University of Southern California Dissertations and Theses

Study of protein deamidation in innate immune signaling

(USC Thesis Other)

Study of protein deamidation in innate immune signaling

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

Study of protein deamidation in innate immune signaling

A Dissertation

by

Jun Zhao

Submitted to the Graduate School of
University of Southern California
in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

May 2016

Major: Genetic, Molecular and Cellular Biology
Mentor: Pinghui Feng, Ph.D.
Committee members: Shoujiang Gao, Ph.D. (Chair)
Jae U Jung, Ph.D.
Ebrahim Zandi, Ph.D.
2

Abstract
Innate immunity is the first line of defense against foreign pathogens. In response
to viral infection, the retinoic acid-inducible gene I (RIG-I) senses viral RNA and
activates the interferon regulatory factor (IRF) and nuclear factor κ of B cell (NF-
κB), leading to the production of type I interferons and pro-inflammatory
cytokines. Protein deamidation is a poorly characterized post-translational
modification. Emerging studies implicate protein deamidation in regulating
fundamental biological processes. We recently reported that gamma-
herpesviruses activate RIG-I via deamidation to evade antiviral cytokine
production. The gamma herpesvirus homologues of glutamine amidotransferase
(vGAT) interact with cellular phosphoribosylformylglycinamidine synthetase
(PFAS) to deamidate RIG-I, indicating that deamidase-catalyzed protein
deamidation plays essential roles in regulating innate immune signaling. In this
thesis, we characterized the regulatory roles of deamidation of the RIG-I receptor
and the NF-κB transcription factor RelA (also known as p65) that are mediated
by herpes simplex virus 1 (HSV-1) UL37 and cellular CAD deamidase,
respectively.
We demonstrated that infection of HSV-1 induced RIG-I deamidation and
prevented RIG-I activation by Sendai virus, a prototype RNA virus known to
activate RIG-I. A focused screen identified the tegument protein UL37 as the
major RIG-I-interacting protein and UL37 was sufficient to induce RIG-I
deamidation and block RIG-I activation. Biochemical analyses identified two
asparigines, N495 and N549, of the central helicase domain (Hel2i) that were
3

deamidated by UL37 expression and HSV-1 infection. Deamidation significantly
diminished the RNA-sensing activity, ATPase activity and antiviral activity of RIG-
I. Conversely, uncoupling RIG-I deamidation via mutations that generated
deamidation-resistant RIG-I or deamidation-deficient UL37 resulted in more
robust antiviral immune response and reduced HSV-1 replication. Altogether, this
study demonstrates that deamidation is a novel mechanism to evade RIG-I-
mediated antiviral immunity by HSV-1, and suggests that UL37 may function as a
protein deamidase.
Cellular PFAS, when complexed with gamma herpesvirus vGAT, deamidates
RIG-I in vitro, demonstrating that cellular metabolic glutamine amidotransferases
(GATs) possess intrinsic protein-deamidating activity. Employing an unbiased
screening targeting the GAT-containing enzymes, we discovered that CAD
(carbamoyl-phosphate synthetase 2, aspartate transcarbamylase, and
dihydroorotase) potently deamidated RelA and diminished NF-κB activation.
Mass spectrometry analysis identified two asparagines within the insertion region
of RelA that were deamidated by CAD. Biochemical analyses indicated that CAD
is a bona fide deamidase of RelA. Interestingly, deamidated RelA demonstrated
normal affinity for classic NF-κB-responsive element in vitro, while failed to
induce pro-inflammatory responses in cells upon stimulation. This work
collectively defined CAD as a novel deamidase that deamidates RelA to dampen
NF-κB activation.
In summary, I have discovered a cellular GAT and a potential viral deamidase
that regulate innate immune signaling via deamidating key components. This
4

study also defines two novel protein deamidations in mammalian cells and their
fundamental roles in host innate defense and viral immune evasion.

