF, Bartunkova S, Wiegand SJ, Radziejewski C,
Compton D, McClain J, Aldrich TH, Papadopoulos N et al: Angiopoietin-2, a natural
antagonist for Tie2 that disrupts in vivo angiogenesis. Science 1997, 277(5322):55-
60.
45. Ahmad SA, Liu W, Jung YD, Fan F, Reinmuth N, Bucana CD, Ellis LM: Differential
expression of angiopoietin-1 and angiopoietin-2 in colon carcinoma. A possible
mechanism for the initiation of angiogenesis. Cancer 2001, 92(5):1138-1143.
46. Ahmad SA, Liu W, Jung YD, Fan F, Wilson M, Reinmuth N, Shaheen RM, Bucana CD,
Ellis LM: The effects of angiopoietin-1 and -2 on tumor growth and angiogenesis in
human colon cancer. Cancer Res 2001, 61(4):1255-1259.
47. Ochiumi T, Tanaka S, Oka S, Hiyama T, Ito M, Kitadai Y, Haruma K, Chayama K:
Clinical significance of angiopoietin-2 expression at the deepest invasive tumor site
of advanced colorectal carcinoma. Int J Oncol 2004, 24(3):539-547.
48. Sfiligoi C, de Luca A, Cascone I, Sorbello V, Fuso L, Ponzone R, Biglia N, Audero E,
Arisio R, Bussolino F et al: Angiopoietin-2 expression in breast cancer correlates
with lymph node invasion and short survival. Int J Cancer 2003, 103(4):466-474.
49. Etoh T, Inoue H, Tanaka S, Barnard GF, Kitano S, Mori M: Angiopoietin-2 is related
to tumor angiogenesis in gastric carcinoma: possible in vivo regulation via
induction of proteases. Cancer Res 2001, 61(5):2145-2153.
148
50. Tanaka S, Mori M, Sakamoto Y, Makuuchi M, Sugimachi K, Wands JR: Biologic
significance of angiopoietin-2 expression in human hepatocellular carcinoma. J
Clin Invest 1999, 103(3):341-345.
51. Mitsuhashi N, Shimizu H, Ohtsuka M, Wakabayashi Y, Ito H, Kimura F, Yoshidome H,
Kato A, Nukui Y, Miyazaki M: Angiopoietins and Tie-2 expression in angiogenesis
and proliferation of human hepatocellular carcinoma. Hepatology 2003, 37(5):1105-
1113.
52. Takanami I: Overexpression of Ang-2 mRNA in non-small cell lung cancer:
association with angiogenesis and poor prognosis. Oncol Rep 2004, 12(4):849-853.
53. Lind AJ, Wikstrom P, Granfors T, Egevad L, Stattin P, Bergh A: Angiopoietin 2
expression is related to histological grade, vascular density, metastases, and
outcome in prostate cancer. Prostate 2005, 62(4):394-399.
54. Hata K, Udagawa J, Fujiwaki R, Nakayama K, Otani H, Miyazaki K: Expression of
angiopoietin-1, angiopoietin-2, and Tie2 genes in normal ovary with corpus luteum
and in ovarian cancer. Oncology 2002, 62(4):340-348.
55. Tammela T, Saaristo A, Lohela M, Morisada T, Tornberg J, Norrmen C, Oike Y,
Pajusola K, Thurston G, Suda T et al: Angiopoietin-1 promotes lymphatic sprouting
and hyperplasia. Blood 2005, 105(12):4642-4648.
56. Suri C, Jones PF, Patan S, Bartunkova S, Maisonpierre PC, Davis S, Sato TN,
Yancopoulos GD: Requisite role of angiopoietin-1, a ligand for the TIE2 receptor,
during embryonic angiogenesis. Cell 1996, 87(7):1171-1180.
57. Angeli V, Randolph GJ: Inflammation, lymphatic function, and dendritic cell
migration. Lymphat Res Biol 2006, 4(4):217-228.
149
58. Douglas JL, Gustin JK, Dezube B, Pantanowitz JL, Moses AV: Kaposi's sarcoma: a
model of both malignancy and chronic inflammation. Panminerva Med 2007,
49(3):119-138.
59. Deitch EA, Xu D, Kaise VL: Role of the gut in the development of injury- and shock
induced SIRS and MODS: the gut-lymph hypothesis, a review. Front Biosci 2006,
11:520-528.
60. Sakamoto I, Ito Y, Mizuno M, Suzuki Y, Sawai A, Tanaka A, Maruyama S, Takei Y,
Yuzawa Y, Matsuo S: Lymphatic vessels develop during tubulointerstitial fibrosis.
Kidney Int 2009, 75(8):828-838.
61. Cui Y: Impact of lymphatic vessels on the heart. Thorac Cardiovasc Surg, 58(1):1-7.
62. Dashkevich A, Bloch W, Antonyan A, Fries JU, Geissler HJ: Morphological and
quantitative changes of the initial myocardial lymphatics in terminal heart failure.
Lymphat Res Biol 2009, 7(1):21-27.
63. Geissler HJ, Dashkevich A, Fischer UM, Fries JW, Kuhn-Regnier F, Addicks K,
Mehlhorn U, Bloch W: First year changes of myocardial lymphatic endothelial
markers in heart transplant recipients. Eur J Cardiothorac Surg 2006, 29(5):767-
771.
64. Machnik A, Neuhofer W, Jantsch J, Dahlmann A, Tammela T, Machura K, Park JK,
Beck FX, Muller DN, Derer W et al: Macrophages regulate salt-dependent volume
and blood pressure by a vascular endothelial growth factor-C-dependent buffering
mechanism. Nat Med 2009, 15(5):545-552.
