CHARACTERIZING THE RETROVIRAL ENVELOPE
GLYCOPROTEIN MEMBRANE PROXIMAL EXTERNAL REGION
AND MEMBRANE-SPANNING DOMAINS FOR THEIR ROLES IN
HELICAL ALIGNMENT, FUSOGENICITY, AND
INCORPORATION INTO VIRAL PARTICLES
A dissertation presented to
the Faculty of the Graduate School
at the University of Missouri-Columbia
________________________________________________________________________
In Partial Fulfillment
Of the Requirements for the Degree
Doctor of Philosophy
________________________________________________________________________
By
DANIEL SALAMANGO
Dr. Marc C. Johnson, Dissertation Supervisor
DECEMBER 2015
The undersigned, appointed by the Dean of the Graduate School, have
examined the dissertation entitled
CHARACTERIZING THE RETROVIRAL ENVELOPE GLYCOPROTEIN
MEMBRANE PROXIMAL EXTERNAL REGION AND MEMBRANE-SPANNING
DOMAINS FOR THEIR ROLES IN HELICAL ALIGNMENT, FUSOGENICITY, AND
INCORPORATION INTO VIRAL PARTICLES
Presented by Daniel Salamango,
a candidate for the degree of Doctor of Philosophy,
and thereby certify that, in their opinion, it is worthy of acceptance
____________________________________________
Dr. Marc C. Johnson
____________________________________________
Dr. Donald H. Burke
____________________________________________
Dr. Bruce A. McClure
_____________________________________________
Dr. Michael J. Petris
ii
ACKNOWLEDGEMENTS
First and foremost, I would like to thank my mentor Dr. Marc Johnson for all of the time
and effort he has put into my scientific development. I cannot thank him enough for giving
me the opportunity to join his lab after leaving another during the middle of my thesis. His
knowledgebase and problem solving abilities have helped me immensely to learn the craft
that we as graduate students attempt to master. I also need to thank him for his behind-the-
scenes effort in helping me find a fantastic postdoctoral research laboratory. I am truly in
debt to Marc for everything he has done for me, and for that I say thank you.
I would like to thank previous lab members Sanath Janaka and Devon Gregory. Their
feedback on my research has been phenomenal and I cannot thank them enough for their
time and effort.
I have to give a shout out to Terri Lyddon. Terri has been a great friend and resource in the
lab and I will miss our bantering conversations about topics on which we share the same
viewpoints (she knows what I mean).
I need to thank my boy Erik Ladomersky for unwavering support in the best of times and
the worst of times during our tenure together at Mizzou. We made a lot of great memories
and things would have been unbearable without him. I would like to thank my parents who
provided an immense amount of emotional support during the times when I struggled in
both science and in life. Last but nowhere near least I need to thank my beautiful,
intelligent, and compassionate girlfriend Mackenzie. I look forward to a fantastic future at
the University of Minnesota with her.
iii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ............................................................................................. ii
LIST OF IMPORTANT ABBREVIATIONS ................................................................ v
LIST OF FIGURES AND TABLES ............................................................................... vi
1. INTRODUCTION......................................................................................................... 1
1.1 Retroviruses .......................................................................................................... 2
1.1.1 Retroviral Proteins and Genome ................................................................... 9
1.2 Retroviral Life Cycle .......................................................................................... 23
1.3 Retroviral Assembly ........................................................................................... 29
1.4 Recruitment of Env to viral particles ................................................................. 36
2. CHARACTERIZING THE MURINE LEUKEMIA VIRUS ENVELOPE
GLYCOPROTEIN MEMBRANE-SPANNING DOMAIN FOR ITS ROLES IN
INTERFACE ALIGNMENT AND FUSOGENICITY ............................................... 41
2.1 Summary ............................................................................................................ 42
2.2 Introduction ........................................................................................................ 43
2.3 Experimental Procedures.................................................................................... 45
2.4 Results ................................................................................................................ 48
2.5 Conclusions ........................................................................................................ 53
3. IN VIVO SELECTIONS REVEAL NOVEL PROTEIN DETERMINANTS FOR
MURINE LEUKEMIA VIRUS (MLV) ENVELOPE GLYCOPROTEIN
PSEUDOTPYING INTO HIV-1 PARTICLES ............................................................ 69
3.1 Summary ............................................................................................................ 70
3.2 Introduction ........................................................................................................ 71
3.3 Experimental Procedures.................................................................................... 74
3.4 Results ................................................................................................................ 79
3.5 Conclusions ........................................................................................................ 88
4. PROBING THE MEMBRANE PROXIMAL EXTERNAL REGION AND
MEMBRANE-SPANNING DOMAINS OF HIV-1 ENVELOPE FOR A ROLE IN
GLYCOPROTEIN INCORPORATION AND FUNCTION .................................... 105
4.1 Summary .......................................................................................................... 106
4.2 Body ................................................................................................................. 107
4.3 Future Directions .............................................................................................. 110
5. SUMMARY AND DISCUSSION ............................................................................ 116
iv
6. FUTURE DIRECTIONS .......................................................................................... 119
APPENDIX .................................................................................................................... 122
I. RECOMBINATION CAN LEAD TO SPURIOUS RESULTS IN RETROVIRAL
TRANSDUCTION WITH DUAL FLUORESCENT REPORTER GENES ........... 123
I.I Summary .......................................................................................................... 123
I.II Body ................................................................................................................. 123
II. IMMUNE SIGNALING COMPETENCY IS REGULATED BY LIGAND-
INDUCED ENDOCYTOSIS AND SUBSEQUENT DE NOVO SYNTHESIS OF
THE FLAGELLIN-RECEPTOR FLS2 ...................................................................... 136
II.I Summary .......................................................................................................... 136
II.II Introduction ...................................................................................................... 137
II.III Experimental Procedures .............................................................................. 139
II.IV Results .......................................................................................................... 141
II.V Conclusions ...................................................................................................... 148
III. CONTINUOUS FLG22 TREATMENT LEADS TO LIGAND-
DESENSITIZATION AND LOSS OF FLS2 PROTEIN ACCUMULATION ....... 167
III.I Summary .......................................................................................................... 167
III.II Introduction .................................................................................................. 167
III.III Experimental Procedures .............................................................................. 169
III.IV Results .......................................................................................................... 171
III.V Conclusions .................................................................................................. 178
III.VI FUTURE DIRECTIONS .............................................................................. 179
IV. IDENTIFYING ARABIDOPSIS EPSINR1 AS A REGULATOR OF FLG22-
INDUCED SIGNALING RESPONSES ...................................................................... 193
IV.I Summary .......................................................................................................... 193
IV.II Introduction .................................................................................................. 194
IV.III Experimental Procedures .............................................................................. 196
IV.IV Results .......................................................................................................... 199
IV.V Conclusions .................................................................................................. 208
IV.VI Future Directions .......................................................................................... 211
REFERENCES .............................................................................................................. 237
VITA............................................................................................................................... 274
v
LIST OF IMPORTANT ABBREVIATIONS
MLV: Murine Leukemia Virus
HIV: Human Immunodeficiency Virus
Env: Envelope
ER: Endoplasmic Reticulum
MSD: Membrane-spanning domain
MPER: Membrane proximal external region
SU: Receptor binding domain of Env
TM: Transmembrane domain of Env that has fusogenic activity
R-peptide: Regulatory peptide at the end of the cytoplasmic tail of Env
CTD: Cytoplasmic tail domain
RT: Reverse transcriptase
CA: Capsid
MA: Matrix
IN: Integrase
PR: Protease
Pseudotype: Recruitment of a glycoprotein to a non-native virus
vi
LIST OF FIGURES AND TABLES
Figure 1.1: Phylogeny of Retroviruses ............................................................................ 3
Figure 1.2: Genomic organization of retroviruses ......................................................... 5
Figure 1.3: Diagram of a retroviral particle. .................................................................. 7
Figure 1.4: Polyprotein Translational Mechanisms .................................................... 13
Figure 1.5: Illustration of an Env Trimer ..................................................................... 14
Figure 1.6: Biosynthesis and trafficking of Env ........................................................... 17
Figure 1.7: Cytoplasmic tail length of retroviral glycoproteins ................................. 19
Figure 1.8: Integrated HIV-1 proviral DNA genome .................................................. 21
Figure 1.9: Depiction of the HIV-1 life cycle ................................................................ 25
Figure 1.10: Reverse transcription mechanism............................................................ 27
Figure 1.11: Depiction of various retroviral assembly sites ........................................ 32
Figure 1.12: ESCRT involvement in HIV-1 and MLV virion budding ..................... 34
Figure 1.13: MLV Env active recruitment to foreign viral assembly sites ................ 39
Figure 2.1: Diagram of predicted trimer interfaces. ................................................... 57
Figure 2.2: Infectivity of full-length deletion glycoproteins in the +1L insertion
background. ..................................................................................................................... 59
Figure 2.3: Leucine insertions upstream and downstream of the respective glycine-
proline pairs do not display a phasing effect. ............................................................... 61
Figure 2.4: Truncation of the cytoplasmic tail restores abrogated infectivity of
certain deletion mutants. ................................................................................................ 63
Figure 2.5: The hydroxyl-containing amino acids in the MSD are crucial for
promoting cell-to-cell fusion........................................................................................... 65
Figure 2.6: Amino acid substitutions in the critical hydroxyl positions display
defective functionality. .................................................................................................... 67
Figure 3.1: Target region of mutagenesis and selection strategy. .............................. 91
Figure 3.2: NGS selection analysis. ............................................................................... 93
Figure 3.3: Two regions in the MPER and MSD are predicted to be important for
Env incorporation. .......................................................................................................... 95
Figure 3.4: Addition of the full-length cytoplasmic tail has a minimal effect on
mutant env functionality. ............................................................................................... 97
Figure 3.5: Several candidate mutants identified in the selection screen display
abrogated infectivity but not fusogenicity or cell surface expression. ....................... 99
vii
Figure 3.6: Specific point mutations in, or near, the membrane-proximal domain
display severely abrogated incorporation into HIV particles. .................................. 101
Figure 3.S1: Mutant Env infectivity is similar between MLV and HIV cores. ....... 103
Figure 4.1: Target region of mutagenesis and selection strategy. ............................ 112
Figure 4.2: Mutational landscape of the doped HIV-1 Env plasmid library .......... 114
Figure I.I: Transduction using retroviral vectors carrying a TdTomato reporter
resulted in two cellular populations with distinct fluorescent intensities. ............... 128
Figure I.II: Retroviral recombination occurs between distinct fluorescent proteins.
......................................................................................................................................... 130
Figure I.III: Fluorescence intensity is independent of TdTom genomic integration
sites. ................................................................................................................................ 132
Figure I.IV: Bimodal fluorescence intensity of the TdTom reporter occurs as a
result of a single nucleotide polymorphism between the TdTom genes. .................. 134
Figure II.I: PF22-induced degradation of endogenous FLS2 is ligand-, dose- and
time-dependent and is inhibited by cycloheximide. ................................................... 150
Figure II.II: Ligand-induced degradation of FLS2 is a means to desensitize cells to
stimuli. ............................................................................................................................ 152
Figure II.III: Ligand-induced de novo synthesis of FLS2 prepared cells for a new
round of PF22-perception. ........................................................................................... 154
Figure II.IV: Regulation of ligand-induced signaling competency of EFR is
regulated similar to PF22-FLS2. ................................................................................. 156
Figure II.S1: Prolonged activation of PF22-signaling in the presence of CHX is
RBOHD-dependent. ...................................................................................................... 158
Figure II.S2: PF22-signaling incompetency is PAMP-, time- and dose-dependent. 160
Figure II.S3: Recovery of signaling competency was due to PF22-induced de novo
synthesis of FLS2........................................................................................................... 162
Figure II.S4: Recovery of elf26-signaling competency using ROS production. ...... 165
Figure III.I: Continuous PAMP Treatment Desensitizes Cells to the Different
PAMPs. .......................................................................................................................... 182
Figure III.II: PAMP-Induced ROS Desensitization Occurs After Six Hours
Following Initial PAMP Treatment in Continuously Elicited Tissue. ..................... 185
Figure III.III: Elicitation using the avirulent bacterial strain Pseudomonas syringae
DC3000 hrcC- Results in a Similar Desensitization Pattern as Shown for PAMP-
peptides. ......................................................................................................................... 187
Figure III.IV: Syringe Infiltration Does Not Simulate Exogenous Continuous
Elicitation. ...................................................................................................................... 189
viii
Figure III.S1: Confirmation Syringe Infiltration of flg22 Elicited an Immune
Response......................................................................................................................... 191
Figure IV.I: EpsinR1 Functions as a Positive Regulator of Early-Induced flg22-
Signaling Responses. ..................................................................................................... 217
Figure IV.II: Role for EpsinR1 in Intermediate and Late-Induced flg22-Signaling
Responses. ...................................................................................................................... 219
Figure IV.III: BAK1-Dependent Signaling Mechanisms May be Disrupted in epsR1.
......................................................................................................................................... 221
Figure IV.IV: Morphological comparison of epsinR1 alleles at 8-days Post-
Germination................................................................................................................... 223
Figure IV.V: The epsR1 Alleles Display Defective Primary Root Cell Elongation and
Pavement Cell Expansion. ............................................................................................ 225
Figure IV.VI: Increased Stomatal Number and Clustering in epsR1 Alleles. ......... 227
Figure IV.S1: Leaf Tissue Display Decreased ROS Production and PAMPs Dose
Dependency. ................................................................................................................... 229
Figure IV.S2: EpsinR1 Functions as a Positive Regulator of Early-Induced flg22-
Dependent MAPK Activation. ..................................................................................... 231
Figure IV.S3: Increase in Stomata Number and Clustering is Independent of
Cellular Expansion........................................................................................................ 233
Table S1: Statistical Analysis of Cellular Pools from Abaxial Cotyledons. ............ 235
1
1. INTRODUCTION
2
1.1 Retroviruses
A retrovirus is an enveloped RNA virus that utilizes a process termed reverse
transcription and integrates a double-stranded DNA copy of its RNA genome into the host
cell chromosome. Reverse transcription is carried out by the viral enzyme reverse
transcriptase (RT) and is the characteristic by which retroviruses are defined. The retroviral
family is comprised of seven members subdivided into two categories (Figure 1.1). Alpha-
, Beta-, Gamma-, and Epsilon- retroviruses are classified as simple, whereas Delta-, Lenti-
, and Spuma- retroviruses are classified as complex. The distinction between simple and
complex retroviruses arises from the repertoire of genes encoded by the viral genome. Both
simple and complex retroviruses maintain three core genes: Gag, Pol, and Envelope (Env).
However, complex retroviruses encode additional accessory proteins that aid in viral
pathogenesis (Figure 1.2). Gag acts as the viral structural protein, Pol encodes genes with
enzymatic function, and Env produces the viral surface glycoproteins necessary for
initiating viral entry and targeting specific cell types (Figure 1.3). The viral accessory
proteins provide vital functions that support viral replication and/or pathogenesis. While
not all accessory genes are necessarily required for replication of the respective virus, some
are instrumental in the virus’ ability to overcome host cell restriction factors.
3
Figure 1.1: Phylogeny of Retroviruses
Depicted is an unrooted phylogenetic tree of the retroviral subfamilies. Simple and
complex classifications are designated based on the presence or absence of accessory genes
encoded from the viral genome. The full names of the abbreviated viruses are listed at left
and the complex or simple designation is listed next to the appropriate family.
4
Figure 1.1: Phylogeny of Retroviruses
5
Figure 1.2: Genomic organization of retroviruses
Figure 1.2 displays the genomic organization of the indicated viruses. Simple
retroviruses encode the three core viral genes of Gag, Pol, and Env, whereas complex
retroviruses encode multiple accessory proteins. Some accessory genes aide in viral
replication through facilitation of viral genome import and export to and from the nucleus
(ex. HIV tat and rev). Others aide in viral replication through suppression of host-cell
restriction factors (ex. HIV vif, vpr, vpu, and nef).
6
Figure 1.2: Genomic organization of retroviruses
7
Figure 1.3: Diagram of a retroviral particle.
All retroviral genomes encode at least three genes: gag, pol, and Env. The cleavage
products of gag, which are matrix (MA), capsid (CA), and Nucleocapsid (NC), are
primarily structural components of the viral particle and responsible for binding to the viral
genome. Pol encodes integrase (IN), protease (PR), and reverse transcriptase (RT). Env is
comprised of a receptor binding domain (SU) and a transmembrane domain (TM).
8
Figure 1.3: Diagram of a retroviral particle.
9
1.1.1 Retroviral Proteins and Genome
Human immunodeficiency virus (HIV) is the most well-studied, and well-
publicized, retrovirus since its discovery in 1983. Rightfully so, seeing as it has been linked
to over 30 million deaths in the past 30 years, with over 35 million people currently living
with the virus (W.H.O., 2013) . HIV, along with all other members of the retroviral family,
encodes three core genes in its viral genome: Gag, Pol, and Env.
HIV-1 Gag, which is the main structural component of the virus, is produced as a
single 55 kDa polyprotein that includes the N-terminal Matrix (MA), Capsid (CA), and
Nucleocapsid (NC) domains. MA targets the viral Gag polyprotein to the inner leaflet of
the plasma membrane during assembly and has been shown to utilize several mechanisms
to facilitate binding to lipids. For all retroviruses, the MA domain is post-translationally
modified by the addition of a myristoyl group at the second residue (glycine), which has
been shown to be critical for protein binding to membranes (1). MA also contains a PIP2
(Phosphatidylinositol 4,5-bisphosphate) binding motif that mediates membrane binding
through interaction with PIP2 lipid head groups. Additionally, a highly basic region (HBR),
consisting of multiple lysine residues, aides in binding MA to the negatively charged
phospholipid head groups of the membrane (2, 3). The viral CA protein forms the
protective cage-like structure surrounding the viral genome (4, 5). CA has an N-terminal
domain that is responsible for forming hexameric and pentameric rings and a C-terminal
domain that forms homodimers allowing for the formation of the core structure (6). The
final domain, NC, recognizes the genomic viral RNA and forms crucial interactions
between Gag molecules when forming the immature particle. Gag expression alone is
sufficient to promote formation of virus like particles (VLPs) that are released from cells.
10
In general, the Gag ORF is transcribed and translated as the Gag protein only,
however; occasionally, a ribosomal frameshift results in the translation of Gag-Pol (7). For
Murine Leukemia Virus (MLV), Gag-Pol production is achieved by stop codon
suppression or read-through (Figure 1.4). In the case of Beta- and Delta-retroviruses,
protease is encoded by a Gag and Pol independent ORF. In this case, two different
frameshifting events are required for production of protease. In rare instances, a non-AUG
initiation codon (CUG) can also be used to initiate translation upstream of Gag (8). In
MLV, this leads to a N-terminal extension of Gag that acts as a modified hydrophobic
signal-peptide that targets Gag to the endoplasmic reticulum (ER), where it is translocated
across the membrane and undergoes glycosylation. However, instead of being cleaved, the
N-terminal extension serves as a membrane anchor, with a majority of the protein
extending into the Lumen of the ER where it is glycosylated (9). Thus, these molecules are
referred to as glyco-Gag proteins.
Pol encodes the enzymatic protease (PR), reverse transcriptase (RT), and integrase
(IN) proteins. PR proteolytically cleaves the individual structural and functional domains
from Gag and Pol after undergoing a self-cleavage event that subsequently releases it from
the rest of the Pol protein. PR activation occurs during a process termed viral maturation,
which happens at, or immediately after, viral release from the host cell. RT is the enzyme
responsible for converting the viral RNA genome to cDNA, while IN is a crucial part of
the complex necessary for integrating a copy of the viral dsDNA genome into the host cell
chromosome (10).
The final core gene, Env, is responsible for producing the viral glycoprotein that
facilitates viral entry through recognition of a cell surface protein. Env is produced from
11
an independent spliced transcript as a polyprotein that is processed into a hetero-trimeric
transmembrane surface protein. Env is synthesized in the endoplasmic reticulum where it
trimerizes and is subsequently cleaved by a host-cell furin or furin-like protease in the
Golgi. Proteolytic cleavage results in the formation of SU, which is the receptor-binding
domain, and TM, which serves as the transmembrane anchor and contains fusogenic
activity (Figure 1.5). In the case of alpha-, gamma-, and deltaretroviruses, SU and TM are
covalently linked through a single disulfide bond (11-15), whereas lentiviruses and
betaretroviruses are non-covalently associated (16, 17). Upon receptor recognition by SU,
numerous molecular rearrangements ensue which promote co-receptor binding, and/or
remodeling of the host cell membrane by the TM fusion peptide. Following trimerization
in the ER, which is thought to be the rate limiting step in biosynthesis, Env undergoes
substantial N- and O-linked glycosylation while trafficking through secretory pathway (10,
18) (Figure 1.6).
Interestingly, lentiviral Env glycoproteins have a very long cytoplasmic tail domain
(CTD) compared to other retroviral Envs (Figure 1.7). It is thought that this may indicate
a functional significance, especially since it has been shown that truncation of the CTD
results in altered trafficking, viral incorporation, fusogenicity, and cell surface expression
(19-21). The CTDs of retroviral Env proteins also contain classical YxxΦ and dileucine
trafficking motifs, which are thought to maintain an optimal level of Env on the cell
surface. Tightly regulating the expression of Env at the cell surface, at least in the case of
HIV-1 Env, it is thought to limit the level of immunogenicity while allowing an adequate
number of trimers to incorporate into assembling particles. Mutations of the YxxΦ motif
for HIV, Human T-lymphotropic virus (HTLV), and MLV can result in aberrant Env
12
distribution and infectivity (22-24). The CTD of retroviral Env glycoproteins also contain
a consensus palmitoylation sequence that is thought to help promote association with
specific lipid microdomains in the membrane (25-27).
A unique feature of Beta- and Gamma-retroviral Envs is that they have a regulatory
peptide at the end of their CTD that negatively regulates fusogenicity. Cleavage of this
peptide (R-peptide) by the viral protease shortly after viral maturation results in
conformational changes within the Env trimer which activates fusogenicity.
The retroviral genome is a dimer of linear, positive-sense, single-stranded RNA
(ssRNA) monomers, each being 7 to 13 kb in length. Repetitive (R) nucleotide sequences
flank either end of the viral genome with untranslated regions residing downstream, and
upstream, of the 5’ and 3’ repeat sequences, respectively (termed U5 and U3). The
integrated proviral DNA contains U3-R-U5 elements at either end of the genome, termed
long terminal repeats (LTRs), which are formed following reverse transcription (Figure
1.6). The viral LTRs serve as a promoter and aide in transcription of viral genes encoded
from the genome. The genome also contains several cis-acting elements that are absolutely
essential for producing infectious particles. These elements are the dimerization
initiation/linkage sequence (DIS or DLS), primer binding site (PBS), which serves as the
initiation site of reverse transcription, a packaging element (Ψ) used by NC to encapsulate
genomic RNA, and a poly purine tract (PPT) used during reverse transcription. The dimeric
interaction between the RNA monomers is maintained by a self-complimentary sequence
(DLS or DIS) located at the 5’ end of the genome.
13
Figure 1.4: Polyprotein Translational Mechanisms
Gag and Gag-Pol expression occurs from the same RNA transcript. Translation of the full-
length poly protein occurs through ribosomal frameshifting, stop codon suppression, or
stop codon read-through. Depicted below the mechanisms that HIV and MLV use to
express Gag and Gag-Pol from a single transcript.
14
Figure 1.4: Polyprotein Translational Mechanisms
15
Figure 1.5: Illustration of an Env Trimer
Env has two domains: the receptor binding domain SU, and the transmembrane domain
TM, which contains fusogenic activity. Viral-to-cell membrane fusion is accomplished
through the fusion peptide located at the N-terminus of TM.
16
Figure 1.5: Illustration of an Env Trimer
17
Figure 1.6: Biosynthesis and trafficking of Env
Env is synthesized in the secretory pathway after translocation to the endoplasmic
reticulum (ER). Env undergoes trimerization in the rough ER and is subsequently trafficked
to the Golgi and trnas-Golgi network (TGN) where it is cleaved into SU and TM by a
cellular furin or furin-like protease. The final destination is the plasma membrane (PM)
where it is incorporated into assembling viral particles. Env is also trafficked from the PM
to early and late endosomal compartments where it is either recycled back to the PM or to
the lysosomes for degradation. Simultaneously, Gag (MA, CA, NC) is being synthesized
by the ribosomes in the cytoplasm as a soluble protein. Once biosynthesis is complete, Gag
is trafficked to the plasma membrane where it begins to multimerizes and for the viral
assembly site. Image from: (28)
18
Figure 1.6: Biosynthesis and trafficking of Env
19
Figure 1.7: Cytoplasmic tail length of retroviral glycoproteins
Lentiviruses have a much longer CTD compared to other retroviral families, which is
thought to impart an functional advantage. The retroviral glycoprotein CT lengths are
depicted below. Lentiviral Env CTs are grouped in one cluster and other retroviral Env
families are grouped in another.
20
Figure 1.7: Cytoplasmic tail length of retroviral glycoproteins
21
Figure 1.8: Integrated HIV-1 proviral DNA genome
Depiction of the integrated form of the DNA provirus from HIV-1. The DNA form contains
the U3-R-U5 long terminal repeats on either end of the viral genome which are absent in
the RNA form. The example used in the depiction is of a HIV-1 provirus with the indicated
core and accessory genes. Gag, Pol, and Env are the core genes and Vif, Vpr, Vpu, Nef are
accessory genes. Tat and Rev are components that are necessary for transportation of the
unspliced viral genome.
22
Figure 1.8: Integrated HIV-1 proviral DNA genome
23
1.2 Retroviral Life Cycle
The viral life cycle begins (or ends depending on one’s perspective) with the Env
glycoprotein binding to the appropriate cell surface receptor, which initiates entry events
(Figure 1.9). In some instances, receptor binding triggers co-receptor binding, as in the case
of HIV-1 Env, or directly leads to activation of the fusion mechanism. Retroviral Envs use
a type I fusion mechanism to fuse the target cell membrane with the viral membrane (29).
This process involves a hydrophobic stretch of amino acids at the end of the TM domain
called the fusion peptide, which embeds into the membrane of the target cell bringing both
membranes into close proximity to facilitate lipid mixing and formation of a fusion pore.
The viral capsid core utilizes the fusion pore to gain entry into the host cell cytoplasm
where reverse transcription of the viral genome takes place. Many of the details of reverse
transcription, capsid uncoating, and nuclear import are still poorly understood.
Initiation of reverse transcription is primed by a cellular tRNA molecule binding to
the PBS. During reverse transcription, several RNA/DNA intermediates are formed with
template switching occurring multiple times. Cis elements in the RNA genome serve as
guide elements during this process, such as the PPT. To synthesize the second strand of the
dsDNA genome, the PPT is used as the primer. RT binds to the PPT
The RNaseH domain of RT degrades these intermediates during the process of dsDNA
synthesis (30) (Figure 1.10).
Following the completion of reverse transcription, the dsDNA genome needs to
enter the nucleus and integrate into the host cell chromosome. This process is completed
by the formation of a pre-integration complex (PIC). The PIC contains viral and cellular
24
proteins that are imported into the nucleus (as in the case of lentiviruses) or gain access to
the genome during mitosis, when the nuclear membrane dissipates. IN aides in integrating
the dsDNA into the cellular genome (31). The various viral transcripts are then trafficked
out of the nucleus to the cytoplasm to be translated by the host cell machinery. Unspliced
genomic RNA is also trafficked into the cytoplasm so that it can be recruited to the viral
assembly site and packaged into nascent virions.
25
Figure 1.9: Depiction of the HIV-1 life cycle
The viral surface glycoprotein binds to the receptor on the host cell which initiates
membrane fusion and creates a pore through which capsid is delivered. Once inside the
host cell reverse transcription is initiated and the dsDNA viral genome is delivered into the
nucleus where is undergoes integration into the host cell chromosome. Following
integration, the viral genes are transcribed and viral proteins are synthesized along with a
RNA copy of the viral genome. These components are then assembled to form new
particles. Image from: (http://www.daviddarling.info/encyclopedia/A/acquired
_immune_deficiency_syndrome.html)
26
Figure 1.9: Depiction of the HIV-1 life cycle
27
Figure 1.10: Reverse transcription mechanism
The complete process of reverse transcription is depicted. The steps of the process are
labeled at the left of the corresponding step. Reverse transcription is begins by a cellular
t-RNA binding to the PBS which initiates minus-strand DNA synthesis. RNasH then
degrades the RNA portion of the RNA-DNA hybrid which frees up the resulting short,
single-stranded DNA fragment. This product is then transferred to the other strand where
it hybridizes with the repeat sequence at the 3’ end of the genome. Synthesis then
continues along the RNA template which is subsequently degraded by RNasH. The PPT
serves as the primer for the plus strand which is then synthesized. The plus strand
synthesis goes until complete replication of the viral genome into DNA. Image from: (32)
28
Figure 1.10: Reverse transcription mechanism
29
1.3 Retroviral Assembly
Assembly of a nascent viral particle requires spatiotemporal coordination of the
viral RNA genome, Gag, Gag-Pol, and Env proteins to coalesce at the same site in the cell.
A brief overview of events is as follows: 1) Gag interacts with the viral RNA genome and
traffics it to the cell membrane where multimerization ensues. 2) Multimerization begins
to define the viral assembly site where Env is recruited. 3) The final step is the budding of
a complete viral particle with the help of cellular trafficking proteins (6, 33).
Nascent retroviral particles assemble in the cytoplasm (termed Intracellular
Cytoplasmic A Particles; ICAPs) or at the plasma membrane. Assembly that occurs on the
plasma membrane is termed C-type assembly, which begins by Gag multimerization at the
inner leaflet of the plasma membrane. Cytoplasmic assembly is termed B/D type and in
this case, Gag multimerizes in the cytoplasm. Although these assembly sites are in two
distinctly different locations in the cell, a point mutation in Mason Pfizer Monkey Virus
(MPMV) Gag changes assembly from B/D to C type assembly, suggesting that assembly
mechanisms might be similar (34). Other types have particles have also been observed
within the cell, such as the Intracisternal A-type particles (IAPs). IAPs predominantly
assemble endogenously due to defective components of Pol and Env (35) (Figure 1.11).
All retroviral particles bud in an immature form, which have a spherical core with an
electron-translucent center, while mature particles have a specific morphology with an
electron dense center. The morphology of the virus is defined by CA-CA interactions that
can also determine the size of the mature viral particle (5).
30
Multiple components of the virus must interact sequentially to form an immature
particle. Although the spatiotemporal timing of these interactions are not well understood,
assembly requires protein-protein, protein-lipid, and protein-RNA interactions to form an
infectious particle. For HIV-1, it has been shown that Gag interacts with the viral RNA
genome in the cytoplasm, forming lower order multimers. Higher order Gag multimers
have been observed on the plasma membrane (36). Gag binding to membrane occurs
through three mechanisms: 1) Interaction with the lipid head groups through the highly
basic region, 2) PIP2 interactions with MA, and 3) Myristoylation of Gag (37-41).
The exact site of HIV-1 assembly is debated. Some imaging studies depict virion
assembly at the inner leaflet of the plasma membrane (42-44), while others demonstrate
assembly at intracellular vesicles (45, 46). Although it is worth noting that these vesicles
were later shown to be continuous invaginations of the plasma membrane (47, 48). For C-
type assembly sites, it has been shown that the lipidome of HIV-1 virions contain an
enrichment of lipid-raft associated phospholipids (49-51). The HIV-1 phospholipid
envelope is enriched for cholesterol and sphingolipids, both of which cluster at ordered
microdomains. Cholesterol depletion in virus producing cells interferes with viral release,
structural stability, and infectivity (52-55). Interestingly, HIV-1 Env has been shown to
contain cholesterol binding domains in the MPER which have been shown to be important
for fusogenic activity. Additionally, some lipid raft associated proteins have been reported
to be incorporated into HIV-1 particles (56). Although a significant amount of correlative
evidence exists for HIV-1 association with lipid rafts, it is not entirely known where
assembly sites occur and if it is always in the same environment. It is plausible that the
presence of Gag multimerization at the plasma membrane induces the formation of specific
31
lipid microdomains that are required to form assembly sites that produce infectious
particles.
Regardless of the nature of the assembly site, the assembled virion needs to bud
from the host cell to complete the process. This process is completely dependent on late
endosomal host cell trafficking machinery, called ESCRTs. An ESCRT component,
Tsg101, is recruited to the viral assembly site through association with the PTAP motif in
p6, which is part of Gag. Most retroviral Gags possess a secondary PPxY motif in the p6
domain that may initiate activation of the ESCRT machinery to aide in budding. Mutation
of these motifs in the p6 domain results in an immature particle phenotype which cannot
complete the budding process (57-61) (Figure 1.12).
Spread of the HIV-1 virus has been shown to be more efficient in a cell-to-cell
manner, as opposed to cell-free viral spread, through formation of virological synapses (62,
63). The virological synapse serves as a secondary assembly site where Gag and Env are
recruited along with several other proteins such as CD4 and adhesion molecules (64, 65).
32
Figure 1.11: Depiction of various retroviral assembly sites
C-type assembly occurs at the inner leaflet of the plasma membrane, while B/D type
assembly occurs in the cytoplasm, followed by subsequent trafficking to the plasma
membrane for budding. IAPs assemble in the lumen of the ER. Image from: (33)
33
Figure 1.11: Depiction of various retroviral assembly sites
34
Figure 1.12: ESCRT involvement in HIV-1 and MLV virion budding
Diagram of the interacting proteins involved in HIV-1 and MLV nascent virion budding
from the host cell. The binding partners are shown with the indicated interaction motif in
the respective Gag protein.
35
Figure 1.12: ESCRT involvement in HIV-1 and MLV virion budding
36
1.4 Recruitment of Env to viral particles
Acquisition of viral Env glycoproteins during assembly is absolutely essential for
forming an infectious particle. Interestingly, retroviruses can incorporate foreign
glycoproteins from other retroviruses and from different viral families. This process is
termed pseudotyping and results in the formation of an infectious particle with an altered
cell tropism (66-68). This phenomenon has been capitalized on by researchers from
numerous fields, especially for applications such as gene therapy; however, very little is
known about the molecular mechanisms that govern this process. Env is synthesized in the
secretory pathway, where is trimerizes in the ER, is trafficked to the Golgi where it is
leaved into SU and TM by a cellular protease, and then is trafficked to the site of viral
assembly where it is incorporated into assembling viral particles.
Pseudotyping does not necessarily apply to all circumstances, it does have
restrictions. For example, some lentiviral glycoproteins are incompatible for incorporation
into non-native particles, presumably because of the steric hindrance of the CTD and the
Gag lattice. The solution to this problem, however, is not as simple as removal of the CTD.
In replication permissive cell lines, HIV-1 ΔCTD Env is incorporated into viral particles
to near normal levels. However, in peripheral blood mononuclear cells and monocyte
derived macrophages, HIV-1 ΔCTD Env is not incorporated into viruses efficiently and as
a result, these cell lines are non-permissive for viral replication. Interestingly, HIV-1
ΔCTD Env is processed normally and is fusogenically active (69-72). Ultimately, the
mechanistic differences for the requirements of the CTD of HIV-1 Env in these cell lines
is not well understood. In the case of MLV Env, removal of the CTD has no discernable
effect on Env incorporation to either native or foreign particles. However, in a competition
37
assay where both MLV and HIV particles are expressed in the same cell, full-length MLV
Env will exclusively partition to MLV particles. Removal of the CTD results in equal
incorporation of MLV Env to both types of particles (73, 74).
In some cases, studies suggest that Env can dictate the site of viral assembly. When
HIV-1 Env is expressed in MDCK cells, HIV viral assembly occurs on the basolateral
surface which may be a result of enrichment of negatively charged phospholipid
headgroups (22-24). In rat neuronal cells, both HIV and MLV Envs cause a redistribution
of their respective Gag molecules (75). MLV Env can also direct MLV Gag recruitment to
viral assembly sites in the virological synapse (76). Additionally, it has been demonstrated
that MLV Env can be actively recruited to non-native viral assembly sites (77). In the
absence of assembling virus, MLV Env displays a random distribution on the cell surface.
However, if assembling virus is present, either MLV or HIV, MLV Env localization
redistributed dramatically to the viral assembly sites (Figure 1.13).
Retroviral Env CTDs also have trafficking motifs that play roles outside of
regulating trafficking of Env, such as in directing virus assembly and pathogenesis. In
polarized epithelial cells, the tyrosine sorting motif determines basolateral targeting of Env
and allows Gag and Env to unite at the virological synapse, assisting in cell-to-cell
transmission of the virus (22-24, 78). An example of a cellular protein that can mediate
incorporation of HIV-1 Env to HIV particles is Tail Interacting Protein (TIP47). TIP47 has
been implicated in lipid biogenesis and protein sorting to those lipids, and has been
implicated as serving as a bridge between hiv-1 Env CTD and HIV-1 MA (79, 80).
However, these findings have been deemed somewhat controversial because of in vitro
studies that showed a non-requirement for TIP47 between HIV-1 Env CTD and HIV MA
38
binding. RNA mediated knockdown of TIP47 also showed no change in infectivity, virion
release, or Env incorporation into HIV-1 particles (81).
The main question we aimed to address in this work was to determine what protein
component(s) in the retroviral Env glycoprotein, specifically MLV Env, dictate
pseudotyping. Additional relevant questions that we would like to address in the future
would be:
Do all glycoproteins that exhibit pseudotyping utilize the same molecular
mechanism? Do these mechanisms differ for retroviral glycoproteins vs.
glycoproteins from other viral families?
If we can determine the protein components that dictate pseudotyping, can we then
engineer glycoproteins to pseudotype that normally do not exhibit that behavior?
For example, full length Rous Sarcoma Virus Env (RSV Env) glycoprotein does
not actively pseudotype with HIV-1 particles. Can we modify RSV Env’s protein
sequence so that it undergoes pseudotyping with HIV-1 particles?
Can we engineer host-cell proteins to pseudotype with HIV-1 particles? Our lab has
demonstrated, using multiple host-cell proteins, that pseudotyping typically only
occurs with viral glycoproteins. This application would be directly relevant for
lentiviral based gene therapy.