5

Table of Contents
Acknowledgements ...................................................................................................................... 8
1. Introduction ................................................................................................................................ 9
1.1 Innate immune signaling ................................................................................................... 9
1.1.1 Pattern recognition receptors .................................................................................... 9
1.1.2 IRF signaling ................................................................................................................ 9
1.1.3 NF-κB signaling ......................................................................................................... 10
1.1.4 Type I Interferon ........................................................................................................ 12
1.1.5 Interferon-stimulated genes .................................................................................... 13
1.2 RIG-I-dependent IFN signaling ...................................................................................... 14
1.2.1 RIG-I ........................................................................................................................... 15
1.2.2 RIG-I structure and activation mechanism ............................................................ 16
1.2.3 RIG-I regulation ......................................................................................................... 18
1.3 Regulation of NF-κB signaling ........................................................................................ 19
1.3.1 Cellular regulation and viral manipulation of NF-κB signaling ........................... 19
1.3.2 Post-translational modification of RelA .................................................................. 20
1.4 Protein Deamidation and Deamidase ........................................................................... 22
1.4.1 Protein Deamidation ................................................................................................. 22
1.4.2 Deamidase ................................................................................................................. 24
1.4.3 Glutamine Amidotransferases ................................................................................ 24
2. Deamidation of RIG-I by Herpes Simplex Virus 1 (HSV-1) UL37 to evade dsRNA
induced IFN-β production .......................................................................................................... 27
2.1 Introduction ....................................................................................................................... 27
2.2 Results ............................................................................................................................... 28
2.2.1 HSV-1 infection inhibits RIG-I activation. .............................................................. 28
2.2.2 UL37 interacts with RIG-I. ....................................................................................... 29
2.2.3 UL37 inhibits RIG-I activation and RIG-I-dependent IFN signaling. ................. 30
2.2.4 UL37 deamidates RIG-I at N495 and N549.......................................................... 32
2.2.5 Deamidation desensitizes RIG-I to dsRNA and impairs its ability to activate
IFN signaling ........................................................................................................................ 33
2.2.6 HSV-1 carrying deamidation-deficient UL37 mutations induce stronger IFN
responses. ........................................................................................................................... 35
6

2.2.7 Deamidation-resistant RIG-I-QQ is activated by HSV-1 and induces stronger
IFN responses. .................................................................................................................... 36
2.3 Discussion ......................................................................................................................... 37
3. Deamidation of RelA by CAD to regulate NF-κB response ............................................. 70
3.1 Introduction ....................................................................................................................... 70
3.2 Results ............................................................................................................................... 70
3.2.1 CAD negatively regulates RelA induced NF-κB responses. .............................. 70
3.2.2 CAD interacts with RelA. ......................................................................................... 72
3.2.3 CAD deamidates RelA. ............................................................................................ 73
3.2.4 Deamidation of RelA at N64 and N139 impairs its ability to transactivate NF-
κB downstream gene expression. .................................................................................... 74
3.3 Discussion ......................................................................................................................... 75
4. Materials and Methods .......................................................................................................... 91
4.1 Cell lines and Viruses ...................................................................................................... 91
4.2 Constructs ......................................................................................................................... 91
4.3 Antibodies and reagents ................................................................................................. 92
4.4 Lentivirus-mediated Stable Cell Line Construction ..................................................... 92
4.5 Dual-Luciferase Reporter Assay .................................................................................... 93
4.6 Protein expression and purification ............................................................................... 93
4.7 RNA mobility shift assay ................................................................................................. 94
4.8 Co-immunoprecipitation (Co-IP) and Immunoblotting ................................................ 94
4.9 IRF3 Dimerization assay ................................................................................................. 95
4.10 RIG-I Purification and Gel Filtration ............................................................................ 95
4.11 Whole cell lysate Gel Filtration .................................................................................... 96
4.12 Two-dimensional Gel Electrophoresis ........................................................................ 96
4.13 Quantitative Real-time PCR (qRT-PCR) .................................................................... 96
4.14 In vitro Deamidation Assay ........................................................................................... 98
4.15 In Vitro ATPase Activity Assay .................................................................................... 99
4.16 Mass Spectrometric Analysis ....................................................................................... 99
4.17 Statistical Analysis ....................................................................................................... 101
5. References ............................................................................................................................ 102

7

List of Figures
Figure 1 ........................................................................................................................................ 43
Figure 2 ........................................................................................................................................ 45
Figure 3 ........................................................................................................................................ 47
Figure 4 ........................................................................................................................................ 49
Figure 5 ........................................................................................................................................ 51
Figure 6 ........................................................................................................................................ 53
Figure 7 ........................................................................................................................................ 55
Figure 8 ........................................................................................................................................ 57
Figure 9 ........................................................................................................................................ 59
Figure 10 ...................................................................................................................................... 61
Figure 11 ...................................................................................................................................... 63
Figure 12 ...................................................................................................................................... 65
Figure 13 ...................................................................................................................................... 67
Figure 14 ...................................................................................................................................... 69
Figure 15 ...................................................................................................................................... 80
Figure 16 ...................................................................................................................................... 82
Figure 17 ...................................................................................................................................... 84
Figure 18 ...................................................................................................................................... 86
Figure 19 ...................................................................................................................................... 88
Figure 20 ...................................................................................................................................... 90