150
65. Honda N, Jinnin M, Kajihara I, Makino T, Fukushima S, Ihn H: Impaired
lymphangiogenesis due to excess VEGF-D/Flt-4 signaling in the skin of patients
with systemic sclerosis. Br J Dermatol.
66. Horra A, Salazar J, Ferre R, Vallve JC, Guardiola M, Rosales R, Masana L, Ribalta J:
Prox-1 and FOXC2 gene expression in adipose tissue: A potential contributory role
of the lymphatic system to familial combined hyperlipidaemia. Atherosclerosis
2009, 206(2):343-345.
67. Harvey NL: The link between lymphatic function and adipose biology. Ann N Y
Acad Sci 2008, 1131:82-88.
68. Adair TH, Guyton AC: Introduction to the Lymphatic System. In: Experimental
Biology of the Lymphatic Circulation. Edited by Johnston MG, vol. 9, 1 edn. New
York: Elsevier Science Publishing Company, Inc.; 1985: 425.
69. Moghimi SM, Rajabi-Siahboomi R: Advanced colloid-based systems for efficient
delivery of drugs and diagnostic agents to the lymphatic tissues. Prog Biophys Mol
Biol 1996, 65(3):221-249.
70. Taylor AE: The lymphatic edema safety factor: the role of edema dependent
lymphatic factors (EDLF). Lymphology 1990, 23(3):111-123.
71. Adair TH, Guyton AC: Lymph Formation and its Modification in the Lymphatic
System. In: Experimental Biology of the Lymphatic Circulation. Edited by Johnston
MG, vol. 9, 1 edn. New York: Elsevier Science Publishing Company, Inc.; 1985.
72. Muthuchamy M, Zawieja D: Molecular regulation of lymphatic contractility. Ann N
Y Acad Sci 2008, 1131:89-99.
151
73. Muthuchamy M, Gashev A, Boswell N, Dawson N, Zawieja D: Molecular and
functional analyses of the contractile apparatus in lymphatic muscle. FASEB J
2003, 17(8):920-922.
74. Gruetter CA, Lemke SM: Comparison of endothelium-dependent relaxation in
bovine intrapulmonary artery and vein by acetylcholine and A23187. J Pharmacol
Exp Ther 1986, 238(3):1055-1062.
75. Moore RM, Sedrish SA, Holmes EP, Koch CE, Venugopal CS: Role of endothelium
and nitric oxide in modulating in vitro responses of colonic arterial and venous
rings to vasodilatory neuropeptides in horses. Can J Vet Res 2005, 69(2):116-122.
76. Norel X, Walch L, Gascard JP, deMontpreville V, Brink C: Prostacyclin release and
receptor activation: differential control of human pulmonary venous and arterial
tone. Br J Pharmacol 2004, 142(4):788-796.
77. Back M, Walch L, Norel X, Gascard JP, Mazmanian G, Brink C: Modulation of
vascular tone and reactivity by nitric oxide in porcine pulmonary arteries and
veins. Acta Physiol Scand 2002, 174(1):9-15.
78. Furchgott RF, Zawadzki JV: The obligatory role of endothelial cells in the relaxation
of arterial smooth muscle by acetylcholine. Nature 1980, 288(5789):373-376.
79. De Mey JG, Vanhoutte PM: Heterogeneous behavior of the canine arterial and
venous wall. Importance of the endothelium. Circ Res 1982, 51(4):439-447.
80. Ferguson MK, DeFilippi VJ: Nitric oxide and endothelium-dependent relaxation in
tracheobronchial lymph vessels. Microvasc Res 1994, 47(3):308-317.
81. Radhakrishnan K, Rockson SG: The clinical spectrum of lymphatic disease. Ann N Y
Acad Sci 2008, 1131:155-184.
152
82. Warren AG, Brorson H, Borud LJ, Slavin SA: Lymphedema: a comprehensive
review. Ann Plast Surg 2007, 59(4):464-472.
83. Casley-Smith JR, Gaffney RM: Excess plasma proteins as a cause of chronic
inflammation and lymphodema: quantitative electron microscopy. J Pathol 1981,
133(3):243-272.
84. Gaffney RM, Casley-Smith JR: Excess plasma proteins as a cause of chronic
inflammation and lymphoedema: biochemical estimations. J Pathol 1981,
133(3):229-242.
85. Ryan TJ: Lymphatics and adipose tissue. Clin Dermatol 1995, 13(5):493-498.
86. Olszewski WL, Engeset A, Romaniuk A, Grzelak I, Ziolkowska A: Immune cells in
peripheral lymph and skin of patients with obstructive lymphedema. Lymphology
1990, 23(1):23-33.
87. Manduch M, Oliveira AM, Nascimento AG, Folpe AL: Massive localised
lymphoedema: a clinicopathological study of 22 cases and review of the literature. J
Clin Pathol 2009, 62(9):808-811.
88. Woo PC, Lum PN, Wong SS, Cheng VC, Yuen KY: Cellulitis complicating
lymphoedema. Eur J Clin Microbiol Infect Dis 2000, 19(4):294-297.
89. Macdonald JM: Wound healing and lymphedema: a new look at an old problem.
Ostomy Wound Manage 2001, 47(4):52-57.
90. Pereira de Godoy JM, Azoubel LM, Guerreiro Godoy Mde F: Erysipelas and
lymphangitis in patients undergoing lymphedema treatment after breast-cancer
therapy. Acta Dermatovenerol Alp Panonica Adriat 2009, 18(2):63-65.