39
Figure 1.13: MLV Env active recruitment to foreign viral assembly sites
SEM imaging of the cell surface with immunogold labeled MLV Env. Images depict
localization of MLV Env in a cell not infected (Left), infected with RSV Gag (Middle), or
infected with HIV-1 Gag (Right). The images depict dramatic redistribution of MLV Env
in the presence of assembling viral particles. Image from: (77)
40
Figure 1.13: MLV Env active recruitment to foreign viral assembly sites
41
2. CHARACTERIZING THE MURINE LEUKEMIA VIRUS
ENVELOPE GLYCOPROTEIN MEMBRANE-SPANNING
DOMAIN FOR ITS ROLES IN INTERFACE ALIGNMENT
AND FUSOGENICITY
42
2.1 Summary
The membrane-proximal region of Murine Leukemia Virus Envelope (MLV Env)
is a critical modulator of its functionality. We have previously shown that the insertion of
one amino acid (+1 leucine) within the membrane-spanning domain (MSD) abolished
protein functionality in infectivity assays. However, functionality could be restored to this
+1 leucine mutant by either inserting two additional amino acids (+3 leucine), or by
deleting the cytoplasmic tail domain (CTD) in the +1 and +2 leucine background. We
inferred that the ectodomain and CTD have protein interfaces that have to be in phase with
each other, or with another protein component in Env, to be functional. Here, we made
single residue deletions to the +1 leucine Env background to restore the phasing (gain of
functionality) and therefore define the boundaries of the two interfaces. We identified the
glycine-proline pairs near the N-terminus (positions 147 and 148) and the C-terminus
(positions 159 and 160) of the MSD as being the boundaries of the two interfaces.
Deletions between these pairs restored function, but deletions outside of them did not.
However, the vast majority of the -1 deletions also regained function if the CTD was
deleted. The exceptions were four hydroxyl-containing amino acid residues (T139, T140,
S143, T144) in the membrane-proximal interface, which were all indispensable for
functionality. We hypothesize that these hydroxyl-containing residues could be a driving
force for stabilizing the ectodomain interface through formation of a hydrogen-bonding
network that directs orientation and oligomerization of the Env monomers to contribute to
glycoprotein trimer formation.
43
2.2 Introduction
Enveloped viruses require membrane-spanning cell surface glycoproteins to
coordinate fusion between viral and host cell membranes. Retroviral envelope (Env)
glycoproteins are produced as precursors that undergo trimerization in the endoplasmic
reticulum. Subsequently, the precursor trimer is cleaved into two subdomains: the receptor
binding surface domain (SU) and the fusion promoting transmembrane domain (82). This
cleavage process is mediated by a host-cell furin, or furin-like protease. In the case of
alpha-, gamma-, and deltaretroviruses, SU and TM are covalently linked through a single
disulfide bond (11-15), whereas lentiviruses and betaretroviruses are non-covalently
associated (16, 17). Upon receptor recognition by SU, numerous molecular rearrangements
ensue that promote co-receptor binding, and/or remodeling of the host cell membrane by
the TM fusion peptide (83).
In the case of Murine Leukemia Virus Envelope (MLV Env), fusogenicity is tightly
controlled by a short peptide (R-peptide) in the cytoplasmic tail domain (CTD) that
negatively regulates fusogenicity (84-87). Cleavage of the R-peptide by the viral protease
during or shortly after viral assembly results in isomerization of the disulfide bond between
SU and TM, subsequently activating fusogenic activity (88, 89). Recent cryo-electron
microscopy (EM) reconstruction data show that the ectodomain of MLV Env is held in a
tight conformation until cleavage of the R-peptide, wherein the TM legs are splayed,
allowing fusogenic activation of Env (90). Thus, R-peptide cleavage is thought to
potentiate fusogenic activation through molecular rearrangements in the MSD and in the
extracellular region of MLV Env.
44
Similarly, the membrane-proximal external region (MPER) and MSD of HIV-1
Env are crucial for promoting fusogenicity. Disruption of hinge regions or the tryptophan-
rich sequence in the MPER as well as polar residues within the MSD all result in non-
functional glycoproteins that are fusogenically inactive (91-96). These types of
mechanisms have parallels with MSDs of voltage-gated ion channels and G-protein-
coupled-receptors (GPCRs). Conformational changes in the transmembrane domains of
GPCRs allows for transmission of signals from the extracellular space to the intracellular
environment (97-101).
Recently, we demonstrated that leucine insertions in the MSD of MLV Env have a
significant effect on glycoprotein functionality. Insertion of +1 (+1L) or +2 leucines (+2L)
in the MSD abrogated function, whereas insertion of +3 leucines (+3L) restored infectivity
to levels comparable to wild-type Env. Interestingly, when the CTD was truncated in the
+1L background, infectivity was restored. Further study revealed that the loss of infectivity
in the +1L Env mutant was due to disruption of the fusion mechanism, and that +3L
restored fusogenicity and became constitutively active. All leucine insertion mutants were
incorporated into viral particles to similar levels compared to wild-type and all were
processed properly into SU and TM domains in cell lysates. We hypothesized that the CTD
forms a coiled-coil interface that disrupts the glycoprotein’s function if it is out of phase
with the trimer interface of the ectodomain. This hypothesis is supported by cryo-EM
reconstruction data wherein proper helical orientation would be required to promote the
molecular rearrangements necessary to release the ectodomain and activate fusogenicity.
Therefore, the +1L insertion would distort the helical orientation of the ectodomain and
CTD by approximately 103º, while the insertion of +3L would restore phasing by adding
45
almost a complete α-helical turn (Fig. 1). Restoration of function of the +1L Env in the
context of the CTD truncation suggested that removal of the CTD interface eliminated the
phasing conflict with the ectodomain, allowing for restoration of function. Here, we sought
to define the borders of these two interfaces by identifying gain-of-function mutations that
restored Env functionality in the +1L background.
2.3 Experimental Procedures
Plasmids and Cell Culture. The amino acid deletions and residue substitutions were
created using oligonucleotide-mediated mutagenesis. Constructs expressing the truncated
version of MLV Env were created by introduction of a stop codon after RLVQFVK, which
removes twenty-five residues from the cytoplasmic tail. Unless indicated, all MLV Env
expression constructs are HA-tagged (YPYDVPDYA) in a proline-rich region in SU (102).
Indicated deletion mutants with the truncated cytoplasmic tail and +1 leucine insertion
were cloned into an MLV genome that contained a puromycin gene upstream of the MLV
Env coding region driven by a cytomegalovirus (CMV) promoter. To ensure each protein
was expressed independently, a self-cleaving T2A peptide was inserted between the open
reading frames of the puromycin and MLV Env genes (103, 104). This construct also
contained a green fluorescent protein (GFP) tag inserted into a proline-rich region in SU.
For all infectivity, a NL4-3 derived HIV-CMV-GFP proviral vector, defective for Vif, Vpr,
Vpu, Nef, and Env, was used (Vineet Kewal-Ramani, National Cancer Institute). This
construct has a CMV immediate-early promoter driving a GFP reporter in place of Nef.
HEK-293FT (Invitrogen), 293T mCAT-1 (Walter Mothes, Yale University), and 293T
mCAT-1 cells stably expressing a tet-off Gaussia Luciferase (G. Luc.) promoter (102)
46
were maintained in DMEM supplemented with 10 % fetal bovine serum, 2 mM glutamine,
1 mM sodium pyruvate, 10 mM non-essential amino acids and 1 % MEM vitamins.
Infectivity Assay. 293FT cells were transfected in six well plate’s with 500 ng of HIV-
CMV-GFP and 500 ng of Env expression plasmid using 3 µg polyethylenimine (PEI) per
microgram of DNA (105). Media was changed 6 to 12 h post transfection to remove
residual transfection reagent. Supernatant was collected 24 h following media exchange
and frozen at -80ºC for at least 3 h to lyse any cells contained within the supernatant. After
thawing supernatants at 37ºC in a water bath, samples were centrifuged at 1,500 X g for 5
minutes to pellet any cellular debris. 500 µL of the supernatant was added to target cells
for 48 hrs. Cells were collected at 48 hrs, fixed with 4% paraformaldehyde, and analyzed
using an Accuri C6 flow cytometer.
Western Blotting. 293FT cells stably expressing the indicated Env proteins were
transfected with 500 ng of HIV-CMV-GFP plasmid in a six well plate following the same
procedure as described for infectivity assays. Viral samples were pelleted through a 20 %
sucrose cushion for 2 h at 20,000 X g at 4ºC. Residual supplement media and sucrose were
aspirated off of the sample pellet and samples were re-suspended in 6x SDS-PAGE loading
buffer. The equivalent of 1 mL of viral supernatant was analyzed by 10 % discontinuous
SDS-PAGE. Cell samples were detached using 10 mM EDTA/PBS solution and pelleted
at 500 x g for 10 minutes. Pellets were re-suspended in RIPA buffer (10 mM Tricl-Cl (pH
8.0), 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1%
SDS, and 140 mM NaCl) and 5-to-10 % of the lysate was combined with 6x SDS-PAGE
loading buffer and analyzed by 10 % discontinuous SDS-PAGE. Proteins were transferred
onto a 0.45 µm polyvinylidene difluoride (PVDF) membrane. The membrane was blocked
47
with 5 % nonfat dried milk in PBS containing Tween 20 (PBS-T) and probed with rabbit
anti-GFP antibody diluted 1:5000 (Sigma) and mouse anti-HIV p24 hybridoma media
diluted 1:500 (AIDS Research and Reference Program, Division of AIDS, NIAID, NIH;
HIV-1 p24 hybridoma [183-H12-5C]). Primary antibody incubations were performed
overnight, on a shaker, at 4ºC. Blots were washed with PBS-T and then probed with
horseradish peroxidase (HRP)-conjugated anti-rabbit and anti-mouse antibodies diluted
1:10,000 (Sigma) for probing Env and p24, respectively. Visualization of the membranes
was performed using Luminata Classico and Cresendo Western HRP chemiluminescence
reagents. Imaging was performed using a LAS3000 image analyzer from Fujifilm.
Cell-to-cell Fusion Assay. 293FT cells stably expressing the indicated Env protein were
transfected with 500 ng of tet-off expression plasmid in a six well plate. Media was changed
6 to 12 h post-transfection to remove residual transfection reagent. Transfected cells were
co-cultured with an equal number of 293T mCAT-1 TRE G. Luc. cells for 48 h (102).
Twenty microliters of sample supernatant from the co-cultured cells were assayed in
duplicate for G. Luc. content with 50 µL of 10 µM coelenterazine in 0.1M Tris (pH 7.4)
and 0.3M sodium ascorbate.
Surface Labeling. 293FT cells stably expressing the indicated Env protein were detached
using 10 mM EDTA/PBS. Cells were centrifuged at 500 X g for 10 minutes at 4ºC and re-
suspended in 1 % bovine serum albumin (BSA)/PBS blocking solution for 20 minutes.
After 20 minutes of blocking, cells were centrifuged at 500 X g for 10 minutes at 4ºC and
re-suspended in 10 mM EDTA/PBS, 1 % goat serum, and 1:1000 primary anti-GFP Alexa-
Fluor 647 antibody (Life Technologies) for 1 h. After primary antibody incubation, cells
were centrifuged at 500 X g for 10 minutes at 4ºC, re-suspended in 4 % paraformaldehyde,
48
and fixed for 20 minutes. Cells were then centrifuged at 500 X g for 10 minutes at 4ºC, re-
suspended in PBS, and analyzed using an Accuri C6 flow cytometer.
Statistical Analysis. Each experiment was done at least three independent times with
similar results. All statistics and graphs were done using Graph Pad Prism4 software.
2.4 Results
Identification of interface boundaries. We observed that proper helical
orientation of MLV Env trimers is crucial for infectivity (102). Disruption of the spacing
between the ectodomain and CTD caused by amino acid insertions within the MSD resulted
in loss of infectivity (102). We hypothesized that we could determine the boundaries of
these interfaces by identifying gain-of-function mutations that restored helical alignment
and, thus, function. In our previous study, we demonstrated that the insertion of a leucine
(+1L) between positions L154 and L155 in the MSD resulted in a non-functional
glycoprotein (Fig. 2.1) (102). If our hypothesis is correct and we perform a deletion scan
across the ectodomain, MSD, and CTD in the context of the +1L Env, we could identify
residues that restore infectivity (Fig. 2.2A). If a residue is located between the ectodomain
and CTD interfaces, removing it would offset the addition of the single leucine and restore
infectivity; however, if the residue is positioned within or beyond the interfaces
themselves, point deletions would not be expected to restore functionality. Furthermore, if
the CTD interface is destroyed, whether through truncation of the CT or deletion of critical
interfacial residues, we would also expect to see restoration of function because conflicting
interfaces would no longer exist. As shown in Fig. 2.2B, deletion of residues in the +1L
Env background upstream of P148, which is positioned within the MSD, did not restore
49
infectivity. However, deletion of residues between P148 and P160 restored infectivity
compared to the +1L Env control. Further, deletion of residues positioned downstream of
P160, within the CTD, did not restore function.
Interestingly, deletion of K171 or D172 in the CTD partially restored infectivity.
We have previously shown that point mutations in this region in the CTD could partially
restore function to a protein with the +1L in the MSD (102). Therefore, it is reasonable to
predict that these residues are critical to the CTD interface and that deleting them is
functionally similar to making truncations in the CTD (102).
Insertions outside of the glycine-proline pairs do not display a phasing effect.
If our hypothesis was correct and the glycine-proline pairs designated the end and
beginning of the ectodomain and CTD interfaces, respectively, then residue insertions
upstream or downstream of these glycine-proline pairs would not display the previously
observed phasing effect. Addition of +1 or +2 leucines between the glycine-proline pairs
abrogates infectivity; however, insertion of +3 leucines in this same region restored
function, presumably because this introduces one turn of an α-helix which effectively
restores helical alignment (Fig. 2.3; MSD insertions). We introduced +1, +2, or +3 leucines
at the N-terminus of the MSD, upstream of the G147/P148 pair, and assayed for infectivity.
As shown in Fig. 2.3, introduction of +1, +2, or +3 leucines upstream of G147/P148
completely abolished infectivity. Likewise, insertions of +1, +2, or +3 leucines in the CTD,
downstream of the G159/P160 pair, did not display a phasing effect (Fig. 2.3; CTD
insertions), suggesting the region between G147/P148 and G159/P160 does not have an
interface critical for function and can therefore tolerate the introduction of a turn of an α-
helix. Conversely, if the ectodomain and CTD have interfacial surfaces, introduction of
50
leucines in these regions would disrupt contacts between residues within the interfaces,
resulting in abrogated function (Fig. 2.3).
Discovery of residues in the MSD crucial for functionality. We previously
showed that deletion of the CTD restored functionality to both +1L and +2L MSD
insertions. Therefore, we sought to determine which of the residues within the ectodomain
and CTD interfaces were crucial for function even in the absence of the CTD. To
accomplish this, we made truncation mutants lacking the final 25 residues of the
cytoplasmic tail (Δ25CT). Truncation of the CTD resulted in restoration of function for
all of the deletion mutants examined except for hydroxyl-containing amino acids in the
ectodomain interface and P148 (Fig. 2.4). If the CTD had an interfacial surface, making
large truncations in this region would eliminate the conflict between trimer interfaces
because only the ectodomain interface would remain intact. As expected, all of the deletion
residues in the CTD displayed restored infectivity when the CT was truncated.
Hydroxyl-containing residues in the MSD are crucial for fusogenicity. Our
finding that P148 was crucial for function was expected since this residue most likely caps
the end of the ectodomain interface and acts as a molecular hinge within the trimer,
however, it was unexpected that the hydroxyl residues were also critical amino acids in the
membrane proximal ectodomain. The loss of functionality resulting from deletion of these
residues could be caused by many factors, therefore, we focused on determining the source
of the defect. To avoid inconsistences in expression levels of Env, we stably expressed the
hydroxyl-residue deletion mutants with CT truncations in 293FT cells. We also engineered
a green fluorescent protein (GFP) tag in the variable proline-rich region of SU for these
constructs. It has been shown that introduction of GFP, or peptides, into this proline-rich
51
variable region in SU is well tolerated with no discernable effects on Env function (106-
108). Additionally, deletions, insertions, or substitutions in, or near, the MSD of Env can
result in improper processing and aberrant trafficking (108). The GFP tag allowed for
visual confirmation of mutant Env expression and provided a conformation-independent
epitope for surface expression experiments. Surface expression studies revealed that all of
the truncated hydroxyl deletion mutants in the +1L background were expressed on the
surface relative to the truncated +1L Env positive control (Fig. 2.5A).
Next, we wanted to determine if these mutants were fusogenically active. Each cell
line stably expressing the indicated Env mutant was transfected with a transcriptional
activating tet-off expression construct (102). Twenty four hours post transfection, an equal
number of cells from each mutant cell line were co-cultured with an equal number of 293T
mCAT-1 cells expressing TRE-driven Gaussia luciferase (Gluc). If the Env mutants are
fusogenic, the transfected cells fuse with the receptor-expressing cells and tet-off-
dependent induction of Gluc occurs (102). Reporter induction can then be correlated
between mutant Env and wild-type controls to assess fusogenic activity. Interestingly, our
results indicated that all four of the hydroxyl deletion mutants displayed reduced cell-to-
cell fusion activity (Fig. 2.5B). To determine if loss of fusogenicity was the sole basis for
the loss of infectivity, we examined incorporation of these mutants into HIV particles. As
shown in Fig. 2.5C, all of the hydroxyl deletion mutant glycoproteins were incorporated
into HIV particles to a similar level compared to the truncated +1L control, indicating loss
of infectivity was due to loss of fusogenicity. Additionally, all of the mutant glycoproteins
accumulated to similar levels in the stable cell lines compared to the +1LΔ25CT control
(Fig. 2.5C; Cell). Taken together, the results depicted in Fig. 2.4 indicated that the loss of
52
infectivity was due to the reduction of fusogenicity, and not due to gross defects in Env
processivity, trafficking, or incorporation into viral particles.
Hydrogen-bonding requirements and steric constraints of the hydroxyl-
containing residues. Polar residues located in MSDs of transmembrane helices have been
shown to be important contributors to protein packing, oligomerization, and TM
stabilization (109-114). To determine if the hydrogen-bonding capacity of the residues we
identified in the ectodomain interface was important for function, we made alanine
substitutions to the wild-type protein in both the presence and absence of the full-length
cytoplasmic tail. We were surprised to observe that only two of the residues, T139 and
S143, displayed greatly reduced infectivity when substituted to alanine (Fig. 2.6A).
We hypothesized that perhaps these residues were positioned in, or near, the core
of the trimer interface and that their ability to hydrogen-bond was crucial for monomer
orientation or trimer stabilization. If this hypothesis were correct, it would also suggest that
positions T139 and S143 would have greater steric restriction to amino acid side-chain size
because of their location toward the trimer core. Conversely, positions T140 and T144
would be able to tolerate amino acid side chains with greater bulk because of their
orientation toward the outside of the trimer. To test this hypothesis, we made substitutions
at these positions with amino acid side chains that were still polar but had increasing bulk.
As shown in Fig. 2.6B, increasing amino acid side-chain size at positions T139 and S143
resulted in a corresponding decrease in infectivity. However, increasing amino acid side-
chain size at positions T140 and T144 showed a minimal effect on infectivity, suggesting
that these positions have less steric restriction (Fig. 2.6B). It is important to highlight that
substitution of T139 and S143 with hydrogen-bonding side chains (asparagine, glutamine,
53
and to a lesser extent, tyrosine) restored function compared to the alanine substitution. This
suggests that the polarity of the side-chain at these positions is important for function.
Substitution of phenylalanine at all of these positions resulted in severely decreased
infectivity. Unlike the alanine substitution, which only eliminated hydrogen-bonding, the
phenylalanine substitution also introduced a bulky side-chain with no polar functional
group. Importantly, as highlighted in Figure 2.6B, tyrosine substitution at each position
significantly restored infectivity compared to the phenylalanine substitution. The only
molecular difference between these two amino acid side chains is the presence of the
hydroxyl group on the benzyl ring of tyrosine, further suggesting that hydrogen-bonding
at these positions is crucial for function.
2.5 Conclusions
We have previously shown that coordination of the membrane-proximal
ectodomain and CTD interfaces is a critical modulator of glycoprotein functionality (102).
Disruption of the helical phasing by the insertion of +1L within the membrane-spanning
domain resulted in conflicting alignment of the ectodomain and CTD interfaces (Fig. 2.1).
Truncation of the cytoplasmic tail restored functionality of the +1L Env by relieving the
conflict between interfaces through removal of a majority of the CTD. This study explored
the boundaries of the ectodomain and CTD interfaces as well as probed the role of this
region in promoting fusogenic activity.
Identifying the boundaries of the ectodomain and the CTD. We identified the
glycine-proline pairs positioned within the MSD as residues that facilitate the transition
from the ectodomain interface to the predicted interface located in the cytoplasmic tail.
54
This observation is consistent with the conformational properties of proline as well as
previous studies that examined positioning of glycine-proline pairs within MSDs and their
roles in forming molecular hinges, packing of TM helices, and inducing α-helical kinks
(97, 110, 115-117). Therefore, the glycine-proline pairs act as ideal interface caps due to
distortions in the secondary structure, and subsequently the tertiary structure, of the trimer.
In addition, mutations to introduce +1, +2, or +3 leucines either upstream or downstream
of the glycine proline pairs did not display a phasing effect (Fig. 2.3). This suggests that
insertion of +3 leucines, and consequently nearly a complete turn of an α-helix, between
the glycine-proline pairs is tolerated because no critical interfacial surface exists. However,
+3L insertions in the ectodomain and CTD are not tolerated because it may disrupt contacts
between residues in these domains (Fig. 2.3).
Hydroxyl-containing residues in the ectodomain are critical for fusogenicity.
To understand why the hydroxyl-containing residues were crucial for Env function, we
probed the role of these residues by creating substitutions that abrogated hydrogen-bonding
or introduced steric hindrance (Fig. 2.6). We were surprised to find that only two of the
positions showed decreased function when hydrogen bonding was disrupted (Fig 2.6A;
T139A and S143A). Interestingly, these two positions also displayed the greatest
sensitivity to polar side-chain substitution with increasing bulk size. These results are
consistent with the hypothesis that positions T139 and S143 are part of the trimer core
while T140 and T144 are peripheral. This interpretation could also explain why deletion of
residues L141 and I142 displayed restored infectivity when the CT was truncated. Deleting
either of these residues would retain the hydrogen-bonding network by shifting T144
towards the core. This model is consistent with a previous study that identified residues in
55
the CTD involved in forming a trimeric coiled-coil interface (118). Using the residues
predicted to reside in the CTD interface, we can determine the location of the hydrogen-
bonding residues T139 and S143 from the α-helical phasing. We predict that T139 and
S143 reside in close proximity to the predicted coiled-coil trimer interface in the
ectodomain.
We suggest that residue T139 and S143 not only contribute to the fusion
mechanism, but may also stabilize the TM trimer or aide in monomer orientation through
formation of an inter-helical hydrogen-bonding network. A previous study focusing on the
MSD of MLV Env and a role in membrane fusion discovered that a W137S point mutant
induced cell-to-cell fusion even in the presence of the full-length CT, which supports the
hypothesis that hydrogen-bonding in this region of Env contributes to the fusion
mechanism (119). Additionally, a recent study using a random peptide library produced in
E. coli. identified two consensus motifs of SxxxSSxxT and SxxSSxxT that drove helical
oligomerization of transmembrane peptides (109). The hydroxyl-containing residues
we’ve identified as being crucial for function reside at the N-terminus of the MSD as part
of a consensus sequence of SxxxTTxxST, which may participate in similar processes as
described above. Other γ-retroviral glycoproteins such as Feline Leukemia Virus and
Gibbon ape Leukemia Virus also have similar consensus sequences, suggesting it may be
a conserved mechanism used to drive trimer oligomerization amongst certain viral
glycoproteins. Interestingly, Vesicular Stomatitis Virus glycoprotein has a similar
sequence of SSxxSSxxS at the external membrane-proximal/MSD interface, suggesting
this mechanism may not be exclusive to only γ-retroviral glycoproteins. Although HIV Env
does not have serine-threonine clustering in this region, it does have a GxxxG helical
56
packing motif in the MSD. It is possible that some glycoproteins evolved to utilize serine-
threonine clustering as a means to not only drive oligomerization but to also contribute to
the fusion mechanism, as in the case of MLV Env, while others use packing motifs
specifically for trimer oligomerization.
Together, these data suggest that the glycine-proline pairs cap the ectodomain and
CTD interfaces and that the hydroxyl-containing residues within the ectodomain interface
contribute significantly to the cell-to-cell fusion mechanism.
57
Figure 2.1: Diagram of predicted trimer interfaces.
(Left) Depiction of ectodomain and CTD trimer interfaces highlighted as white segments
within the trimer core of wild-type MLV Env. (Middle) Depiction of the +1L insertion
within the MSD and the resulting disruption of the CTD interface by approximately 103º.
(97) Depiction of the +3L insertions within the MSD and the subsequent restoration of
phasing of the CTD interface. The introduction of 3 leucines almost completely restores
alignment of the CTD interface within the trimer core.
58
Figure 2.1: Diagram of Predicted Trimer Interfaces
59
Figure 2.2: Infectivity of full-length deletion glycoproteins in the +1L insertion
background.
(A) Diagram of the extracellular, membrane-spanning, and cytoplasmic tail domains in
MLV Env targeted for the -1 deletion scan. The arrow indicates the insertion of +1L within
the MSD (B) Infectivity of the full-length +1L insertion is shown as a control and its
insertion position within the membrane-spanning domain is indicated with an arrow. All
filled bars are full-length Env in the +1L background with the indicted residue deleted. The
hashed open bar is full-length wild-type Env shown as a positive control. All data are
normalized relative to the wild-type Env control. The dashed line shown across the data
indicates the relative level of infectivity of the +1L Env. Data shown here are the average
of three independent experiments.
60
Figure 2.2: Infectivity of full-length deletion glycoproteins in the +1L insertion
background.
61
Figure 2.3: Leucine insertions upstream and downstream of the respective glycine-
proline pairs do not display a phasing effect.
(A) (Top) Diagram of the leucine insertions upstream of the glycine-proline pair positioned
at the N-terminus of the membrane-spanning domain and downstream of the glycine-
proline pair positioned at the C-terminus of the membrane-spanning domain. Leucine
insertions in the middle of the MSD, between the glycine-proline pairs, are shown as a
control. (Bottom) Infectivity of leucine insertion mutants shown relative to wild-type MLV
Env. Data shown in this figure are the average of three independent experiments.
62
Figure 2.3: Leucine insertions upstream and downstream of the respective glycine-
proline pairs do not display a phasing effect.
63
Figure 2.4: Truncation of the cytoplasmic tail restores abrogated infectivity of certain
deletion mutants.
The cytoplasmic tail of the specified deletion mutants, in the context of the +1L
glycoprotein background, were truncated and infectivity was assessed. Filled bars indicate
full-length glycoprotein deletion mutants with +1L inserted in the MSD. Open bars indicate
glycoprotein deletion mutants with +1L in the MSD missing 25 residues of the CT. The
hashed open bar is wild-type Env Δ25CT shown as a positive control. The dashed line
between positions P148 and L157 indicates these residues were not tested in this
experiment. All data are normalized relative to the wild-type Env Δ25CT control. Data
shown here are the average of three independent experiments.
64
Figure 2.4: Truncation of the cytoplasmic tail restores abrogated infectivity of
certain deletion mutants.
65
Figure 2.5: The hydroxyl-containing amino acids in the MSD are crucial for
promoting cell-to-cell fusion.
(A) Stable cell lines were created in 293FT cells expressing constructs for the indicated
Env mutants. Cells were surface labeled using an anti-GFP AlexaFluor 647 antibody
(Sigma) and analyzed via flow cytometry. Mean fluorescence intensity was normalized to
the +1L Δ25CT control. 293FT cells not expressing MLV Env were included as a negative
control in the labeling process and are indicated as Env (-). (B) Stable cell lines were
assayed for cell-to-cell fusogenicity. Stable cells lines were transfected with tet-off
expression plasmid and co-cultured with a permissive tre-G.Luc. cell line. Results are
depicted as relative light units normalized to the +1L Δ25CT control. 293FT cells not
expressing MLV Env were transfected with the tet-off expression plasmid and were
included in the cell-to-cell fusion assay as a negative control (Env (-)). (C) Western blot
analysis of Env incorporation into viral particles. Stable cell lines were transfected with
HIV GagPol and supernatants and cell lysates were analyzed for Env incorporation and
cellular expression, respectively.
66
Figure 2.5: The hydroxyl-containing amino acids in the MSD are crucial for
promoting cell-to-cell fusion.
67
Figure 2.6: Hydroxyl substitutions display defective functionality.
(A) Infectivity of alanine substitutions in the context of full-length (filled bars) and
truncated Env (open bars) relative to wild-type. (B) Infectivity of amino acid substitutions
in the context of full-length Env relative to wild-type. The hydroxyl-containing positions
were substituted to either: N, asparagine; Q, glutamine; Y, tyrosine; A, alanine; or F,
phenylalanine and infectivity was assessed. Asterisks indicate a significant difference
between the infectivity of tyrosine and phenylalanine substitutions at all four positions
tested.
68
Figure 2.6: Amino acid substitutions in the critical hydroxyl positions display
defective functionality.
69
3. IN VIVO SELECTIONS REVEAL NOVEL PROTEIN
DETERMINANTS FOR MURINE LEUKEMIA VIRUS
(MLV) ENVELOPE GLYCOPROTEIN PSEUDOTPYING
INTO HIV-1 PARTICLES
70
3.1 Summary
Enveloped viruses utilize membrane-spanning cell surface glycoproteins to
mediate fusion between viral and host cell membranes. In a process termed pseudotyping,
glycoproteins from diverse families exhibit incorporation into non-native viral particles by
means of an unknown mechanism. For example, Murine Leukemia Virus envelope (MLV
Env) glycoprotein actively incorporates into foreign viral particles both in the presence,
and absence, of its C-terminal tail. To identify the region(s) of MLV Env required for
efficient incorporation into assembling particles, we developed an in vivo selection system
coupled with Illumina next generation sequencing (NGS) to generate comprehensive
libraries of mutant glycoproteins which were 1) enriched in viral particles, 2) able to
promote viral infection, and 3) fusogenically active. We created a large cell library
(approximately 4x105 cells) with each cell stably expressing a unique MLV Env viral
amplicon with randomly mutagenized membrane-proximal external region (MPER) and
membrane-spanning domains (MSD). NGS analysis revealed that a large number of
incorporation defective substitutions clustered within two regions: A 26 amino acid region
spanning the MSD into the MPER (P136-I162) and a 3 amino acid region in the MPER
(W127-E129). These substitutions had abrogated infectivity but displayed normal
fusogenic activity, suggesting that these regions are important for MLV Env incorporation.
Direct testing of a panel of mutants selected from these two regions identified W127R,
W137L, and I142R as defective for Env incorporation, all of which correlated with the
NGS analysis. Our data suggest that these two regions are essential for dictating Env
recruitment to assembling viral particles.
71
3.2 Introduction
Retroviral particle assembly requires a symphony of viral and host cell proteins
to coordinate in a spatial and temporal manner to form an infectious particle. Typically,
this is an exclusive process wherein viral components only assemble with viral components
from the same or closely related viruses, such as HIV-1 and -2 (120-124). However, an
exception to this typical exclusivity is the incorporation of viral glycoproteins. This
phenomenon is termed pseudotyping, which is defined as the ability of foreign
glycoproteins to incorporate into non-native viral particles (125-127). Researchers from
numerous fields have capitalized on the ability to manipulate viral tropism by exploiting
glycoprotein pseudotyping; however, very little is known about the molecular mechanisms
that govern this process.
There have been at least four models proposed to describe the pseudotyping
mechanism (67, 77). The first model suggests that glycoprotein incorporation is random
and that all surface expressed proteins would be passively incorporated into assembling
particles barring steric hindrance. In this case, viral glycoproteins would be randomly
distributed at the plasma membrane both in the presence and absence of viral assembly
sites. We recently demonstrated that vesicular stomatitis virus glycoprotein (VSV-G) and
murine leukemia virus envelope glycoprotein (MLV Env) are actively recruited to foreign
viral assembly sites (73, 77). This indicates, at least in the case of these two viral surface
proteins, that passive incorporation is not the mechanism used for pseudotyping.
72
The second model, similar targeting, proposes that all viral structural and surface
glycoproteins are independently targeted to specific microdomains at the plasma
membrane, such as lipid rafts. Lipid rafts have been implicated as assembly sites for some
viruses and could serve as a common platform to facilitate viral pseudotyping (125).
However, a surface glycoprotein commonly used for pseudotyping, VSV-G, is not raft
associated (128). Additionally, clustering of MLV Env and VSV-G observed when
assembling particles are present in the cell does not occur in the absence of assembling
virus, suggesting that similar targeting is not the mechanism used for pseudotyping these
viral proteins.
The third model proposes that viral structural proteins directly contact viral
glycoproteins and subsequently recruit them to viral assembly sites. It is unlikely that
pseudotyped glycoproteins form direct physical contacts during incorporation, particularly
because many glycoproteins can be incorporated into virions in the absence of their C-
terminal domains. However, Incorporation of native glycoproteins likely involves specific
interactions between the cytoplasmic tail (CT) of Env and the viral gag protein. A well
characterized example of this is between the CT of HIV-1 and the N-terminal membrane-
binding matrix (MA) domain of HIV-1 gag (129-132). Not only do the CT of HIV-1 Env
and the MA domain of HIV-1 gag interact, but this interaction seems to be required for
efficient Env incorporation into viral particles (133-135). In the case of MLV Env, it has
been shown that: 1) interactions between the CT and MLV gag are required for polarized
budding (78) and 2) a cleavable portion of the CT has been found in association with viral
cores (136). Additionally, we have shown that the CT of MLV Env contains a protein
component that dictates specific recruitment of MLV Env to MLV assembly sites. When
73
both MLV and HIV particles are expressed in the same cell, full-length MLV Env will
exclusively partition to MLV assembly sites. However, if Env is truncated, it is equally
recruited to both HIV and MLV cores (73). This suggests that the CT of MLV Env
modulates a specific recruitment mechanism which results in exclusive partitioning to
MLV cores. In the absence of the CT, Env undergoes generic pseudotyping into foreign
viral particles.
The final model suggests indirect interactions drive incorporation of viral
glycoproteins into assembling particles through a cellular intermediate. We have
previously shown that active recruitment of MLV Env into assembling viral particles
occurs robustly in the absence of the CT, suggesting that an alternative protein
component(s) specifies this recruitment mechanism (73). It has been demonstrated that the
cholesterol recognition amino acid consensus (CRAC) domain in the membrane-proximal
external region (MPER) of HIV Env has a propensity to interact with cholesterol rich
domains in the plasma membrane (137-141). It has been postulated that this region has a
role in remodeling of the HIV-1 lipid membrane to facilitate fusion (95, 96, 142) and
regulates HIV Env lateral sorting (143). We hypothesized that the MPER, and even
membrane-spanning domain (MSD), of MLV Env may be involved in similar behaviors.
Importantly, these domains may be the protein determinants that facilitate Env recruitment
to viral assembly sites though interactions with lipid microdomains that form during viral
assembly. Here, we coupled large-scale in vivo selections with next generation sequencing
(NGS) of a combinatorial mutagenic library of MLV Env glycoproteins to identify mutants
that could 1) be incorporated into viral particles, 2) promote viral infectivity, and 3) retain
fusogenic activity. In addition to probing novel aspects of Env, we suggest that this
74
selection strategy could be applied to other targets under diverse selection pressures and
may minimize difficulties associated with targets highly sensitive to traditional
mutagenesis.
3.3 Experimental Procedures
Plasmids and Cell Culture. The amino acid substitutions were created using
oligonucleotide-mediated mutagenesis. Constructs expressing the truncated version of
MLV Env were created by introducing a stop codon after the sequence encoding
RLVQFVK, which removes twenty-five residues from the cytoplasmic tail. For HIV
infectivity, a NL4-3 derived HIV-CMV-GFP proviral vector, defective for Vif, Vpr, Vpu,
Nef, and Env, was used (Vineet Kewal-Ramani, National Cancer Institute). This construct
has a CMV immediate-early promoter driving a GFP reporter in place of Nef. Creation of
the tet-off/TRE-Gluc system is described here (102). Briefly, the gene encoding the tTA
(tet-off) protein was cloned into pQCXIP vector downstream of the CMV immediate early
promoter. The TRE-driven Gluc-inducible expression system was created by introduction
of the Gluc gene into the retro-tight-X-hygro retroviral transfer vector.
HEK-293FT (Invitrogen), 293T mCAT-1 (Walter Mothes, Yale University), 293T TVA,
and 293T mCAT-1 cells stably expressing a tet-off Gaussia Luciferase (G. Luc.) promoter
(102) were maintained in DMEM supplemented with 10 % fetal bovine serum, 2 mM
glutamine, 1 mM sodium pyruvate, 10 mM non-essential amino acids and 1 % MEM
vitamins.
75
Infectivity Assay. 293FT cells were transfected in six well plates with 700 ng of HIV-
CMV-GFP, or 500 ng of CMV-MLV-GagPol and 200 ng CMV-GFP-MLV Genome, and
500 ng of Env expression plasmid using 3 µg polyethylenimine (PEI) per microgram of
DNA (105). Media was changed 6 to 12 h post-transfection to remove residual transfection
reagent. Supernatant was collected 24 h following media exchange and frozen at -80ºC for
at least 3 h to lyse cells contained within the supernatant. After thawing supernatants at
37ºC in a water bath, samples were centrifuged at 1,500 X g for 5 minutes to pellet any
cellular debris. 500 µL of the supernatant was added to target cells for 48 hrs. Cells were
collected at 48 hrs, fixed with 4% paraformaldehyde, and analyzed using an Accuri C6
flow cytometer.
For infectivity using stable cell lines expressing the indicate MLV Env, cells were
transfected with 700 ng of HIV-CMV-GFP, or 500 ng of CMV-MLV-GagPol and 200 ng
CMV-GFP-MLV genome. Following transfection, the above procedure was used to assess
infectivity. In the case of the stable cell lines expressing both Env and HIV-GagPol,
supernatant was collected, frozen at -80ºC for at least 3 h, and 500 µL was added to target
cells.
Western Blotting. 293FT cells stably expressing the indicated viral proteins were used for
Western Blotting analysis. Viral samples were pelleted through a 20 % sucrose cushion for
2 h at 20,000 X g at 4ºC. Residual supplement media and sucrose were aspirated off of the
sample pellet and samples were re-suspended in 6x SDS-PAGE loading buffer. The
equivalent of 1 mL of viral supernatant was analyzed by 10 % discontinuous SDS-PAGE.