8

Acknowledgements
This dissertation would not have been possible without the help of so many
people. It is a great pleasure to take this opportunity to express my gratitude for
them. First of all, I would like to express the deepest appreciation to my advisor
Professor Pinghui Feng, who has provided me with mentorship in designing and
completing the projects. I would never forget his continuous support and
encouragement throughout my doctoral years. Pinghui has been my inspiration
as I hurdle all the obstacles in the completion of this dissertation. I would like to
thank my committee members, Professor Jae Jung, Professor Shoujiang Gao
and Professor Ebrahim Zandi, who have helped and advised me throughout the
different phases of this dissertation study. And I really appreciate Professor
Weiming Yuan, Professor Takeshi Saito and Professor Chengyu Liang for
providing me with valuable comments and suggestions. I would like to express
my gratitude to my parents and my wife for their unconditional support and love,
without which I would have not been able to accomplish this dissertation. Last but
not least, I would also like to thank Dr. Shanping He, Dr. Zhiheng He, Dr. Junjie
Zhang, Dr. Junhua Li, Arlet Minassian, Yuqi Wang, Jackson Cabo, Katie Lee, Jie
Chen and Simin Xu who helped me in my dissertation work.

9

1. Introduction
1.1 Innate immune signaling
1.1.1 Pattern recognition receptors
Innate immunity is the first line of defense against foreign pathogens. Upon
microbial infection, innate immune signaling utilizes pattern recognition receptors
(PRR) to recognize pathogen-associated molecular pattern (PAMPs) and
activate interferon-regulatory factor (IRF) and nuclear factor kappa-light-chain-
enhancer of activated B cells (NF-κB) signaling, thereby coordinate to produce
type I interferon (IFN) to elicit strong antiviral defense and pro-inflammatory
cytokines to guide the subsequent adaptive immune responses. These receptors
have the characteristics of distinguishing between 'self' and 'non-self' structural
components
1, 2
. PRRs discovered so far could be classified into membrane
anchored receptors (Toll-like receptors, C-type lectin receptors, etc) and cytosolic
receptors (NOD-like receptors, RIG-I-like receptors and DNA sensors)
3, 4
.
Besides IRF and NF-κB signaling, certain PRRs can induce activation of
inflammsome (e.g. NLRP3)
5
, pyroptosis (e.g. caspase 4/5)
6
and apoptosis (e.g.
RIG-I/MDA5 in melanoma)
7
.
1.1.2 IRF signaling
Interferon-regulatory factor (IRF) family members are master transcription factors
of interferon-induced genes. Mammalian IRF family comprises nine members,
from IRF1 to IRF9
8
. These IRFs contain an N-terminal DNA binding domain and
C-terminal regulatory regions responsible for binding to their transcription
partners. The N-terminal DNA binding domain recognizes a consensus DNA
sequence known as the interferon-stimulated response element (ISRE)
8
. ISRE is
10

found in the promoter region of most type I interferon-induced antiviral genes and
also in the type I interferons (IFN-α/β) themselves
9
. IRF was brought into sight
with the discovery of pattern-recognition receptors, as it was shown to be
activated in parallel to NF-κB. The coordination of these two transcription factors
enable the initial production of type I interferons, while IRFs continue to serve as
transcription factors in response to type I interferon and transactivate the
expression of thousands of interferon-stimulated genes, eliciting and amplifying a
potent host antiviral defense.
IRF3 and IRF7, which are homologous to each other, represent two key
regulators of type I interferon production induced by virus. IRF3 is constitutively
expressed and activated by C-terminal phosphorylation via TBK1/IKKε complex
(as will be discussed later) upon viral infection
10, 11
. Phosphorylation of IRF3
enables its homodimerization and nuclear translocation
12
. In the nucleus, IRF3
dimer would further recruit histone acetyltransferase p300 or CBP which
increases the accessibility of chromatin to allow the binding of IRF3 and NF-κB to
their cognate sites
13
. Unlike IRF3, the protein level of IRF7 is low in resting cells
but induced by ISGF3 following type I interferon stimulation
14-16
. These
observations lead to the model that IRF3 is responsible for the initial round
induction of type I IFNs, where IRF7 mediates and amplifies the later phase via a
positive-feedback loop
17
.
1.1.3 NF-κB signaling
Nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) proteins
are transcription factors (TFs) sharing Rel homology domain which regulate the
11