153
91. Damstra RJ, van Steensel MA, Boomsma JH, Nelemans P, Veraart JC: Erysipelas as a
sign of subclinical primary lymphoedema: a prospective quantitative scintigraphic
study of 40 patients with unilateral erysipelas of the leg. Br J Dermatol 2008,
158(6):1210-1215.
92. Shih YC, Xu Y, Cormier JN, Giordano S, Ridner SH, Buchholz TA, Perkins GH, Elting
LS: Incidence, treatment costs, and complications of lymphedema after breast
cancer among women of working age: a 2-year follow-up study. J Clin Oncol 2009,
27(12):2007-2014.
93. Pearce VJ, Mortimer PS: Hand dermatitis and lymphoedema. Br J Dermatol 2009,
161(1):177-180.
94. Chen HC, Gharb BB, Salgado CJ, Rampazzo A, Xu E, di Spilimbergo SS, Su S:
Elective amputation of the toes in severe lymphedema of the lower leg: rationale
and indications. Ann Plast Surg 2009, 63(2):193-197.
95. Watanabe T: Congenital penoscrotal lymphedema complicated by sepsis associated
with a streptococcal infection. Pediatr Emerg Care 2000, 16(6):423-425.
96. Person B, Addiss D, Bartholomew LK, Meijer C, Pou V, Gonzalvez G, Borne BV:
"Can it be that god does not remember me": a qualitative study on the
psychological distress, suffering, and coping of Dominican women with chronic
filarial lymphedema and elephantiasis of the leg. Health Care Women Int 2008,
29(4):349-365.
97. Gloviczki P: Principles of surgical treatment of chronic lymphoedema. Int Angiol
1999, 18(1):42-46.
154
98. Rockson SG: Current concepts and future directions in the diagnosis and
management of lymphatic vascular disease. Vasc Med, 15(3):223-231.
99. Rockson SG: The unique biology of lymphatic edema. Lymphat Res Biol 2009,
7(2):97-100.
100. Rockson SG, Granger DN, Skeff KM, Chaite W: Lymphatic biology and disease: is it
being taught? Who is listening? Lymphat Res Biol 2004, 2(2):86-95.
101. Cliff S, Bedlow AJ, Stanton AW, Mortimer PS: An in vivo study of the
microlymphatics in psoriasis using fluorescence microlymphography. Br J
Dermatol 1999, 140(1):61-66.
102. Kovi J, Duong HD, Hoang CT: Ultrastructure of intestinal lymphatics in Crohn's
disease. Am J Clin Pathol 1981, 76(4):385-394.
103. Kaipainen A, Korhonen J, Mustonen T, van Hinsbergh VW, Fang GH, Dumont D,
Breitman M, Alitalo K: Expression of the fms-like tyrosine kinase 4 gene becomes
restricted to lymphatic endothelium during development. Proc Natl Acad Sci U S A
1995, 92(8):3566-3570.
104. Stacker SA, Hughes RA, Achen MG: Molecular targeting of lymphatics for therapy.
Curr Pharm Des 2004, 10(1):65-74.
105. Banerji S, Ni J, Wang SX, Clasper S, Su J, Tammi R, Jones M, Jackson DG: LYVE-1,
a new homologue of the CD44 glycoprotein, is a lymph-specific receptor for
hyaluronan. J Cell Biol 1999, 144(4):789-801.
106. Wigle JT, Oliver G: Prox1 function is required for the development of the murine
lymphatic system. Cell 1999, 98(6):769-778.
155
107. Hirakawa S, Hong YK, Harvey N, Schacht V, Matsuda K, Libermann T, Detmar M:
Identification of vascular lineage-specific genes by transcriptional profiling of
isolated blood vascular and lymphatic endothelial cells. Am J Pathol 2003,
162(2):575-586.
108. Oppenheim JJ, Zachariae CO, Mukaida N, Matsushima K: Properties of the novel
proinflammatory supergene "intercrine" cytokine family. Annu Rev Immunol 1991,
9:617-648.
109. Schall TJ, Bacon K, Toy KJ, Goeddel DV: Selective attraction of monocytes and T
lymphocytes of the memory phenotype by cytokine RANTES. Nature 1990,
347(6294):669-671.
110. Ando T, Jordan P, Wang Y, Jennings MH, Harper MH, Houghton J, Elrod J, Alexander
JS: Homogeneity of mesothelial cells with lymphatic endothelium: expression of
lymphatic endothelial markers by mesothelial cells. Lymphat Res Biol 2005,
3(3):117-125.
111. Attout T, Hoerauf A, Denece G, Debrah AY, Marfo-Debrekyei Y, Boussinesq M, Wanji
S, Martinez V, Mand S, Adjei O et al: Lymphatic vascularisation and involvement of
Lyve-1+ macrophages in the human onchocerca nodule. PLoS One 2009,
4(12):e8234.
112. Blanchet MR, Maltby S, Haddon DJ, Merkens H, Zbytnuik L, McNagny KM: CD34
facilitates the development of allergic asthma. Blood 2007, 110(6):2005-2012.
113. McKenney JK, Weiss SW, Folpe AL: CD31 expression in intratumoral
macrophages: a potential diagnostic pitfall. Am J Surg Pathol 2001, 25(9):1167-
1173.
156
114. Breiteneder-Geleff S, Soleiman A, Kowalski H, Horvat R, Amann G, Kriehuber E,
Diem K, Weninger W, Tschachler E, Alitalo K et al: Angiosarcomas express mixed
endothelial phenotypes of blood and lymphatic capillaries: podoplanin as a specific
marker for lymphatic endothelium. Am J Pathol 1999, 154(2):385-394.