Cell samples were detached using 10 mM EDTA/PBS solution and pelleted at 500 x g for
10 minutes. Pellets were re-suspended in RIPA buffer (10 mM Tricl-Cl (pH 8.0), 1 mM
76
EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, and 140
mM NaCl) and 5-to-10 % of the lysate was combined with 6x SDS-PAGE loading buffer
and analyzed by 10 % discontinuous SDS-PAGE. Proteins were transferred onto a 0.45 µm
polyvinylidene difluoride (PVDF) membrane. The membrane was blocked with 5 % nonfat
dried milk in PBS containing Tween 20 (PBS-T) and probed with rabbit anti-GFP antibody
diluted 1:5000 (Sigma) and mouse anti-HIV p24 hybridoma media diluted 1:500 (AIDS
Research and Reference Program, Division of AIDS, NIAID, NIH; HIV-1 p24 hybridoma
[183-H12-5C]). Primary antibody incubations were performed overnight, on a shaker, at
4ºC. Blots were washed with PBS-T and then probed with horseradish peroxidase (HRP)-
conjugated anti-rabbit and anti-mouse antibodies diluted 1:10,000 (Sigma) for probing Env
and p24, respectively. Visualization of the membranes was performed using Luminata
Classico and Cresendo Western HRP chemiluminescence reagents. Imaging was
performed using a LAS3000 image analyzer from Fujifilm.
Cell-to-cell Fusion Assay. 293FT cells stably expressing the indicated Env protein were
transfected with 500 ng of tet-off expression plasmid in a six well plate. Media was changed
6 to 12 h post-transfection to remove residual transfection reagent. Transfected cells were
co-cultured with an equal number of 293T mCAT-1 TRE Gluc cells for 48 h. Twenty
microliters of sample supernatant from the co-cultured cells were assayed in duplicate for
Gluc. content with 50 µL of 10 µM coelenterazine in 0.1M Tris (pH 7.4) and 0.3M sodium
ascorbate.
Surface Labeling. 293FT cells stably expressing the indicated Env protein were detached
using 10 mM EDTA/PBS. Cells were centrifuged at 500 X g for 10 minutes at 4ºC and re-
suspended in 1 % bovine serum albumin (BSA)/PBS blocking solution for 20 minutes.
77
After 20 minutes of blocking, cells were centrifuged at 500 X g for 10 minutes at 4ºC and
re-suspended in 10 mM EDTA/PBS, 1 % goat serum, and 1:1000 primary anti-GFP Alexa-
Fluor 647 antibody (Life Technologies) for 1 h. After primary antibody incubation, cells
were centrifuged at 500 X g for 10 minutes at 4ºC, re-suspended in PBS, and analyzed
using an Accuri C6 flow cytometer.
RNA harvesting and cDNA synthesis. Cells grown on 10 cm culture dishes were
harvested from each of the four libraries using 10 mM EDTA/PBS. Cells were pelleted at
500 x g for 10 minutes at 4ºC and re-suspended in Trizol reagent (Sigma). Following a 5
minute incubation period, 200 µL chloroform was added per 1 mL Trizol and vortexed.
After a 3 minute incubation period at room temperature, samples were centrifuged for 15
minutes at 12,000 x g at 4ºC. The upper phase was then transferred to a new
microcentrifuge tube and combined with 500 µL isopropanol per 1 mL Trizol overnight at
4ºC. Samples were then centrifuged 15 minutes at 12,000 x g at 4ºC. RNA pellets were
washed one time using 1 mL of 75% ethanol for every 1 mL Trizol. Following a 20 minute
ethanol incubation, samples were centrifuged for 5 minutes at 7,500 x g at 4ºC. Pellets were
air dried and re-suspended in RNA/DNAase free water.
For cDNA synthesis, 2 µg of the purified RNA was mixed with 1 µL of 100 µM poly dT
primers and incubated at 70ºC for 5 minutes, chilled on ice, and combined with 4 µL
reverse transcription reaction buffer (Promega), 1 µL of 10 mM dNTP mix, 2 µL of 25 mM
MgCl2, 0.5 µL of 40 U/µL RNase inhibitor (Promega), and 1 µL of 200 U/µL M-MLV
reverse transcriptase. The reaction was incubated at 42ºC for 2 h and then 85ºC for 5
minutes.
78
Illumina sequencing and analysis. cDNA libraries were diluted 1:2 with RN/DNase free
water and 5 µL of the diluted samples were used as template for PCR amplification to
introduce Illumina adapters and sequencing indices for multiplexing. Samples were
analyzed using Illumina MiSeq (University of Missouri DNA Core Facility) 2 x 150 nt
paired-end reads. Reads were paired using FLASh (http://ccb.jhu.edu/software/FLASH/).
Data processing was performed using a locally installed FASTX-Toolkit
(http://hannonlab.cshl.edu/fastx_toolkit/). Fastx clipper was used to trim the 3’ constant
region from sequences and a stand-alone script was used to trim 5’ constant regions. During
this process, sequences shorter than 150 nucleotides, or non-trimmed sequences, were
discarded. Trimmed sequences were then filtered for high-quality reads using FASTQ
quality filter. Sequences with a Phred quality score less than 30 (99.9% base calling
accuracy) at any position were eliminated. Finally, nucleotide sequences were translated to
amino acid sequences using the TranSeq program in a locally installed EMBOSS package
(www.ebi.ac.uk/Tools/st/emboss_transeq).
Pre-processed sequences were then further analyzed using the FASTAptamer toolkit (144).
FASTAptamer-Count was used to count the number of times each sequence was sampled
from the population. Each sequence was then ranked and sorted based on overall
abundance, normalized to the total number of reads in each population, and directed into
FASTAptamer-Enrich. FASTAptamer-Enrich calculates the fold-enrichment ratio from a
starting population to a selected population by using the normalized reads-per-million
value for each sequence. After generating the enrichment file, nucleotide changes in each
sequence were determine by using FASTAptamer-Distance, which identifies how many
mutations each sequence has relative to a WT input sequence. Sequences with less than 5
79
RPM in the starting population, or with greater than 6 nt mutations were filter removed
from the population. Only sequences with a single non-synonymous amino acid
substitution with up to 6 nt substitutions were considered for the final analysis. The fitness
index for each substitution was determine by taking the average enrichment or depletion
value of all sequences with that specific mutation and plotting it as a heat map (Figure 2).
Maximal enrichment was set to that of WT for the specific selection and was colored gold.
Maximal depletion was set to the average fitness index of all of the randomly introduced
stop codons that occurred in the MPER or MSD for the specific selection and was set to
blue.
3.4 Results
Creating a stable cell library expressing mutant MLV Envelopes. To determine
if the MPER or MSD of MLV Env contribute to selective recruitment of MLV Env to viral
assembly sites, we generated a highly dense mutagenic library targeting 50 contiguous
amino acid positions through random mutagenesis (Figure 3.1A). The goal was to disrupt,
and subsequently identify, motifs or amino acids necessary for Env incorporation. We
achieved random mutagenesis by synthesizing and assembling oligonucleotide primers that
contained a positional mutation rate of 3%, correlating to approximately 4.5 nucleotide
changes per assembled amplicon. Double stranded DNA templates of the randomized
region were generated by primer extension and the sequence was subcloned in the retroviral
amplicon illustrated in Fig. 3.1B using unique restriction sites that flank the randomization
region. The amplicon was then transformed into E. coli and between 105-106 colonies were
harvested. This amplicon was engineered with a puromycin selection cassette, complete
80
RNA packaging components, and flanking LTRs. Therefore, this amplicon can be used to
generate stable cells lines through retroviral transduction. Transduced cells not only
produce the mutant glycoprotein, but also the packageable RNA encoding that
glycoprotein. The amplicons isolated from the pooled library were transfected into 293FT
cells along with plasmids expressing viral packaging components (MLV Gag/GagPol) and
the promiscuous viral glycoprotein VSV-G. Media from this transfection was used to
transduce fresh 293FT cells at a multiplicity of infection (MOI) <0.1. Because a low MOI
was used, 99 % of transduced cells would be infected only once, and therefore express only
a single mutant MLV Env. Following transduction, the cells were selected using
puromycin to create the cell library used in subsequent experiments described below.
It is important to note that the amplicon was constructed with a version of MLV
Env with a truncated CTD for two reasons: 1) we have previously shown that the
cytoplasmic tail of MLV Env contains a protein component that dictates exclusive
recruitment of MLV Env to MLV assembly sites (73) thus, removal of this sequence
ensures that we are only studying the pseudotyping mechanism, and 2) removal of these
residues results in a constitutively active glycoprotein population due to the loss of the
negative regulation provided by the R-peptide (89, 145, 146), allowing for study of
fusogenic activity.
Mutant library selection and next generation illumina sequencing. We coupled
next generation illumina sequencing to large scale biological selections to identify mutant
glycoproteins defective for incorporation into viral particles. Our library was subjected to
three independent selection procedures (described in each section) where the amplicon
sequence was maintained only if the mutant MLV Env that it encoded met the particular
81
criteria. Following these biological selections, the sequence containing the randomized
sequence were recovered and sequenced by Illumina MiSeq analysis using paired-end 2 x
150 read parameters. The population of sequences present following selection was
compared with that of sequences prior to selection to determine the fitness of each mutant
glycoproteins.
Fitness was determined as follows. First, mutations with a low number of input
sequence counts (< 5 reads per million) in the pre-selection library were discarded. Many
of these rare sequences were likely the result of sequencing errors and could skew results.
Second, sequences that contains greater than 6 nucleotide mutations per sequence were
also removed. Amplicons with numerous mutations tended to be quite unfit and were less
informative than amplicons with few mutations. We calculated a fitness index for each
non-synonymous point mutation based on the average of the enrichment or depletion
values that occurred for all of the sequences with that specific substitution. To avoid
complications from compensatory mutations, which may occur in sequences that contained
more than one non-synonymous amino acid substitution, we focused on sequences with
only one altered amino acid. In all cases, there were multiple sequences in the library
representing each individual non-synonymous change. For instance, the starting cell
library contained 108 different amplicons that produced a protein with no amino acid
changes (i.e. only silent mutations). For each unique sequence, the ratio of sequence
density in reads per million between the starting library and the selected library was
calculated. The ratios of all of the unique sequences that led to each single amino acid
change was then averaged. Maximal enrichment values for each sequence were adjusted
to the wild-type Env enrichment for the specified selection. In output illustrations (Fig 3.2
82
and 3.3) sensitivity values were colored so that substitutions detrimental to the biological
selection were colored blue, and substitutions that had no effect on the biological selection
were colored gold. Shown above each of the selections is the maximal sensitivity value for
each position and the location of randomly introduced stop codons (Figure 3.2A-C).
Functional selection. The first and most rigorous selection was for fully functional
MLV Env. For this selection the cell library was transfected with a plasmid expressing the
viral packaging components, but not a viral glycoprotein. In this situation, infectious
particles were produced only if the glycoprotein encoded in the amplicon was able to
perform this function. Media from this transfection were collected and used to transduce
293T mCAT cells, which express the receptor for MLV Env. Following transduction and
selection, the mutant region was recovered by PCR and deep sequenced. In this selection
scheme the wildtype sequences had an average enrichment of ~3.0, whereas sequences that
introduced a stop codon prior to the MSD had enrichment of 0.06. The fact that
introduction of stop codons resulted in dramatic and consistent loss of sequence following
selection criteria provided validation that the selection procedure was effective.
This selection was the most stringent of the three we examined. Every single
residue displayed at least one change that noticeably reduced functionality. The least
sensitive sequence was the last 5 amino acids of the protein. As one would expect,
introduction of a charged residue (R, K, E or D) anywhere in the membrane spanning
domain completely impaired function. The MPER was less sensitive to change than the
MSD, but there were clear hot spots that were very sensitive to change.
Fusion Selection. To identify mutants that retained fusogenic activity, the starting
library was co-cultured with mCAT cells that expressed resistance to hygromycin. Because
83
the library of Env mutants was constructed with a truncated CT, the R-peptide was no
longer able to negatively regulate fusogenicity, which resulted in a constitutively fusogenic
glycoprotein population. Therefore, any cells expressing mutant MLV Env that could
promote cell-to-cell fusion would form syncytia with the mCAT cells and express
resistance to both puromycin and hygromycin. After co-cultured cells were dually treated
with both puromycin and hygromycin antibiotics, the mutagenized region was recovered
using PCR and deep sequenced. The wildtype sequences had an average enrichment of
~1.2, while sequences with a stop codon preceding the MSD had an enrichment value of
0.75. Again, in this instance, introduction of a stop codon resulted in depletion of these
sequences from the selected population, confirming that the selection was successful.
In general, many of the substitutions in the selected fusion population had a
minimal impact on fusogenic activity. However, some residues in the MSD displayed
greatly reduced fusogenic activity. Interestingly, the substitutions that showed the greatest
decrease in fusogenicity were the hydroxyl-containing residues and the proline near the N-
terminus of the MSD.
Incorporation Selection. To select for mutant glycoproteins that were
incorporated into viral particles, the starting library was transfected with expression
constructs for MLV packaging components and VSV-G. Our approach was to utilize a viral
ELISA (147) to identify MLV Env mutants that were incorporated into viral particles.
Briefly, an ELISA plate adsorbed with anti-GFP antibody was used to immobilize virus
that had incorporated GFP-tagged MLV Env mutants. Importantly, viral particles were also
pseudotyped with VSV-G, allowing for the transduction of 293FT cells that were overlaid
on the immobilized virus. This approach generated a selectable cell population that could
84
be used for the extraction of the mutant region, which was subsequently amplified using
PCR and deep sequenced. Wildtype sequence had an enrichment of ~2.0, while sequences
with a stop codon preceding the MSD had an enrichment value of 0.2. Consistent with the
previous selections, introduction of stop codons in this region resulted in sequence
depletion in the selected population, validating the selection.
Analysis of the selected incorporation population revealed that mutations resulting
in the greatest reduction of Env incorporation clustered within the MSD and in two
segments in the MPER from W127-E129 and R134-W137.
NGS analyses identify regions in Env important for incorporation. Our
selection strategy allowed for correlation of mutant fitness across multiple biological
selections. The most informative mutants would retain fusogenic activity (suggesting the
mutation was not detrimental to overall protein folding, surface expression, or
processivity), abrogate infectivity (suggesting the mutant was not incorporated into viral
particles), and exhibit decreased incorporation into viral particles. We hypothesized that
by filter removing mutant sequences from our analysis that did not fulfill these criteria, we
would reliably identify residues in Env important for incorporation. We focused on
substitutions with the lowest functional fitness indices, and then overlaid the fitness indices
from the incorporation and fusion selections at those positions, respectively (Figure 3.3).
This identified two regions where substitutions that fulfilled all three criteria clustered: 3
amino acids from W127-E129 and 46 amino acids from P136-I162. Importantly, these
results correlated well with the results for the maximal sensitivity value from the
incorporation selection (Figure 3.3, top).
85
Characterizing mutants that displayed striking phenotypes based on NGS
analysis. We selected 14 substitutions that covered a range of performances based on the
NGS analysis in Figure 3. Our aim was to identify substitutions from both regions (W127-
E129 and P136-I162) that were incorporation defective, but to also validate that this
method of data analysis remained accurate. Initially, point mutants were transiently
expressed in 293FT cells and assayed for their ability to promote viral infectivity. We have
previously reported that the length of the CT can impact infectivity and influence Env
incorporation into MLV particles (73). With this in mind, we wanted to assess the ability
of the mutants to promote viral infectivity in both the presence and absence (Δ25 CT) of
the full-length cytoplasmic tail (Figure 3.4). Addition of the full-length cytoplasmic tail
had a minimal impact on infectivity for all of the candidates except for T139A.
Surprisingly, the presence of the full-length cytoplasmic tail resulted in a dramatic
restoration of infectivity for this particular point mutant (Figure 3.4).
In general, these single-round infectivity data largely recapitulated the results from
the functional biological selection; however, two of the point mutants, W127R and I142R,
displayed an opposite phenotype when compared to the functional NGS analysis (Figure
3.2 and Figure 3.3). Based on our analysis from Figure 3.3, we reasoned that perhaps these
point mutants were defective for incorporation yet retained fusogenic activity, but because
of high expression levels at the cell surface, they were non-specifically incorporated into
viral particles. To test this hypothesis, we generated 293FT cell lines that stably expressed
the indicated point mutation and assayed their ability to promote viral infectivity (Figure
3.5A). As we expected, W127R and I142R displayed abrogated infectivity, suggesting that
these mutants were likely defective for their ability to incorporate into viral particles.
86
As discussed previously, mutants that disassociate incorporation from fusogenic
activity would display abrogated infectivity. Because we identified several mutants from
our NGS analyses that displayed decreased infectivity, we tested whether any of these
mutants still retained fusogenic activity. Since these mutants have a truncated cytoplasmic
tail, they remain constitutively active and could be tested for fusogenic activity. Briefly,
each cell line stably expressing the indicated Env mutant was transfected with a tet-off
expression construct and was co-cultured with an equal number of 293T mCAT-1 cells
expressing TRE-driven Gaussia luciferase (Gluc) (102). If the Env mutants retain
fusogenic activity, the transfected cells fuse with the receptor-expressing cells induction of
Gluc. occurs in a tet-off dependent manner. Gluc. induction can then be correlated between
mutant Env and wild-type Δ25CT controls to assess fusogenic activity (Figure 3.5B). It is
important to note that the fusogenic activity of these mutants generally reproduced the
trends found in the NGS analysis for the fusion selection (Figure 3.2B). Interestingly, we
identified three point mutations, W127R, W137L, and I142R, that displayed fusogenic
activity but had abrogated infectivity, suggesting that these mutants might be defective for
incorporation into viral particles. To ensure all mutants were similarly on the cell surface
compared to wild-type Δ25CT, stable cell lines were surface labelled and analyzed using
flow cytometry. As depicted in Figure 3.5C, there were no significant differences observed
between the surface expressions of the mutants compared to controls.
Identification of mutants defective for incorporation into viral particles. Next,
we wanted to test whether any of our candidate mutants were defective for incorporation
into viral particles. However, Figures 3.4 and 3.5 show that viral protein expression levels
have a significant effect on mutant Env phenotypes. Therefore, we stably expressed HIV
87
GagPol in the cell lines that also stably expressed the Env mutants. To ensure that the MLV
Env mutants behaved similarly when using HIV cores compared to MLV cores, we
transfected the stable Env cell lines with 500 ng of either HIV or MLV GagPol expression
constructs (Supp. Figure 3.S1A). The results show no discernable difference in infectivity
between HIV and MLV cores. Thus, creating a stable cell line expressing both Env and
HIV GagPol can reproduce previously observed phenotypes. To confirm that cells stably
expressing both Env and HIV GagPol were producing viral particles, media was collected
and used to transduce 293T mCAT-1 cells (Figure 3.S1B). The infectivity results shown in
Figure 3.S1B closely resembled previous infectivity data shown in Figures 3.5A and Figure
S.1A.
After verifying that these cells were producing infectious particles, we collected
viral supernatants and cell lysates to assess mutant Env incorporation into viral particles
via western blotting. In the case of W127R, W137L, and I142R there was a significant
reduction in the amount of gp70 present in viral particles (Figure 3.6A, Virus). MLV Env
is produced as a gp85 precursor protein and is processed by a cellular protease into gp70
(SU) and p15E (TM) (148). Analysis of cell lysates revealed that the reduction of gp70 in
viral pellets for W127R, W137L, and I142R was not due to a reduction of processed gp70
(Figure 3.6A, Cell). To verify that these mutants had a reproducible reduction of
incorporation into viral particles, several independent Western Blots were quantified and
the results are shown in Figure 6B. These data indicate that all three mutants reproducibly
display severely reduced incorporation into viral particles.
.
88
3.5 Conclusions
We have previously shown that MLV Env is actively recruited to HIV-1 assembly
sites (77) and that the CTD is dispensable for this mechanism (73). Additionally, further
studies to probe the hydrophobicity of the MSD as a protein component necessary for
active recruitment have shown no correlation between hydrophobic index and
incorporation (102). Attempts to characterize the domains in gag necessary to facilitate
active incorporation have also been obscure. Interestingly, the MA domain of gag is not
required for active Env recruitment to assembly sites (74). Studies to identify host-cellular
factors that influence glycoprotein recruitment to viral assembly sites have indicated that
membrane composition is a crucial determinant for infectivity and fusogenicity (96, 137-
140, 143). For HIV-1 Env, these studies indicate that the MPER retains critical protein
determinants necessary for Env fusogenicity. Here, we evaluated the role of MPER and
MSD of MLV Env for facilitating pseudotyping into HIV-1 particles.
Identification of residues in the MSD crucial for infectivity and fusogenicity.
We provide evidence that the MSD of MLV Env contributes to fusogenicity. Introducing
specific amino acid substitutions at positions T139, T140, S143, and P148 reduced both
infectivity (over 95% decrease) and fusogenicity (70-98% decrease) (Figures 3.3 and 3.4).
Position T144 had a similar effect, displaying decreased infectivity (90% decrease) and
moderately decreased fusogenicity (40% decrease). In the case of P148, it is possible that
this residue, in combination with glycine at position 147, forms a molecular hinge that
allows for a conformational change within the TM stalk during activation of the fusion
mechanism (97, 110, 115, 116, 149). Recent cryo-electron microscopy imaging data
89
indicates that the CTD of MLV Env holds the ectodomain in a tight conformation;
however, cleavage of the R-peptide results in a conformational rearrangement in TM
wherein the trimer helices are splayed apart, allowing for fusogenic activation of Env (90).
It is possible that the glycine-proline pair (residues 147 and 148) at the N-terminus of the
MSD facilitates the molecular reordering of TM during this process. The hydroxyl-
containing residues located in the same region (T139, T140, S143, and T144) may also
contribute to this mechanism by creating a hydrogen-bonding network that potentially
stabilizes the TM trimer (109-111, 149-151). It is possible that these residues form
hydrogen bonds inter-helically within the trimer, resulting in stabilization of the TM
domain. This stabilization may allow for the conformational change necessary for
transmitting the signal of R-peptide cleavage up the TM stalk. It could also be that these
hydrogen-bonds aide in orientation of the monomers within the trimer, establishing a
helical interface (112, 117, 152, 153). Importantly, mean surface fluorescence and protein
blotting analysis indicate that processivity and trafficking of these mutants are not
impaired, supporting the idea that the loss of infectivity is due to the loss of the fusogenicity
and not due to an overall gross defect in protein folding (Figures 3.4 and 3.5).
Identification of residues in the MPER and MSD that affect Env pseudotyping.
Substitutions W127R, W137L, and I142R were defective for infectivity but remained
fusogenically active (Figure 3.5). This suggested that these mutations would display
reduced incorporation into viral particles, which would be consistent with the analyses in
Figure 3.3. Protein blotting analysis showed that mutants W127R, W137L, and I142R had
significantly decreased gp70 incorporation into viral particles (Figure 3.5A and 3.5B), but
accumulated to wild-type levels in cell lysates (Figure 3.5, lysate). As shown in Figures
90
1A and 3, a majority of the MPER in our target region is predicted to be unstructured,
except for a short helical stretch from positions F128-L131. Perhaps this region folds
towards the membrane and interacts with specific lipids that cluster at the viral assembly
site once Gag begins to multimerizes. If this were the case, this region, along with the
region from P136-I162, could act in concert to direct Env trimers to the assembly site
through interactions with the lipid environment. Both of these regions are primarily
hydrophobic, and two of the residues that we directly tested and demonstrated as
incorporation defective were tryptophan residues. Studies performed on the MPER of HIV-
1 Env, show that the tryptophan residues in this region play a crucial role in Env
functionality through direct interactions with cholesterol molecules that cluster at the
assembly site. Further, it has been shown that a molecular hinge region in the MPER of
HIV-1 Env consists of a pair of molecular helices that immerse themselves in the lipid
environment (93). Therefore, it is reasonable to suggest that the tryptophan residues that
we identified could engage in similar molecular interactions. Taken together, our data
suggest that the MSD has a role in MLV Env fusogenicity, and that two regions from
positions W127-E129 and P136-I162 are important for Env incorporation into assembling
viral particles.
Our selection strategy can be applied to other targets under diverse selection
pressures. It will be particularly valuable in minimizing difficulties associated with
targets highly sensitive to traditional mutagenesis, such as the viral capsid protein.
Coupling next-generation sequencing technologies with in vivo selection allows
generating large data sets capable of sampling entire selected populations, providing
insight into previously unidentified features of viral protein biology.
91
Figure 3.1: Target region of mutagenesis and selection strategy.
(A) Diagram of the region in MLV Env targeted by random mutagenesis. The 50 amino
acid cassette depicted above was randomly mutagenized at a rate of three percent,
correlating to an average of 4.5 nucleotide mutations per individual Env mutant. The
sequence upstream of the target cassette was maintained as wild-type sequence, while the
cytoplasmic-tail has a truncation removing 25 amino acids if the cytoplasmic tail. (B)
Strategy used for creating the stable mutant cell population and in vivo selection
procedures.
92
Figure 3.1: Target region of mutagenesis and selection strategy.
93
Figure 3.2: NGS selection analysis.
Blue indicates a non-synonymous substitution at the indicated position was detrimental to
the specific selection while gold indicates the substitution had no effect. The panel
indicated as stop depicts the location of randomly introduced stop codons for each of the
libraries. Maximal sensitivity displays the lowest fitness index for each position within the
indicated library. (A) Analyzed NGS results from the large-scale infectivity selection. (B)
Analyzed NGS results from the large-scale cell-to-cell fusion selection. (C) Analyzed NGS
results from the large-scale incorporation selection.
94
Figure 3.2: NGS selection analysis.
95
Figure 3.3: Two regions in the MPER and MSD are predicted to be important for
Env incorporation.
Depicted are substitutions that were defective for incorporation and function but not
fusion. NGS results were first filtered for the substitutions that were most detrimental to
infectivity. The incorporation and fusion fitness indices were then plotted for each of
those functionally defective substitutions. The incorporation fitness index is shown fist
and the fusion index is shown second for each of the substitutions above. The dashed line
represents the locations with the largest concentration of substitutions that resulted in
fulfillment of the aforementioned criteria. The predicted secondary structure is shown
above the heat map with the maximal fitness values from the incorporation selection
shown above that. Blue indicates that a substitution was detrimental to the selection while
gold indicates no effect
96
Figure 3.3: Two regions in the MPER and MSD are predicted to be important for
Env incorporation.
97
Figure 3.4: Effect of C-terminal tail on infectivity of 14 selected mutants.
Infectivity results from 293FT cells transfected with 500 ng of the indicated Env mutant
and 500 ng of MLV GagPol expression plasmids. Filled bars represent Envs with the full-
length tail and open bars indicated Envs with a truncated cytoplasmic tail. Infectivity is
shown relative to the appropriate wild-type (WT) control. Data is a composite from three
independent experiments.
98
Figure 3.4: Addition of the full-length cytoplasmic tail has a minimal effect on
mutant env functionality.
99
Figure 3.5: Several candidate mutants identified in the selection screen display
abrogated infectivity but not fusogenicity or cell surface expression.
(A) 293FT cell lines stably expressing the indicted point mutant were transfected with
MLV GagPol and a MLV-CMV-GFP genome and assayed for infectivity. Infectivity is
shown relative to WT Δ25CT Env. (B) 293FT cell lines stably expressing the indicated
point mutant were assayed for cell-to-cell fusion activity. Cell lines were transfected with
a tet-off expression plasmid and co-cultured with a permissive cell line expressing a
Gaussia luciferase reporter. Luminescence output is depicted as percent relative to WT
Δ25CT Env. 293FT cells not expressing MLV Env were transfected with the tet-off
expression plasmid and were included in the cell-to-cell fusion assay as a negative control
(Env (-)). (C) Surface expression of Env point mutants. Cells stably expressing the
indicated Env mutant were surface labeled using an anti-GFP Alexa-Fluor 647 antibody
and analyzed via flow cytometry. Mean fluorescent intensify is shown relative to WT
Δ25CT Env. 293FT cells not expressing MLV Env were included as a negative control in
the labeling process and are indicated as Env (-). Data shown for all three figure panels are
the average of three independent experimental replicates.
100
Figure 3.5: Several candidate mutants identified in the selection screen display
abrogated infectivity but not fusogenicity or cell surface expression.
101
Figure 3.6: Specific point mutations in, or near, the membrane-proximal domain
display severely abrogated incorporation into HIV particles.
(A) (Top) Immnoblot analysis of HIV particles pseudotyped with the indicated MLV Env
point mutant. Particles were collected from supernatants of 293FT cells stably expressing
HIV GagPol and the indicated Env mutant. (Bottom) Immunoblot analysis of cell lysates
stably expressing both HIV GagPol and the indicated MLV Env point mutant. (B)
Immunoblot quantification of four independent replicates of Env incorporation into HIV
particles. Particles were isolated from supernatants of cells stably expressing HIV GagPol
and the indicated Env mutant. The gp70 signal was normalized to the corresponding capsid
p24 band and are shown as the average of all four experiments relative to the WT Δ25CT
normalized signal. Data shown here are representative of at least three independent
experiments
102
Figure 3.6: Specific point mutations in, or near, the membrane-proximal domain
display severely abrogated incorporation into HIV particles.
103
Figure 3.S1: Mutant Env infectivity is similar between MLV and HIV cores.
(A) Stable cell lines expressing the indicated Env mutant were transfected with either
MLV (Open Bars) or HIV (Grey Bars) cores and infectivity was assessed. Infectivity is
shown relative to WT Δ25CT controls of the appropriate core. (B) Infectivity output from
cells stably expressing the indicated Env mutant and HIV GagPol. Infectivity shown
relative to WT Δ25CT controls. Data shown in both figure panels are the average of three
independent replicates.
104
Figure 3.S1: Mutant Env infectivity is similar between MLV and HIV cores.
105
4. PROBING THE MEMBRANE PROXIMAL EXTERNAL
REGION AND MEMBRANE-SPANNING DOMAINS OF
HIV-1 ENVELOPE FOR A ROLE IN GLYCOPROTEIN
INCORPORATION AND FUNCTION
106
4.1 Summary
Enveloped viruses utilize transmembrane surface glycoproteins to mediate fusion
between viral and host cell membranes. However, the acquisition of these glycoproteins
into assembling viral particles is not well understood. Molecular studies on the MPER and
MSD have partially defined the roles of these regions in infectivity and fusogenicity, but
are deficient regarding protein determinants necessary for incorporation. Here, we created
a cell library (approximately 6x104 unique mutants) stably expressing a replication
competent NL4-3 derived proviral construct with mutagenized HIV-1 Env. The MPER and
MSD of HIV-1 Env were targeted for random mutagenesis at a rate of 3%, which correlates
to approximately 3 nucleotide mutations per glycoprotein. The cell library will be
independently selected for mutant Envs that are able to promote infectivity and are
incorporated into viral particles. These selected populations will be subjected to NGS
analysis and compared to the starting population to determine which mutants were not
capable of fulfilling selection.
107
4.2 Body
For HIV-1 Env, it has been demonstrated that a molecular hinge region in the
MPER, consisting of a pair of structurally conserved helices that immerse themselves in
the lipid environment, play a crucial role in fusogenicity (93). Disruption of this region
through mutational analysis, or through application of neutralizing antibodies, which are
predicted to disrupt the hinge region’s ability to interact with the lipid environment,
abrogates fusogenicity (93, 154, 155). The MPER also contains a tryptophan-rich region
that has been shown to interact with cholesterol molecules (137, 139, 142). Deletions,
insertions, or substitutions within this tryptophan rich region abrogated the ability of HIV-
1 Env to facilitate membrane fusion, but maintained normal processivity and trafficking
(95). It has been proposed that the tryptophan-rich region in HIV-1 Env destabilized the
HIV-1 envelope allowing promotion of membrane fusion and it has been shown that
depletion of cholesterol in target cell membranes, in specific cell types, reduced their
susceptibility to membrane fusion by HIV-1 Env (96, 138, 156). It has further been
postulated that the tryptophan-rich region induces formation of these cholesterol-
containing lipid environments (140, 141, 157). The MSD has also been implicated in
promoting fusogenicity. Substitutions in the core of the MSD, or chimeric mutants that
exchange the HIV-1 Env MSD with another viral glycoproteins MSD, result in a
fusogenically inactive Env (91, 92, 158). Interestingly, none of these studies implicates the
MPER or MSD of HIV-1 Env as being important for glycoprotein incorporation into
assembling particles. One study demonstrated that the tryptophan-rich region of HIV-1 Env
is important for lateral sorting and oligomerization of HIV-1 Env (143).
108
Here, we randomly mutagenized the MPER and MSD of HIV-1 Env to identify any
amino acids or motifs necessary for Env incorporation into viral particles, as well as to
further define their roles in promoting infectivity. We targeted 36 residues in this region
for random mutagenesis (Figure 4.1A). To accomplish this, we assembled oligonucleotide
primers with a mutation rate of 3%, which correlates to approximately 3 nucleotide
mutations per amplicon. These amplicons were cloned into a NL4-3 derived proviral
plasmid defective for Vif, Vpr, and Nef. In the place of Nef, a blasticidin resistance gene
driven by a CMV promoter was introduced. Our goal was to create a large cell library
stably expressing these proviral constructs with mutagenized HIV-1 Env. We would then
collect viral particles released into the supernatant and test them for mutants that could still
promote infectivity, and determine which mutants were still incorporated into viral
particles. After selecting these libraries, we would submit the samples for next generation
sequencing to identify mutants that could, or could not, fulfill selection (Figure 4.1B).
The advantage of this approach is that the NL4-3 derived provirus is replication
competent, therefore, the first round of mutant Envs that were selected for infectivity can
be continually passaged to determine how the library changes over time. Additionally,
different target cell types can be used to determine whether the selected population changes
due to cell type specificity. This can be done for the initial infectivity selection, or, become
part of the study for passaging the library over time.
For this study, the mutagenized HIV-1 Env assembled amplicons were cloned into
the NL4-3 derived proviral plasmid and subjected to 2 x 150 nt paired-end Illumina MiSeq
analysis. The resulting NGS analysis was filtered based on sequence read representation in
the plasmid population (< 3 reads per million were filter removed from the population),
109
and the substitutions were cataloged based on position within the target region (Figure 4.2).
293FT cells were transfected with the mutagenized NL4-3 derived plasmid library along
with MLV Env and viral supernatant was collected. These viral particles were used to
infect 293T mCAT-1 cells, which are permissive to viral particles that contain MLV Env.
Once the cells had been infected with the pseudotyped lentiviral library, the cells were
treated with blasticidin. Any cells not infected with the viral particles are then removed
from the population, resulting in a pure cell population expressing the NL4-3 derived
genome. The following future directions section will outline the experiments that will be
performed to utilize the mutagenic cell library to accomplish the goals of this study.
110
4.3 Future Directions
1.) Identify mutants that survive a single round of infectivity. This selection is
simple. Because the proviral construct is replication competent, the stable cell
library will continuously produce virus that can be collected and used to infect
permissive cell lines. A limitation of the cell type selection is that the provirus is
defective for Vif and Nef, so cell type selection will play a role because of host cell
restriction factors. To ensure a single round of infection, a reverse transcription
inhibitor would be used to inhibit spread of the selected virus. Once these have
been addressed, the cells can be selected with blasticidin and prepared for NGS
analysis.
2.) Identify mutants that are still incorporated into viral particles. The stable cell
library would be transfected with Rous Sarcoma Virus (RSV) Env ΔCTD. This
glycoprotein would co-package with the HIV-1 Env mutants into the viral
particles. The viral supernatant would then be used in a viral ELISA based assay
to determine which mutant HIV-1 Envs are still incorporated into viral particles.
Briefly, anti-HIV antibodies would be adsorbed to an ELISA plate so that any
virions that contain a mutant HIV-1 Env would be immobilized. Because RSV Env
ΔCTD pseudotypes into HIV particles, it would be present in the virion as well.
Therefore, a RSV Env permissive cell line can be overlaid onto the virus and the
cells become infected. After selection, the cells can be prepared for NGS analysis
and mutant HIV-1 Envs that could, and importantly, could not fulfill selection can
be identified.
111
3.) Passaging the library on permissive cell lines. This experiment will collect virus
produced by the primary cell library and use it to infect a permissive cell line. Once
that line is infected and selected with blasticidin to select a pure population and
inhibit super infection, it will begin to produce replication competent virus. The
viral particles produced by the primary selected population can then be used to
infect another round of permissive cells. Repeating this cycle will assess how the
functional mutant population changes over time. Each round, as well as the starting
population, can be deep sequenced and analyzed to identify global, and local,
changes in the population.
4.) Identify mutants that can and cannot fulfill selection. Once the library has been
selected for both initial and passaged function and incorporation, the NGS results
will be analyzed to identify interesting residues or motifs. Once candidate
substitutions have been identified, they would then be cloned into expression
vectors and tested to ensure reproducibility. Single round infectivity assays and
western blotting of viral particles and cell lysates to assess incorporation would be
used to test mutant glycoproteins. One important modification would be that,
ideally, mutants would be cloned into a replication incompetent provirus, which
makes these experiments much safer and easier to perform.
112
Figure 4.1: Target region of mutagenesis and selection strategy.
(A) Diagram of the region in HIV Env targeted by random mutagenesis. The 36 amino acid
cassette depicted above was randomly mutagenized at a rate of three percent, correlating
to an average of 3 nucleotide mutations per individual Env mutant. The sequence upstream,
and downstream, of the target cassette was maintained as wild-type sequence. The
mutagenesis was performed in the presence of the full-length cytoplasmic tail (B) Strategy
used for creating the stable mutant cell population and in vivo selection procedures. The
NL4-3 derived construct is depicted with the indicated accessory genes deleted. Nef has
been replaced with a blasticidin resistance gene driven by a CMV promoter. The red
asterisk indicates the location of the mutations in HIV Env.
113
Figure 4.1: Target region of mutagenesis and selection strategy.
114
Figure 4.2: Mutational landscape of the doped HIV-1 Env plasmid library
Depiction of the introduced substitutions at each position within the HIV Env target
cassette. Filled (Grey) boxes indicate the presence of that specific substitution in the HIV
Env doped plasmid library. Depicted above are the locations of randomly introduced stop
codons within the target cassette.