transcription of genes downstream of κB promoters
18-20
. The NF-κB-responsive
genes encode a large array of effectors in the host antiviral immune system.
Most of these genes play pro-inflammatory roles in which they upregulate of
adhesion molecules, mediate the recruitment and activation of inflammatory
effector cells, drive the proliferation of B-cells and so on
21
. Thus, NF-κB signaling
is the central hub of initiating host inflammatory responses. Precise activation,
termination and regulation of NF-κB is the key to maintain a normally functional
immune system while aberrant activation/termination of NF-κB signaling is tightly
associated with all types of autoimmune diseases and cancers
22
. NF-κB signaling
is classified into canonical (inflammatory related) and non-canonical pathways,
with the later associating with cell differentiation and development
23
. The thesis
project focuses on the canonical pathway.
The mammalian NF-κB family consists of NF-κB1 (p50), NF-κB2 (p52), RelA
(p65), RelB and c-Rel, all of which contain a Rel homology domain (RHD). NF-κB
proteins usually appear as homo/hetero-dimers and the combination of RelA
(p65) - NF-κB1 (p50) is the most common prototype. RelA, RelB and c-Rel
contains transactivation domain (TAD) at their c-terminus and are all considered
to be trans-activator of NF-κB genes
24
, while p50 and p52, lacking the TAD
domain, either behave as a repressor (such as p50-p50 homodimer) or alter the
binding specificity when forming heterodimers with TAD containing members
25
.
In resting cells, NF-κB TFs are kept as an inactive dimer sequestered by
inhibitors of NF-κB (IκBs) in the cytoplasm
26
. NF-κB signaling is activated in
response to different stimuli, including reactive oxygen species (ROS)
27
, tumor
12

necrosis factor alpha (TNF-α)
28
, interleukin 1-beta (IL-1β)
29
, bacterial
lipopolysaccharides (LPS)
30
and so on. Upon viral infection, PRRs, such as Toll-
like receptors (TLRs), RIG-like receptors (RLRs) and DNA sensors, recognize
pathogen-associated molecular pattern (PAMPs). When bound to cognate
PAMPs, TLR3 will recruit TRIF, while the other TLRs signal through myeloid
differentiation primary response gene 88 (MyD88)
1
. RIG-I-like receptors, RIG-I
and MDA5, signal via MAVS
31-34
. DNA sensors, including cyclic GMP-AMP
synthase (cGAS)
35
, interferon-γ inducible protein 16 (IFI16)
36
, DDX41
37
and DAI
38
,
senses viral DNA and activating downstream NF-κB signaling via STING
adaptor
4
. All these upstream signaling events merge at the activation of IKK
kinase complex, consisting of IKK , IKK  and IKKγ (also known as NEMO)
39
.
Activated IKK phosphorylates IκBs at Ser32, e.g., IκB , which primes IκBs for
ubiquitination by SCFβ-TrCP complex and subsequent degradation by the 26S
proteasome
40
. Degradation of IκBs will release the NF-κB dimer which then
translocates into the nucleus and bind to κB sites with several transcription co-
activators, initiating the transcription of hundreds of pro-inflammatory genes.
1.1.4 Type I Interferon
Activation of innate immune signaling triggers the transcription of IRF and NF-κB
downstream genes, resulting in the production of cytokines, chemokines and
interferons (IFNs). IFNs could be classified into type I (IFNα/β), type II (IFNγ) and
less-characterized type III
41
. Compared with IFNγ which has weak antiviral
effects and serves as regulators of adaptive immune responses
42
, type I
interferons elicit strong antiviral activities as they induce the transcription of
13