115. Schacht V, Dadras SS, Johnson LA, Jackson DG, Hong YK, Detmar M: Up-regulation
of the lymphatic marker podoplanin, a mucin-type transmembrane glycoprotein,
in human squamous cell carcinomas and germ cell tumors. Am J Pathol 2005,
166(3):913-921.
116. Baluk P, McDonald DM: Markers for microscopic imaging of lymphangiogenesis
and angiogenesis. Ann N Y Acad Sci 2008, 1131:1-12.
117. Shibuya M, Claesson-Welsh L: Signal transduction by VEGF receptors in regulation
of angiogenesis and lymphangiogenesis. Exp Cell Res 2006, 312(5):549-560.
118. Lohela M, Bry M, Tammela T, Alitalo K: VEGFs and receptors involved in
angiogenesis versus lymphangiogenesis. Curr Opin Cell Biol 2009, 21(2):154-165.
119. Thurston G: Role of Angiopoietins and Tie receptor tyrosine kinases in angiogenesis
and lymphangiogenesis. Cell Tissue Res 2003, 314(1):61-68.
120. Woolf AS: Angiopoietins: vascular growth factors looking for roles in glomeruli.
Curr Opin Nephrol Hypertens, 19(1):20-25.
121. Morisada T, Oike Y, Yamada Y, Urano T, Akao M, Kubota Y, Maekawa H, Kimura Y,
Ohmura M, Miyamoto T et al: Angiopoietin-1 promotes LYVE-1-positive lymphatic
vessel formation. Blood 2005, 105(12):4649-4656.
122. Kim I, Kim JH, Moon SO, Kwak HJ, Kim NG, Koh GY: Angiopoietin-2 at high
concentration can enhance endothelial cell survival through the
157
phosphatidylinositol 3'-kinase/Akt signal transduction pathway. Oncogene 2000,
19(39):4549-4552.
123. Teichert-Kuliszewska K, Maisonpierre PC, Jones N, Campbell AI, Master Z, Bendeck
MP, Alitalo K, Dumont DJ, Yancopoulos GD, Stewart DJ: Biological action of
angiopoietin-2 in a fibrin matrix model of angiogenesis is associated with activation
of Tie2. Cardiovasc Res 2001, 49(3):659-670.
124. Mochizuki Y, Nakamura T, Kanetake H, Kanda S: Angiopoietin 2 stimulates
migration and tube-like structure formation of murine brain capillary endothelial
cells through c-Fes and c-Fyn. J Cell Sci 2002, 115(Pt 1):175-183.
125. Harfouche R, Hussain SN: Signaling and regulation of endothelial cell survival by
angiopoietin-2. Am J Physiol Heart Circ Physiol 2006, 291(4):H1635-1645.
126. Carlson TR, Feng Y, Maisonpierre PC, Mrksich M, Morla AO: Direct cell adhesion to
the angiopoietins mediated by integrins. J Biol Chem 2001, 276(28):26516-26525.
127. Pitera JE, Woolf AS, Gale NW, Yancopoulos GD, Yuan HT: Dysmorphogenesis of
kidney cortical peritubular capillaries in angiopoietin-2-deficient mice. Am J Pathol
2004, 165(6):1895-1906.
128. Hackett SF, Wiegand S, Yancopoulos G, Campochiaro PA: Angiopoietin-2 plays an
important role in retinal angiogenesis. J Cell Physiol 2002, 192(2):182-187.
129. Gordon JA: Use of vanadate as protein-phosphotyrosine phosphatase inhibitor.
Methods Enzymol 1991, 201:477-482.
130. Leak LV, Jones M: Lymphatic endothelium isolation, characterization and long-
term culture. Anat Rec 1993, 236(4):641-652.
158
131. Johnston MG, Walker MA: Lymphatic endothelial and smooth-muscle cells in tissue
culture. In Vitro 1984, 20(7):566-572.
132. Tammela T, Petrova TV, Alitalo K: Molecular lymphangiogenesis: new players.
Trends Cell Biol 2005, 15(8):434-441.
133. Herzog Y, Kalcheim C, Kahane N, Reshef R, Neufeld G: Differential expression of
neuropilin-1 and neuropilin-2 in arteries and veins. Mech Dev 2001, 109(1):115-119.
134. Barton WA, Tzvetkova-Robev D, Miranda EP, Kolev MV, Rajashankar KR, Himanen
JP, Nikolov DB: Crystal structures of the Tie2 receptor ectodomain and the
angiopoietin-2-Tie2 complex. Nat Struct Mol Biol 2006, 13(6):524-532.
135. Bogdanovic E, Nguyen VP, Dumont DJ: Activation of Tie2 by angiopoietin-1 and
angiopoietin-2 results in their release and receptor internalization. J Cell Sci 2006,
119(Pt 17):3551-3560.
136. Yoon MJ, Cho CH, Lee CS, Jang IH, Ryu SH, Koh GY: Localization of Tie2 and
phospholipase D in endothelial caveolae is involved in angiopoietin-1-induced
MEK/ERK phosphorylation and migration in endothelial cells. Biochem Biophys
Res Commun 2003, 308(1):101-105.
137. Kanda S, Miyata Y, Mochizuki Y, Matsuyama T, Kanetake H: Angiopoietin 1 is
mitogenic for cultured endothelial cells. Cancer Res 2005, 65(15):6820-6827.
138. Hu B, Jarzynka MJ, Guo P, Imanishi Y, Schlaepfer DD, Cheng SY: Angiopoietin 2
induces glioma cell invasion by stimulating matrix metalloprotease 2 expression
through the alphavbeta1 integrin and focal adhesion kinase signaling pathway.