115
Figure 4.2: Mutational landscape of the doped HIV-1 Env plasmid library
116
5. SUMMARY AND DISCUSSION
117
Retroviruses readily form pseudotyped particles with a diverse panel of viral glycoproteins
from similar and unrelated families (73, 77). This phenomenon has been exploited by
researchers from various fields for manipulation of retroviral vectors to target specific cell
types. Lentiviral vectors can infect both dividing and non-dividing cells and therefore, are
appealing for use in gene therapy. One limitation, however, is that very little is known
about the pseudotyping mechanism. If we can gain insight into the molecular mechanism
of this process, then potentially non-viral surface proteins could be engineered to
incorporate into viral particles. This could greatly broaden the range of cell types that could
be specifically targeted by lentiviral gene delivery vectors. My work has shown that the
MSD of MLV Env has critical protein components necessary for fusogenicity and that the
MPER and MSD both contribute to MLV Env incorporation into viral particles.
Previous work showed that insertion of a leucine residue in the MSD of MLV Env
resulted in a non-functional glycoprotein, while insertions of 3 leucine residues restored
function. Suspecting that this observation had to do with alignment of interfaces within the
ectodomain and CTD of the trimer, mutational analyses were performed to identify the
boundaries of these potential interfaces. This approach lead to the identification of the
boundaries of these potential interfaces and also indicated that fusogenicity is greatly
influences by the hydroxyl-containing residues at the N-terminus of the MSD. Further
study revealed that these residues may be part of a SxxxTTxxS motif previously observed
to influence oligomerization of membrane helices (109). Interestingly, other gamma-
retroviral glycoproteins such as FLV and GaLV Env also have this serine/threonine
clustering. The same motif can also be found in VSV-G and in many different strains of
the influenza virus, both A and B. Interestingly, these glycoproteins all have a mechanism
118
to negatively regulate fusogenic activity. In the case of the gamma-retroviruses, that is
achieved by the presence of the R-peptide at the end of the CTD. In the case of VSV-G and
Influenza glycoproteins, fusogenicity is regulated by the pH of the surrounding
environment. It is plausible that this motif has an evolutionary advantage to aide in
regulation fusogenic activity.
To better understand the protein determinants in MLV Env dictate pseudotyping,
we randomly mutagenized the MPER and MSD to identify residues or motifs necessary
for incorporation. Previously, our lab established that the CTD is dispensable for this
recruitment, seeing as in the complete absence of the CTD Env is still actively recruited to
HIV-1 assembly sites (73). Because of this, we hypothesized that the driving force for
recruitment was located in the MPER and MSD. Our large scale biological selection study
revealed that two regions in the MPER and MSD showed increased sensitivity to mutation
and were predicted to be crucial for Env incorporation into viral particles. After directly
testing mutants identified in our screen, we demonstrated that W167R, W137L, and I142R
were incorporation defective. The phenotypes of all three of these substitutions matched
our NGS analysis and correlate with both regions that we identified. We hypothesize that
these regions interact with the lipid environment during viral assembly. It is plausible that
the multimerization of Gag creates a specific lipid environment that interacts with these
two domains in Env, thus resulting in active incorporation.
119
6. FUTURE DIRECTIONS
120
The body of work described in this dissertation provides a basis for further studies. With
the creation of distinct cell libraries that stably express both MLV and HIV Env mutants,
we are in possession of unique tools that can be used to understand multiple facets of Env
biology. In addition, each biological selection performed from either the MLV or HIV
starting library can be maintained and put under another type of selective pressure as the
need arises. Importantly, this methodology can be applied to various targets under diverse
selection pressures. This is particularly useful for manipulating proteins sensitive to
traditional mutagenesis, such as capsid, or in defining protein-protein interfaces, such as
between Vif and the host cell restriction factor APOBEC.
It will be interesting to determine if HIV-1 Env incorporation follows a similar
mechanism that we have discovered for MLV Env. Selection of the HIV-1 Env
mutagenized library for glycoproteins that are unable to incorporate into viral particles will
shed some light on this mechanism. It has been shown that there are short helical segments
in the HIV-1 MPER that fold towards the lipid membrane and interact with membrane
lipids. Although these studies did not define a role for this region in Env incorporation, it
may be that the MSD is also involved in this process and that by leaving it intact it can
compensate for the loss of the MPER hinge region. Our study will provide an answer for
this hypothesis, and hopefully will determine if these regions are important for
incorporation.
If it is determined that the recruitment mechanism is similar between MLV and
HIV it would be useful to attempt and apply this to glycoproteins known to be incompatible
for pseudotyping, such as full length RSV Env. Or, to apply these findings to host cell
proteins in an effort to induce incorporation into assembling viral particles. Progress in this
121
area will bring retroviral based gene therapy one step closer to achieving the goal of being
able to target any desired cell type will minimal off target effects. With these hopes for
future applications, this dissertation is, thus, concluded.
122
APPENDIX
123
I. RECOMBINATION CAN LEAD TO SPURIOUS RESULTS IN RETROVIRAL
TRANSDUCTION WITH DUAL FLUORESCENT REPORTER GENES
I.I Summary
Fluorescent proteins are routinely employed as reporters in retroviral vectors. Here,
we demonstrate that transduction with retroviral vectors carrying a tandem-dimer Tomato
(TdTom) reporter produces two distinct fluorescent cell populations following template
jumping due to a single nucleotide polymorphism between the first and second Tomato
genes. Template jumping also occurs between repeated sequences in the Tomato and green
fluorescent protein (GFP) genes. Thus proper interpretation of the fluorescence intensity
of transduced cells requires caution.
I.II Body
Fluorescent reporter proteins are frequently used in retroviral vectors for
identifying transduced cells (159-162). The fluorescent proteins green fluorescent protein
(GFP) and tandem-dimer tomato (TdTomato) are particularly useful because they are very
amenable to detection by flow cytometry. These two proteins are derivatives of genes
acquired from organisms of the genera Aequorea (163) and Discosoma but as a
consequence of its development as a molecular tool, TdTomato (as well as many other
fluorescent proteins) contains short segments of GFP sequence at its N- and C- terminus
(164-166). We have observed that transduction of cells with retroviral vectors containing
TdTomato results in two distinct populations with either bright or dim fluorescent
intensities, as detected by fluorescent light microscopy (Fig. 1A) and flow cytometry (Fig
124
1B). This phenomenon was observed using both gammaretroviral and lentiviral vectors
and was not dependent on the MOI of transduction (data not shown).
Retroviral recombination occurs during reverse-transcription of the retroviral
genome due to template-switching at stretches of homologous nucleotide sequence (167,
168). Recombination also occurs frequently in retroviral vectors containing short stretches
of repeated nucleotide sequences (169-172), such as the repeated sequences at the N- and
C- termini of TdTomato. To determine if the bimodal distribution of the distinct fluorescent
populations arose from one of the copies of Tomato being deleted due to template jumping,
we generated a retroviral vector with only the first copy of the Tomato gene (Fig. 1C). This
construct produced a single population of transduced cells (Fig. 1C and 1D), which was
distinctly brighter than the dim population produce from the TdTomato vectors depicted in
Fig. 1A. Thus, the dim fluorescent population did not appear to arise simply from the
deletion of one copy of the TdTomato gene as a result of recombination.
We next wished to determine if the dim Tomato fluorescent population resulted
from alterations in the Tomato gene itself, or alternatively from variations in Tomato
mRNA expression. To address this, we generated a TdTomato vector that also expressed
GFP from the downstream internal ribosomal entry site (IRES) in place of the puromycin
resistance gene (Fig. 2A). If the dim Tomato fluorescence resulted from alterations in the
Tomato gene itself, we expected the dim Tomato expressing cells to display normal GFP
expression levels (Fig. 2A, left panel). If the dim Tomato fluorescence resulted from
differences in mRNA expression, we expected the dim Tomato expressing cells to also
express dim GFP (Fig. 2A, center panel). Surprisingly, we did not observe dominant cell
populations of either of these phenotypes, but instead observed various fluorescent
125
populations with various combinations of Tomato and GFP fluorescence, including cells
expressing only GFP and cells expressing only Tomato (Fig. 2A, right panel). To
determine the nature of these single fluorescent populations, we sorted the fluorescent
green-only and red-only cells by flow cytometry, and then recovered the reporter sequences
by PCR and sequenced them. The results showed that differential fluorescent populations
resulted from template jumping between various locations in the vector where short
stretches of homologous sequence were present (Fig. 2B-E). Notably, template jumping
occurred between the short identical sequences at both the N-terminal and C-terminal
sequences in GFP and Tomato (Fig. 2C-E). This result implies that template jumping is
very prevalent in this system, as might be expected from the principles of viral reverse
transcription, but fails to fully explain the nature of the dim fluorescent Tomato population.
To determine the cause of the bright and dim Tomato fluorescent populations, we
performed single cell isolations of bright and dim red fluorescent cells by limiting dilution
(Fig 3A). It is well established that retroviral integration often occurs near regions of
chromatin associated with active transcription (173-177). To rule out genomic integration
as a contributing factor for the differential fluorescent intensity observed in Fig. 1A, we
analyzed bright and dim fluorescent single-cell isolates for TdTomato integration position
within the genome. Genomic DNA was extracted from the six populations and integration
locations were determined using the genome walker technique (Clontech) for five of these
populations. No obvious correlation was observed between integration site and gene
brightness (Fig. 3B). Next, we recovered the Tomato sequence of the bright and dim
populations to assess any alterations in the Tomato gene. All three of the dim Tomato
populations expressed a single copy of Tomato, and notably, in each case it was the second
126
copy of the Tomato gene (Fig. 3B). Some versions of the TdTomato gene have a single
nucleotide polymorphism in which an adenine is substituted with a guanine, subsequently
resulting in a histidine to arginine mutation within the second Tomato gene (Fig. 3C). This
polymorphism arose from a PCR mutation during the creation of TdTomato but was
initially overlooked because it did not appear to alter the phenotype of Tomato (N. Sharer,
Personal communication).
To confirm that the polymorphism was responsible for the Tomato dimness, we
generated a retroviral vector with only the second copy of Tomato (Fig. 4E). This construct
indeed produced dim red transduced cells (Fig. 4F). A parallel vector that expressed
TdTomato without the polymorphism did not change the brightness of Tomato, suggesting
that the polymorphism is detrimental only if expressed alone (Fig. 4G and 4H). Finally,
because sequence homology directly influences retroviral recombination (171, 178), we
also tested if the dim red population could be eliminating by reducing the sequence
homology between the two copies of Tomato. To accomplish this, we produced a
TdTomato gene where the codons in the first copy of Tomato were changed so that the
amino acid sequence was not changed, but the nucleotide homology with the second copy
of Tomato was dramatically reduced (Fig. 4I and 4J). This vector produced only bright red
cells.
In summary, we have shown that retroviral recombination among and between
fluorescent genes can lead to spurious phenotypes. Although template jumping is a well-
known phenomenon in retroviruses and retroviral vectors, short stretches of sequence
homology such as those found in fluorescent proteins are not always taken into
consideration when vectors are designed. Lack of attention to possible template jumping
127
could lead to misinterpretation of results, particularly when multiple reporters are used to
study retroviral phenomenon such as latency (179, 180). This potential problem can be
prevented by eliminating all repeated sequence, for example by scrambling the codon
sequence in the regions where amino acid sequences are repeated (Fig. 4J).
128
Figure I.I: Transduction using retroviral vectors carrying a TdTomato reporter
resulted in two cellular populations with distinct fluorescent intensities.
(A and B) TdTomato was cloned into the Murine Leukemia Virus (MLV) transfer vector
pQCXIP (Clontech) and used to generate non-clonal stable cell lines form HEK 293FT
cells. (A) Fluorescent images of the stable cell line transduced with TdTomato, white
arrows highlight bright (top) and dim (bottom) cells. (B) Total cell populations depicted in
(A) were analyzed using flow cytometry to confirm the presence of two distinct fluorescent
populations. (C and D) Parallel analysis of cells expressing a monomeric Tomato gene.
129
Figure I.I: Transduction using retroviral vectors carrying a TdTomato reporter
resulted in two cellular populations with distinct fluorescent intensities.
130
Figure I.II: Retroviral recombination occurs between distinct fluorescent proteins.
(A) Schematic of double-labeled reporter. The left and center flow plot panels depict
expected distributions of transduced cells if dimness was due to a transcriptional defect or
a gene specific defect, respectively. The right flow panel depicts observed flow cytometry
results following transduction of HEK 293FT cells. (B-E) Red-only and green-only cells
were sorted by flow cytometry. DNA was extracted from sorted cells and the region
between the CMV promoter and the end of GFP was PCR amplified and sequenced.
Control PRCs were performed on the parent plasmid to ensure the observed changes were
not caused by PCR. Template jumping observed within populations of green-only (B and
C) and red-only (D and E) populations are shown. Underlined regions indicate areas of
homologous nucleotide sequence.
131
Figure I.II: Retroviral recombination occurs between distinct fluorescent proteins.
132
Figure I.III: Fluorescence intensity is independent of TdTom genomic integration
sites.
(A) Cell pellets of bright and dim fluorescent single-cell isolated from HEK 293FT cells
transduced with MLV carrying the TdTom reporter. Cells transduced with the tandem-
dimer tomato (TdTom) and monomeric-tomato (mTom) reporters are included for
reference, along with non-transduced HEK 293FT cells (-). (B) Genomic DNA from bright
and dim fluorescent single-cell isolates, depicted in (A), was extracted and analyzed to
determine genomic integration sites and Tomato gene alteration. (C) Schematic
representation of the nucleotide and corresponding amino acid polymorphism between the
first and second Tomato gene within the TdTomato reporter.
133
Figure I.III: Fluorescence intensity is independent of TdTom genomic integration
sites.
134
Figure I.IV: Bimodal fluorescence intensity of the TdTom reporter occurs as a
result of a single nucleotide polymorphism between the TdTom genes.
(A-J) HEK 293FT cells were infected with MLV carrying the modified retroviral Tomato
reporters depicted in the left column. Resulting cell populations were analyzed by flow
cytometry to assess fluorescent cell populations and are depicted in histograms. Dashed
vertical lines in the right column highlight fluorescence intensity of the bright and dim
fluorescent cell populations present following transduction with virus carrying the
TdTomato reporter.
135
Figure I.IV: Bimodal fluorescence intensity of the TdTom reporter occurs as a result
of a single nucleotide polymorphism between the TdTom genes.
136
II. IMMUNE SIGNALING COMPETENCY IS REGULATED BY LIGAND-
INDUCED ENDOCYTOSIS AND SUBSEQUENT DE NOVO SYNTHESIS OF THE
FLAGELLIN-RECEPTOR FLS2
II.I Summary
FLAGELLIN-SENSING 2 (FLS2) is the plant cell surface receptor for the
perception of the bacterial pathogen-associated molecular pattern (PAMP) flagellin or
flg22 peptide. Flg22 recognition initiates immune signaling responses that contribute to
bacterial growth restriction. However, flg22 elicitation also leads to ligand-induced
endocytosis and degradation of FLS2 within 1 hour. Why plant cells remove an important
receptor precisely at the time during which its function is required remains mainly
unknown. To address this question, we assessed flg22- signaling competency of
endogenous FLS2 in the context of ligand-induced endocytic degradation of FLS2. Within
the first hour, cells were unable to re-elicit two independent flg22- signaling responses in
a ligand-, time- and dose-dependent manner. These experiments indicate that flg22-
induced degradation of FLS2 is a means to desensitize cells to stimuli, likely by removing
FLS2 receptors from the site of stimulus perception and to attenuate continuous flg22-
signaling. At later times (> 2hr) after flg22-treatment, de novo protein synthesis restored
signaling-competent FLS2 receptors, suggesting that new synthesis of FLS2 protein
functions in preparing cells for a new round of PAMP perception. Similar mechanisms
applied for regulating signaling competency of EF-TU receptor (EFR), an unrelated plant
PAMP receptor, in response to its bacterial ligand. Our study supports the hypothesis that
ligand-induced endocytic degradation of FLS2, and other defense-related perception
mechanisms, functions in preventing potential detrimental effects of prolonged immune
137
signaling on growth and development. Subsequent flg22-induced new synthesis of FLS2
restores receptor levels to enable cells to monitor its environment for the continued
presence of flg22-containing pathogens.
II.II Introduction
Eukaryotic host cell surface receptors are key components of innate immunity in
that they act as the first line of defense against invading microbial pathogens (181, 182).
These receptors detect extracellular pathogen-associated molecular patterns (PAMPs) as
“non-self” to alert host cells of these microbes. PAMP perception activates host immune
responses that ultimately contribute to cessation of microbial infection. In plants, the best
studied PAMP-triggered immunity (PTI) system is perception of bacterial flagellin or its
active peptide-derivative flg22 by FLAGELLINSENSING 2 (FLS2) (181, 183, 184). FLS2
is the plasma membrane (PM) localized receptor-like kinase (RLK) detecting apoplastic
flg22 through its extracellular leucine-rich repeat (LRR) domain (185, 186). Flg22
perception initiates downstream immune responses including production of reactive
oxygen species (ROS), activation of mitogen-activated protein kinases (MAPKs) and
transcriptional changes including up-regulation of FLS2 mRNA (183, 184), and these
responses fall into multiple parallel pathways rather than a single linear pathway (184, 187,
188). FLS2-induced responses are important for plant immunity as the absence of
functional FLS2 leads to impaired flg22-signaling and contributes to enhanced
susceptibility to bacterial infection (189). FLS2 is one of only few plant RLKs shown to
undergo ligand-induced endocytosis and subsequent degradation by mostly undefined
mechanism(s) (190-192). Based on a live cell imaging study (191), elicitation with active
138
flg22 (derived from Pseudomonas syringae; PF22), but not inactive flg22 (from
Agrobacterium tumefaciens; AF22) or elf26, a bacterial PAMP recognized by the plant
EF-TU receptor (EFR) (193), results in internalization and degradation of ectopically
expressed FLS2-tagged green fluorescent protein (FLS2-GFP). So far, only a limited
number of components are known to affect endocytosis and degradation of FLS2 (192,
194, 195). These observations lead to an apparent paradox whereby a functional FLS2 is
required for a full immune response, yet perception of the ligand removes FLS2 from the
cell surface at the time during which PAMP perception is required.
Constitutive activation of immune signaling diverts valuable resources from
growth and development to defense mechanisms, thus termination of PAMP-signaling
must be tightly controlled. Signal attenuation is likely achieved by diverse mechanisms
that include regulating activity or abundance of FLS2 itself and of components functioning
in PF22-signaling (192, 196-198). In animals, one of the established roles of ligand-
induced endocytosis is to down-regulate receptors from the site of perception to desensitize
cells to stimuli, thereby preventing continuous initiation of signaling (182, 199); but so far,
it is unclear whether this also holds true for Toll-Like Receptor 5 (TLR5), the mammalian
receptor of flagellin. In plants, ligand-induced endocytic degradation of RLKs has not been
formally investigated as a mode of regulating PAMP signaling (199, 200). Comparison
between endocytosis of FLS2 and another well-studied LRR-RLK, the brassinosteroid
receptor BRI1, highlights differences in endocytic regulation or mechanisms between plant
RLKs. FLS2 undergoes ligand-induced endocytic degradation that is relatively fast (~ 0.5
hr) (191) and requires functional BRI1-associated kinase BAK1 (192, 194). In contrast,
139
BRI1 is endocytosed constitutively and comparatively slow (> 5 hrs) in a ligand- and
BAK1-independent manner (201, 202).
To examine potential regulatory roles of ligand-induced down-regulation of FLS2
in PF22- signaling, we establish here re-elicitation assays to correlate FLS2 signaling
competency with receptor abundance. Our results indicate that PF22-induced degradation
of FLS2 serves as a means to desensitize cells to stimuli. In addition, subsequent PF22-
elicited de novo FLS2 protein synthesis prepares cells for a new round of PF22-perception.
Similar patterns of signaling competencies were observed in response to elf26, the
unrelated bacterial PAMP perceived by EFR (193), suggesting that such desensitization
mechanisms may be a conserved strategy for plant RLKs involved in PAMP perception.
II.III Experimental Procedures
Chemicals and PAMPs. Unless otherwise specified, all chemicals were from Fisher
Scientific (St. Louis, MO). Cyclohexamide (CHX; Sigma, St. Louis, MO) was used at a
final concentration of 50µM for Col0 leaves and 100µM for Ler seedlings at indicated
times. PAMP peptides (187) were made by GenScript (Scotch Plains, NJ) and used at
indicated concentrations and times.
Plant Material and Growth. Arabidopsis seedling and plant growth was at 22.C as
described (186) except that plants were grown in an 8-h light/16-h dark cycle. fls2-24 is in
Ler background (203); fls2., efr-2 and rbohD are in Col0 background (193, 204, 205).
Immunoblot Analysis and Antibodies. For sample elicitation for immunoblot analysis,
three leaf disks (each 1.5cm2; cut into 5 strips) of 4-5 week old plants or 7-8 d-old seedlings
were floated on 1 mL dH2O overnight at 22.C in continuous light to reduce wounding
140
response. Samples were elicited with indicated concentrations of specified PAMP for 45-
50 min, washed twice with 1mL dH2O to remove excess PAMP and incubated in dH2O at
22.C until flash frozen in liquid nitrogen at indicated times. For re-elicitation of MAPK
phosphorylation, samples were initially elicited and washed as indicated above, but then
incubated with the indicated PAMP for an additional 10 minutes prior to freezing. All
samples were stored at -80ºC. Sample preparation and immunoblot analyses of total
proteins using antibody concentrations were done as described (204). An exception is that
aP-p44/42 MAPK (1:3000) (#4370; Cell Signaling Tech, Danvers, MA) was used to detect
phosphorylated MAPKs.
Apoplastic ROS Production. Half of a cut leaf disk (1.1cm2) of 4-5 week old plants were
used for individual ROS production measurements using a luminol-based assay (204). For
ROS re-elicitation, samples were first elicited for 40 minutes with indicated PAMP at
specified concentration, washed twice with 150µL of dH2O, and placed at 22.C under
continuous light until indicated times of re-elicitation. To allow for correct comparison,
ROS experiments shown within the same figure were set-up at the same time and
performed in the same 96-well plate. Although absolute values of ROS production
[displayed as Relative Light Units; RLUs] varied from experiment to experiment, the actual
trends were always the same.
Quantitative Real-time PCR Analysis. For pulse elicitation, four leaf disks (each 1.1cm2,
cut into 5 strips) of 4-5 week old plants were elicited with 1µM PF22 for 40 min, washed
twice and kept in dH2O at 22ºC until flash frozen in liquid nitrogen at indicated times.
Total RNA was isolated and real-time PCR reactions were performed and analyzed as
described in (187, 206) with FLS2-f 5’-TCTGATGAAACTTAGAGGCAAAGCG-3’ and
141
FLS2-r 5’-CGTAACAGAGTTTGGCAAAGTCG-3’ primers using the expression of
At2G28390 (SAND family protein gene) to normalize all qRT-PCR results (187, 206).
Statistical Analysis. Each experiment was done at least 3 independent times with similar
results. Statistical significances based on unpaired two sample t-test were determined with
Graph Pad Prism4 software.
II.IV Results
Endogenous FLS2 undergoes ligand-induced degradation that is ligand-,
time- and dose-dependent. A previous live-cell imaging study using emerging young
leaves (191) showed that in response to 10 µM PF22, FLS2-GFP is internalized from the
PM into small vesicles at 30 min followed by loss of FLS2-GFP fluorescence at 60 min,
consistent with ligand-induced endocytosis and subsequent degradation of FLS2-GFP. To
eliminate the possibility that reduced FLS2-GFP signal was due to cleavage of GFP from
FLS2-GFP followed by degradation of GFP alone or to a change in local pH affecting
GFP’s ability to fluoresce (207), we confirmed first that endogenous, non-tagged FLS2
underwent PF22-induced degradation. Furthermore, a recent report underlines the
importance in studying endogenous proteins for functional analyses of RLK-based events
because C-terminal tags can affect RLK functions (208). Using Arabidopsis ecotypes
expressing functional endogenous FLS2 (209), we assessed FLS2 accumulation using an
αFLS2-peptide antibody that detects FLS2 in wild-type (WT) but not fls2∆ plants (187,
204). When wild-type Ler seedlings were treated with 10 µM PF22, concentration used in
the live-cell imaging study (191), FLS2 levels decreased after 40 min of PF22-elicitation
(Fig. 1A) consistent with PF22-induced degradation of endogenous FLS2. In contrast, fls2-
24 seedlings expressing FLS2 mutant protein impaired in PF22-binding and signaling
142
(209) did not show altered FLS2 levels (Fig. 1A), confirming the specificity of the
degradation response. Similarly, FLS2 underwent ligand-induced degradation in a dose-
dependent manner in 4-5 week old Col0 leaves (Fig. 1C), a developmental tissue stage
commonly used for PF22-signaling and pathogen infection assays. Importantly, FLS2
degradation occurred at a concentration as low as 0.1 µM, commonly used for PF22-
signaling response assays (187, 194, 204). Confirming the specificity for the ligand-
receptor pair, FLS2 degradation was specific to active flg22 (PF) but not to water, inactive
flg22 (186) or the unrelated bacterial PAMP elf26 (E) (Fig. 1A-C). Protein levels of
Mitogen-Associated Protein Kinase 6 (MPK6; M6) or the ER membrane-bound calnexin
(Caln) were not significantly altered by any treatment (Fig. 1A-D). These results
demonstrate that endogenous, non-tagged FLS2 underwent PF22-induced degradation in a
ligand-, dose- and time-dependent manner in tissues of different developmental stages.
Cycloheximide treatment blocks ligand-induced degradation of FLS2 and
attenuation of PF22-signaling. PF22-elicitation resulted in a significant decrease but not
complete loss of FLS2 protein accumulation (Fig. 1). Because PF22 induces FLS2 mRNA
accumulation at 1 h (189, 210) (see below), Col0 leaf strips were treated with the protein
translation inhibitor cycloheximide (CHX) to examine whether FLS2 detected at 1 h after
PF22-elicitation represented newly synthesized FLS2. We consistently noticed, however,
that in the presence of CHX, FLS2 did not undergo significant ligand-induced degradation
at 1 h in response to PF22 (Fig. 1D). Similar results were obtained for Ler seedlings
(Supporting information Fig. S1A). CHX treatment also inhibited attenuation of PF22-
signaling responses. MAPKs remained phosphorylated at 60 min post-PF22-elicitation in
the presence (+CHX/+PF) but not the absence of CHX (-CHX/+PF) (Fig. 1D; Fig. S1A).
143
Similarly, the duration of PF22-induced ROS production was greatly prolonged in the
presence (Fig. 1F) but not absence of CHX (Fig. 1E). Treatment with CHX alone did not
significantly induce MAPK phosphorylation or ROS production (Fig 1D or Fig 1F,
respectively). In contrast to duration, the amplitude of PF22-induced MAPK activation or
ROS production was not significantly affected. Elevated ROS levels after treatment with
PF22+CHX were not observed in rbohD null mutant plants (Fig. S1B) lacking functional
RBOHD, the PM- located NADPH oxidase responsible for PAMP-dependent extracellular
ROS production (211). Thus, prolonged ROS production in CHX-treated tissue was
unlikely due to cell death-associated ROS production. Our results indicate that de novo
synthesis of yet unknown protein(s) contribute to ligand-induced degradation of FLS2. It
is unclear whether the block in ligand-induced degradation of FLS2 directly or indirectly
affects attenuation of PF22-signaling responses. Our data are consistent with the hypothesis
that ligand-induced degradation of FLS2 may serve as a means to attenuate constitutive
signaling of innate immune responses. We cannot exclude, however, the possibility that
distinct de novo synthesized proteins such as phosphatases (212) and/or factors involved
in superoxide radical detoxification contribute to attenuation of these PF22-induced
responses.
Ligand-induced degradation of FLS2 desensitized cells to stimulus. To address
whether PF22-dependent degradation of FLS2 is a means to desensitize host cells to PF22,
we utilized two independent PF22-signaling responses (ROS production and MAPK
activation) because a) they are implicated in two parallel PF22-signaling pathways (184)
and b) both responses are rapid and transient, in that PF22-induced activities peak 10-15
144
min post-elicitation and return to nearly basal levels within 40-60 min (Fig. 2A-D). Leaf
tissue was first treated at 0 min with the indicated PAMP, incubated for 45-50 min and
washed to remove excess ligand. To evaluate PAMP-signaling competency, the same
samples were re-elicited with specified PAMPs at 60 min. Responses were monitored at
indicated times post-elicitation. After the initial fast and transient PF22-induced ROS
response (Fig. 2A), re-elicitation of the same tissue with PF22 at 60 min did not result in
significant ROS production (Fig. 2A). Thus, at 60 min post-elicitation, these cells were
PF22-signaling incompetent. Loss of PF22-induced signaling at 60 min correlated with
PF22-induced degradation of FLS2 at that time (Fig 1C). In contrast, re-elicitation with
elf26 led to ROS production indicating that cells treated initially with PF22 were capable
of mounting PAMP-dependent ROS responses at 60 min per se (Fig. 2B). Cells initially
elicited with inactive flg22 (186) (Fig. 2C) or the unrelated elf26 (E) (Fig. S2A) remained
signaling competent for PF22, consistent with FLS2 not being degraded in response to
these peptides (Fig. 1C). We conclude that the inability to re-elicit cells with PF22 was
specific to the receptor-ligand pair FLS2-PF22. PF22-signaling incompetency was
observed as early as 20 min after the initial PF22- treatment (Fig. S2B); but re-elicitation
with elf26 resulted in significant ROS production at these times (Fig. S2B). This argues
against the possibility that this PF22-signaling incompetency between 20 and 60 min may
be solely due to an extended refractory period of cellular components required for PAMP-
induced ROS production such as RBOHD. Loss of PF22 signaling competency at 60 min
was dose-dependent, in that the amplitude of the re-elicitation ROS response was
dependent on the initial PF22-concentration to which cells were exposed to at 0 min. Cells
treated first with 0.001 or 0.01 µM PF22 remained signaling competent to 0.1 µM PF22 at
145
60 min (Fig. S2C or S2D). In contrast, cells initially exposed to 0.1 µM or 1 µM PF22 were
PF22-signaling incompetent at 60 min (Fig. S2E or S2F). The dose-dependency of the re-
elicitation response showed an inverse correlation with the dose-dependent degradation of
FLS2 in response to PF22 (Fig. 1C). Thus, concentrations between 0.1-1 µM PF22 were
used in subsequent experiments.
To assess the cell’s ability to re-elicit an independent PAMP-response, namely
MAPK activation, samples were collected at 0 and 60 min as well as 10 min post-elicitation
treatments (10 and 70 min, respectively) with indicated PAMPs and analyzed for
phosphorylated MAPKs (7, 26). Consistent with ROS results, cells initially treated with
PF22 were PF22-signaling incompetent at 60 min (Fig. 2D; PF/PF, 70*; P-M6, P-M3) but
remained signaling competent to elf26 (PF/E, 70*). The inability to re-elicit PF22-
dependent MAPK phosphorylation was not due to reduced MAPK protein levels at any
times (Fig. 2D; M6). Taken together, re-elicitation data using two independent PAMP-
signaling assays support the hypothesis that ligand-induced degradation of FLS2 is a
regulatory reponse to desensitize host cells to the stimulus.
PF22-induced de novo synthesis of FLS2 is a means to prepare cells for a new
round of PF22-perception. Using quantitative real-time (qRT)-PCR (Fig. S3A), we
confirmed previous microarray studies showing that PF22-elicitation leads to an increase
in FLS2 gene expression at 1 h (189, 210). It is, however, unknown whether an increase in
FLS2 mRNA correlates with increased FLS2 protein accumulation and, if so, whether
newly synthesized receptors are signaling competent. To this end, samples were pulse-
treated with PF22 for 45-50 min and incubated for up to 24 h in the absence of ligand. After
the initial PF22-dependent FLS2 degradation (1h), FLS2 protein levels began to be restored
146
until reaching pre-treatment levels at 24 h (Fig. 3A). No protein of similar apparent
molecular weight was detected in fls2. null mutants (Fig. S3B), confirming that the re-
appearing protein is FLS2. To verify that FLS2 accumulation at later times was due to de
novo protein synthesis, we adjusted our experimental design because CHX inhibited PF22-
degradation of FLS2 (Fig. 1D). Tissue was first incubated with PF22 for 1 hr in the absence
of CHX. At 1 h post-elicitation corresponding to the time at which FLS2 was mostly
degraded, samples were incubated for an additional 2 h in the absence (- ) or presence (+)
of CHX. When comparing FLS2 protein accumulation at 3h (Fig. S3C), CHX-treated
samples (+CHX/+PF) did not accumulate additional FLS2 protein whereas in the absence
of CHX (- CHX/+PF), FLS2 protein levels increased at 3 h post-elicitation. These results
are in agreement with PF22-treatment leading to de novo FLS2 protein synthesis at later
times after an initial ligand-induced degradation of FLS2 (1 h).
To assess whether newly synthesized FLS2 was signaling competent, samples were
first elicited at 0h with PF22 for 45 min, washed and kept in water until 1, 2, 3, 4 or 24 h.
For MAPK activation, control samples (only 1st elicitation) were collected at indicated
times after initial PF22 elicitation. Samples from tissue re-elicited with PF22 were
collected 10 min after the second elicitation. A 10 min sample (i.e. initial 10 min response
to PF22) served as a positive control. Consistent with Fig. 2D, MAPK phosphorylation was
detected at 10 min after the 1st elicitation (Fig. 3A, 10m) but only minimally upon re-
elicitation at 1h (1h+10m). In contrast, PF22 was able to re-elicit strong MAPK
phosphorylation at later times between 2 and 24 h, and the intensity of MAPK
phosphorylation correlated with increased FLS2 protein accumulation. Using a similar
experimental design for ROS re-elicitation, tissue that was initially treated with PF22 also
147
displayed significant PF22-induced ROS production after re-elicitation at 3 and 16 h (Fig.
3C-D). This gradual increase in ROS production over time, reflecting recovery of PF22-
signaling competency, is highlighted in Fig. S3D. No ROS production was observed in
fls2. (Fig. S3E) or rbohD (Fig. S3F) null mutants indicating that ROS re-elicitation
responses were dependent on functional FLS2 and RBOHD. The RBOHD requirements
also confirmed that ROS detected at later times were not due to other mechanisms of ROS
production. In conclusion, results from two independent re-elicitation assays correlated
restoration of PF22-signaling competency with increased accumulation of newly
synthesized FLS2. Our results are consistent with the hypothesis that after the initial PF22-
dependent endocytic degradation of FLS2 (at 1 h), ligand-induced de novo synthesis of
FLS2 at later times (> 1 h) is a means to prepare cells for a new round of PF22-perception.
Early ligand-induced desensitization and subsequent re-establishment of
signaling competency extends to elf26-EFR. Next, we examined whether the principles
of desensitization in immune-signaling may hold as a general paradigm for other bacterial
PAMP-RLKs. We focused on elf26, the bacterial PAMP perceived by EFR, because it
initiates PAMP-responses similar to PF22 (183, 193). After elf26-elicitation at 0 h, Col0
was unable to re-elicit elf26-dependent ROS and MAPK activation at 1 h (Fig. 4A and 4C)
indicating that the initial elf26 treatment led to desensitization of EFR to its ligand. In
contrast, tissue treated first with elf26 was able to re-elicit subsequent PF22-dependent
responses (Fig. 4B-C) showing that samples were able to mount PAMP-responses per se.
Signaling competency to elf26 was not impaired when samples were first treated with PF22
(Fig. 2B) confirming that the inability to re-elicit with elf26 was specific to the ligand-
receptor elf26-EFR pair combination. EFR signaling competency to elf26 recovered over
148
time as assessed by MAPK phosphorylation (Fig. 4D) and ROS production (Fig. S4A-E).
In control experiments for ROS-re-elicitation, re-establishment of EFR signaling
competency was EFR- and RBOHD-dependent (Fig. S4D-E). At least for ROS production,
signaling recovery appeared to occur more slowly compared to that to PF22. Such variation
may be due to the fact that requirements for receptor biosynthesis, glycosylation and
trafficking through the secretory pathway are different for FLS2 and EFR (197, 198, 213,
214).
II.V Conclusions
We provide evidence that endogenous, non-tagged FLS2 protein was degraded in
a ligand-, time and dose-dependent manner followed by a subsequent de novo protein
synthesis of FLS2. Re-elicitation assays for two independent PAMP-responses allowed us
to correlate FLS2 signaling competency with receptor accumulation. Our results indicate
that at early times (< 2h) after initial elicitation, PF22-induced degradation of FLS2 serves
as a means to desensitize cells to stimuli, possibly by removing FLS2 receptor molecules
from the site of stimulus perception and helping to prevent continuous initiation of PF22-
signaling responses. Elicitation with PF22 resulted consistently in a significant decrease
but not a complete loss in FLS2 protein accumulation at 60 min in a dose-dependent
manner (Fig. 1). Based on our signaling- competency data, a subset of FLS2 molecules
may be in an altered conformational state or sequestered to a subcellular location that
prevents these receptors from perceiving PF22 and/or inducing PF22-events. In the latter
scenario, FLS2 may be present in endocytic vesicles en route for protein degradation or in
parts of the secretory pathway en route to the PM, the site of flg22 perception. Consistent
149
with this hypothesis, lack of signal initiation at 20 min after PF22- elicitation correlated
with the time at which FLS2 is reported in intracellular vesicles (191).
At later times (>2h) after PF22-treatment, increased accumulation of de novo
synthesized FLS2 protein correlated with time-dependent recovery of signaling
competency. We propose that new synthesis of FLS2 protein serves as a means to prepare
cells for a new round of PF22- perception. This hypothesis is congruent as cells are likely
to encounter flagellated bacteria multiple times throughout their lifecycle. During an initial
encounter, ligand-induced degradation of FLS2 may aid in preventing continuous PF22-
signaling known to be detrimental for plants because it diverts valuable cellular resources
from growth and development to signaling. Subsequent replenishment of signaling
competent FLS2 is conceivable because cells unable to reinstate degraded receptors with
newly synthesized FLS2 would lose their ability to perceive flagellin and initiate immune
responses during subsequent infections. In agreement, cells lacking functional FLS2
exhibit enhanced susceptibility to bacterial infection (189). Such replenishment scenario
may provide insight into why pathogenic bacteria evolved mechanisms to inject effectors
into the host cell that target components of the FLS2 receptor complex for degradation
(190, 215, 216), thus suppressing establishment of host PTI (217). Because effector
delivery is assumed to occur after PTI response initiation (217), effectors that target the
FLS2-complex components may interfere with the re-establishment of PF22-signaling
competency of cells rather than with the initial detection of bacteria and PF22-response
initiation. Furthermore, our results using an unrelated bacterial PAMP elf26 suggest that
similar mechanisms may apply for regulating signaling competency of other plant PAMP-
receptors in response to their ligands.
150
Figure II.I: PF22-induced degradation of endogenous FLS2 is ligand-, dose- and time-
dependent and is inhibited by cycloheximide.
(A) Time-dependency and specificity in Ler seedlings. Ler WT and fls2-24 seedlings were
elicited in the presence (+) or absence (-) of 10μM PF22 or in (B) with 1μM PAMPs for
indicated times. (C) Dose- and ligand-dependency in Col0 leaves. Col0 leaf strips were
elicited with indicated μM concentration (+) or without (-) specified PAMP for 0 or 60
min. (D) Cycloheximide (CHX) inhibits PF22-induced degradation of FLS2 and
attenuation of MAPK activation. Col0 leaf strips were elicited in plus (+) or minus (-) 1μM
PF22 or 50μM CHX for indicated times in minutes. For immunoblot analyses (A-D), total
protein extracts were probed with αFLS2 or αP-p44/42 MAPK to assess FLS2 protein
degradation or phosphorylation of MPK6 (P-M6), MPK3 (P-M3) and an unknown MKP
(P-M?). Individual MAPKs were identified by mass. Immunoblots probed with αCalnexin
(Caln) or αMPK6 served as loading controls or confirmed MPK6 accumulation. (E-F) Col0
leaf disk halves (n=20/ treatment) were elicited plus (+) or minus (-) 0.1μM PF22 or 50μM
CHX at 0 minutes. ROS production was monitored over 40 minutes shown as Relative
Light Units (RLU). To allow for correct comparisons, ROS experiments in (E-F) were
performed in the same 96-well plate at the same time. Values are mean ± SE. PF, PF22;
AF, AF22, E, elf26.
151
Figure II.I: PF22-induced degradation of endogenous FLS2 is ligand-, dose- and time-
dependent and is inhibited by cycloheximide.
152
Figure II.II: Ligand-induced degradation of FLS2 is a means to desensitize cells to
stimuli.
(A-C) Specificity of ROS re-elicitation. For first elicitation, Col0 leaf disks (n=20/
treatment) were elicited with 0.1μM of indicated PAMP at 0 min, washed and re-elicited
with indicated PAMP at 60 min (arrow) (n=20/ treatment). ROS experiments in (A-C) were
set-up in the same 96-well plate at the same time. (D) Specificity of re-elicitation of MAPK
phosphorylation. Col0 leaf strips were first elicited with 0.1μM PF22 for 0, 10 or 60
minutes. Samples denoted 70* were re-elicited at 60 min for 10 min with 0.1μM PAMP.
Total protein extracts were probed as in Fig 1. Values are mean ± SE. PF, PF22; AF, AF22;
E, elf26; RLU, Relative Light Unit.
153
Figure II.II: Ligand-induced degradation of FLS2 is a means to desensitize cells to
stimuli.
154
Figure II.III: Ligand-induced de novo synthesis of FLS2 prepared cells for a new
round of PF22-perception.
(A) Increased PF22-induced FLS2 protein accumulation and recovery of MAPK-signaling
competency. After a 1st elicitation with 0.1 μM PF22 at 0 min for 45-50 min, Col0 leaf
strips were washed and incubated in the absence of PF22 until indicated times [in min (m)
or hours (h)]. For reelicitation (2nd), samples were elicited with 0.1 μM PF22 at indicated
hours for 10 min. Immunoblot analysis of total protein extracts was done as in Fig. 1. (B-
D) Recovery of ROS signaling competency. Col0 leaf disk halves were pulsed with 0.1 μM
PF22 at 0h for 45 min (n=60/treatment), washed and re-elicited with 0.1 μM PF22 at 1, 3
or 16 h (n=20/treatment). Arrows indicate times of second elicitation. Bar graph
representation shows RLU at peak (10-12 min) after elicitation at indicated hr. ROS
experiments in (B-D) were set-up in the same 96-well plate. Values are mean ± SE. PF,
PF22; RLU, Relative Light Unit.
155
Figure II.III: Ligand-induced de novo synthesis of FLS2 prepared cells for a new
round of PF22-perception.
156
Figure II.IV: Regulation of ligand-induced signaling competency of EFR is regulated
similar to PF22-FLS2.
(A-B) Specificity of ROS re-elicitation. At 0 min, Col0 leaf disks (n=20) were treated with
0.1μM elf26, washed after 40-45 min and re-elicited with indicated PAMP at 60 min
(arrow) (n=20/treatment). (C) Specificity of re-elicitation of MAPK phosphorylation. Col0
leaf strips were first elicited with 0.1μM E for 0, 10, or 60 min. Samples denoted by 70*
were re-elicited at 60 min for an extra 10 min with indicated 0.1μM PAMP. (D) Elf26-
induced recovery of MAPK-signaling competency at later times. Ability to induce MAPK-
phosphorylation at different times after first elicitation was assessed as described in Fig.
3A except that 0.1μM elf26 was used as eliciting PAMP. ROS experiments in (A-B) were
set-up in the same 96-well plate at the same time. Values are mean ± SE. PF, PF22; E,
elf26.
157
Figure II.IV: Regulation of ligand-induced signaling competency of EFR is
regulated similar to PF22-FLS2.
158
Figure II.S1: Prolonged activation of PF22-signaling in the presence of CHX is
RBOHD-dependent.
(A) Prolonged PF22-induced MAPK phosphorylation in Ler seedlings. Ler WT seedlings
were elicited in the presence (+) or absence (-) of 10μM PF22 or 100μM CHX at the
indicated times in minutes. Immunoblot analyses, total protein extracts were probed with
αFLS2, αP-p44/42 MAPK or αMPK6 antibodies. αMPK6 served as loading controls and
confirmed MPK6 accumulation. PF, PF22; min, minutes. (B) Prolonged PF22-induced
ROS production is RBOHD-dependent. Col0 and rbohD leaf disk halves were elicited with
(+) or without (-) 0.1μM PF22 in the presence or absence of 50μM CHX at 0 min
(n=10/treatment). In this bar graph representation, ROS production is shown at the peak of
ROS production 10-12 min postelicitation. ROS experiments were performed in the same
96-well plate at the same time. Values are mean ± SE. PF, PF22; P-M6 or P-M3,
phosphorylated MPK6 or MPK3, respectively; CHX, cycloheximide; min, minutes, RLU,
Relative Light Units. Each experiment was done at least 3 times with similar results.
159
Figure II.S1: Prolonged activation of PF22-signaling in the presence of CHX is
RBOHD-dependent.
160
Figure II.S2: PF22-signaling incompetency is PAMP-, time- and dose-dependent.
(A) PAMP-dependency of ROS-re-elicitation. Col0 leaf disk halves (n=48) were elicited
at 0 min (1st), washed at 45 min and reelicited (2nd) at 60 min with 0.1μM of indicated
PAMP (n= 48/ treatment). ROS production is shown at peak elicitation (10-12 min post-
elicitation). (B) Signaling competency between 20 and 60 min. Col0 leaf disk halves (n=96)
were elicited at 0 min (1st) for 15 min with 0.1μM PF22. Tissue was then washed and re-
elicited (2nd) at 20, 40 or 60 min, respectively, with 0.1μM of indicated PAMP (n=
16/treatment). ROS production is shown at peak elicitation (10-12 min post-elicitation).
(C-F) PF22-signaling incompetency is dose-dependent. For their first elicitation (1st), Col0
leaf disk halves were elicited with the indicated concentration of PF22 at 0 min
(n=24/treatment), washed and re-elicited at 60 minutes (arrow) with 0.1μM PF22
(n=24/treatment). PF, PF22; E, elf26; min, minutes; RLU, Relative Light Units. To allow
for comparisons, all ROS experiments in (A), (B), and (C-F), respectively, were performed
in the same 96-well plate at the same time. Values are mean ± SE and each experiment was
done at least 3 independent times with similar results.
161
Figure II.S2: PF22-signaling incompetency is PAMP-, time- and dose-dependent.
162
Figure II.S3: Recovery of signaling competency was due to PF22-induced de novo
synthesis of FLS2.
(A) PF22-induced expression levels of FLS2 mRNA. Col0 leaf strips were elicited for 40-
45 min with 1μM PF22, washed and incubated in the absence of PF22 until indicated time
(hours). Samples were processed for qRT-PCR using At2g28390 as the reference gene. For
each timepoint, results of at least three independent experiments containing one biological
and three technical repeats are shown. (B) Increased PF22-induced FLS2 protein
accumulation. After a 1st elicitation with 0.1 μM PF22 at 0 min for 45 min, Col0 or fls2Δ
leaf strips were washed and incubated in the absence of PF22. Samples were collected at
indicated times in hours. (C) PF22-induced de novo synthesis of FLS2. Col0 leaf strips
were treated for 40 min with 1μM PF22 and washed to remove access ligand. At 60 min, a
subset of samples were treated with (+) or without (-) 50μM CHX for an additional 2 hours
(3h). Samples were collected at indicated times. For immunoblot analyses in (B-C), total
protein extracts were probed with αFLS2 or αMPK6 to assess FLS2 protein accumulation
or MPK6 protein levels as a loading control. (D) Time-dependent recovery of ROS
signaling competency in response to PF22. In a first elicitation (1st), Col0 leaf disk halves
were pulsed with 0.1 μM PF22 at 0h for 45 min (n=60). Samples were washed to remove
access ligand and then re-elicited (2nd) with 0.1 μM PF22 at indicated times
(n=10/treatment). Bar graph representation shows RLU at peak (10-12 min) after elicitation
at indicated hour (h). (E-F) Recovery of ROS signaling competency is FLS2-and RBOHD-
dependent. Samples of WT (Col0), fls2 and rbohD plants were treated as in (D) at indicated
times. ROS experiments shown in the same graph were set-up in the same 96-well plate.
163
Values are mean ± SE. PF, PF22; h, hours; m, min; RLU, Relative Light Unit. Each
experiment was done at least 3 independent times with similar results.
164
Figure II.S3: Recovery of signaling competency was due to PF22-induced de novo
synthesis of FLS2.
165
Figure II.S4: Recovery of elf26-signaling competency using ROS production.
(A-C) Timecourse of ROS re-elicitation. For the first elicitation, Col0 leaf disks (n=60)
were elicited with 0.1μM elf26.at 0 min, washed at 40 min and re-elicited with 0.1μM elf26
at 1, 3, or 16 h (n=20/treatment). (D) Recovery of elf26-signaling-competency is EFR-
dependent. (E) Recovery of elf26-signaling-competency is RBOHD-dependent. Re-
elicitation assays in (E-F) were done as in (A-C) except that in addition to Col0, efr-2 or
rbohD leaf disks, respectively, were used and that data are shown in bar graph format at
12-14 min post-elicitation. ROS experiments were set-up in the same 96-well plate at the
same time for (A-C), (D) or (E), respectively. Values are mean ± SE. PF, PF22; E, elf26;
min, minutes; h, hours, RLU, Relative Light Units.
166
Figure II.S4 Figure II.S4: Recovery of elf26-signaling competency using ROS
production.
167
III. CONTINUOUS FLG22 TREATMENT LEADS TO LIGAND-
DESENSITIZATION AND LOSS OF FLS2 PROTEIN ACCUMULATION
III.I Summary
In plants, innate immune signaling is the primary mechanism for defense against
invading microbial pathogens. Perception of microbial pathogens occurs through
recognition of conserved microbial features by plant cell-surface receptors. In the best
studied plant PAMP-receptor interaction system, the cell surface receptor Flagellin
Sensitive 2 (FLS2) perceives bacterial flagellin (or a conserved peptide derivative, flg22).
Perception of flg22 by FLS2 leads to activation of a multitude of downstream innate
immune signaling responses. However, these signaling responses must be tightly regulated
due to a limited pool of cellular signaling components shared between immune signaling
and developmental signaling pathways. Here, using two independent pathogen-associated
molecular patterns (PAMP)-ligands flg22 and elf26 I demonstrate continuous PAMP-
stimulation results in ligand-desensitization. Continuous treatment using flg22 at time
points greater than 8h resulted in decreased reactive oxygen species production and
decreased FLS2 protein accumulation. Furthermore, I compare the commonly utilized
syringe-infiltration technique to continuous exogenous application of flg22-peptide and
determine that syringe-infiltration does not recapitulate exogenous PAMP treatment.
III.II Introduction
Cell surface receptors are the front-line sentries that defend against pathogen
invasion and initiate innate immunity in both plants and animals. Perception of microbial
PAMPs initiates a multitude of downstream signaling events which trigger immunity and
168
inflammation to halt microbial infection (181, 218, 219). In the best studied plant PTI-
system, the host cell surface receptor Flagellin Sensitive 2 (FLS2) perceives bacterial
flagellin (or its peptide derivative flg22) leading to activation of immune responses which
contribute to restriction of pathogen growth (189, 220). FLS2 is a plasma membrane
leucine-rich repeat receptor-like kinase which perceives apoplastic flg22 through its
extracellular leucine-rich repeat domain (185, 214). We, and others, have demonstrated
that the cell surface PAMP-receptor FLS2 undergoes ligand-induced degradation
following perception of its cognate ligand, which likely occurs through endocytic
trafficking of activated receptor to a degradative compartment, possibly the vacuole (191,
221) (Salamango et al, in prep.). Recently, our lab demonstrated ligand-induced
degradation of FLS2 may serve as a means to desensitize cells to the flg22 stimulus
(Salamango et al, in prep.). Constitutive immune signaling redirects valuable resources
from growth and developmental pathways to sustain immune signaling pathways;
therefore, strict regulation of immune signaling amplitude and duration must be maintained
for activated immune signaling pathways. Several reports have established that prolonged
innate immune signaling has detrimental effects on organism fitness; and therefore,
activated innate immune signaling mechanisms require tight regulation (222-225).
We previously developed re-elicitation assays using ROS production and MAPK
phosphorylation to correlate FLS2 degradation, and subsequent ligand-induced new
synthesis of FLS2, with cellular signaling competency (Salamango et al, in prep.). Utilizing
these assays, we established that initial FLS2 degradation (~1h following ligand
perception) correlated with desensitizing cells to a stimulus (active flg22) and that over-
time, newly synthesized FLS2 protein accumulation restored cellular signaling competency
169
to near initial elicitation levels (~24h following initial elicitation) (Salamango et al, in
prep.). This type of pulse-chase assay was performed by eliciting tissues with a specific
PAMP for 45-50 minutes, washing with dH2O to remove excess PAMP-ligand, and re-
eliciting with the same PAMP at subsequent time points. This experimental approach
allowed for correlation of FLS2 activation, degradation, and subsequent de novo
accumulation to cellular signaling competency. However, this experimental approach
unlikely reflects what occurs during a natural bacterial infection wherein cells are exposed
to the flagellin-flg22 epitope continuously. Thus, it is unlikely that cessation of PAMP-
induced signaling would occur after 45-50 minutes due to persistence of bacterial infection;
therefore, I sought to determine cellular signaling competency after continual PAMP
stimulation.
III.III Experimental Procedures
Immunoblot Analysis and Antibodies. For sample elicitation, three leaf disks (each
1.5cm2; cut into 5 strips) of 4-5 week old plants were floated on 1 mL dH2O overnight at
22◦C in continuous light to reduce wounding response. The following day, samples were
elicited with indicated concentrations of flg22, or with 1x108 CFU Pseudomonas syringae
DC3000 HcrC- in dH2O. For pulse treatment, elicitations were performed for 45-50 min,
then the tissue was washed twice with dH2O to remove excess PAMP ligand, and flash
frozen at the indicated time. For continuous treatment, samples were elicited in the same
manner with the exception that PAMP solution was not removed after 45-50 min and tissue
was continuously exposed to the same PAMP solution. Samples were flash frozen at the
indicated time point. Sample preparation and immunoblot analyses of total proteins using
antibody concentrations were done as described (204).
170
Apoplastic ROS Production. Half of a cut leaf disk (1.1cm2) of 4-5 week old plants were
used for individual ROS production measurements using a luminol-based assay (204). For
pulse ROS re-elicitation, samples were first elicited for 40 minutes with indicated PAMP
at specified concentration, washed twice with 150μL of dH2O, and placed at 22◦C under
continuous light until indicated times of re-elicitation. For continuous ROS re-elicitation,
samples were treated as described for pulse samples; however, samples were not washed
after the initial elicitation and remained in the presence of PAMP until re-elicited using
new PAMP solution at the indicated time. For syringe-infiltrated ROS re-elicitation, Col0
adult leaf tissue was syringe infiltrated with 1μM flg22 solution until leaves were saturated.
After ~1h when the leaf tissue had dried, leaf disk punches were taken (1.1cm2) from the
center of the infiltrated area and leaf disk halves were floated on 150μL dH2O to allow
wounding responses to subside before re-elicitation at 24h. To allow for direct comparison,
ROS experiments shown within the same figure panel were initiated at the same time and
performed in the same 96-well plate. Although absolute values of ROS production
[displayed as Relative Light Units; RLUs] varied from experiment to experiment, the actual
trends were always the same.
Bacterial Preparation and ROS Elicitation. Pseudomonas syringae DC3000 hrcC- was
grown for two days on agar plates under rifamycin selection. On the third day, bacteria
were transferred to a new plate to ensure elicitation was performed using actively growing
bacteria. At the time of elicitation, bacteria were isolated, washed in dH2O, and used for
elicitation at an optical density of 0.1 at 600nm wavelength (approximately1x108 CFU) for
171
elicitation. For the bacterial wash, bacteria were suspended in dH2O and centrifuged for 5
minutes at 3000xg. Col0 cut leaf disks were floated on bacterial solution and ROS
production was monitored using a luminol-based assay (204) at 0h. Samples were either
washed twice with dH2O (pulse), or left in bacterial solution until the time of re-elicitation
(continuous). For re-elicitation, samples were treated with 100nM flg22 solution and ROs
production was monitored.
Callose Deposition. For syringe infiltration, adult leaf tissue 4 to 5 weeks in soil was
syringe infiltrated using 1μM flg22 or dH2O and collected directly into 95% ethanol
solution 24 hours post-infiltration. After overnight incubation in 95% ethanol, tissue was
rinsed in 50% ethanol and incubated in 67 mM potassium phosphate (K2HPO4; pH 12)
buffer for 1 hour. Tissue was stained using 0.01% analine blue/potassium phosphate buffer
solution for 1h in the dark at room temperature. Stained tissue was mounted in 70%
glycerol, 30% 67mM K2HPO4 solution (204).
Plant Material and Growth. Arabidopsis plant growth was at 22◦C as described (204)
except that plants were grown in an 8-h light/16-h dark cycle.
III.IV Results
Continuous PAMP-Signaling Desensitizes Cells to the Stimulus. To determine
if continuous PAMP stimulation resulted in negative regulation of cellular signaling
responses, I compared pulse and continuous elicitation of Col0 leaf tissue at 1, 3, and 16
hours following an initial PAMP-treatment at 0h. Four-to-five week old cut leaf disks of
172
wild-type Arabidopsis plants (Col-0) were elicited using 100nM flg22. ROS production
was monitored for 40 minutes using a luminol based assay and is shown as relative light
units (RLU) (Figure 1A, 10-12 minutes post-elicitation). At 45-50 minutes following
treatment, tissue was either washed twice with dH2O (pulse treatment; P), or retained in
PAMP-containing solution (continuous treatment; C). For re-elicitation, both pulsed and
continuously treated tissue was exposed to fresh PAMP-elicitation solution at indicated
times. Lack of significant ROS production for pulsed and continuously treated tissues re-
elicited at 1h correlated with previous findings that FLS2 protein is degraded at this time
(Figure 1A; P, C, 1h) (191) (Salamango et al, in prep.). Recovery of cellular signaling
competency after re-elicitation of pulsed and continuously treated tissues at 3h using
100nM flg22 correlates with our previous findings that ligand-induced de novo synthesis
of FLS2 protein accumulates in a signaling competent state (Figure 1A; P, C, 3h)
(Salamango et al, in prep.). Interestingly, continuously treated cut leaf tissue re-elicited at
16h using 100nM flg22 exhibited almost no ROS production, while pulsed tissue
demonstrated nearly full recovery of signal amplitude (Figure 1A; P, C, 16h). Lack of
cellular signaling competency at 16h for continuously treated tissues suggests FLS2 protein
may be degraded, or in a signaling-incompetent conformation. To assess FLS2 protein
accumulation in continuously treated tissues, I used immunoblot analysis on cut Col0 4-to-
5 week old leaf strips to determine FLS2 protein accumulation following treatment using
100nM flg22. As depicted in Figure 1B, Col0 cut leaf strips continuously treated (C) with
flg22 (+) show almost no FLS2 protein accumulation 8h following flg22 treatment
compared to pulse-treated samples. This preliminary experiment correlates the lack of
cellular responsiveness towards flg22-peptide with decreased FLS2 protein accumulation.
173
To determine if lack of cellular signaling competency at 16h was specific to
FLS2-flg22 signaling pathways, I continuously treated tissue using elf26, an independent
PAMP-ligand for the cell-surface receptor EFR (193). Following the treatment procedure
described for flg22, I initially elicited cut Col0 leaf disks with 100nM elf26 and re-elicited
tissue at 1, 3, and 16h with the same stimulus (Figure 1B). Re-elicitation of pulsed and
continuously treated leaf tissue at 1h using elf26 resulted in low amounts of ROS
production, similar to levels previously described for flg22 re-elicitation at 1h. This may
suggest EFR receptor undergoes ligand-induced degradation in a similar fashion as
previously described for FLS2 (191) (Salamango et al, in prep.). However, we also cannot
rule out the possibility that EFR is sequestered to the plasma membrane in a signaling
incompetent state. To verify decreased ROS production following re-elicitation of
continuously stimulated tissue was specific to the PAMP-receptor combination, we
continuously elicited Col0 leaf tissue with independent ligands AF (an inactive flg22
peptide derivative), flg22, or elf26 and assessed FLS2 signaling competency 16h following
initial treatment. As shown in Figure 1C, continuous elicitation with a non-specific peptide
had no effect on FLS2 signaling competency following re-elicitation at 16h. Tissues
continuously treated with elf26 were re-elicited using flg22 and, although ROS production
was slightly decreased compared to initial treatment (0h), FLS2 signaling competency was
significantly higher compared to tissues continuously treated with flg22, indicating
desensitization is specific to the ligand-receptor combination. This suggests decreased
FLS2 signaling competency in elf26 continuously treated tissue could be the result of
depletion of common signaling component(s), and/or signaling mechanisms, utilized
between EFR and FLS2
174
Ligand-desensitization during continuous treatment occurs after 6h following
initial PAMP-treatment. Examination of cellular responsiveness following PAMP-
treatment was assessed at time points between 3 and 16h following initial treatment to
determine when ligand-desensitization began to occur. Pulsed and continuously treated
tissue was re-elicited using 100nM flg22 every 2h spanning 4-to-14h after initial treatment
at 0h with 100nM flg22. I consistently observed ligand-desensitization following re-
elicitation of continuously treated tissue after 6h compared to pulsed samples (Figure 2A).
To ensure continuously treated cells had not undergone cell death and were capable of
responding to stimuli in general, I elicited 14h samples using 100nM elf26 following 14h
flg22 re-elicitation. I observed an increased ROS production in both pulsed and
continuously treated tissues following elicitation with 100nM elf26 at 14.5h compared to
elicitations at 14h using 100nM flg22 (Figure 3A; 14.5h) indicating continuously treated
cells were able to perceive PAMP-ligands other than flg22. Tissue continuously treated
using elf26 demonstrated a similar re-elicitation profile compared to flg22 re-elicitations
(Figure 2B), indicating observed results may be due to a common desensitization
mechanism. This strategy may be employed to negatively regulate immune signaling
pathways in situations wherein continuous signaling is occurring, effectively averting
depletion of crucial nutrients required for growth and development.
Continuous Avirulent Bacterial Treatment Leads to Ligand-Desensitization.
Next, I wanted to determine if ligand-induced desensitization could be recapitulated
utilizing the bacterial strain Pseudomonas syringae DC3000 hrcC- as a stimulus in lieu of
flg22 and elf26 peptides. Pto DC3000 hrcC- mutant is deficient in its type III secretion
system (T3SS). This bacterial strain was used specifically because it lacks a T3SS and is
175
unable to deliver bacterial effectors; therefore, pathogen recognition exclusively generates
a PAMP-triggered immune response in the absence of bacterial effectors known to target
PAMP-recognition receptors including FLS2 and EFR (226). Col0 cut leaf disks were
initially incubated with 1x108 CFU Pto DC3000 hcrC- for 1h and ROS production was
monitored. At 65-70 minutes following bacterial application, tissue was either washed
twice with dH2O (pulse), or left in elicitation solution containing bacteria (continuous) until
the indicated time of re-elicitation. For re-elicitation, pulsed and continuously treated
tissues were re-elicited with 100nM flg22 and ROS production was monitored. As shown
in Figure 3A, initial ROS amplitude produced following incubation with Pto DC3000 hrcC-
at 0h was lower compared to previous elicitations performed using 100nM flg22 peptide.
For the mammalian bacterial pathogen Caulobacter cresentus, the flagellin-epitope is not
exposed to the external environment until after the bacterial flagellum is shed (227).
Therefore, low ROS amplitude could be attributed to low levels of available flg22-epitope
in my experimental setup. Flg22 re-elicitation of pulsed and continuously treated tissue
over time generated a similar signaling profile as described previously. At 8h following
initial treatment, continuously treated samples began to display decreased ROS production
compared to time-matched pulsed samples. Ten-hours following initial bacterial
incubation, the ROS profile of continuously treated tissue resembled ligand desensitization
previously described for flg22 and elf26 treatments in that it was significantly decreased
compared to pulse treated samples. To determine if the desensitization to flg22-peptide at
8h was due to lack of FLS2 protein accumulation, I continuously treated Col0 cut leaf strips
with 1x108 CFU Pto DC3000 hrcC- and collected tissue for immunoblot analysis at the
indicated times (Figure 3B). In my preliminary experiments, samples continuously treated
176
with (+) Pto DC3000 hcrC- for 1h displayed significant FLS2 protein degradation
compared to samples at 0h. Analysis of samples collected at 4 and 6h following initial
bacterial treatment indicated FLS2 protein accumulated to similar levels compared to
untreated tissue, while continuously treated tissue displayed significantly decreased FLS2
protein accumulation 8h following initial treatment (Figure 3A; C, 8h). These results
indicate flg22-desensitization occurring in tissues continuously incubated in the presence
of the avirulent bacteria DC3000 hrcC- correlate to lack of FLS2 protein accumulation.
Syringe-Infiltration Does Not Recapitulate Ligand-Desensitization
Demonstrated for Continuous Exogenous PAMP-Treatment. A common strategy used
to assess PAMP-triggered immunity is to inoculate plant tissue via syringe- or vacuum-
infiltration of PAMP-containing solution directly into the apoplastic environment, during
which the ligand is assumed to be present continuously for the duration of the experiment.
I wanted to compare ROS re-elicitation and FLS2 protein accumulation using flg22-
infiltration with my previously used methods in which flg22-containing solution is added
to leaf strips either as a pulse for 1h (P) or continuously for 24h (C) to determine if ligand-
desensitization could be recapitulated in syringe-infiltrated tissues. Col0 4 to 5 week old
leaf tissue was cut into leaf disks and leaf disk halves were floated on 150μL dH2O for 24h
to allow for wounding responses to subside. After 24h, the cut leaf disks were elicited using
1μM flg22 while simultaneously, Col0 4 to 5 week old leaves were syringe-infiltrated with
the same 1μM flg22 containing solution (IPF), or with dH2O (IH2O). After 1h of flg22-
treatment, Col0 cut leaf disks were either washed twice with dH2O, or retained in PAMP
solution. At the same time, infiltrated leaf tissue was cut into leaf disk halves and floated
on 150μL dH2O in the same ROS plate containing the elicited cut leaf disks. Because all
177
tissues perceived flg22-peptide at the same time, all were re-elicited using 1μM flg22 24 h
after initial treatment. As shown in Figure 4A, ROS production following re-elicitation
using 1μM flg22 at 24h for cut leaf disks syringe-infiltrated with either 1μM flg22 or with
dH2O showed no significant difference compared to pulse treated Col0 cut leaf disks
elicited at 0h . However, re-elicitation of continuously treated Col0 cut leaf disks using
1μM flg22 at 24h displayed significantly decreased ROS production compared to pulse
treated and syringe-infiltrated cut leaf disks. To ensure syringe-infiltrated flg22 was
perceived and capable of generating PAMP-triggered immune responses in infiltrated Col0
leaf tissue, a sampling of the syringe-infiltrated tissues used for the ROS experiment
depicted in Figure 4A were collected and analyzed for callose deposition 24h after flg22-
infiltration (i.e. at the same time of the 24h re-elicitation time point shown for ROS
production). Leaf tissues syringe-infiltrated with 1μM flg22 displayed strong callose
deposition 24h following infiltration (at the same time of ROS re-elicitation), whereas
tissue infiltrated with dH2O exhibited almost no callose deposition (Fig S1), indicating
flg22-ligand was perceived in flg22-infiltrated leaf tissue. To determine if signaling
competency at 24h correlated with FLS2 protein accumulation, I compared pulsed or
continuously treated tissue against syringe-infiltrated tissue 24h following initial treatment.
In preliminary experiments, immunoblot analysis revealed that FLS2 protein accumulated
to almost pre-elicitation levels (0h; - PF) for both pulsed leaf strips and syringe-infiltrated
leaves but not in continuously treated leaf strips. These results suggest that desensitization
mechanism responsible for loss of ROS signaling competency is specific to continuous
exogenous PAMP-treatment.
178
III.V Conclusions
Previously, we developed re-elicitation techniques utilizing flg22-peptide to
correlate cellular signaling competency with FLS2 protein accumulation. We demonstrated
that treatment with concentrations of flg22 ranging from 0.1-10μM resulted in ligand-
induced degradation of FLS2 protein 1h following treatment. We were able to correlate
degradation of FLS2 protein with the loss of cellular signaling-competency in response to
flg22 peptide during this period. Furthermore, were could demonstrate that accumulation
of ligand-induced new synthesis of FLS2 protein resulted in restoration of cellular-
signaling competency following flg22-peptide treatment over time (Salamango et al, in
prep.). Here, I build on those methods and probe the effects of continuous PAMP-
stimulation on signaling competency and receptor protein accumulation over extended
periods of treatment to simulate the persistence of bacterial infection. Continuous PAMP-
stimulation resulted in ligand-desensitization at time points after 6h following initial
PAMP-treatment at 0h, which in preliminary experiments correlated with lack of
accumulation of FLS2 receptor. Utilizing two independent PAMP-ligands, namely, flg22
and elf26, which recognize different PRRs, I demonstrated that ligand-induced
desensitization mechanisms may be utilized by multiple PAMP-receptors to avert
continuous signaling. This desensitization mechanism could be used as a means to conserve
vital cellular signaling components used for both innate immune signaling and
developmental pathways. I also show that Col0 leaf tissue treated with 100nM elf26 at 0h
displays a lack of cellular signaling competency following re-elicitation using 100nM elf26
at 1h for ROS production. It is unknown whether EFR undergoes ligand-induced endocytic
trafficking and degradation as described for FLS2. It is possible that cellular desensitization
179
to elf26-ligand at 1h parallels a similar mechanism described for FLS2 trafficking wherein
activated receptor undergoes endocytic degradation. However, I cannot rule out the
possibility that EFR is sequestered to the plasma membrane in a signaling-incompetent
conformation. Interestingly, continuous bacterial treatment of Col0 leaf strips recapitulated
PAMP-peptide ligand-induced desensitization (Figure 3A) as well as the correlation to the
level of receptor accumulation (Figure 3B). To further explore PAMP-induced continuous
signaling and ligand-desensitization, I compared syringe-infiltration to continuous
exogenous PAMP-treatment. Syringe-infiltration did not demonstrate ligand-
desensitization or lack of FLS2 protein accumulation at 24h (Figure 4A and 4B) compared
to continuous exogenous PAMP treatment (Figure 4A and 4B). It may be that ligand-
induced desensitization observed in tissues continuously treated with exogenous PAMP
arises from the quantity of flg22 peptide present in the elicitation solution. Syringe-
infiltration allows for only a finite amount of flg22-containing solution to be introduced
into the apoplastic space of the leaf tissue; however, the Col0 cut leaf disks used in these
experiments are floated on a much greater volume of elicitation solution containing flg22-
peptide. Therefore, even though the concentration of flg22-peptide is held constant for both
experimental approaches, the molar amount of flg22-peptide present in continuous
exogenous treatment is much greater, which may result in the ligand-desensitization
described here.
III.VI Future Directions
An interesting question that arises from these data is what is the molecular
mechanism responsible for FLS2 degradation at 8h following continuous stimulation using
180
flg22? Two very plausible and testable hypotheses are that: 1) FLS2 mRNA is degraded,
possibly through small interfering RNA (228), after extended periods of continuous
stimulation, and 2) FLS2 is post-translationally targeted to degradative organelles in a
manner that it prevents de novo synthesis to accumulate FLS2 protein that is in a signaling
competent state.
Decreased FLS2 mRNA would eventually lead to decreased FLS2 protein
accumulation because of ligand-independent trafficking and degradation of FLS2
associated with generalized plasma membrane turnover. A straightforward method to test
this hypothesis would be to pulse and continuously elicit Col0 leaf tissue using 100nM
flg22. At 0, 1, 3, 6, 9, and 12h post-treatment leaf disks would be collected for both qPCR
and immunoblot analysis. This would ensure for a direct comparison between FLS2 protein
levels and transcript accumulation of FLS2 mRNA. Although this would not answer if
siRNA was involved per se, it would give weight to the idea that there is an active
mechanism down-regulating activated receptor, rather than a potential artifact of the
treatment. It would also be interesting to screen Arabidopsis mutants such as argonaut-1
and -2 and dicer-like-1 since both are known contributors to siRNA mediated pathways
and argonaut-2 has been implicated in regulating innate immune signaling pathways (229,
230).
Unfortunately, it may be more difficult to directly determine if FLS2 is being
actively degraded following continuous stimulation. We, and others, have shown that
ligand-induced degradation of FLS2 is extremely sensitive to pharmacological inhibitors
(191, 221). For instance, co-treatment of flg22 with 50μM cyclohexamide, a protein
181
translational inhibitor, results in the disruption of FLS2 degradation through unknown
mechanisms (Salamango et al, in prep.).
182
Figure III.I: Continuous PAMP Treatment Desensitizes Cells to the Different
PAMPs.
(A) and (C) Continuous PAMP treatment results in desensitization to the stimuli. For the
first elicitation (0h), Col0 4 to 5 week old cut leaf disks were elicited with either 100nM
flg22 (open bar) or 100nM elf26 (grey bar with horizontal stripes) and ROS production
was monitored using a luminol based assay. For subsequent elicitations (filled black bars),
samples were either washed with dH2O after 40 min of treatment ( P; pulse) or
continuously elicited with PAMP ( C; continuous). At the indicated time point, both sample
treatments were re-elicited with the same fresh aliquot of the respective 100nM PAMP
solution (1h, 3h, and 16h). (experimental, n>3; samples, n=16) (B) ROS desensitization
correlates with a loss of FLS2 protein accumulation. Col0 4 to 5 week old cut leaf strips
were elicited (+) with 1μM flg22, or without ( (-); dH2O) and either washed after 40
minutes using dH2O, or left in the presence of flg22 to simulate continuous treatment.
Samples were flash frozen at 8h post treatment and total proteins were isolated and
subjected to immunoblot analysis. Blots were probed with αFLS2 to assess FLS2 protein
accumulation, or αMPK6 to assess equal loading. h, hours; PF, flg22; P, pulse; C,
continuous, RLU; relative light units. (experimental, n=1). (D) Desensitization is specific
to the PAMP used. For the first elicitation (0h), Col0 4 to 5 week old cut leaf disks were
elicited with 100nM AF (inactive flg22 peptide derivative), PF (active flg22 peptide), or E
(elf26 peptide) and ROS production was monitored. Samples were continuously elicited
with indicated PAMP-solution (cont. elicitation). For re-elicitation, all samples were
exposed to 100nM flg22 at 16 hours post-initial treatment. Open bars represent flg22
treatment, grey bars with horizontal strips represent elf26 treatment, diagonal stripes
183
represent AF treatment, and filled black bars represent re-elicitation with flg22. ROS peaks
(10-12 minutes post-elicitation) are shown for all indicated measurements. The respective
ROS measurements within the individual figure panels were performed on the same 96-
well ROS plate for direct comparison. h, hours; AF, inactive peptide; PF, flg22; E, elf26,
RLU; relative light units, P; pulse, C; continuous. (experimental, n=3; samples, n=32)
184
Figure III.I: Continuous PAMP Treatment Desensitizes Cells to the Different
PAMPs.
185
Figure III.II: PAMP-Induced ROS Desensitization Occurs After Six Hours Following
Initial PAMP Treatment in Continuously Elicited Tissue.
(A) and (B) ROS desensitization in continuously elicited samples occurs after six hours
following initial PAMP treatment. For the first elicitation (0h), Col0 4 to 5 week old cut
leaf disks were elicited with either 100nM flg22 (open bar) or 100nM elf26 (grey bar with
horizontal stripes) and ROS production was monitored using a luminol based assay. For
subsequent elicitations (filled black bars), samples were either washed with dH2O after 40
min of treatment ( P; pulse) or continuously elicited with indicated PAMP ( C; continuous).
At the indicated time point, both sample treatments were re-elicited with fresh aliquots of
the same100nM PAMP used for the initial elicitation (4h, 6h, 8h, 10h, 12h and 14h). In
(A), the 14h flg22-treated samples were re-elicited using 100nM elf26 at 14.5h post-initial
elicitation to demonstrate samples were able to generate ROS in response to elicitation
using an independent PAMP ligand. ROS peaks (10-12 minutes post-elicitation) are shown
for all indicated measurements. (experimental, n=3; samples, n=8).
186
Figure III.II: PAMP-Induced ROS Desensitization Occurs After Six Hours Following
Initial PAMP Treatment in Continuously Elicited Tissue.
187
Figure III.III: Elicitation using the avirulent bacterial strain Pseudomonas syringae
DC3000 hrcC- Results in a Similar Desensitization Pattern as Shown for PAMP-
peptides.