Cancer Res 2006, 66(2):775-783.
159
139. Murray BW, Padrique ES, Pinko C, McTigue MA: Mechanistic effects of
autophosphorylation on receptor tyrosine kinase catalysis: enzymatic
characterization of Tie2 and phospho-Tie2. Biochemistry 2001, 40(34):10243-10253.
140. Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: improving the sensitivity of
progressive multiple sequence alignment through sequence weighting, position-
specific gap penalties and weight matrix choice. Nucleic Acids Res 1994,
22(22):4673-4680.
141. Witzenbichler B, Maisonpierre PC, Jones P, Yancopoulos GD, Isner JM: Chemotactic
properties of angiopoietin-1 and -2, ligands for the endothelial-specific receptor
tyrosine kinase Tie2. J Biol Chem 1998, 273(29):18514-18521.
142. Liu W, Reinmuth N, Stoeltzing O, Parikh AA, Fan F, Ahmad SA, Jung YD, Ellis LM:
Antiangiogenic therapy targeting factors that enhance endothelial cell survival.
Semin Oncol 2002, 29(3 Suppl 11):96-103.
143. Kontos CD, Stauffer TP, Yang WP, York JD, Huang L, Blanar MA, Meyer T, Peters
KG: Tyrosine 1101 of Tie2 is the major site of association of p85 and is required for
activation of phosphatidylinositol 3-kinase and Akt. Mol Cell Biol 1998, 18(7):4131-
4140.
144. Jones N, Master Z, Jones J, Bouchard D, Gunji Y, Sasaki H, Daly R, Alitalo K, Dumont
DJ: Identification of Tek/Tie2 binding partners. Binding to a multifunctional
docking site mediates cell survival and migration. J Biol Chem 1999, 274(43):30896-
30905.
145. Kwak HJ, So JN, Lee SJ, Kim I, Koh GY: Angiopoietin-1 is an apoptosis survival
factor for endothelial cells. FEBS Lett 1999, 448(2-3):249-253.
160
146. Hayes AJ, Huang WQ, Mallah J, Yang D, Lippman ME, Li LY: Angiopoietin-1 and its
receptor Tie-2 participate in the regulation of capillary-like tubule formation and
survival of endothelial cells. Microvasc Res 1999, 58(3):224-237.
147. Kim I, Kim HG, So JN, Kim JH, Kwak HJ, Koh GY: Angiopoietin-1 regulates
endothelial cell survival through the phosphatidylinositol 3'-Kinase/Akt signal
transduction pathway. Circ Res 2000, 86(1):24-29.
148. Papapetropoulos A, Fulton D, Mahboubi K, Kalb RG, O'Connor DS, Li F, Altieri DC,
Sessa WC: Angiopoietin-1 inhibits endothelial cell apoptosis via the Akt/survivin
pathway. J Biol Chem 2000, 275(13):9102-9105.
149. Babaei S, Teichert-Kuliszewska K, Zhang Q, Jones N, Dumont DJ, Stewart DJ:
Angiogenic actions of angiopoietin-1 require endothelium-derived nitric oxide. Am
J Pathol 2003, 162(6):1927-1936.
150. Murakami T, Takagi H, Suzuma K, Suzuma I, Ohashi H, Watanabe D, Ojima T,
Suganami E, Kurimoto M, Kaneto H et al: Angiopoietin-1 attenuates H2O2-induced
SEK1/JNK phosphorylation through the phosphatidylinositol 3-kinase/Akt
pathway in vascular endothelial cells. J Biol Chem 2005, 280(36):31841-31849.
151. Zagzag D, Hooper A, Friedlander DR, Chan W, Holash J, Wiegand SJ, Yancopoulos
GD, Grumet M: In situ expression of angiopoietins in astrocytomas identifies
angiopoietin-2 as an early marker of tumor angiogenesis. Exp Neurol 1999,
159(2):391-400.
152. Cohen B, Barkan D, Levy Y, Goldberg I, Fridman E, Kopolovic J, Rubinstein M:
Leptin induces angiopoietin-2 expression in adipose tissues. J Biol Chem 2001,
276(11):7697-7700.
161
153. Nag S, Papneja T, Venugopalan R, Stewart DJ: Increased angiopoietin2 expression is
associated with endothelial apoptosis and blood-brain barrier breakdown. Lab
Invest 2005, 85(10):1189-1198.
154. Daly C, Pasnikowski E, Burova E, Wong V, Aldrich TH, Griffiths J, Ioffe E, Daly TJ,
Fandl JP, Papadopoulos N et al: Angiopoietin-2 functions as an autocrine protective
factor in stressed endothelial cells. Proc Natl Acad Sci U S A 2006, 103(42):15491-
15496.
155. Gnepp DR, Chandler W: Tissue culture of human and canine thoracic duct
endothelium. In Vitro Cell Dev Biol 1985, 21(4):200-206.
156. Bowman CA, Witte MH, Witte CL, Way DL, Nagle RB, Copeland JG, Daschbach CC:
Cystic hygroma reconsidered: hamartoma or neoplasm? Primary culture of an
endothelial cell line from a massive cervicomediastinal hygroma with bony
lymphangiomatosis. Lymphology 1984, 17(1):15-22.
157. Witte MH, Witte CL: Lymphatics and blood vessels, lymphangiogenesis and
hemangiogenesis: from cell biology to clinical medicine. Lymphology 1987,
20(4):257-266.
158. Sauter B, Foedinger D, Sterniczky B, Wolff K, Rappersberger K: Immunoelectron
microscopic characterization of human dermal lymphatic microvascular
endothelial cells. Differential expression of CD31, CD34, and type IV collagen with
lymphatic endothelial cells vs blood capillary endothelial cells in normal human
skin, lymphangioma, and hemangioma in situ. J Histochem Cytochem 1998,
46(2):165-176.
162
159. Chi JT, Chang HY, Haraldsen G, Jahnsen FL, Troyanskaya OG, Chang DS, Wang Z,
Rockson SG, van de Rijn M, Botstein D et al: Endothelial cell diversity revealed by
global expression profiling. Proc Natl Acad Sci U S A 2003, 100(19):10623-10628.
160. Sironi M, Conti A, Bernasconi S, Fra AM, Pasqualini F, Nebuloni M, Lauri E, De
Bortoli M, Mantovani A, Dejana E et al: Generation and characterization of a mouse
lymphatic endothelial cell line. Cell Tissue Res 2006, 325(1):91-100.
161. Bianchi L, Lorenzoni P, Bini L, Weber E, Tani C, Rossi A, Agliano M, Pallini V,
Sacchi G: Protein expression profiles of Bos taurus blood and lymphatic vessel
endothelial cells. Proteomics 2007, 7(10):1600-1614.
162. Wick N, Saharinen P, Saharinen J, Gurnhofer E, Steiner CW, Raab I, Stokic D,
Giovanoli P, Buchsbaum S, Burchard A et al: Transcriptomal comparison of human
dermal lymphatic endothelial cells ex vivo and in vitro. Physiol Genomics 2007,
28(2):179-192.
163. Amatschek S, Kriehuber E, Bauer W, Reininger B, Meraner P, Wolpl A, Schweifer N,
Haslinger C, Stingl G, Maurer D: Blood and lymphatic endothelial cell-specific
differentiation programs are stringently controlled by the tissue environment.
Blood 2007, 109(11):4777-4785.
164. Nelson GM, Padera TP, Garkavtsev I, Shioda T, Jain RK: Differential gene expression
of primary cultured lymphatic and blood vascular endothelial cells. Neoplasia 2007,
9(12):1038-1045.
165. Kawai Y, Minami T, Fujimori M, Hosaka K, Mizuno R, Ikomi F, Kodama T, Ohhashi
T: Characterization and microarray analysis of genes in human lymphatic
163
endothelial cells from patients with breast cancer. Lymphat Res Biol 2007, 5(2):115-
126.
166. Roesli C, Mumprecht V, Neri D, Detmar M: Identification of the surface-accessible,
lineage-specific vascular proteome by two-dimensional peptide mapping. FASEB J
2008, 22(6):1933-1944.
167. Nguyen VP, Chen SH, Trinh J, Kim H, Coomber BL, Dumont DJ: Differential
response of lymphatic, venous and arterial endothelial cells to angiopoietin-1 and
angiopoietin-2. BMC Cell Biol 2007, 8:10.
168. Veikkola T, Alitalo K: Dual role of Ang2 in postnatal angiogenesis and
lymphangiogenesis. Dev Cell 2002, 3(3):302-304.
169. Weber E, Lorenzoni P, Lozzi G, Sacchi G: Cytochemical differentiation between
blood and lymphatic endothelium: bovine blood and lymphatic large vessels and
endothelial cells in culture. J Histochem Cytochem 1994, 42(8):1109-1115.
170. Veikkola T, Lohela M, Ikenberg K, Makinen T, Korff T, Saaristo A, Petrova T, Jeltsch
M, Augustin HG, Alitalo K: Intrinsic versus microenvironmental regulation of
lymphatic endothelial cell phenotype and function. FASEB J 2003, 17(14):2006-
2013.
171. Ando T, Jordan P, Joh T, Wang Y, Jennings MH, Houghton J, Alexander JS: Isolation
and characterization of a novel mouse lymphatic endothelial cell line: SV-LEC.
Lymphat Res Biol 2005, 3(3):105-115.
172. Pruitt KD, Tatusova T, Maglott DR: NCBI reference sequences (RefSeq): a curated
non-redundant sequence database of genomes, transcripts and proteins. Nucleic
Acids Res 2007, 35(Database issue):D61-65.
164
173. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski
K, Dwight SS, Eppig JT et al: Gene ontology: tool for the unification of biology. The
Gene Ontology Consortium. Nat Genet 2000, 25(1):25-29.
174. Harel A, Inger A, Stelzer G, Strichman-Almashanu L, Dalah I, Safran M, Lancet D:
GIFtS: annotation landscape analysis with GeneCards. BMC Bioinformatics 2009,
10:348.
175. Oliver G, Srinivasan RS: Lymphatic vasculature development: current concepts.
Ann N Y Acad Sci 2008, 1131:75-81.
176. Qian WJ, Jacobs JM, Liu T, Camp DG, 2nd, Smith RD: Advances and challenges in
liquid chromatography-mass spectrometry-based proteomics profiling for clinical
applications. Mol Cell Proteomics 2006, 5(10):1727-1744.
177. Wick G, Knoflach M, Kind M, Henderson B, Bernhard D: Heat shock proteins and
stress in atherosclerosis. Autoimmun Rev 2004, 3 Suppl 1:S30-31.
178. Wagner M, Hermanns I, Bittinger F, Kirkpatrick CJ: Induction of stress proteins in
human endothelial cells by heavy metal ions and heat shock. Am J Physiol 1999,
277(5 Pt 1):L1026-1033.
179. Kalmar B, Greensmith L: Induction of heat shock proteins for protection against
oxidative stress. Adv Drug Deliv Rev 2009.