(A) ROS production displays desensitization following continuous treatment using P.
syringae DC3000 hrcC-. For the first elicitation (0h), Col0 4 to 5 week old cut leaf disks
were elicited using 1x108 CFU P. syringae DC3000 HrcC-, and ROS production was
monitored using a luminol based assay (at peak, 30-40 minutes following addition of
bacterial solution) . For subsequent elicitations (filled black bars), samples were either
washed with dH2O after 40 min of treatment ( P; pulse) or continuously elicited with P.
syringae DC3000 HrcC- ( C; continuous). At the indicated time point, both sample
treatments were re-elicited with the same 100nM flg22 solution ( 4h, 6h, 8h, and 10h).
(experimental, n=3; samples, n=12) (B) FLS2 protein accumulation correlates with ROS
desensitization. Col0 4 to 5 week old cut leaf strips were elicited (+) with 1x108 CFU
HrcC-, or without ( (-); dH2O) and either washed after 40 minutes using dH2O, or left in
the presence of P. syringae DC3000 HrcC- to simulate continuous treatment. Samples were
flash frozen at the indicated time post treatment. Total proteins were isolated and subjected
to immunoblot analysis. Blots were probed with αFLS2 to assess receptor accumulation,
or αMPK6 to assess equal loading. RLU; relative light units, h; hour, PF; flg22, HrcC-; P.
syringae DC3000 HrcC-, P; pulse, C; continuous. (experimental, n=1)
188
Figure III.III: Elicitation using the avirulent bacterial strain Pseudomonas syringae
DC3000 hrcC- Results in a Similar Desensitization Pattern as Shown for PAMP-
peptides.
189
Figure III.IV: Syringe Infiltration Does Not Simulate Exogenous Continuous
Elicitation.
(A) ROS desensitization does not occur in leaf tissue syringe infiltrated with flg22.
Representative tissue from Col0 4 to 5 week old leaves were cut into leaf disks and floated
on dH2O overnight. After ~24h, the leaf disks were elicited with 1μM flg22 and monitored
for ROS production (P; 0h, and C; 0h). Simultaneously at the same time of elicitation for
ROS production, Col0 leaf tissue was syringe infiltrated with the same 1μM flg22
elicitation solution used to generate ROS. After 40 minutes, a subset of the leaf disks used
in the ROS assay were washed with dH2O (P; pulse), or left in elicitation solution (C;
continuous). At 1h post-infiltration, once the leaf tissue had dried, leaf disks were removed
from the center of the infiltrated leaf tissue and cut in half and placed into the ROS plate
containing the 0h cut leaf disk samples. Twenty-four hours later, all tissue samples were
re-elicited with 1μM flg22 and monitored for ROS production (P; 24h, C; 24h, IPF; 24h,
and IH2O; 24h). (experimental, n=3; samples, n=24) (B) FLS2 protein accumulation is not
reduced at late time points in tissues syringe infiltrated with flg22. During each treatment
described in (A), a subset of tissue was flash frozen for immunoblot analysis. Total proteins
from those samples were isolated and probed with αFLS2 to assess protein accumulation
24h after initial exposure to its cognate ligand, and αMPK6 to determine equal loading.
RLU; relative light units, h; hour, P; pulse, C; continuous, PF, flg22 (experimental, n=1)
190
Figure III.IV: Syringe Infiltration Does Not Simulate Exogenous Continuous
Elicitation.
191
Figure III.S1: Confirmation Syringe Infiltration of flg22 Elicited an Immune
Response.
(A) Callose deposition from samples obtained during the experimental approach described
in Figure 4. Samples were taken from the same syringe infiltrated plants used for ROS
production to confirm that the infiltration using 1μM flg22 induced an immune response.
At about 24h post-infiltration, leaf disks from leaves infiltrated with flg22 or with water
were processed to visualize callose deposition using analine blue stain. (experimental, n=2)
192
Figure III.S1 Confirmation Syringe Infiltration of flg22 Elicited an Immune
Response.
193
IV. IDENTIFYING ARABIDOPSIS EPSINR1 AS A REGULATOR OF FLG22-
INDUCED SIGNALING RESPONSES
IV.I Summary
Cell surface receptors are the front-line sentries which contribute to defending
against microbial pathogen invasion and initiate innate immunity in both plants and
animals. In plants, perception of conserve features harbored on whole classes of microbial
pathogens activates pathogen-associated molecular pattern (PAMP)-triggered immunity
(PTI). In the best studied plant PTI-system, perception of bacterial flagellin (or its active
peptide derivative flg22) by the host cell surface receptor flagellin sensitive 2 (FLS2) leads
to activation of immune-signaling pathways contributing to restriction of pathogen growth.
Previously in collaboration with Scott Peck (Univ. of Missouri-Columbia), the Heese Lab
identified EpsinR1 as novel protein phosphorylated in response to flg22. Here, I provide
evidence that EpsinR1 has a role(s) as a positive regulator of early, intermediate, and late
flg22-induced parallel signaling responses using two independent epsinR1 mutant alleles.
Furthermore, I identified two morphological phenotypes in epsR1 alleles indicating that in
addition to its function in flg22-signaling, EpsinR1 has roles as a general cellular signaling
and vesicular trafficking component.
194
IV.II Introduction
In plants, innate immune signaling is the primary defense against invading
microbes. Unlike mammalian organisms, plants do not possess a known acquired immune
system and therefore rely on perception of microbes through recognition of conserved
pathogen-associated molecular patterns (PAMPs) or pathogen-effector proteins to initiate
immunity. Activation of PAMP-triggered immunity occurs through cell surface receptors
which recognize molecular structures conserved in whole classes of microbial pathogens.
Microbial recognition mediated by these PAMP-recognition receptors initiates a plethora
of downstream immune signaling cascades which cause cessation of microbial infection
(181, 219, 231). In the best studied plant PAMP-receptor interaction, a conserved region
of the bacterial flagellum is perceived by the extracellular leucine-rich repeats of the
PAMP-recognition receptor Flagellin Sensitive 2 (FLS2) (185, 214). Perception of flagellin
(or its 22-amino acid peptide derivative flg22) by FLS2 leads to activation of several
parallel immune signaling pathways which contribute to restriction of pathogen growth
(185, 220, 232, 233). Within seconds after ligand perception, FLS2 rapidly hetero-
dimerizes with another receptor-like kinase BAK1 (Brassinosteroid insensitive-1
associated kinase) (194, 204, 234). In the absence of BAK1, all early and late flg22-induced
signaling responses are strongly impaired (194, 204). BAK1 has also been shown to be
required for signaling of additional cell-surface receptors including EF-Tu (or its peptide
derivative elf26) (185, 193, 215), and the brassinosteroid receptor BRI (235, 236). Recent
reports have established that FLS2 undergoes ligand-induced endocytosis through early
195
and late endosomes to a degradative compartment, most likely the vacuole (191, 221).
Endocytic vesicle formation requires recruitment of specific coat and accessory proteins to
initiate membrane invagination and facilitate vesicle scission from the target membrane.
Mammalian and yeast Epsins function at two distinct subcellular locations, the plasma
membrane and trans-golgi network (TGN). Plasma membrane localized Epsins participate
in cargo selection and recruitment of vesicle coat-forming components such as clathrin and
AP-complexes to initiate the internalization step of endocytosis (237-239). Epsins
localized to the TGN initiate clathrin-coated vesicle budding and trafficking of cargo to the
lysosome/vacuole for degradation. Epsin proteins utilize the epsin N-terminal homology
(238) domain for membrane targeting and induction of membrane curvature (237-239).
Binding of PIP(4,5)2 to the ENTH domain stimulates folding of an N-terminal amphipathic
helix, which intercalates into the target membrane to induce invagination and recruitment
of vesicle forming components (237, 240).
In Arabidopsis, three genes encode for Epsin-related proteins (AtEpsinR1, R2, and
R3), which is comparative to yeast and most vertebrate organisms which contain at least 2
epsin paralogs (241, 242). In plants however, there is a paucity of knowledge regarding the
role of Epsin proteins in regulating temporal and spatial recruitment of components
required for clathrin-coated vesicle formation and subsequent trafficking. Arabidopsis
EpsinR1 has been shown to localize to the TGN/PVC and is suggested to have a role in
soluble protein trafficking to the vacuole (243). In collaboration with Scott Peck (Univ. of
Missouri-Columbia), our lab had previously identified EpsinR1 in a large scale phospho-
proteomics screen as a novel vesicular trafficking component phosphorylated in response
196
to flg22-treatment. Prior to my starting to work on the EpsinR1 project, the Heese lab had
isolated two independent mutant alleles containing T-DNA insertions in EpsinR1 in the
Arabidopsis thaliana ecotype Columbia (Col0) background. More specifically, epsR1-2 is
a knock-out, null mutant line (SAIL_394G02, insertion in the fourth intron of the gene)
and epsR1-1 is a knock-down mutant line (SALK_049204, insertion in the EPSINR1
promoter region). Two different peptide-specific antibodies were produced and used to
show that epsR1-2 mutant plants did not accumulate any full-length EpsinR1 proteins
whereas epsR1-1 accumulated reduced levels of full length EpsinR1 protein.
Here, I utilize these two independent mutant alleles to investigate a possible role of
EpsinR1 in flg22-responses. I provide evidence that epsinR1 knock-out and knock-down
mutant alleles show reduced flg22-elicited immune signaling responses. Furthermore, I
identified defective cellular elongation, expansion, and stomatal patterning in epsinR1
alleles.
IV.III Experimental Procedures
Immunoblot Analysis and Antibodies. For sample elicitation, 8-to-10 seedlings which
were 8-days old were floated on 1 mL dH2O overnight in continuous light at 22◦C to reduce
wounding response. The following day, samples were elicited with indicated
concentrations of specified PAMP for 45-50 min, then flash frozen in liquid nitrogen at
indicated times. Sample preparation and immunoblot analyses of total proteins using
antibody concentrations were done as described (204). An exception is that αP-p44/42
MAPK (1:3000) (#4370; Cell Signaling Tech, Danvers, MA) was used to detect
phosphorylated MAPKs. For assessment of EpsinR1 protein accumulation primary
antibodies αEpsin1 #131 were used at a concentration of 1:1000. For PR1 protein
197
accumulation, a polyclonal rabbit antibody raised against a synthetic peptide
(QDSPQDYLRVHNQARC) of the amino terminus of the Arabidopsis PR1 protein (kindly
provided by Xinnian Dong, Duke Univ.) was used at a concentration of 1:1000.
Plant Material and Developmental Measurements. Arabidopsis seedling and plant
growth was at 22◦C as described (204), except that plants were grown in an 8-h light/16-h
dark cycle. epsR1-1, epsR1-2, and bak1-4 are in Col0 background. For primary root length
measurements, seedlings were imbibed in dH2O for three days prior to plating on 0.5% MS
1% sucrose plates. Seedlings were grown horizontally for the indicated number of days in
continuous light. To directly compare developmental measurements, plant genotypes
indicated in respective figure panels were grown on the same plates to eliminate the
possibility of nutrient inconsistencies. 8-day old seedlings were gently removed from MS-
plates, and primary roots were aligned vertically and measured using a ruler. Seedlings
used for primary epidermal root cell length, cotyledon pavement cell expansion, stomata
number and clustering, and cell-type quantification were grown as indicated above except
whole seedlings were stained with Toluidine blue. Overall primary root length, primary
epidermal root cell length, stomata number and clustering, and cellular pools were
quantified using ImageJ and plotted using Graph Pad Prism4 software. Cotyledon
pavement cell expansion was quantified using both Image J and Graph Pad Prism4
software. Morphological phenotypes were imaged using an Olympus vanox upright
microscope with DIC. PAMP-induced callose measurements were imaged using a Leica
stereomicroscope with a ultra-violet filter.
198
Toluidine Blue O (TBO) Staining. Whole seedlings were fixed in 9:1 ethanol:acetic acid
solution at room temperature, in the dark, overnight. After ~24 hours, tissue was washed
in consecutive ethanol at 85% and 70% concentration for 30 minutes each. Ethanol was
replaced with TBO (0.5% w/v TBO; Sigma #198161 in 50% ethanol) stain for 3 minutes.
Stained tissue was washed several times with 30% ethanol solution and mounted in 15%
glycerol solution.
Apoplastic ROS Production. For seedling elicitations, cotyledons were cut into two,
roughly equal, pieces and placed into one well of a 96-well plate. For leaf tissue elicitations,
half of a cut leaf disk (1.1cm2) of 4-5 week old plants was placed into one well of a 96-
well plate. Samples were assayed for individual ROS production measurements using a
luminol-based assay (204). Although absolute values of ROS production [displayed as
Relative Light Units; RLUs] varied from experiment to experiment, the actual trends were
always the same.
Callose Deposition. 8-day old seedlings were removed from 0.5 MS 1% sucrose plates
and floated in dH2O overnight to reduce wounding responses. After 24 hours, dH2O was
exchanged with dH2O containing the appropriate concentration of indicated PAMP.
Seedlings were incubated in PAMP solution for 24 hours before being placed directly into
95% ethanol overnight at room temperature. For syringe infiltration, adult leaf tissue of 4
to 5 weeks plants grown in soil were syringe infiltrated using 10nM flg22, 10nM AF
(inactive flg22 peptide derivative), or with dH2O, and collected directly into 95% ethanol
solution 24 hours post-infiltration. After the 95% ethanol incubation, tissue was rinsed in
50% ethanol and incubated in 67 mM potassium phosphate (K2HPO4; pH 12) buffer for 1
199
hour. Tissue was stained using 0.01% analine blue/potassium phosphate buffer solution for
1h in the dark at room temperature. Stained tissue was mounted in 70% glycerol, 30%
67mM K2HPO4 solution.
Quantitative Real-time PCR Analysis. For sample elicitation, 8-to-10 seedlings which
were 8-days old were floated on 1 mL dH2O overnight at 22◦C in continuous light to reduce
wounding response. Samples were elicited with 10nM flg22 for 1 or 4hours, then flash
frozen in liquid nitrogen. Total RNA was isolated and real-time PCR reactions were
performed and analyzed as described in (206) using the following primers WRKY29-f 5’-
AAGGATCTCCATACCCAAGGAGT-3’ and WRKY29-r 5’-
TCGACTTGTTTTCTTGCCAAACAC -3’ and FRK1-f 5’-
ATCTTCGCTTGGAGCTTCTC -3’ and FRK1-r 5’-TGCAGCGCAAGGACTAGAG -3’
primers using the expression of At2G28390 (SAND family protein) SAND-f 5’-
AACTCTATGCAGCATTTGATCCACT-3’ and SAND-r 5’-
TGATTGCATATCTTTATCGCCATC-3’ to normalize all qRT-PCR results (206).
Statistical Analysis. Statistical significances based on unpaired two sample t-test were
determined with Graph Pad Prism4 software.
IV.IV Results
EpsinR1 is a Positive Regulator of Early-Induced flg22-Signaling Responses.
Before testing whether EpsinR1 was required for flg22-induced signaling responses I first
needed to confirm that in my hands both epsR1 mutant lines were impaired in EpsinR1
protein accumulation. Using immunoblot analysis, 8-day old seedling total proteins were
200
analyzed to determine the amount of EpsinR1 protein accumulation. No full length
EpsinR1 protein was detectable in epsR1-2 seedlings, and protein accumulation was greatly
reduced in epsR1-1 seedlings, compared to Col0 (Figure 1A). These results are consistent
with a previous report which determined EpsinR1 protein accumulation in epsR1-1 mature
leaf tissue to be roughly ~10-20% of wild-type EpsinR1 levels (243).
To delineate the role(s) of EpsinR1 in flg22-induced signaling pathways, we
examined well characterized early, intermediate, and late responses in epsR1-1 and epsR1-
2 alleles. Early (~10 minutes post-elicitation) flg22-induced signaling responses were
assessed by monitoring reactive oxygen species (ROS) production and mitogen-activated
protein kinase (MAPK) phosphorylation. Using a luminol-based assay, I monitored ROS
production for 40 minutes following 10nM flg22-treatment of 8-day old cut cotyledons. As
depicted in Figure 1B, peak ROS production (10 minutes post-elicitation) was significantly
decreased in both epsR1 alleles. A similar trend was confirmed in 4 to 5 week old leaf
tissue (Fig S1A). Elicitation using dosages of flg22 and elf26 peptides (an independent
ligand for the PAMP receptor EFR) as low as 1nM on 8-day old cut epsR1-2 cotyledons
indicated significantly decreased ROS production compared to Col0 (Fig S1B and S1C).
Phosphorylation of MAPKs 6 and 3 in epsR1 alleles was also decreased following flg22-
treatment. We examined MAPK phosphorylation in response to flg22 elicitation using a
αP-p44/42 MAPK antibody, which detects only phosphorylated forms of MPK 6 and 3.
PAMP-induced MAPK phosphorylation is transient, with peak phosphorylation occurring
about 10 minutes post-elicitation, and de-phosphorylation occurring from 30-60 minutes
post-elicitation (232). Immunoblot analysis using total proteins isolated from 8-day old
201
seedlings indicated decreased MAPK phosphorylation in both epsR1 alleles 10 minutes
following flg22-treatment (Figure 1C, lanes 2, 4, and 6; P-MPK panel). Preliminary results
indicate this was not due to a shift in peak MAPK-phosphorylation. As indicated by Fig
S2A, flg22-treated Col0 seedlings displayed peak MAPK-phosphorylation 10-15 minutes
following treatment, with significant de-phosphorylation occurring by 30 minutes.
However in epsR1-2 seedlings, no well-defined peak occurred following flg22-treatment,
rather, it appeared there was only a slight increase in flg22-stimulated MAPK-
phosphorylation. Additionally, we determined that the lack of strong MAPK-
phosphorylation was not due to decreased MPK6 protein accumulation by using an anti-
MPK6 antibody as a loading control (Figures 1C and S1A, MPK6 panels).
Role for EpsinR1 in Intermediate and Late-Induced flg22-Signaling
Responses. Further investigation revealed that downstream flg22-induced signaling
responses display decreased activation in both epsR1 alleles. Preliminary analysis of flg22-
induced gene induction using two independent downstream markers of MAPK-cascade
activation revealed that inducible expression was decreased in both mutant alleles
compared to wild-type (232, 244, 245). WRKY29 has been previously reported to have
induced mRNA expression ~1h following flg22-treatment, while FRK1 is reported to have
induced expression ranging from 2-4 hours following flg22-treatment (232). To maintain
consistency with our previous findings, we treated Col0, epsR1-1, and epsR1-2 8-day old
seedlings with 10nM flg22 to examine both WRKY29 and FRK1 mRNA transcript
accumulation. As shown in Figures 2A and 2B, both epsR1 mutant alleles demonstrated
highly significant decreases in WRKY29 and FRK1 transcript accumulation in response to
202
flg22-elicitation. These results are consistent with our previous findings in which MAPK
phosphorylation is significantly decreased in epsR1 mutant alleles following flg22-
treatment and suggests signal abrogation occurs upstream of MAPK 6 and 3 in the absence
of EpsinR1.
Preliminary results also indicate callose deposition and PR1 protein accumulation,
which are significantly induced ~24 hours following PAMP-perception in Col0, are
decreased in the absence of EpsinR1. These assays were performed through syringe-
infiltration of 100nM flg22 directly into the apoplastic space of 4 to 5 week old adult leaf
tissue. ROS results indicated flg22-induced signaling responses displayed decreased
amplitude in epsR1 adult leaves compared to Col0, recapitulating phenotypic results
obtained from experiments utilizing seedling tissues (Fig S1A). Callose is a β-1,3-glucan
polymer which is believed to serve as a matrix for antimicrobial deposition at the apoplast-
extracellular environment interface during pathogen infection (246). Using aniline blue
stain, which is visualized as punctate bodies, we detected significantly decreased callose
accumulation in epsR1-2 leaf tissue 24 hours following infiltration using 10nM flg22
compared to Col0 (Figure 2C). Controls infiltrated with 10 nM AF, an inactive flg22-
peptide derivative, displayed little-to-no callose deposition in either genotype (data not
shown). The plasma membrane callose synthase PMR4 is thought to be responsible for a
majority of callose synthesis and deposition in response to bacterial and fungal pathogens
in Arabidopsis (247-249). Although the mechanism of signal transduction is not fully
understood, callose deposition may be downstream of ROS production based on the fact
that rbohD mutants exhibit fewer callose deposits after flg22 treatment (250). Therefore,
203
decreased flg22-induced callose deposition is consistent with decreased ROS production
shown in Figure 1A.
SA-signaling leads to transcript accumulation of PR genes, and is considered to be
an important molecular signal for PTI and systemic resistance against pathogens. Mutants
deficient in SA-biosynthesis, or PR gene-product accumulation, display increased
susceptibility against bacterial infection (251-253). Considerable PR1 protein
accumulation occurs as early as ~24 hours following PAMP-perception (254), with
significant transcript accumulation occurring at ~12 hours (255). We wanted to determine
whether PR1 protein accumulation is impaired following flg22-treatment in epsR1 mutant
alleles leading up to 24 hours post flg22-treatment. Immunoblot analysis probing total
proteins from 4 to 5 week old epsR1-2 leaf tissues that were syringe-infiltrated with 100nM
flg22 detected decreased PR1 protein accumulation using a synthetic peptide raised against
a N-terminal portion of PR1 compared to Col0 (Figure 2D). We were able to detect slight
PR1 protein accumulation as early as 16 hours in Col0 leaf tissue, with increase PR1
accumulation at 24 hours post-infiltration. However, preliminary results using epsR1-2
infiltrated leaves indicated almost no PR1 protein accumulation until approximately 20
hours post-infiltration.
BAK1-Dependent, but not BAK1-Independent Signaling Mechanisms
Disrupted in epsR1. To determine if decreased flg22-induced signaling responses were
due to lack of FLS2 protein accumulation in epsR1 alleles, we assessed total FLS2 protein
accumulation using 8 day old seedlings. Immunoblot analysis using total proteins from 8-
204
day old seedling samples displayed similar levels of steady-state FLS2 protein
accumulation in epsR1-2 compared to Col0 (Figure 3A). These results indicate that
decreased flg22 signaling responses are not due to reduced, or lack of, FLS2 protein
accumulation in the absence of EpsinR1. To further understand the mechanism(s)
attributing to decreased flg22-induced signaling responses in epsR1 seedlings, I compared
signaling competency of independent bacterial and fungal PAMP receptors FLS2, EFR and
CERK1. EFR perceives the bacterial PAMP EF-Tu (or its peptide derivative elf26) and
signals primarily through BAK1-dependent mechanisms (193), whereas CERK1, which
recognizes the fungal PAMP chitin (256, 257), signals through BAK1-independent
mechanisms. Preliminary results shown in Figure 3B demonstrate decreased MAPK
phosphorylation using 8-day old epsR1-2 and bak1-4 seedlings following treatment with
10nM flg22, but not after treatment with 100μg/mL chitin, compared to 8-day old Col0
seedlings. As a positive control, bak1-4 null seedlings were included to compare relative
strength of PAMP-induced MAPK phosphorylation. To determine if this same trend
occurred in an independent signaling pathway, we performed experiments focused on
callose deposition. Examination of callose deposition using 8-day old cotyledons treated
with 10nM flg22, 10nM elf26, or 100μg/mL chitin revealed results consistent with MAPK-
phosphorylation (Figure 3C; controls treated with an inactive peptide show no deposition,
data not shown). Both flg22 and elf26-induced callose deposition were decreased in epsR1-
2 and bak1-4 compared to Col0; however, callose deposition in response to chitin
elicitation appeared increased compared to flg22- and elf26-treatments (Figure 3C). These
results suggest epsR1 mutants may be compromised in PAMP-induced-BAK1-dependent
signaling pathways, but not in BAK1-independent signaling pathways. Because BAK1
205
serves as the co-activator of both FLS2 and BRI1, a defect in BAK1 functionality would
result in decreased BR-induced-signaling subsequently inhibiting cellular elongation and
expansion, as well as changing stomatal patterning. Additionally, flg22- and elf26-induced
signaling pathways would display decreased activation following PAMP treatment.
Abrogated Cellular Elongation and Expansion Revealed in epsR1 Mutant
Alleles. To ensure tissue surface areas exposed to flg22-peptide were relatively similar
between epsR1 and Col0 seedlings, we examined various facets of growth and development
for gross morphological alterations. Initial observations revealed no significant differences
in cotyledon size and weight when comparing Col0, epsR1-1, and epsR1-2 8-day old
seedlings (Figure 4A, and data not shown). However, closer examination revealed epsR1
seedlings displayed significantly stunted primary root lengths when grown on horizontal
plates. Seedlings were imbibed in water for 3 days, plated on 0.5% MS 1% sucrose plates,
and grown horizontally for 8 days in 24 hour light. At 8-days post-germination, seedlings
were extracted and primary root lengths were compared. As shown in Figures 4B and 4C,
epsR1-1 and epsR1-2 seedlings displayed significantly reduced primary root lengths
compared to Col0 controls. Primary root length stunting in epsR1-1 seedlings was not as
severe as epsR1-2, most likely due to low levels of EpsinR1 protein accumulation.
Importantly, our flg22-induced signaling assays primarily utilized seedling cotyledons at a
developmental stage shown in Figure 4A. As shown, cotyledon surface area exposed to
flg22 peptide appeared similar between mutant and wild-type seedlings.
206
To gain a better understanding of how epsR1 seedlings developed over time, we performed
time course analyses focusing on root and aerial tissue development. Comparative analysis
between Col0, epsR1-2, and epsR1-1 seedlings at 5, 7, and 10 days post-germination
revealed primary root length stunting may be due to a defect in cellular elongation and
tissue expansion. To ensure results were not biased by variations in nutrient content of
plates used throughout the growth period, epsR1 and Col0 seedlings were grown on the
same plates and extracted at the indicated number of days post-germination. As depicted
in Figures 5A and 5B, analysis of primary epidermal root cell lengths over time indicated
that epsR1-2 and epsR1-1 root cells did not elongate as extensively as Col0 controls.
Interestingly, epsR1-2 and epsR1-1 primary epidermal root cells displayed differential
phenotypes at 5 days post-germination, which is most likely attributed to low levels of
EpsinR1 protein accumulation in the epsR1-1 mutant allele. At 7 days post-germination,
epsR1-1 primary root lengths displayed significant stunting compared to Col0. This growth
pattern was documented multiple times and may suggest the requirement for EpsinR1 is
necessitated during this growth period, and regardless of low level EpsinR1 accumulation,
cells fail to undergo appropriate cellular elongation. Loss of EpsinR1 may constrain
movement of specific components, or cargoes, required for adequate signal propagation
within these pathways; nevertheless, restricted root cell elongation would cumulatively
result in global primary root length stunting observed in Figures 4B and 4C.
To determine if similar developmental abnormalities occurred in aerial tissues,
seedlings used for primary epidermal root cell elongation measurements were analyzed for
pavement cell expansion by examining the cotyledon abaxial surface (Figures 5C and 5D).
207
My results indicate that over time, epsR1 cotyledon pavement cells do not expand as
extensively as Col0 pavement cells at the same developmental stage. Additionally,
cotyledon pavement cell expansion over time displayed an identical trend compared to
primary root elongation in both epsR1 alleles (Figures 5A and 5B). Although overall
cotyledon size is similar between 8-day old epsR1 and Col0 seedlings (Figure 4A), we
observed significant rosette stunting in adult epsR1 plants 4 to 5 weeks in soil, compared
to Col0 (Fig S2A).
Stomatal Patterning and Distribution is Altered in epsR1 Alleles. To further
explore developmental defects in epsR1 alleles, I examined stomatal patterning on the
abaxial surface of epsR1 8-day old cotyledons using toluidine blue stain. Interestingly, I
discovered increased number of absolute stomata and even identified stomatal clusters in
epsR1 seedlings, compared to Col0 (Figure 6A). As a comparison, I quantified the number
and classes of stomatal clusters observed for epsR1 seedlings to bak1-4 null seedlings,
which are lacking BAK1, the co-activator of FLS2 and multiple other cell surface receptors
(258, 259). As shown in Figures 3B and 3C, the absolute number of stomata, as well as the
number of double and triple stomatal clusters in epsR1 cotyledons, is strikingly similar to
bak1-4 cotyledons. By taking a genetic approach, previous work has categorized similar
stomatal clustering phenotypes in various BR-defective and -insensitive mutants (260,
261). Additionally, as discussed for Figures 2C and 2D, decreased cellular expansion in
epsR1 alleles may produce an increased number of stomata within a defined area of tissue;
therefore, to determine if the alteration in stomatal patterning was due to a lack of cellular
expansion, we assessed the total number of cells: pavement cells, guard cell pairs, and
208
meristemoid precursor cells, in epsR1 and Col0 cotyledons at 5, 7, and 10-days post-
germination. Comparison between these cellular pools revealed that the increased number
of stomata was not due to a lack of cellular expansion (Fig S2B and S2C and Table S1);
rather, our data suggest the change in stomatal patterning may arise from mis-regulation of
precursor stem cells which differentiate into guard cell pairs.
IV.V Conclusions
I provide evidence that EpsinR1 has a role(s) in promoting tissue growth and
development and functions as a positive regulator of distinct PAMP-induced signaling
pathways. Using PAMP-induced and morphological phenotypes in epsR1 seedlings, we
gained insights into determining possible signaling pathways abrogated in the absence of
EpsinR1. Analysis of developing epsR1 seedlings revealed decreased primary epidermal
root cell elongation and cotyledon pavement cell expansion at 5, 7, and 10 days post-
germination compared to Col0. The loss of EpsinR1 may affect developmental intracellular
signaling pathways which have been shown to be important for the localization and
intracellular trafficking of activated copper, iron, and boron transporters utilized in
developmental pathways (262-264). Additionally, auxin and brassinosteroid hormonal
signaling pathways have been shown to be regulated in a similar fashion (201, 235, 265,
266). Hormonal signaling pathways in plants are rather complex and defective tissue
elongation and expansion could arise from multiple hormonal signaling imbalances;
therefore, we sought to identify additional developmental defects which may shed light on
possible hormonal imbalances in the absence of EpsinR1. New evidence has identified a
role for BR-signaling in regulation of stomata formation in newly developing tissues.
209
Brassinosteroid-deficient, and –insensitive, mutants display an increase in stomatal
clusters, which has led to the proposition that BR-signaling negatively regulates stomatal
development through BIN2-mediated inhibition of YODA. When YODA is active, it
suppresses activity of transcription factors responsible for stomatal differentiation such as
SPEECHLESS, FAMA, TMM, and MUTE (267-270). Although the complete mechanism
is unresolved, it is clear that BR-signaling influences activity of the transcription factor
SPEECHLESS and the kinase YODA (260, 261, 271). Analysis of the abaxial surface of
epsR1 cotyledons revealed increased stomata number and stomatal clustering compared to
Col0. Interestingly, we observed epsR1 stomatal numbers and patterning were strikingly
similar to bak1-4 null mutants. Recent evidence implicates an antagonistic relationship
between BR-signaling and PAMP-induced signaling pathways (272, 273). Although the
mechanism of action is currently debated, initial speculation focused on the obvious
candidate BAK1, seeing as it serves as the co-activator to BRI1 and multiple PAMP
receptors, including FLS2. In the absence of BAK1 all early and late flg22-induced
signaling responses are strongly impaired (194, 204). Treatment of 8-day old seedling
tissues using flg22 indicated early, (ROS production; 10 min post-elicitation) and late
(callose deposition; 24h post-elicitation) signaling responses associated with the RBOHD-
branch of PAMP-induced signaling showed decreased activation in epsR1 seedlings
compared to Col0. Similarly, early (MAPK-phosphorylation; 10 min post-treatment) and
intermediate (WRKY29 and FRK1 transcript accumulation; 1h and 4h post-treatment,
respectively) flg22-induced signaling responses associated with the MAPK branch of
PAMP-induced signaling also exhibited decreased activation in 8-day old epsR1 seedlings,
compared to Col0 following flg22-treatment. Because signal transduction within the
210
respective RBOHD and MAPK branches of PAMP-induced signaling is believed to occur
independently, we reasoned that decreased flg22-signaling may stem from abrogated
receptor-complex formation, dysfunction of specific receptor-complex component(s),
and/or a lack of accumulation of receptor component(s) at the plasma membrane. Analysis
of total FLS2 protein levels using 8-day old epsR1-2 and Col0 seedlings revealed that
decreased flg22-induced signaling was not due to lack of FLS2 protein accumulation;
rather, it may be due to dysfunction of a critical receptor-complex component such as
BAK1, or due to lack of accumulation of FLS2 or other receptor-complex components at
the plasma membrane. We provide support suggesting BAK1-dependent signaling
pathways, but not BAK1-independent signaling pathways, may not function correctly in
epsR1. Examination of signaling competency for independent bacterial and fungal PAMP
receptors EFR (EF-Tu Receptor) and CERK1, respectively, revealed that elf26-treatment
demonstrated decreased signaling for ROS production and callose deposition in epsR1-2;
however, elicitation using chitin displayed no significant difference in callose deposition
and MAPK-phosphorylation in epsR1-2, compared to Col0 seedlings. These preliminary
results suggest that signaling defects in epsR1 mutants are not due to a general cell-
signaling defect but rather are specific to a class of elicitors (bacterial vs. fungal), specific
to signal a transduction mechanism (BAK1-dependent vs. BAK1-independent), or specific
to an EpsinR1-dependent vesicular trafficking pathway. Collectively, our PAMP-induced
signaling results taken together with our analysis of morphological defects in epsR1
suggests BAK1-dependent signaling pathways may be compromised in the absence of
EpsinR1. As described below, further experimentation is needed to identify if BAK1
211
dysfunction, or lack of accumulation, is a potential causative agent for the collective
experimental observations.
IV.VI Future Directions
All tested flg22-induced signaling responses in epsR1 mutants displayed decreased
activation compared to WT controls. Decreased amplitude for early, intermediate, and late
stage PAMP-induced signaling responses in epsR1 are consistent with abrogated receptor-
complex formation, defective initiation of signal transduction, or mis-localization of
receptor-complex components. Because total FLS2 protein accumulation is similar
between epsR1 and WT, one explanation may be that other factor(s), possibly BAK1, as
being the causative component(s) producing the documented phenotypes. It is also
important to note that observed signaling defects may originate from abrogation of
activated receptor internalization and endocytic trafficking. In mammalian systems, the
innate immune receptor TLR4 has been shown to have compartmentalized signaling
following perception of its cognate ligand (274). Although this signaling phenomenon has
not been demonstrated for FLS2, it could be an explanation for the decreased signaling
amplitude in epsR1 mutants. Activated FLS2 protein may enter endocytic trafficking
pathways, but in the absence of EpsinR1, may be unable to reach a unique endosomal
signaling compartment wherein amplification of signaling occurs through specific adaptor
molecules, resulting in decreased activation of ROS and MAPK signaling pathways.
212
1. Because all documented flg22- and elf26-induced signaling responses display
decreased activation, it is a logical extension to propose that immunity against both
virulent and avirulent bacterial pathogens would not be as robust in epsR1 mutant
plants, compared to Col0. Inoculation using the isogenic bacterial mutant
Pseudomonas syringae DC3000 hrcC- would assess if bacteria lacking a functional
type-three secretion system could multiply to a higher titer in epsR1 plants because
of decreased PAMP-induced signaling pathways. Inoculation would need to be
carried out directly using syringe-infiltration to circumvent experimental bias from
other inoculation techniques, such as spray inoculation, due to the increase of
stomatal number and clustering in epsR1 plants. As controls, Col0 and sid2-2 would
be included to assess infection in WT plants and infection of mutant line known to
have increased susceptibility to bacterial infection, respectively. Additionally, to
further test the ability of epsR1 plants to mount a robust immune response,
inoculation using the virulent bacterial strain Pseudomonas syringae DC3000 could
be used. This bacterial strain retains its type three secretion system and is able to
inject roughly 30-50 effectors into the host tissue (275). Inoculation would be
carried out in the same manner as described above using the same plant lines.
2. As discussed above, BAK1 could be a possible target responsible for the observed
signaling phenotypes. Serving as the co-activator of FLS2, impaired function, lack
of BAK1 protein accumulation at the plasma membrane, or disrupted FLS2-BAK1
complex formation, would result in decreased signaling through all flg22-induced
signaling pathways. A direct way to test this hypothesis would be to use
213
immunoblot analysis on microsomal, or total, protein isolations from both epsR1
and Col0 seedlings, or leaf tissue, and probe using the lab’s anti-BAK1 antibody.
If BAK1 protein does not accumulate, accumulates to reduced levels, or does not
localize to the plasma membrane in epsR1 compared to Col0, this would likely
explain all observed phenotypes.
Two experimental approaches to determine if BAK1 function is abrogated would
be to cross the epsR1 plants with the bak1-4 plants to generate double-mutants and
to use antagonists of the BR-signaling pathway, such as BL and BRZ, and use
qPCR analysis to look at induction or repression of specific genetic markers in the
BR-signaling pathway. Briefly, I will go into further detail regarding both of these
approaches. The genetic approach is relatively straight forward. The epsR1-2
knock-out line would be crossed with the bak1-4 null mutant line to create a double
mutant. Using the same signaling assays shown and described in the results section
(ROS production and MAPK-phosphorylation) one could determine if the double-
mutant displayed addative signaling defects, or if no change occurred. For
comparison, both epsR1-2 and bak1-4 single mutants and Col0 would be included
in these analyses. If no additive or synergistic change in signaling amplitude
occurred, then one could suggest that the protein affecting in the absence of
EpsinR1 is either BAK1 or receptor-complex formation.
For the second approach, BR-signaling would be used to determine if BAK1
function is abrogated in epsR1 because BAK1 is the co-activator of the BR-receptor
214
BRI1. Several morphological phenotypes were described in the results section (cell
expansion, elongation, and stomata formation) which cumulatively suggest BR-
signaling is decreased in epsR1 plants. Two well characterized genetic markers
used for analysis of BR-signaling are CPD and DWF4 (82, 276). We would analyze
both treated and non-treated samples of epsR1-2 and Col0 8-day old seedlings to
maintain consistency with our previous findings. For treatment, brassinolide (BL)
and/or brassinazole (BRZ) would be used on respective samples to effectively
suppress the transcript accumulation of CPD and DWF4. We would treat seedlings
for 3h (the time point most consistently used in the literature) with BL/BRZ (+), or
without (-, dH2O) and perform qPCR analysis comparing CPD and DWF4 in both
epsR1-2 and Col0 seedlings at 0h and 3h following treatment. The 0h time point
will be crucial because, based on literature and observed phenotypes in epsR1
seedlings, our data suggest that CPD and DWF4 may be elevated at 0h in epsR1
seedlings compared to Col0. We would expect that analysis of the 3h time point
would reveal suppression (decreased mRNA accumulation) of these two markers
in Col0 samples, but none, or minimal, suppression in epsR1-2 samples because
signaling through the BR pathways would not occur as strongly due to possible
BAK1 dysfunction.