180. Preville X, Salvemini F, Giraud S, Chaufour S, Paul C, Stepien G, Ursini MV, Arrigo
AP: Mammalian small stress proteins protect against oxidative stress through their
ability to increase glucose-6-phosphate dehydrogenase activity and by maintaining
optimal cellular detoxifying machinery. Exp Cell Res 1999, 247(1):61-78.
165
181. Kim YH, Park EJ, Han ST, Park JW, Kwon TK: Arsenic trioxide induces Hsp70
expression via reactive oxygen species and JNK pathway in MDA231 cells. Life Sci
2005, 77(22):2783-2793.
182. Franklin TB, Krueger-Naug AM, Clarke DB, Arrigo AP, Currie RW: The role of heat
shock proteins Hsp70 and Hsp27 in cellular protection of the central nervous
system. Int J Hyperthermia 2005, 21(5):379-392.
183. Steurer J, Hoffmann U, Dur P, Russi E, Vetter W: Changes in arterial and
transcutaneous oxygen and carbon dioxide tensions during and after voluntary
hyperventilation. Respiration 1997, 64(3):200-205.
184. Ho HY, Cheng ML, Cheng PF, Chiu DT: Low oxygen tension alleviates oxidative
damage and delays cellular senescence in G6PD-deficient cells. Free Radic Res
2007, 41(5):571-579.
185. Rhee SG, Chae HZ, Kim K: Peroxiredoxins: a historical overview and speculative
preview of novel mechanisms and emerging concepts in cell signaling. Free Radic
Biol Med 2005, 38(12):1543-1552.
186. Shang F, Gong X, Taylor A: Activity of ubiquitin-dependent pathway in response to
oxidative stress. Ubiquitin-activating enzyme is transiently up-regulated. J Biol
Chem 1997, 272(37):23086-23093.
187. Atkuri KR, Herzenberg LA: Culturing at atmospheric oxygen levels impacts
lymphocyte function. Proc Natl Acad Sci U S A 2005, 102(10):3756-3759.
188. Atkuri KR, Herzenberg LA, Niemi AK, Cowan T: Importance of culturing primary
lymphocytes at physiological oxygen levels. Proc Natl Acad Sci U S A 2007,
104(11):4547-4552.
166
189. Ji RC: Lymphatic endothelial cells, lymphedematous lymphangiogenesis, and
molecular control of edema formation. Lymphat Res Biol 2008, 6(3-4):123-137.
190. Jurisic G, Detmar M: Lymphatic endothelium in health and disease. Cell Tissue Res
2009, 335(1):97-108.
191. Liersch R, Biermann C, Mesters RM, Berdel WE: Lymphangiogenesis in cancer:
current perspectives. Recent Results Cancer Res, 180:115-135.
192. Dorrell C, Abraham SL, Lanxon-Cookson KM, Canaday PS, Streeter PR, Grompe M:
Isolation of major pancreatic cell types and long-term culture-initiating cells using
novel human surface markers. Stem Cell Res 2008, 1(3):183-194.
193. Mernaugh RL, Yan H, Chen D, Edl J, Hanley G, Pozzi A, Zent R: Production and
characterization of mouse ureteric bud cell-specific rat hybridoma antibodies
utilizing subtractive immunization and high-throughput screening. J Immunol
Methods 2005, 306(1-2):115-127.
194. Ning Y, Wang Y, Li Y, Hong Y, Peng D, Liu Y, Wang J, Hao W, Tian X, Wu F et al:
An alternative strategy for high throughput generation and characterization of
monoclonal antibodies against human plasma proteins using fractionated native
proteins as immunogens. Proteomics 2006, 6(2):438-448.
195. Khan SA, Suryawanshi AR, Ranpura SA, Jadhav SV, Khole VV: Identification of
novel immunodominant epididymal sperm proteins using combinatorial approach.
Reproduction 2009, 138(1):81-93.
196. Hooper JD, Zijlstra A, Aimes RT, Liang H, Claassen GF, Tarin D, Testa JE, Quigley JP:
Subtractive immunization using highly metastatic human tumor cells identifies
167
SIMA135/CDCP1, a 135 kDa cell surface phosphorylated glycoprotein antigen.
Oncogene 2003, 22(12):1783-1794.
197. Boukerche H, Baril P, Tabone E, Berard F, Sanhadji K, Balme B, Wolf F, Perrot H,
Thomas L: A new Mr 55,000 surface protein implicated in melanoma progression:
association with a metastatic phenotype. Cancer Res 2000, 60(20):5848-5856.
198. Testa JE, Brooks PC, Lin JM, Quigley JP: Eukaryotic expression cloning with an
antimetastatic monoclonal antibody identifies a tetraspanin (PETA-3/CD151) as an
effector of human tumor cell migration and metastasis. Cancer Res 1999,
59(15):3812-3820.
199. Rasmussen N, Ditzel HJ: Scanning the cell surface proteome of cancer cells and
identification of metastasis-associated proteins using a subtractive immunization
strategy. J Proteome Res 2009, 8(11):5048-5059.
200. Sleeman JP, Kim U, LePendu J, Howells N, Coquerelle T, Ponta H, Herrlich P:
Inhibition of MT-450 rat mammary tumour growth by antibodies recognising
subtypes of blood group antigen B. Oncogene 1999, 18(31):4485-4494.
201. Brooks PC, Lin JM, French DL, Quigley JP: Subtractive immunization yields
monoclonal antibodies that specifically inhibit metastasis. J Cell Biol 1993,
122(6):1351-1359.
202. Trefzer U, Rietz N, Chen Y, Audring H, Herberth G, Siegel P, Reinke S, Koniger P, Wu
S, Ma J et al: SM5-1: a new monoclonal antibody which is highly sensitive and
specific for melanocytic lesions. Arch Dermatol Res 2000, 292(12):583-589.
168
203. Williams CV, Stechmann CL, McLoon SC: Subtractive immunization techniques for
the production of monoclonal antibodies to rare antigens. Biotechniques 1992,
12(6):842-847.
204. Sleister HM, Rao AG: Strategies to generate antibodies capable of distinguishing
between proteins with >90% amino acid identity. J Immunol Methods 2001, 252(1-
2):121-129.
205. Sleister HM, Rao AG: Subtractive immunization: a tool for the generation of
discriminatory antibodies to proteins of similar sequence. J Immunol Methods 2002,
261(1-2):213-220.
206. Zijlstra A, Testa JE, Quigley JP: Targeting the proteome/epitome, implementation of
subtractive immunization. Biochem Biophys Res Commun 2003, 303(3):733-744.
207. Nelson PN, Reynolds GM, Waldron EE, Ward E, Giannopoulos K, Murray PG:
Monoclonal antibodies. Mol Pathol 2000, 53(3):111-117.
208. Song KS, Scherer PE, Tang Z, Okamoto T, Li S, Chafel M, Chu C, Kohtz DS, Lisanti
MP: Expression of caveolin-3 in skeletal, cardiac, and smooth muscle cells.
Caveolin-3 is a component of the sarcolemma and co-fractionates with dystrophin
and dystrophin-associated glycoproteins. J Biol Chem 1996, 271(25):15160-15165.
209. Dancik V, Addona TA, Clauser KR, Vath JE, Pevzner PA: De novo peptide
sequencing via tandem mass spectrometry. J Comput Biol 1999, 6(3-4):327-342.
210. Wang HU, Chen ZF, Anderson DJ: Molecular distinction and angiogenic interaction
between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-
B4. Cell 1998, 93(5):741-753.
169
211. Srinivasan RS, Dillard ME, Lagutin OV, Lin FJ, Tsai S, Tsai MJ, Samokhvalov IM,
Oliver G: Lineage tracing demonstrates the venous origin of the mammalian
lymphatic vasculature. Genes Dev 2007, 21(19):2422-2432.
212. Song KS, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP: Co-purification
and direct interaction of Ras with caveolin, an integral membrane protein of
caveolae microdomains. Detergent-free purification of caveolae microdomains. J
Biol Chem 1996, 271(16):9690-9697.
213. Foster LJ, De Hoog CL, Mann M: Unbiased quantitative proteomics of lipid rafts
reveals high specificity for signaling factors. Proc Natl Acad Sci U S A 2003,
100(10):5813-5818.
214. Mirza SP, Olivier M: Methods and approaches for the comprehensive
characterization and quantification of cellular proteomes using mass spectrometry.
Physiol Genomics 2008, 33(1):3-11.
215. Shevchenko A, Sunyaev S, Loboda A, Bork P, Ens W, Standing KG: Charting the
proteomes of organisms with unsequenced genomes by MALDI-quadrupole time-
of-flight mass spectrometry and BLAST homology searching. Anal Chem 2001,
73(9):1917-1926.
216. Dagenais SL, Hartsough RL, Erickson RP, Witte MH, Butler MG, Glover TW: Foxc2 is
expressed in developing lymphatic vessels and other tissues associated with
lymphedema-distichiasis syndrome. Gene Expr Patterns 2004, 4(6):611-619.
217. Fujii T, Zen Y, Sato Y, Sasaki M, Enomae M, Minato H, Masuda S, Uehara T,
Katsuyama T, Nakanuma Y: Podoplanin is a useful diagnostic marker for epithelioid
hemangioendothelioma of the liver. Mod Pathol 2008, 21(2):125-130.
170
218. Becker J, Wang B, Pavlakovic H, Buttler K, Wilting J: Homeobox Transcription
Factor Prox1 in Sympathetic Ganglia of Vertebrate Embryos: Correlation With
Human Stage 4s Neuroblastoma. Pediatr Res.
219. Wicker N, Carles A, Mills IG, Wolf M, Veerakumarasivam A, Edgren H, Boileau F,
Wasylyk B, Schalken JA, Neal DE et al: A new look towards BAC-based array CGH
through a comprehensive comparison with oligo-based array CGH. BMC Genomics
2007, 8:84.
220. Troxler RF, Iontcheva I, Oppenheim FG, Nunes DP, Offner GD: Molecular
characterization of a major high molecular weight mucin from human sublingual
gland. Glycobiology 1997, 7(7):965-973.
221. Levine MJ, Reddy MS, Tabak LA, Loomis RE, Bergey EJ, Jones PC, Cohen RE,
Stinson MW, Al-Hashimi I: Structural aspects of salivary glycoproteins. J Dent Res
1987, 66(2):436-441.
222. Ezaki T, Kuwahara K, Morikawa S, Shimizu K, Sakaguchi N, Matsushima K, Matsuno
K: Production of two novel monoclonal antibodies that distinguish mouse
lymphatic and blood vascular endothelial cells. Anat Embryol (Berl) 2006,
211(5):379-393.
223. Ball LE, Oatis JE, Jr., Dharmasiri K, Busman M, Wang J, Cowden LB, Galijatovic A,
Chen N, Crouch RK, Knapp DR: Mass spectrometric analysis of integral membrane
proteins: application to complete mapping of bacteriorhodopsins and rhodopsin.
Protein Sci 1998, 7(3):758-764.
224. Porter CJ, Charman WN: Transport and absorption of drugs via the lymphatic
system. Adv Drug Deliv Rev 2001, 50(1-2):1-2.