Another readily utilized analysis which could be used in conjunction with the
aforementioned BR-treatment is immunoblot analysis of BES1 protein
phosphorylation. BES1 is a protein in the BR-signaling pathway which is upstream
of CPD and DWF4 and its phosphorylated and de-phosphorylated forms are readily
215
detectable using immunoblot analysis. Samples would be treated with (+) BL/BRZ,
or without (-, dH2O) for 0h and 3h and total proteins would be isolated and
subjected to immunoblot analysis. Again, the 0h time point is of importance
because there may be an equilibrium shift of the phosphorylation/de-
phosphorylation state of BES1 in epsR1-2 seedlings compared to Col0. In the
presence of BL/BRZ we would expect BES1 to shift to the de-phosphorylated state
in Col0 samples, but not as strong of a shift, or no shift, in epsR1-2 samples.
3. Another question is whether FLS2 is degraded in the absence of EpsinR1? Because
EpsinR1 has been shown to localize to the EE/TGN and PVC compartments (243),
one may hypothesize that in epsR1 plants FLS2 may not degraded following flg22-
perception. To determine if FLS2 is degraded following ligand-perception, epsR1-
2 and Col0 seedlings would be treated with 100nM flg22 and samples would be
collected at 0, 10, 20, 30, 40, and 60 minutes following treatment. These time points
were chosen specifically because FLS2 has been shown to be internalized and
degraded from 30-60 minutes following ligand perception (191, 221). Total
proteins would be isolated and subjected to immunoblot analysis and probed using
αFLS2 antibody to assess degradation over time. If EpsinR1 has a direct, or indirect,
role(s) in trafficking FLS2 to a degradative compartment, we would expect to see
less, or no, FLS2 degradation in epsR1-2 compared to Col0 during this time course
in response to flg22. We have also established that following degradation FLS2
accumulated in a signaling competent state over time (Salamango, in prep.). It
would also be interesting to determine if EpsinR1 has a role in trafficking newly
216
synthesized FLS2 protein to the plasma membrane. Following a similar
experimental approach as listed above, one would expand the time course to focus
on times greater than 1h, including an intermediate time at 3h and a late time at
16h. Initially, immunoblot analysis of total FLS2 protein could be used to assess
degradation and accumulation following flg22-treatment. However, if this
methodology does not yield conclusive results, one could test the role of EpsinR1
in FLS2 protein secretion using two-phase partitioning to determine if FLS2 protein
co-localizes with known plasma membrane markers following flg22-treatment.
217
Figure IV.I: EpsinR1 Functions as a Positive Regulator of Early-Induced flg22-
Signaling Responses.
(A) EpsinR1 protein accumulation in allelic T-DNA insertion lines. Total proteins from
Col0, epsR1-1 (SALK_049204), and epsR1-2 (SAIL_394G02) 8-day old seedlings were
isolated and subjected to immunoblot analysis. Samples were probed using αEpsinR1 #131
to determine EpsinR1 protein accumulation, and αMPK6 to assess equal loading
(repeats=3). (B) flg22-induced ROS production is decreased in epsR1. Cut cotyledons from
8-day old seedlings were elicited with 10nM flg22 and ROS production was monitored for
40 minutes using a luminol based assay (RLU= relative light units). Filled squares
represent Col0, open squares represent epsR1-1, and grey squares represent epsR1-2
(n=32/genotype) (repeats≥10). Statistical difference between epsR1 alleles and Col0 is ***
p<0.0001 (C) flg22-induced peak MAPK-phosphorylation is decreased in epsR1 alleles.
Whole 8-day old seedlings were elicited with (+), or without (-), 10nM flg22 for 10 minutes
and total proteins were subjected to immunoblot analysis. Samples were probed with αP-
p44/42 MAPK to assess flg22-induced phosphorylation of MPK6 (P-M6) and MPK3 (P-
M3), and with αMPK6 to determine MAPK6 accumulation and equal loading. (repeats=3)
218
Figure IV.I: EpsinR1 Functions as a Positive Regulator of Early-Induced flg22-
Signaling Responses.
219
Figure IV.II: Role for EpsinR1 in Intermediate and Late-Induced flg22-Signaling
Responses.
(A) and (B) flg22-induced transcript accumulation of WRKY29 and FRK1 mRNA is
significantly reduced in epsR1 alleles. Whole 8-day old seedlings were treated with 10nM
flg22 (1h, or 4h), or without (0h, dH2O), for the indicated time and flash frozen. Samples
were processed for qRT-PCR using At2g28390 as the reference gene. Filled bar represents
Col0, open bar represents epsR1-1, and grey bar represents epsR1-2. Results depicted as
relative expression normalized to At2g28390. Asterisks represent the degree of significant
difference (**; p<0.001, ***; p<0.0001) between epsR1-2 and Col0. (repeats= 1 biological
rep with an n=3) (C) Callose deposition following flg22-treatment is reduced in the absence
of EpsinR1. Adult 4 to 5 week old leaf tissue was syringe infiltrated with10nM flg22 and
analyzed for callose deposition 24 hours post-infiltration using analine blue stain. Scale bar
represents 50nm and ** represents p<0.001. (repeats=2) (D) flg22-indcued PR1 protein
accumulation is reduced in epsR1. 4 to 5 week old leaf tissue was syringe infiltrated using
100nM flg22 (+), or water (-) and flash frozen at the indicated time post-infiltration. Total
proteins from samples were subjected to immunoblot analysis and probed using αPR1 to
assess PR1 protein accumulation, or αMPK6 to assess equal loading. h, hour; PF, flg22;
MPK6, MAPK6. (repeats=1).
220
Figure IV.II: Role for EpsinR1 in Intermediate and Late-Induced flg22-Signaling
Responses.
221
Figure IV.III: BAK1-Dependent Signaling Mechanisms May be Disrupted in epsR1.
(A) Total FLS2 protein accumulation is similar between Col0 and epsR1. Total proteins
from un-elicited 8-day old seedlings were subjected to immunoblot analysis and probed
using αFLS2 to assess FLS2 protein accumulation, or αMPK6 to show equal loading.
(repeats≥3) (B) Decreased PAMP-induced MAPK-Phosphorylation in epsR1 is PAMP-
dependent. Whole 8-day old seedlings were elicited with 10nM flg22 or 100μg/mL chitin
for 10 minutes and flash frozen. Total proteins from samples were subjected to immunoblot
analysis and probed with αP-p44/42 MAPK to assess PAMP-induced phosphorylation of
MPK6 (P-M6) and MPK3 (P-M3), and with αMPK6 to assess equal loading. (repeats=1)
(C) Decreased PAMP-induced callose deposition in epsR1 is PAMP-dependent. 8-day old
seedlings were submerged in either 10nM flg22, 10nM elf26, or 100μg/ml chitin for 24
hours. Samples were stained using aniline blue to assess callose deposition. Scale bar
represents 50nm. (repeats=1).
222
Figure IV.III: BAK1-Dependent Signaling Mechanisms May be Disrupted in epsR1.
223
Figure IV.IV: Morphological comparison of epsinR1 alleles at 8-days Post-
Germination.
(A) Comparison of 8-day old seedling aerial tissue. Representative Col0, epsR1-1, and
epsR1-2 8-day old seedlings were removed from 0.5% MS 1% sucrose plates and aerial
tissue was imaged for comparison. (repeats=3) (B) and (C) Primary root length is stunted
in epsR1 alleles. Col0, epsR1-2, and epsR1-1 8-day old seedlings were grown horizontally
on 0.5% MS 1% sucrose plates and gently extracted for primary root length measurements.
Image depicts representative seedlings from an individual growth period. White bars
represent primary root tips to highlight overall lengths. (repeats ≥5) (C) Root lengths were
statistically analyzed (n=30) and filled bars represent Col0, grey bars represent epsR1-2,
and open bars represent epsR1-1. Asterisks represent a significant difference between
mutants and wild-type primary root lengths (**; p<0.001, ***; p<0.0001). (repeats ≥5).
224
Figure IV.IV: Morphological comparison of epsinR1 alleles at 8-days Post-
Germination.
225
Figure IV.V: The epsR1 Alleles Display Defective Primary Root Cell Elongation and
Pavement Cell Expansion.
(A) and (B) Mutant epsR1 seedlings exhibit shortened primary epidermal root cells.
Primary epidermal root cell length measurements for seedlings 5, 7, and 10 days post-
germination (n=20 for each time point) Asterisks *** represent p<0.0001. Filled bars
represent Col0, Grey bars represent epsR1-2, and open bars represent epsR1-1.
(repeats=3). (C) and (D) Mutant epsR1 pavement cells exhibit inhibited cellular
expansion. Cotyledon pavement cell area measurements for seedlings 5 (n≥130), 7
(n≥100), and 10 (n≥70) days post-germination (repeats=3). Filled bars represent Col0,
grey bars represent epsR1-2, and open bars represent epsR1-1. Asterisks **,*** indicate
a highly-significant difference between mutant and wild-type measurements p<0.001 and
p<0.0001 respectively. μm; micrometers, μm2; micrometers squared (surface area).
226
Figure IV.V: The epsR1 Alleles Display Defective Primary Root Cell Elongation and
Pavement Cell Expansion.
227
Figure IV.VI: Increased Stomatal Number and Clustering in epsR1 Alleles.
(A) Mutant epsR1 seedlings display increased stomatal number and clustering.
Representative images of the abaxial surface of toluidine blue stained 8-day old cotyledons.
* Asterisks highlight the presence of stomata on the surface of stained cotyledons. Scale
bar represents 50nm. (repeats=3) (B) Quantification of absolute stomata number per ocular
field of view present on the abaxial surface of 8-day old cotyledons following toluidine
blue staining (n=total from 5 independent fields of view). Filled bar represents Col0, grey
bar represents epsR1-2, open bar represents epsR1-1, diagonally striped bar represents
bak1-4. Asterisks **, *** indicate a significant difference between wild-type and mutant
stomatal counts p<0.001 and p<0.0001 respectively. (repeat=1). (C) Similarity between
epsR1 and bak1-4 stomatal clustering on the abaxial surface of 8-day old cotyledons.
Doubly and triply clustered stomata were quantified on the abaxial surface of 8-day old
seedlings following toluidine blue staining. The number depicted represents the sum across
3 independent images taken from one cotyledon. (repeat=1).
228
Figure IV.VI: Increased Stomatal Number and Clustering in epsR1 Alleles.
229
Figure IV.S1: Leaf Tissue Display Decreased ROS Production and PAMPs Dose
Dependency.
(A) Adult leaf tissue displays decreased ROS production following flg22-treatment. 4 to
5 week old cut leaf disks were elicited with 10nM flg22 and ROS production was
monitored over 40 minutes using a luminol based assay (RLU=relative light units). Filled
squares represent Col0, grey squares represent epsR1-2, and open squares represent
epsR1-1 (n=32/genotype) (repeats=3). (B) and (C) Seedlings display decreased ROS
production using titrations of PAMPs as low as 1nM. 8-day old cut cotyledons were
elicited with either 1, 5, or 10nM of the indicated PAMP and ROS production was
monitored over 40 minutes using a luminol based assay (RLU= relative light units). ROS
production depicted occurred at 10-12 min post-elicitation and was taken at the peak of
ROS production for each treatment. Filled bars represent Col0, grey bars represent
epsR1-2. Each set of PAMP titrations were run on the same 96-well ROS plate for direct
comparison (n=16/treatment/genotype). Asterisks represent the degree of significant
difference (*; p<0.05, **; p<0.001) between epsR1-2 and Col0 cut cotyledons (repeats
for flg22=3, for elf26=1).
230
Figure IV.S1: Leaf Tissue Display Decreased ROS Production and PAMPs Dose
Dependency.
231
Figure IV.S2: EpsinR1 Functions as a Positive Regulator of Early-Induced flg22-
Dependent MAPK Activation.
Flg22-indcued MAPK-phosphorylation profile is severely abrogated in the absence of
EpsinR1. Whole 8-day old seedlings were elicited with (+), or without (-), 10nM flg22 for
the indicated amount of time and flash frozen. Total proteins were subjected to immunoblot
analysis and probed with αP-p44/42 MAPK to assess flg22-induced phosphorylation of
MPK6 (P-M6) and MPK3 (P-M3), and with αMPK6 to determine MAPK6 accumulation
and to assess loading. min, minutes; PF, flg22; MPK6, MAPK6; P-MPK 6/3,
phosphorylated MAPK6/3 (repeats=1).
232
Figure IV.S2: EpsinR1 Functions as a Positive Regulator of Early-Induced flg22-
Dependent MAPK Activation.
233
Figure IV.S3: Increase in Stomata Number and Clustering is Independent of
Cellular Expansion.
(A) Adult rosettes display stunting in epsR1 alleles. 4 to 5 week old plants were grown in
8h light/16h dark and imaged to compare relative rosette diameters between Col0 and
epsR1 alleles. (repeat=1) (B) and (C) Decreased cellular expansion is not the causative
agent for increased stomata number in epsR1 alleles. Abaxial surfaces of 8-day old
cotyledons were imaged at 5, 7, and 10-days post-germination. Seedlings were stained
using toluidine blue and cellular pools were quantified at each individual time point. Guard
cell pairs are represented by light green bars, meristemoid precursor cells are represented
by dark green bars, and pavement cells are represented by open bars. The total number of
cells was determined from a fixed area of the cotyledon, which was used to normalize the
individual cellular pools shown as a percentage of the total cells. (repeat=3).
234
Figure IV.S3: Increase in Stomata Number and Clustering is Independent of
Cellular Expansion.
235
Table S1: Statistical Analysis of Cellular Pools from Abaxial Cotyledons.
Breakdown of the quantification of cellular pools ascertained for seedlings 5, 7, and 10-
day post-germination. The total for each cell type from the mutants was directly compared
to Col0 controls for the indicated time point. Asterisks represent the degree of significant
difference between mutants and Col0 controls for the indicated cellular pool at the
indicated time point (*; p<0.05, **; p<0.001, and ***; p<0.0001). Non-significant
differences are represented with “ns.” (repeats=3).
236
Table IV.S1: Statistical Analysis of Cellular Pools from Abaxial Cotyledons
237
REFERENCES
1. Provitera P, El-Maghrabi, R, and Scarlata, S. 2006. The effect of HIV-1 Gag
myristoylation on membrane binding. . Biophys Chem 119:23-32.
2. Chukkapalli V, and Ono, A. . 2011. Molecular determinants that regulate
plasma membrane association of HIV-1 Gag. J Mol Biol 410:512-524.
3. Hogue I, Llewellyn, G, and Ono, A. . 2012. Dynamic Association between HIV-
1 Gag and Membrane Domains. Mol Biol Intl.
4. Yufenyuy EaA, C. 2013. The NTD-CTD intersubunit interface plays a critical
role in assembly and stabilization of the HIV-1 capsid. Retrovirology 10:29.
5. Ako-Adjei D, Johnson, M and Vogt, V. 2005. The Retroviral Capsid Domain
Dictates Virion Size, Morphology, and Coassembly of Gag into Virus-Like
Particles. Journal of Virology 79:13463–13472.
6. Bell N, and Lever, A. 2013. HIV Gag polyprotein: processing and early viral
particle assembly. Trends Microbiol 21:136-144.
7. Tazi J, Bakkour, N, Marchand, V, Ayadi, L, Aboufirassi, A, and Branlant, C.
2010. Alternative splicing: regulation of HIV-1 multiplication as a target for
therapeutic action. FEBS J 277:867-876.
8. Corbin A, Prats, A, Darlix, J, and Sitbon, M. 1994. A nonstructural gag-
encoded glycoprotein precursor is necessary for efficient spreading and
pathogenesis of murine leukemia viruses. Journal of Virology 68:3857-3867.
9. Pillemer E, Kooistra, D, Witte, O, and Weissman, I 1986. Monoclonal
antibody to the amino-terminal L sequence of murine leukemia virus glycosylated
238
Gag polyproteins demonstrates their unusual orientation in the cell membrane.
Journal of Virology 57:413-421.
10. VM V. 1997. Retroviral Virions and Genomes. Retroviruses, eds Coffin JM,
Hughes SH, Varmus HE (Cold Spring Harbor Lab Press, Cold Spring Harbor,
NY).
11. Fass D, Harrison, S, and Kim, P. 1996. Retrovirus envelope domain at 1.7
angstrom resolution. Nat Struct Biol 3:465-469.
12. Johnston E, and Radke, K. 2000. The SU and TM envelope protein subunits of
bovine leukemia virus are linked by disulfide bonds, both in cells and in virions
Journal of Virology 74:2930-2935.
13. Leamnson R, and Halpern, M. 1976. Subunit structure of the glycoprotein
complex of avian tumor virus. Journal of Virology 18:956-968.
14. Opstelten D, Wallin, M., and Garoff, H. 1998. Moloney murine leukemia virus
envelope protein subunits, gp70 and Pr15E, form a stable disulfide-linked
complex. Journal of Virology 72:6537-6545.
15. Pinter A, and Fleissner, E. 1977. The presence of disulfide-linked gp70-p15(E)
complexes in AKR murine leukemia viruses. Virology 83:417-422.
16. Henzy J, and Coffin, J. 2013. Betaretroviral envelope subunits are noncovalently
associated and restricted to the mammalian class. Journal of Virology 87:1937-
1946.
17. Kowalski M, Potz, J., Basiripour, L., Dorfman, T., Goh, W., Terwilliger, E.,
Dayton, A., Rosen, C., Haseltine, W., and Sodrosk,i J. 1987. Functional
239
regions of the envelope glycoprotein of human immunodeficiency virus type 1.
Science 237.
18. Hunter E. 1997. Viral Entry and Receptors. Retroviruses, eds Coffin JM, Hughes
SH, Varmus HE (Cold Spring Harbor Lab Press, Cold Spring Harbor, NY).
19. Celma C, Paladino, M, González, S, and Affranchino, J. 2007. Importance of
the short cytoplasmic domain of the feline immunodeficiency virus
transmembrane glycoprotein for fusion activity and envelope glycoprotein
incorporation into virions. Virology 366:405-414.
20. Côté M, Zheng, Y, Albritton, L, and Liu, S. 2008. Fusogenicity of Jaagsiekte
sheep retrovirus envelope protein is dependent on low pH and is enhanced by
cytoplasmic tail truncations. Journal of Virology 82:2543-2554.
21. Postler T, and Desrosiers, R. 2013. The tale of the long tail: the cytoplasmic
domain of HIV-1 gp41. Journal of Virology 87:2-15.
22. Deschambeault J, Lalonde, J, Cervantes-Acosta, G, Lodge, R, Cohen, E, and
Lemay, G. 1999. Polarized human immunodeficiency virus budding in
lymphocytes involves a tyrosine-based signal and favors cell-to-cell viral
transmission. Journal of Virology 73:5010-5017.
23. Lodge R, Delemarre, L, Lalonde, J, Alvarado, J, Sanders, D, Dokhélar, M,
Cohen, E, and Lemay, G. 1997. Two distinct oncornaviruses harbor an
intracytoplasmic tyrosine-based basolateral targeting signal in their viral envelope
glycoprotein. Journal of Virology 71:5696-5702.
240
24. Lodge R, Lalonde, J, Lemay, G, and Cohen, E. 1997. The membrane-proximal
intracytoplasmic tyrosine residue of HIV-1 envelope glycoprotein is critical for
basolateral targeting of viral budding in MDCK cells. EMBO J 16:695-705.
25. Hensel J, Hintz, M, Karas, M, Linder, D, Stahl, B, and Geyer, R. 1995.
Localization of the palmitoylation site in the transmembrane protein p12E of
Friend murine leukaemia virus. Eur J Biochem 232:373-380.
26. Ochsenbauer-Jambor C, Miller, D, Roberts, C, Rhee, S, and Hunter, E. 2001.
Palmitoylation of the Rous sarcoma virus transmembrane glycoprotein is required
for protein stability and virus infectivity. Journal of Virology 75:11544-11554.
27. Yang C, Spies, C, and Compans, R. 1995. The human and simian
immunodeficiency virus envelope glycoprotein transmembrane subunits are
palmitoylated. PNAS 92:9871-9875.
28. Murakami T. 2012. Retroviral env glycoprotein trafficking and incorporation
into virions. Mol Biol Intl.
29. White J. 1992. Membrane Fusion. Science 258:917-924.
30. Telesnitsky A, and Goff, S. 1997. Reverse Transcriptase and the Generation of
Retroviral DNA. Retroviruses, eds Coffin JM, Hughes SH, Varmus HE (Cold
Spring Harbor Lab Press, Cold Spring Harbor, NY).
31. Po B. 1997. Integration. Retroviruses, eds Coffin JM, Hughes SH, Varmus HE
(Cold Spring Harbor Lab Press, Cold Spring Harbor, NY).
32. Iwatani Y, Chan, D, Wang, F, Maynard, K, Sugiura, W, Gronenborn, A,
Rouzina, I, Williams, M, Musier-Forsyth, K, and Levin, J. 2007. Deaminase-
241
independent inhibition of HIV-1 reverse transcription by APOBEC3G. Nuc Acids
Res 35:7096-7108.
33. Swanstrom R. 1997. Synthesis, Assembly, and Processing of Viral Proteins.
Retroviruses, eds Coffin JM, Hughes SH, Varmus HE (Cold Spring Harbor Lab
Press, Cold Spring Harbor, NY).
34. Rhee S, and Hunter, E. 1990. A single amino acid substitution within the matrix
protein of a type D retrovirus converts its morphogenesis to that of a type C
retrovirus. Cell 63:77-86.
35. Kuff E, and Lueders, K. 1988. The intracisternal A-particle gene family:
structure and functional aspects. Adv Cancer Res 51:183-276.
36. Kutluay S, and Bieniasz, P. 2010. Analysis of the initiating events in HIV-1
particle assembly and genome packaging. PLos Pathol 6.
37. Bryant M, and Ratner, L. 1990. Myristoylation-dependent replication and
assembly of human immunodeficiency virus 1. PNAS 87:523-527.
38. Chukkapalli V, Hogue, I, Boyko, V, Hu, W, and Ono, A. 2008. Interaction
between the human immunodeficiency virus type 1 Gag matrix domain and
phosphatidylinositol-(4,5)-bisphosphate is essential for efficient gag membrane
binding. Journal of Virology 82:2405-2417.
39. Dalton A, Ako-Adjei, D, Murray, P, Murray, D, and Vogt, V. 2007.
Electrostatic interactions drive membrane association of the human
immunodeficiency virus type 1 Gag MA domain. Journal of Virology 81:6434-
6445.
242
40. Dalton A, Murray, P, Murray, D, and Vogt, V. 2005. Biochemical
characterization of rous sarcoma virus MA protein interaction with membranes.
Journal of Virology 79:6227-6238.
41. Luttge B, and Freed, E. 2010. FIV Gag: virus assembly and host-cell
interactions. Vet Immunol Immunopathol 134:3-13.
42. Ono A. 2009. HIV-1 Assembly at the Plasma Membrane: Gag Trafficking and
Localization. Future Virol 4:241-257.
43. Ono A. 2010. Relationships between plasma membrane microdomains and HIV-1
assembly. Biol Cell 102:335-350.
44. Jouvenet N, Bieniasz, P, and Simon, S. 2008. Imaging the biogenesis of
individual HIV-1 virions in live cells. Nature 454:236-240.
45. Pelchen-Matthews A, and Marsh, M. 2007. Electron microscopy analysis of
viral morphogenesis. Methods Cell Biol 79:515-542.
46. Deneka M, Pelchen-Mathews, A, Byland, R, Ruiz-Mateos, E, and Marsh, M.
2007. In macrophages, HIV-1 assembles into an intracellular plasma membrane
domain containing the tetraspanins CD81, CD9, and CD53. Journal of Cell
Biology 177:329-341.
47. Bennett A, Narayan, K, Shi, D, Hartnell, L, Gousset, K, He, H, Lowekamp,
B, Yoo, T, Bliss, D, Freed, E, and Subramaniam, S. 2009. Ion-abrasion
scanning electron microscopy reveals surface-connected tubular conduits in HIV-
infected macrophages. PLos Pathol 5.
243
48. Welsch S, Keppler, O, Habermann, A, Allespach, I, Krijnse-Locker, J, and
Kräusslich, H. 2007. HIV-1 buds predominantly at the plasma membrane of
primary human macrophages. PLos Pathol 3.
49. Pessin J, and Glaser, M. 1980. Budding of Rous sarcoma virus and vesicular
stomatitis virus from localized lipid regions in the plasma membrane of chicken
embryo fibroblasts. Journal of Biological Chemistry 255:9044-9050.
50. Quigley J, Rifkin, D, and Reich, E. 1972. Lipid studies of Rous sarcoma virus
and host cell membranes. Virology 50:550-557.
51. Quigley J, Rifkin, D, and Reich, E. 1971. Phospholipid composition of Rous
sarcoma virus, host cell membranes and other enveloped RNA viruses. Virology
46:106-116.
52. Aloia R, Tian, H, and Jensen, F. 1993. Lipid composition and fluidity of the
human immunodeficiency virus envelope and host cell plasma membranes. PNAS
90:5181-5185.
53. Brügger B, Glass, B, Haberkant, P, Leibrecht, I, Wieland, F, and Kräusslich,
H. 2006. The HIV lipidome: a raft with an unusual composition. PNAS
103:2641-2646.
54. Campbell S, Gaus, K, Bittman, R, Jessup, W, Crowe, S, and Mak, J. 2004.
The raft-promoting property of virion-associated cholesterol, but not the presence
of virion-associated Brij 98 rafts, is a determinant of human immunodeficiency
virus type 1 infectivity. Journal of Virology 78:10556-10565.
244
55. Campbell S, Crowe, S, and Mak, J. 2002. Virion-associated cholesterol is
critical for the maintenance of HIV-1 structure and infectivity. AIDS 16:2253-
2261.
56. Nguyen D, and Hildreth, J. 2002. Evidence for budding of human
immunodeficiency virus type 1 selectively from glycolipid-enriched membrane
lipid rafts. Journal of Virology 74:3264-3272.
57. Blot V, Perugi, F, Gay, B, Prévost, M, Briant, L, Tangy, F, Abriel, H, Staub,
O, Dokhélar, M, and Pique, C. 2004. Nedd4.1-mediated ubiquitination and
subsequent recruitment of Tsg101 ensure HTLV-1 Gag trafficking towards the
multivesicular body pathway prior to virus budding. Journal of Cell Sci 117:2357-
2367.
58. Bouamr F, Melilo, J, Wang, M, Nagashima, K, de Los Santos, M, Rein, A,
and Goff, S. 2003. PPPYVEPTAP motif is the late domain of human T-cell
leukemia virus type 1 Gag and mediates its functional interaction with cellular
proteins Nedd4 and Tsg101. Journal of Virology 77:11882-11895.
59. Chung H, Morita, E, von Schwedler, U, Müller, B, Kräusslich, H, and
Sundquist, W. 2008. NEDD4L overexpression rescues the release and infectivity
of human immunodeficiency virus type 1 constructs lacking PTAP and YPXL late
domains. Journal of Virology 82:4884-4897.
60. Gottwein E, Bodem, J, Müller, B, Schmechel, A, Zentgraf, H, and
Kräusslich, H. 2003. The Mason-Pfizer monkey virus PPPY and PSAP motifs
both contribute to virus release. Journal of Virology 77:9474-9485.
245
61. Heidecker G, Lloyd, P, Fox, K, Nagashima, K, and Derse, D. 2004. Late
assembly motifs of human T-cell leukemia virus type 1 and their relative roles in
particle release. Journal of Virology 78:6636-6648.
62. Jolly C. 2010. T cell polarization at the virological synapse. Viruses 2:1261-
1278.
63. Jolly C, and Sattentau, Q. 2004. Retroviral spread by induction of virological
synapses. Traffic 5:643-650.
64. Jolly C, Welsch, S, Michor, S, and Sattentau, Q. 2007. The regulated secretory
pathway in CD4(+) T cells contributes to human immunodeficiency virus type-1
cell-to-cell spread at the virological synapse. PLos Pathol 7.
65. Sattentau Q. 2010. Cell-to-Cell Spread of Retroviruses. Viruses 2:1306-1321.
66. Cronin J, Zhang, X, and Reiser, J. 2005. Altering the tropism of lentiviral
vectors through pseudotyping. Curr Gene Ther 5:387-398.
67. M. J. 2011. Mechanisms for Env glycoprotein acquisition by retroviruses. AIDS
Res Hum Retrovir 27:239-247.
68. Mazari PaR, M. 2013. Library screening and receptor-directed targeting of
gammaretroviral vectors. Future Microbiol:107-121.
69. Akari H, Fukumori, T, and Adachi, A. 2000. Cell-dependent requirement of
human immunodeficiency virus type 1 gp41 cytoplasmic tail for Env
incorporation into virions. Journal of Virology 74:4891-4893.
70. Dubay J, Roberts, S, Hahn, B, and Hunter, E. 2002. Truncation of the human
immunodeficiency virus type 1 transmembrane glycoprotein cytoplasmic domain
blocks virus infectivity. Journal of Virology 66:6616-6625.
246
71. Gabuzda D, Lever, A, Terwilliger, E, and Sodroski, J. 1992. Effects of
deletions in the cytoplasmic domain on biological functions of human
immunodeficiency virus type 1 envelope glycoproteins. Journal of Virology
66:3306-3315.
72. Murakami T, and Freed, E. 2000. The long cytoplasmic tail of gp41 is required
in a cell type-dependent manner for HIV-1 envelope glycoprotein incorporation
into virions. PNAS 97:343-348.
73. Lucas TM, Lyddon TD, Grosse SA, Johnson MC. 2010. Two distinct
mechanisms regulate recruitment of murine leukemia virus envelope protein to
retroviral assembly sites. Virology 405:548-555.
74. Gregory D. LT, and Johnson MC. 2013. Multiple gag domains contriobute to
selective recruitment of meurine leukemia virus (MLV) Env to MLV virions.
Journal of Virology 87:1518-1527.
75. Weclewicz K, Ekström, M, Kristensson, K, and Garoff, H. 1998. Specific
interactions between retrovirus Env and Gag proteins in rat neurons. Journal of
Virology 72:2832-2845.
76. Zhong P, Agosto, L, Munro, J, and Mothes, W. 2013. Cell-to-cell transmission
of viruses. Curr Opin Virol 3:44-50.
77. Jorgenson R. VV, and Johnson M. 2009. Foreign Glycoproteins Can Be
Actively Recruited to Virus Assembly Sites during Pseudotyping. Journal of
Virology 83:4060-4067.
78. Jin J. LF, Mothes W. 2011. Viral determinants of polarized assembly for the
murine leukemia virus. Journal of Virology 85.
247
79. Bauby H, Lopez-Verges, S, Hoeffel, G, Delcroix-Genête, D, Janvier, K,
Mammano, F, Hosmalin, A, and Berlioz-Torrent, C. 2010. TIP47 is required
for the production of infectious HIV-1 particles from primary macrophages.
Traffic 11:455-467.
80. Blot G, Janvier, K, Le Panse, S, Benarous, R, and Berlioz-Torrent, C. 2003.
Targeting of the human immunodeficiency virus type 1 envelope to the trans-
Golgi network through binding to TIP47 is required for env incorporation into
virions and infectivity. Journal of Virology 77:6931-6945.
81. Checkley M, Luttge, B, Mercredi, P, Kyere, S, Donlan, J, Murakami, T,
Summers, M, Cocklin, S, and Freed, E. 2013. Reevaluation of the requirement
for TIP47 in human immunodeficiency virus type 1 envelope glycoprotein
incorporation. Journal of Virology 87:3561-3570.
82. Miklós Szekeres KN, Zsuzsanna Koncz-Kálmán, Jaideep Mathur, Annette
Kauschmann, Thomas Altmann, George P Rédei, Ferenc Nagy, Jeff Schell,
and Csaba Koncz. 1996. Cytochrome P450, Controlling Cell Elongation and De-
etiolation in Arabidopsis. Cell 85:171-182.
83. Grove J, and Marsh, M. 2011. The cell biology of receptor-mediated virus
entry. Journal of Cell Biology 195:1071-1082.
84. Loving R, Li, K., Wallin, M., Sjoberg, M., and Garoff, H. 2008. R-peptide
cleavage potentiates fusion-controlling isomerization of the intersubunit disulfide
in Moloney murine leukemia virus Env. Journal of Virology 82:2594-2597.
248
85. Ragheb J, and Anderson, W. 1994. pH-independent murine leukemia virus
ecotproic envelpe-mediated cell fusion: implications for the role of the R peptide
and p12E TM in viral entry. Journal of Virology 68:3220-3231.
86. Kubo Y, Tominaga, C., Yoshii, H., Kamiyama, H., Mitani C., Amanuma H.,
and Yamamoto, N. 2007. Characterization of R-peptide of murine leukemia
virus envelope glycoprotein in syncytium formation and entry Arch Virology
152:2169-2182.
87. Li M, Li, Z., Yao, Q., Yang, C., Steinhauer, D., and Compans, R. 2006.
Murine leukemia virus R peptide inhibits influenza virus hemagglutinin-induced
membrane fusion. Journal of Virology 80:6106-6114.
88. Rein A, Mirro, J., Haynes, J., Ernst, S., and Nagashima, K. 1994. Function of
the cytoplasmic domain of a retroviral transmembrane protein: p15E-p2E
cleavage activates the membrane fusion capability of the murine leukemia virus
Env protein. Journal of Virology 68:1773-1781.
89. Henderson L, Sowder, R., Copeland, T., Smythers, G., and Oroszlan, S. 1984.
Quantitative separation of murine leukemia virus proteins by reverse-phase high-
pressure liquid chromatography reveals newly described gag and env cleavage
products. Journal of Virology 52:492-500.
90. Loving R, Wu, S., Sjoberg, M., Lindqvist, B., and Garoff, H. 2012. Maturation
cleavage of the meurine leukemia virus Env precursor separates the
transmembrane subunits to prime it for receptor triggering. PNAS 109:7735-
7740.
249
91. Owens R, Burke, C., and Rose, J. 1994. Mutations in the Membrane-Spanning
Domain of the Human Immunodeficiency Virus Envelope Glycoprotein That
Affect Fusion Activity. Journal of Virology 68:570-574.
92. Apellániz B, Rujas, E., Serrano, S., Morante, K., Tsumoto, K., Caaveiro, J.,
Jiménez, M., and Nieva, J. 2015. The Atomic Structure of the HIV-1 gp41
Transmembrane Domain and Its Connection to the Immunogenic Membrane-
proximal External Region. Journal of Biological Chemistry 290:12999-13015.
93. Sun Z, Cheng, Y., Kim, M., Song, L., Choi, J., Kudahl, U, Brusic, V,
Chowdhury, B, Yu, L, Seaman, M, Bellot, G, Shih, W, Wagner, G, and
Reinherz, E. 2014. Disruption of Helix-Capping Residues 671 and 674 Reveals a
Role in HIV-1 Entry for a Specialized Hinge Segment of the Membrane Proximal
External Region of gp41. J Mol Biol 426:1095-1108.
94. Yi H, Diaz-Rohrer, B., Saminathan, P., and Jacobs, A. 2015. The Membrane
Proximal External Regions of gp41 from HIV‑1 Strains HXB2 and JRFL Have
Different Sensitivities to Alanine Mutation. Biochemistry 54:1681-1693.
95. Salzwedel K, West, J. and Hunter, E. 1999. A Conserved Tryptophan-Rich
Motif in the Membrane-Proximal Region of the Human Immunodeficiency Virus
Type 1 gp41 Ectodomain is Important for Env-Mediated Fusion and Virus
Infectivity. Journal of Virology 73:2469-2480.
96. Apellaniz B, Rujas, E., Carravilla, P., Requejo-Isidro, J., Huarte, N.,
Domene, C., and Nieva, J. 2014. Cholesterol-Dependent Membrane Fusion
Induced by the gp41 Membrane-Proximal External Region-Transmembrane
250
Domain Connection Suggests a Mechanism for Broad HIV-1 Neutralization.
Journal of Virology 88:13367-13377.
97. Cordes F, Bright, J, and Samson, M. 2002. Proline-induced distortions of
transmembrane helices. J Mol Biol 323:951-960.
98. Jardetzky O. 1996. Simple allosteric model for membrane pumps. Nature
211:696-670.
99. Sansom M, and Weinstein, H. 2000. Hinges, swivels & switches: the role of
prolines in signalling via transmembrane a-helices. Trends Pharmacol Sci 21:445-
451.
100. Govaerts C, Blanpain, C., Deupi, X., Ballet, S., Ballesteros, J. A., Wodak, S.
J. et al. 2001. The TXP motif in the second transmembrane helix of CCR5—a
structural determinant of chemokine-induced activation. Journal of Biological
Chemistry 276:13217-13225.
101. Ri Y, Ballesteros, J., Abrams, C., Oh, S., Verselis, V., Weinstein, H., and
Bargiello, T. 1999. The role of a conserved proline residue in mediating
conformational changes associated with voltage gating of Cx32 gap junctions.
Biophysical J 76:2887-2898.
102. Janaka S, Gregory, D., and Johnson, M. 2013. Retrovirus Glycoprotein
Functionality Requires Proper Alignment of the Ectodomain and the Membrane-
Proximal Cytoplasmic Tail. Journal of Virology 87:12805-12813.
103. Kim J, Lee, S., Li, L., Park, H., Park, J., Lee, K., Kim, M., Shin, B., and
Choi, S 2011. HIgh Cleavage Efficiency of a 2A Peptide Derived from Porcine
Teschovirus-1 in Human Cell Lines, Zebrfish and Mice. Plos One 6:e18556.
251
104. Trichas G, Begbie, J., and Srinivas, S. 2008. Use of the viral 2A peptide for
bicistronic expression in transgenic mice. BMC Biol 6.
105. Boussif O, Lezoualc'h, F., Zanta, M., Scherman, D., Demeneix, B., and Behr,
J. 1995. A versitle vector for gene and oligonucleotide transfer into cells in
culture and in vivo: polyethylenimine. PNAS 92:7297-7301.
106. Wu B, Lu, J., Gallaher, T., Anderson, W., and Cannon, P. 2000. Identification
of regions in the Moloney murine leukemia virus SU protein that tolerate the
insertion of an integrin-binding peptide Virology 269:7-17.
107. Erlwien O, Buchholz, C., and Schnierle, B. 2003. The proline rich region of
ecotropic Moloney murine leukemia virus envelope protein tolerates the insertion
of the green fluorescent protein and allows the generation of replication-
competent virus. J Gen Virol 84:369-373.
108. Rothenberg S, Olsen, M., Laurent, L., Crowly, R., and Brown, P. 2001.
Comprehensive mutational analysis of the Moloney murine leukemia virus
envelope glycoprotein. Journal of Virology 75:11851-11862.
109. Dawson J, Weinger, J., and Engleman, D. 2002. Motifs of Serine and
Threonine can Drive Association of Transmembrane Helices J Mol Biol 316:799-
805.
110. Deupi X, Olivella, M., Govaerts, C., Ballesteros, J., Campillo, M., and Pardo,
L. 2004. Ser and Thr residues modulate the comformation of Pro-kinked
transmembrane aplha-helices. Biophysical J 86:105-115.
252
111. del Val C, White, S., and Bondar, A. 2012. Ser/Thr Motifs in Transmembrane
Proteins: Conservation Patterns and Effects on Local Protein Structure and
Dynamics. J Membrane Biol 245:717-730.
112. Kurochkina N. 2007. Amino acid composition of parallel helix-helix interfaces.
Journal of Theoretical Biology 247:110-121.
113. Bowie J. 2011. Membrane Protein Folding: How Important are Hydrogen Bonds?
Curr Opin Struc Biol 21:42-49.
114. Gray T, and Matthews, B. 1984. Intrahelical hydrogen bonding of serine,
threonine and cysteine residues within alpha-helices and its relevance to
membrane-bound proteins. J Mol Biol 175:75-81.
115. Sansom M, and Weinstein, H. 2000. Hinges, swivels and switches: the role of
prolines in signalling via transmembrane alpha-helices. Trends Pharmacol Sci
21:445-451.
116. Tieleman P, Shrivastava, I., Ulmschneider, M., and Sansom, M. 2001.
Proline-induced hinges in transmembrane helices: possible roles in ion channel
gating. PROTEINS 44:63-72.
117. Adamian L, and Liang, J. 2001. Helix-Helix Packing and Interfacial Pairwise
Interactions of Residues in membrane Proteins. J Mol Biol 311:891-907.
118. Taylor G, and Sanders, A. 2003. Structural criteria for regulation of membrane
fusion and virion incorporation by the murine leukemia virus TM cytoplasmic
domain. Virology 312:295-305.
253
119. Taylor G, and Sanders, A. 1999. The Role of the Membrane-spanning Domain
Sequence in Glycoprotein-mediated Membrane Fusion. Mol Biol of the Cell
10:2803-2815.
120. Boyko V. LM, Gorelick R., Fu W., Nikolaitchik O., Pathak V. 2006.
Coassembly and complementation of Gag proteins from HIV-1 and HIV-2, two
distinct human pathogens. Mol Cell 23:281-287.
121. Browning M. SR, Lew K., Rizvi T. . 2001. Primate and feline lentivirus vector
RNA packaging and propagation by heterologous lentivirus virions. Journal of
Virology 75:5129-5140.
122. Embretson J. TH. 1987. Lack of competition results in efficient packaging of
heterologous murine retroviral RNAs and reticuloendotheliosis virus
encapsidation-minus RNAs by the reticuloendotheliosis virus helper cell line.
Journal of Virology 61:2675-2683.
123. Lee S. BV, Hu W. 2007. Capsid is an important determinant for functional
complementation of murine leukemia virus and spleen necrosis virus Gag
proteins. Virology 360:388-397.
124. Rizvi T. PA. 1993. Simian immunodeficiency virus RNA is efficiently
encapsidated by human immunodeficiency virus type 1 particles. . Journal of
Virology 67:26981-26988.
125. Briggs J, Wilk T., and Fuller S. . 2003. Do lipid rafts mediate virus assembly
and pseudotyping? . J Gen Virol 84:757-768.
254
126. Granoff A. aHG. 1954. Experimental production of combination forms of virus.
IV. Mixed influenza A-Newcastle disease virus infections. PNAS Proc. Soc. Exp.
Biol. Med.:84-88.
127. J. Z. 1982. The pseudotypic paradox. J Gen Virol 63:15-24.
128. Scheiffele P. RA, Wilk T., and Simons K. 1999. Influenza viruses select ordered
lipid domains during budding from the plasma membrane. Journal of Biological
Chemistry 274:2038-2044.
129. Bhattacharya J. RA, Clapham P. 2006. Gag regulates association of human
immunodeficiency virus type 1 envelope with detergent-resistant membranes.
Journal of Virology 80:5292-5300.
130. Bugelski P. MB, Klinkner A., Ventre J., and Hart T. . 1995. Ultrastructural
evidence of an interaction between Env and Gag proteins during assembly of HIV
type 1. AIDS Res Hum Retrovir 11:55-64.
131. P. C. 1996. Direct interaction between the envelope and matrix proteins of HIV-1.
EMBO J 15:5783-5788.
132. Dorfman T. MF, Haseltine W., Gottlinger H. 1994. Role of the matrix protein
in the virion association of the human immunodeficiency virus type 1 envelope
glycoprotein. . Journal of Virology 68:1689-1696.
133. Freed E. MM. 1996. Domains of the human immunodeficiency virus type 1
matrix and gp41 cytoplasmic tail required for envelope incorporation into virions.
Journal of Virology 70:341-351.
134. Lee Y. TX, Cimakasky L., Hildreth J., and Yu X. 1997. Mutations in the
matrix protein of human immunodeficiency virus type 1 inhibit surface expression
255
and virion incorporation of viral envelope glycoproteins in CD4+ T lymphocytes.
Journal of Virology 71:1443-1452.
135. Mammano F. KE, Sodroski J., Bukovsky A., Gottlinger H. . 1995. Rescue of
human immunodeficiency virus type 1 matrix protein mutants by envelope
glycoproteins with short cytoplasmic domains. Journal of Virology 69:3824-3830.
136. Andersen K. DH, Zedeler A. 2006. Murine leukemia virus transmembrane
protein R-peptide is found in small virus core-like complexes in cells. J Gen Virol
87:1583-1588.
137. Vincent N, Genin, C., and Malvoisin, E. 2002. Identification of a conserved
domain of the HIV-1 transmembrane protein gp41 which interacts with
cholesteryl groups. Biochimica et Biophysica Acta 1567:157-164.
138. Mahfoud R. MM, Lingwood C., and Fantini J. 2002. A novel soluble analog of
the HIV-1 fusion cofactor globotriaosylceramide (Gb3) eliminates the cholesterol
requirement for high affinity gp120/Gb3 interaction. Journal of Lipid Rresearch
43:1670-1679.
139. Greenwood A. PJ, Mills T., Nagle F., Epand R., and Nagle-Tristram S. 2008.
CRAC motif peptide of the HIV-1 gp41 protein thins SOPC membranes and
interacts with cholesterol Biochimica et Biophysica Acta 1778:1120-1130.
140. Epand R. SB, and Epand R. 2005. The Tryptopphan Rich Region of HIV gp41
and the Promotion of Cholesterol-Rich Domains. Biochemistry 44:5525-5531.
141. Epand R. SB, and Epand R. 2002. Peptide-Induced Formation of Cholesterol-
Rich Domains. Biochemistry 42:14677-14689.
256
142. Schibli D. MR, and Vogel H. 2001. The Membrane-Proximal Tryptophan-Rich
Region of the HIV Glycoprotein gp41 Forms a Well-Defined Helix in
Dodecylphosphocholine Micelles. Biochemistry 40:9570-9578.
143. Schwarzer R. LI, Gramatica A., Scolari S., Buschmann V., Veit M., and
Herrmann A. 2014. The cholesterol-binding motif of the HIV-1 glycoprotein
gp41 regulates lateral sorting and oligomerization. Cellular Microbiology
16:1565-1581.
144. Alam KK, Chang JL, Burke DH. 2015. FASTAptamer: A Bioinformatic
Toolkit for High-throughput Sequence Analysis of Combinatorial Selections. Mol
Ther Nucleic Acids 4:e230.
145. Loving R. KM, Sjober M., and Garoff H. 2011. Cooperative clevage of the R
peptide in the Env trimer of Moloney murine leukemia virus facilitates its
maturation for fusion competence. Journal of Virology 85:3262-3269.
146. Green N, Shinnick, T., Witte, O., Ponticelli, A., Sutcliffe, J., and Lerner, R.
1981. Sequence-specific antibodies show that maturation of Moloney leukemia
virus envelope polyprotein involves removal of a COOH-terminal peptide. PNAS
78:6023-6027.
147. Gregory D. OGLT, and Johnson M. 2014. Diverse viral glycoproteins as well
as CD4 co-package into the same human immunodeficiency virus (HIV-1)
particles. Retrovirology 11.
148. Ng V. WT, and Arlinghaus R. 1982. Processing of the env gene products of
Moloney murine leukemia virus J Gen Virol 59:329-343.
257
149. Senes A. ED, DeGrado W. 2004. Folding of helical membrane proteins: the role
of polar, GxxxG-like, and proline motifs. Curr Opin Struc Biol 14:465-479.
150. J. ALaL. 2002. Interhelical Hydrogen Bonds and Spatial Motifs in Membrane
Proteins: Polar Clamps and Serine Zippers. PROTEINS 47:209-218.
151. Tatko C. NV, Lear J., and DeGrado W. 2006. Polar Networks Control
Oligomeric Assembly in Membranes. JACS 128:4170-4171.
152. Nordholm J. dSD, Damjanovic J., Dou D., and Daniels R. 2013. Polar
Residues and Their Positional Context Dictate the Transmembrane Domain
Interactions of Influenza A Neuraminidases. Journal of Biological Chemistry
288:10652-10660.
153. Wan W. aM-WJ. 1999. A recurring two-hydrogen-bond motif incorporating a
serine or threonine residues is found at both at alpha-helical N termini and other
situations. J Mol Biol 286:1651-1662.
154. Song L, Sun, Z, Coleman, K, Zwick, M, Gach, J, Wang, J, Reinherz, E,
Wagner, G, and Kim, M. 2009. Broadly neutralizing anti-HIV-1 antibodies
disrupt a hinge-related function of gp41 at the membrane interface. PNAS
106:9057-9062.
155. Sun Z, Oh, K, Kim, M, Yu, J, Brusic, V, Song, L, Qiao, Z, Wang, J, Wagner,
G, and Reinherz, E. 2008. HIV-1 broadly neutralizing antibody extracts its
epitope from a kinked gp41 ectodomain region on the viral membrane. Immunity
28:52-63.
156. Viard M, Parolini, I, Sargiacomo, M, Fecchi, K, Ramoni, C, Ablan, S,
Ruscetti, F, Wang, J, and Blumenthal, R. 2002. Role of cholesterol in human
258
immunodeficiency virus type 1 envelope protein-mediated fusion with host cells.
Journal of Virology 76:11584-11595.
157. Epand R. 2004. Do proteins facilitate the formation of cholesterol-rich domains?
Biochimica et Biophysica Acta 1666:227-238.
158. Shang LaH, E. 2010. Residues in membrane-spanning domain core modulate
conformation and fusogenicity of the HIV-1 envelope glycoprotein. Journal of
Virology 404:158-167.
159. Persons DA AJ, Allay ER, Smeyne RJ, Ashmun RA, Sorrentino BP, Nienhuis
AW. 1997. Retroviral-mediated transfer of the green fluorescent protein gene into
murine hematopoietic cells facilitates scoring and selection of transduced
progenitors in vitro and identification of genetically modified cells in vivo. Blood
1:1777-1786.
160. Klein D IS, von Rombs K, Amadori A, Salmons B, Günzburg WH. 1997.
Rapid identification of viable retrovirus-transduced cells using the green
fluorescent protein as a marker. Gene Ther 4:1256-1260.
161. Welsh S KS. 1997. Reporter gene expression for monitoring gene transfer. Curr
Opin Biotechnol 8:617-622.
162. Cheng L FJ, Tsukamoto A, Hawley RG. 1996. Use of green fluorescent protein
variants to monitor gene transfer and expression in mammalian cells. Nat
Biotechnol 14:606-609.
163. RY. T. 1998. The green fluorescent protein. Annu Rev Biochem 67:509-544.
259
164. Matz MV FA, Labas YA, Savitsky AP, Zaraisky AG, Markelov ML,
Lukyanov SA. 1999. Fluorescent proteins from nonbioluminescent Anthozoa
species. Nat Biotechnol 17:969-973.
165. Shaner NC CR, Steinbach PA, Giepmans BN, Palmer AE, Tsien RY. 2004.
Improved monomeric red, orange and yellow fluorescent proteins derived from
Discosoma sp. red fluorescent protein. Nature Biotechnology 22:1567-1572.
166. Robert E. Campbell OT, Amy E. Palmer, Paul A. Steinbach, Geoffrey S.
Baird, David A. Zacharias,, Tsien aRY. 2002. A monomeric red fluorescent
protein. PNAS 99:7877-7882.
167. Onafuwa-Nuga A TA. 2009. The remarkable frequency of human
immunodeficiency virus type 1 genetic recombination. Microbiol Mol Biol Rev
73:451-480.
168. Galetto R NM. 2005. Mechanistic features of recombination in HIV. AIDS
Review 7:92-102.
169. Krista Delviks-Frankenberry AG, Olga Nikolaitchik, Helene Mens,, Hu
VKPaW-S. 2011. Mechanisms and Factors that Influence High Frequency
Retroviral Recombination. Viruses 3:1650-1680.
170. Li T, Zhang, J. 2000. Determination of the frequency of retroviral recombination
between two identical sequences within a provirus. Journal of Virology 74:7646-
7650.
171. Hu WS BE, Delviks KA, Pathak VK. 1997. Homologous recombination occurs
in a distinct retroviral subpopulation and exhibits high negative interference. J
Virol 71:6028-6036.
260
172. Li TaZ, J. 2001. Retroviral recombination is temperature dependent. Journal of
General Virology 82:1359-1364.
173. Wu X LY, Crise B, Burgess SM. 2003. Transcription start regions in the human
genome are favored targets for MLV integration. Science 300:1749-1751.
174. Lewinski MK, Yamashita, M., Emerman, M., et al. 2006. Retroviral DNA
integration: viral and cellular determinants of target-site selection. PLoS
Pathogens 2:e60.
175. Felice B, Cattoglio, C., Cittaro, D., et al. 2009. Transcription factor binding
sites are genetic determinants of retroviral integration in the human genome. .
PLos One 4:e4571.
176. Bushman FD. 2003. Targeting Survival: Integration Site Selection by
Retroviruses and LTR-Retrotransposons. Cell 115:135-138.
177. Astrid R.W. Schröder PS, Huaming Chen, Charles Berry, Joseph R. Ecker,
Frederic Bushman. 2002. HIV-1 Integration in the Human Genome Favors
Active Genes and Local Hotspots. Cell 110:521-529.
178. An W, Telesnitsky, A. 2002. Effects of varying sequence similarity on the
frequency of repeat deletion during reverse transcription of a human
immunodeficiency virus type 1 vector. J Virol 76:7897-7902.
179. Dahabieh MS OM, Simon V, Sadowski I. 2013. A doubly fluorescent HIV-1
reporter shows that the majority of integrated HIV-1 is latent shortly after
infection. J Virol 87:4716-4727.
261
180. McNamara LA GJ, Collins KL. 2012. Latent HIV-1 infection occurs in multiple
subsets of hematopoietic progenitor cells and is reversed by NF-κB activation. J
Virol 86:9337-9350.
181. Boller T, Felix G. 2009. A renaissance of elicitors: perception of microbe-
associated molecular patterns and danger signals by pattern-recognition receptors.
Annu Rev Plant Biol 60:379-406.
182. Kumar H KT, & Akira S 2011. Pathogen Recognition by the Innate Immune
System. Internat Rev Immunol 30:16-34.
183. Nicaise V, Roux M, Zipfel C. 2009. Recent advances in PAMP-triggered
immunity against bacteria: pattern recognition receptors watch over and raise the
alarm. Plant Physiol 150:1638-1647.
184. Tena G BM, & Sheen J. 2011. Protein kinase signaling networks in plant innate
immunity Curr Opin Plant Biol 14:519-529.
185. Chinchilla D, Bauer Z, Regenass M, Boller T, Felix G. 2006. The Arabidopsis
receptor kinase FLS2 binds flg22 and determines the specificity of flagellin
perception. Plant Cell 18:465-476.
186. Dunning FM SW, Jansen KL, Helft L, & Bent AF 2007. Identification and
Mutational Analysis of Arabidopsis FLS2 Leucine-Rich Repeat Domain Residues
That Contribute to Flagellin Perception. Plant Cell 19:3297-3313.
187. Korasick DA ea. 2010. Novel Functions of Stomatal Cytokinesis-Defective 1
(SCD1) in Innate Immune Responses against Bacteria. J Biol Chem 285:23342-
23350.
262
188. Lu X ea. 2009. Uncoupling of sustained MAMP receptor signaling from early
outputs in an Arabidopsis endoplasmic reticulum glucosidase II allele. . Proc Natl
Acad Sci U S A 106:22522-22527.
189. Zipfel C, Robatzek S, Navarro L, Oakeley EJ, Jones JD, Felix G, Boller T.
2004. Bacterial disease resistance in Arabidopsis through flagellin perception.
Nature 428:764-767.
190. Göhre V ea. 2008 Plant Pattern-Recognition Receptor FLS2 Is Directed for
Degradation by the Bacterial Ubiquitin Ligase AvrPtoB. Curr Biol 18:1824-1832.
191. Robatzek S C, D, and Boller T. 2006. Ligand-induced endocytosis of the pattern
recognition receptor FLS2 in Arabidopsis. Genes and Dev 20:537-542.
192. Lu D, Lin W, Gao X, Wu S, Cheng C, Avila J, Heese A, Devarenne TP, He P,
Shan L. 2011. Direct ubiquitination of pattern recognition receptor FLS2
attenuates plant innate immunity. Science 332:1439-1442.
193. Zipfel C KG, Chinchilla D, Caniard A, Jones JD, Boller T, Felix G. 2006.
Perception of the bacterial PAMP EF-Tu by the receptor EFR restricts
Agrobacterium-mediated transformation. Cell 125:749-760.
194. Chinchilla D, Zipfel C, Robatzek S, Kemmerling B, Nurnberger T, Jones JD,
Felix G, Boller T. 2007. A flagellin-induced complex of the receptor FLS2 and
BAK1 initiates plant defence. Nature 448:497-500.
195. Serrano M ea. 2007. Chemical Interference of Pathogen-associated Molecular
Pattern-triggered Immune Responses in Arabidopsis Reveals a Potential Role for
Fattyacid Synthase Type II Complex-derived Lipid Signals. J Biol Chem
282:6803-6811.
263
196. Trujillo M IK, Casais C, and Shirasu K 2008. Negative Regulation of PAMP
Triggered Immunity by an E3 Ubiquitin Ligase Triplet in Arabidopsis. . Curr Biol
18:1396-1401.
197. Lee H, Bowen, C, Popescu, G, Kang, H, Kato, N, Ma, S, Dinesh-Kumar, S,
Snyder, M, and Popescu, S. (2011) Arabidopsis RTNLB1 and RTNLB2
Reticulon-Like Proteins Regulate Intracellular Trafficking and Activity of the
FLS2 Immune Receptor. . Plant Cell 23:3374-3391.
198. Saijo Y. 2010. ER quality control of immune receptors and regulators in plants. .
Cellular Microbiol Res 12:716-724.
199. Robatzek S. 2007. Vesicle trafficking in plant immune responses. Cell Microbiol
9:1-8.
200. S GNR. 2008. Plant Receptors Go Endosomal: A Moving View on Signal
Transduction. . Plant Physiol 147:1565-1574.
201. Geldner N HD, Wang X, Schumacher K, Chory J. 2007. Endosomal signaling
of plant steroid receptor kinase BRI1. Genes and Dev 21:1598-1602.
202. Irani NG ea. 2012 Fluorescent castasterone reveals BRI1 signaling from the
plasma membrane. . Nat Chem Biol 8:583-589.
203. Gómez-Gómez L FG, Boller T. 1999. A single locus determines sensitivity to
bacterial flagellin in Arabidopsis thaliana. Plant J 18:277-284.
204. Heese A, Hann DR, Gimenez-Ibanez S, Jones AM, He K, Li J, Schroeder JI,
Peck SC, Rathjen JP. 2007. The receptor-like kinase SERK3/BAK1 is a central
regulator of innate immunity in plants. Proc Natl Acad Sci U S A 104:12217-
12222.
264
205. Torres MA, Dangl J, Jones J. 2002. Arabidopsis gp91 phox homologues
Atrbohd and Atrbohf are required for accumulation of reactive oxygen
intermediates in the plant defense response. Proc Natl Acad Sci U S A 99:517-
522.
206. Anderson JC ea. 2011. Arabidopsis MAP Kinase Phosphatase 1 (AtMKP1)
negatively regulates MPK6-mediated PAMP responses and resistance against
bacteria. Plant Journal 67:258-268.
207. Kneen M FJ, Li Y, & Verkman AS 1998. Green Fluorescent Protein as a
Noninvasive Intracellular pH Indicator. Biophysical J 74:1591-1599.
208. Ntoukakis V SB, Segonzac C, and Zipfel C. 2011 Cautionary Notes on the Use
of C-Terminal BAK1 Fusion Proteins for Functional Studies. Plant Cell 23:3871-
3878.
209. Bauer Z G-GL, Boller T, & Felix G 2001. Sensitivity of Different Ecotypes and
Mutants ofArabidopsis thaliana toward the Bacterial Elicitor Flagellin Correlates
with the Presence of Receptor-binding Sites J Biol Chem 276:45669-45676.
210. Denoux C ea. 2008 Activation of Defense Response Pathways by OGs and Flg22
Elicitors in Arabidopsis Seedlings. Mol Plant 1:423-445.
211. Nühse TS B, A.R., Jones, A.M.E., Peck, S.C. . 2007 Quantitative
phosphoproteomic analysis of plasma membrane proteins reveals regulatory
mechanisms of plant innate immune responses. Plant J 51:931-940.
212. Schweighofer A ea. (2007) The PP2C-Type Phosphatase AP2C1, Which
Negatively Regulates MPK4 and MPK6, Modulates Innate Immunity, Jasmonic
Acid, and Ethylene Levels in Arabidopsis. Plant Cell 19:2213-2224.
265
213. Roux M, Schwessinger B, Albrecht C, Chinchilla D, Jones A, Holton N,
Malinovsky FG, Tor M, de Vries S, Zipfel C. 2011. The Arabidopsis leucine-
rich repeat receptor-like kinases BAK1/SERK3 and BKK1/SERK4 are required
for innate immunity to hemibiotrophic and biotrophic pathogens. Plant Cell
23:2440-2455.
214. Sun W, Cao Y, Jansen Labby K, Bittel P, Boller T, Bent AF. 2012. Probing
the Arabidopsis flagellin receptor: FLS2-FLS2 association and the contributions
of specific domains to signaling function. Plant Cell 24:1096-1113.
215. Shan L, He P, Li J, Heese A, Peck SC, Nurnberger T, Martin GB, Sheen J.
2008. Bacterial effectors target the common signaling partner BAK1 to disrupt
multiple MAMP receptor-signaling complexes and impede plant immunity. Cell
Host Microbe 4:17-27.
216. Zhang J, Li W, Xiang T, Liu Z, Laluk K, Ding X, Zou Y, Gao M, Zhang X,
Chen S, Mengiste T, Zhang Y, Zhou JM. 2010. Receptor-like cytoplasmic
kinases integrate signaling from multiple plant immune receptors and are targeted
by a Pseudomonas syringae effector. Cell Host Microbe 7:290-301.
217. Block A, Alfano JR. 2011. Plant targets for Pseudomonas syringae type III
effectors: virulence targets or guarded decoys? Curr Opin Microbiol 14:39-46.
218. Iwasaki A aMR. 2012. Regulation of Adaptive Immunity by the Innate Immune
System. Science 327:291-295.
219. Monaghan J ZC. 2012. Plant pattern recognition receptor complexes at the
plasma membrane. Curr Opin Plant Biol 15:349-357.
266
220. Gomez-Gomez L BT. 2000. FLS2: An LRR Receptor–like Kinase Involved in
the Perception of the Bacterial Elicitor Flagellin in Arabidopsis. Molecular Cell
5:1003-1011.
221. Beck M ZJ, Faulkner C, MacLean D, Robatzek S. 2012. Spatio-temporal
cellular dynamics of the Arabidopsis flagellin receptor reveal activation status-
dependent endosomal sorting. Plant Cell 24:4205-4219.
222. Andersson ER. 2012. The role of endocytosis in activating and regulating signal
transduction. Cell Mol Life Sci 69:1755-1771.
223. Ichimura KC, C. Peck, S. C. Shinozaki, K. Shirasu, K. 2006. MEKK1 is
required for MPK4 activation and regulates tissue-specific and temperature-
dependent cell death in Arabidopsis. J Biol Chem 281:36969-36976.
224. Petersen M, Brodersen, Peter, Naested, Henrik, Andreasson, Erik, Lindhart,
Ursula, Johansen, Bo, Nielsen, Henrik B., Lacy, Michelle, Austin, Mark J.,
Parker, Jane E., Sharma, Sashi B., Klessig, Daniel F., Martienssen, Rob,
Mattsson, Ole, Jensen, Anders B., Mundy, John. 2000. Arabidopsis MAP
Kinase 4 Negatively Regulates Systemic Acquired Resistance. Cell 103:1111-
1120.
225. He P, Shan L, Sheen J. 2007. Elicitation and suppression of microbe-associated
molecular pattern-triggered immunity in plant-microbe interactions. Cell
Microbiol 9:1385-1396.
226. Jens Boch VJ, Lisa Gao, Tara L. Robertson, Melisa Lim, Barbara N. Kunke.
2002. Identification of Pseudomonas syringae pv. tomato genes induced during
infection of Arabidopsis thaliana. Molecular Microbiology 44:73-88.
267
227. Shapiro UJaL. 1996. Cell cycle-controlled proteolysis of a flagellar motor
protein that is asymmetrically distributed in the Caulobacter predivisional cell.
EMBO J 15:2393-2406.
228. Nicaise V, Joe A, Jeong BR, Korneli C, Boutrot F, Westedt I, Staiger D,
Alfano JR, Zipfel C. 2013. Pseudomonas HopU1 modulates plant immune
receptor levels by blocking the interaction of their mRNA with GRP7. EMBO J
32:701-712.
229. Baulcombe NBaDC. 2005. Arabidopsis ARGONAUTE1 is an RNA Slicer that
selectively recruits microRNAs and short interfering RNAs. PNAS 102:11928-
11933.
230. Xiaoming Zhang HZ, Shang Gao, Wei-Chi Wang, Surekha Katiyar-Agarwal,
Hsien-Da Huang, Natasha Raikhel, and Hailing Jin. 2011. Arabidopsis
Argonaute 2 regulates innate immunity via miRNA393*-mediated silencing of a
Golgi-localized SNARE gene MEMB12. Mol Cell 42:356-366.
231. Iwasaki A, Medzhitov R. 2010. Regulation of adaptive immunity by the innate
immune system. Science 327:291-295.
232. Asai T TG, Plotnikova J, Willman M, Chiu W et al. 2002. MAP Kinase
Signaling Cascade in Arabidopsis Innate Immunity. Nature 415:977-983.
233. Vetter MM, Kronholm I, He F, Haweker H, Reymond M, Bergelson J,
Robatzek S, de Meaux J. 2012. Flagellin perception varies quantitatively in
Arabidopsis thaliana and its relatives. Mol Biol Evol 29:1655-1667.
234. Schulze B MT, Jehle A, Mueller K, Beeler S, Boller T, Felix G, Chinchilla D.
2010. Rapid Heteromerization and Phosphorylation of Ligand-activated Plant
268
Transmembrane Receptors and Their Associated Kinase BAK1. JBC 285:9444-
9451.
235. Eugenia Russinova J-WB, Mark Kwaaitaal, Ana Can Delgado, Yanhai Yin,
Joanne Chory,and Sacco C. de Vries. 2004. Heterodimerization and
Endocytosis of Arabidopsis Brassinosteroid Receptors BRI1 and AtSERK3
(BAK1). Plant Cell 16:3216-3229.
236. Wang X KU, He K, Blackburn K, Li J, Goshe MB, Huber SC, Clouse SD.
2008. Sequential transphosphorylation of the BRI1/BAK1 receptor kinase
complex impacts early events in brassinosteroid signaling. Dev Cell 15:220-235.
237. Horvath CA, Vanden Broeck D, Boulet GA, Bogers J, De Wolf MJ. 2007.
Epsin: inducing membrane curvature. Int J Biochem Cell Biol 39:1765-1770.
238. Rosenthal J CH, Slepnev V, Pellegrini L, Salcini A, Di Fiore P, and Camilli P.
1999. The epsins define a family of proteins that interact with components of the
clathrin coat and contain a new protein module. JBC 274:33959-33965.
239. D W. 2002. Epsins: adaptors in endocytosis? Nature Reviews 3:971-977.
240. Ford M MI, Peter B, Vallis Y, Praefcke G, Evans P, McMahon H. 2002.
Curvature of clathrin-coated pits driven by epsin. Nature 419:361-366.
241. Sen A MK, Mukherjee D, Aguilar C. 2012. The epsin protein family:
coordinators of endocytosis and signaling. Biomolecular Concepts 3:117-126.
242. Holstein SE OP. 2005. Sequence analysis of Arabidopsis thaliana E/ANTH-
domain-containing proteins: membrane tethers of the clathrin-dependent vesicle
budding machinery. Protoplasma 226:13-21.
269
243. Song J, Lee MH, Lee GJ, Yoo CM, Hwang I. 2006. Arabidopsis EPSIN1 plays
an important role in vacuolar trafficking of soluble cargo proteins in plant cells
via interactions with clathrin, AP-1, VTI11, and VSR1. Plant Cell 18:2258-2274.
244. S Z, D.F. K. 2001. MAPK cascades in plant defense signaling. Trends Plant Sci
6:520-527.
245. Andrea Pitzschke ASaHH. 2009. MAPK cascade signalling networks in plant
defence. 12:1-6.
246. Luna E PV, Robert J, Flors V, Mauch-Mani B, Ton J. 2011. Callose
deposition: a multifaceted plant defense response. Mol Plant Microbe Interact
24:183-193.
247. Jacobs AK LV, Burton RA, Panstruga R, Strizhov N, Schulze-Lefert P,
Fincher GB. 2003. An Arabidopsis callose synthase, GSL5, is required for
wound and papillary callose formation. Plant Cell 15:2503-2513.
248. Kim MG dCL, McFall AJ, Belkhadir Y, DebRoy S, Dangl JL, Mackey D.
2005. Two Pseudomonas syringae type III effectors inhibit RIN4-regulated basal
defense in Arabidopsis. Cell 121:749-759.
249. Nishimura MT SM, Hou BH, Vogel JP, Edwards H, Somerville SC. 2003.
Loss of a callose synthase results in salicylic acid-dependent disease resistance.
Science 301:969-972.
250. Zhang J SF, Li Y, Cui H, Chen L, Li H, Zou Y, Long C, Lan L, Chai J, et al.
2007. A Pseudomonas syringae effector inactivates MAPKs to suppress PAMP-
induced immunity in plants. Cell Host Microbe 1:175-185.
270
251. Mary C. Wildermuth JD, Gang Wu, Frederick M. Ausubel. 2001.
Isochorismate synthase is required to synthesize salicylic acid for plant defence.
Nature 414:562-565.
252. Nawrath C MJ. 1999. Salicylic acid induction-deficient mutants of Arabidopsis
express PR-2 and PR-5 and accumulate high levels of camalexin after pathogen
inoculation. Plant Cell 11:1393-1404.
253. N Zhou TLT, F Tsui, D F Klessig, and J Glazebrook. 1998. PAD4 functions
upstream from salicylic acid to control defense responses in Arabidopsis. Plant
Cell 10:1021-1030.
254. Jiyoung Lee JN, Hyeong Cheol Park, Gunnam Na, Kenji Miura, Jing Bo Jin,
Chan Yul Yoo, Dongwon Baek, Doh Hoon Kim, Jae Cheol Jeong, Donggiun
Kim, Sang Yeol Lee, David E. Salt, Tesfaye Mengiste, Qingqiu Gong, Shisong
Ma, Hans J. Bohnert, Sang-Soo Kwak, Ray A. Bressan, Paul M. Hasegawa
and Dae-Jin Yun. 2006. Salicylic acid-mediated innate immunity in Arabidopsis
is regulated by SIZ1 SUMO E3 ligase. The Plant Journal 49:79-90.
255. Arup K. Mukherjeea M-JC, Rina Zuchmanc, Tamar Zivc, Benjamin A.
Horwitza, Shimon Gepsteina. 2010. Proteomics of the response of Arabidopsis
thaliana to infection with Alternaria brassicicola. J Proteomics 73:709-720.
256. Gimenez-Ibanez S HD, Ntoukakis V, Petutschnig E, Lipka V, and Rathjen P.
2009. AvrPtoB Targets the LysM Receptor Kinase CERK1 to Promote Bacterial
Virulence on Plants. Current Biology 19:423-429.
271
257. Miya A AP, Shinya T, Desaki Y, Ichimura K, Shirasu K, Narusaka Y,
Kawakami N, Kaku H, and Shibuya N. 2007. CERK1, a LysM receptor kinase,
is essential for chitin elicitor signaling in Arabidopsis. PNAS 104:19613-19618.
258. Li J, Wen, J., Lease, K.A., Doke, J.T., Tax, F.E., and Walker, J.C. 2002.
BAK1, an Arabidopsis LRR receptor-like protein kinase, interacts with BRI1 and
modulates brassinosteroid signaling. Cell 110:213-222.
259. Nam KH, and Li, J. 2002. BRI1/BAK1, a receptor kinase pair mediating
brassinosteroid signaling. Cell 110:203-212.
260. S. Casson AH. 2012. GSK3-Like Kinases Integrate Brassinosteroid Signaling and
Stomatal Development. Perspec in Plant Biol 5:e30.
261. Tae-Wuk Kim MM, Dominique C. Bergmann & Zhi-YongWang. 2012.
Brassinosteroid regulates stomatal development by GSK3-mediated inhibition of
a MAPK pathway. Nature 482:419-423.
262. Nuria Andrés-Colás AP-G, Sergi Puig and Lola Peñarrubia. 2010.
Deregulated Copper Transport Affects Arabidopsis Development Especially in
the Absence of Environmental Cycles. Plant Physiology 153:170-184.
263. Marie Barberona EZ, Stéphanie Robertb, Geneviève Conéjéroa, Cathy
Curiea, Jìří Friml and Grégory Verta. 2011. Monoubiquitin-dependent
endocytosis of the IRON-REGULATED TRANSPORTER 1 (IRT1) transporter
controls iron uptake in plants. PNAS 108:E450-E458.
264. Enric Zelazny MB, Catherine Curie, and Grégory Vert. 2011. Ubiquitination
of transporters at the forefront of plant nutrition. Plant Signal Behav 6:1597-1599.
272
265. Marhavý P BA, Abas L, Abuzeineh A, Duclercq J, Tanaka H, Pařezová M,
Petrášek J, Friml J, Kleine-Vehn J, Benková E. 2011. Cytokinin modulates
endocytic trafficking of PIN1 auxin efflux carrier to control plant organogenesis.
Dev Cell 21:796-804.
266. Jürgen Kleine-Vehna PD, Ranjan Swarupb, Malcolm Bennett and Jiří Friml.
2006. Subcellular Trafficking of the Arabidopsis Auxin Influx Carrier AUX1
Uses a Novel Pathway Distinct from PIN1. Plant Cell 18:3171-3181.
267. Gregory R Lampard DCB. 2007. A Shout-Out to Stomatal Development: How
the bHLH Proteins SPEECHLESS, MUTE and FAMA Regulate Cell Division
and Cell Fate. Plant Signal Behav 2:290-292.
268. Hetherington JEGaAM. 2004. Plant Development: YODA the Stomatal Switch.
Curr Biol 14:488-490.
269. Neela S. Bhave KMV, Jeanette A. Nadeau, Jessica R. Lucas, Sanjay L.
Bhave, and Fred D. Sack. 2009. TOO MANY MOUTHS promotes cell fate
progression in stomatal development of Arabidopsis stems. Planta 229:357-367.
270. Serna L. 2009. Cell fate transitions during stomatal development. BioEssays
31:865-873.
271. Gudesblat GE. 2012. SPEECHLESS integrates brassinosteroid and stomata
signalling pathways. Nature cell biology 14.
272. Albrecht C, Boutrot F, Segonzac C, Schwessinger B, Gimenez-Ibanez S,
Chinchilla D, Rathjen JP, de Vries SC, Zipfel C. 2012. Brassinosteroids inhibit
pathogen-associated molecular pattern-triggered immune signaling independent of
the receptor kinase BAK1. Proc Natl Acad Sci U S A 109:303-308.
273
273. Belkhadir Y JY, Epple P, Balsemao-Pires E, Dangl J.L., Chory J. 2012.
Brassinosteroids modulate the efficiency of plant immune responses to microbe-
associated molecular patterns PNAS 109:297-302.
274. Lu Y-C YW-C, Ohashi P. 2008. LPS/TLR4 signal transduction pathway.
Cytokine 42:145-151.
275. Petnicki-Ocwieja T, Schneider DJ, Tam VC, Chancey ST, Shan L, Jamir Y,
Schechter LM, Janes MD, Buell CR, Tang X, Collmer A, Alfano JR. 2002.
Genomewide identification of proteins secreted by the Hrp type III protein
secretion system of Pseudomonas syringae pv. tomato DC3000. Proc Natl Acad
Sci U S A 99:7652-7657.
276. Sunghwa Choe BPD, Shozo Fujioka, Suguru Takatsuto, Akira Sakurai, and
Kenneth A. Feldmann. 1998. The DWF4 Gene of Arabidopsis Encodes a
Cytochrome P450 That Mediates Multiple 22 a-Hydroxylation Steps in
Brassinosteroid Biosynthesis. The Plant Cell 10:231-243.
274
VITA
Daniel Salamango was born March 9, 1987 and has always been fascinated by
science. In the beginning, Dan knew he wanted to pursue a career in science and set out
to earn a bachelor’s degree in chemistry from Hope College. Along the way he was
exposed to the field of biochemistry and immediately fell in love with it. Soon after, he
was exposed to research and the rest was history. Dan has research experience in both
plant innate immunity and retroviral assembly.