hundreds of interferon-stimulated genes (ISGs)
43
. These ISGs either exert direct
inhibition on viral life cycle (Viral entry, replication, particle assembly, egress, etc)
or further reinforce the IFN responses. In this study, we focus on type I interferon
signaling and term it ‘IFN’ in the result part.
Mechanistically, pattern recognition receptor activation triggers the production of
IFNα and IFNβ in an autocrine manner, which these type I interferons bind to
their cognate receptors IFNAR1 and IFNAR2 on cell surface. This binding event
brings tyrosine kinases Jak1 and Tyk2 in close proximity, enabling
autophosphorylation and trans-activation of the two kinases. STAT1 is then
recruited to the complex and forms STAT1-STAT2 heterodimer with the receptor-
bound STAT2
44-46
. The heterodimer gets released from the receptor and
translocates into the nucleus where it associates with the IFN regulatory factor 9
(IRF9) to form the heterotrimeric complex termed IFN-stimulated gene factor 3
(ISGF3)
47, 48
. ISGF3 binds to the interferon-stimulated responses elements (ISRE)
region of the ISGs and trans-activate their expression.
1.1.5 Interferon-stimulated genes
Interferon-stimulated genes (ISGs) form a diversified yet overlapping antiviral
network
49
. Typically, 200-500 genes are induced by the IFN signaling. However,
the functions of most ISGs remain poorly studied. Several well-characterized
ISGs, including PKR
50, 51
, MX protein
52
, OAS1
53
, APOBEC3G
54
, TRIM5
55
,
ISG15
56
, IFITM
57
, etc, have been reported to significantly inhibit viral life cycle
58
.
PKR, MX1, OAS1 are classical examples of ISGs which have potent antiviral
activities.
14

Upon stimulation by viral double-stranded RNA (dsRNA), PKR get activated
through dimerization and auto-phosphorylation
59-63
. Activated PKR will then
phosphorylates the eukaryotic translation initiation factor EIF2A, resulting in
globally inhibition on mRNA translation, thereby thwarting viral protein synthesis
64
.
Mx proteins (encoded by MX1/2 gene) are dynamin-like GTPases induced by
interferons. It’s been reported to target a wide range of RNA viruses (IAV, HBV,
LACV, etc) and DNA viruses
52, 58
. Mx protein antagonizes viral trafficking into the
nucleus, sequesters viral nucleoprotein required for viral genome amplification
and enhances ER stress-mediated cell death in response to IAV infection
65-67
.
OAS1 is another interferon-induced and dsRNA-activated protein. When
activated, OAS1 catalyzes the formation of 2’-5’-oligoadenylates (2-5A)
50, 68
. 2-5A
binds to the inactive monomeric form of ribonuclease L (RNase L), allowing its
dimerization and activation
69
. Activated RNase L degrades cellular mRNA and
viral RNA, terminating viral replication. The digested RNAs are also reported to
activate dsRNA sensor RIG-I to further amplify the interferon responses
70
.
Apart from direct inhibition on viral life cycle, type I interferon also induces ISGs
to further reinforce its own signaling. This includes interferons regulatory factors
(IRF1, IRF7, etc), pattern recognition receptors (RIG-I, MDA5, IFI16, etc) and so
on
43
.
1.2 RIG-I-dependent IFN signaling
In the context of virus, viral nucleotides are the main PAMPs being sensed by
these receptors. RIG-I-like receptors, e.g., retinoic acid-inducible gene 1 (RIG-I)
71

15

and melanoma differentiation-associated protein 5 (MDA-5)
72
, sense the
presence of virus derived double stranded RNA (dsRNA) bearing terminal
signature structures and signal via mitochondria antiviral-signaling protein
(MAVS)
31-34
. MAVS then recruits different adaptor proteins to activate IRF and
NFκB signaling. On one hand, MAVS recruits TNF receptor-associated factors
(TRAF) and transforming growth factor β-activated kinase 1 (TAK1)
73
. TRAFs
such as TRAF6, catalyzes the synthesis of K63-linked polyubiquitin chains. K63
polyUb will then recruit and activate the IKK kinase complex which consists of
IKKα, IKKβ and IKKγ (also known as NEMO)
39
. Activated IKK phosphorylates IκB
inhibitors, which primes IκBs for degradation by the 26S proteasome
40
.
Destruction of IκBs releases the NF-κB dimer (usually p65 and p50 heterodimer)
which then translocates into the nucleus. On the other hand, MAVS recruits and
activates TBK1/IKKε kinase complex by unknown mechanism, which requires
TRAF molecules including TRAF3
74
. TBK1/IKKε then phosphorylates MAVS at its
conserved serine and threonine clusters, allowing MAVS to recruit IRF3 via
binding to its positively charged surface
75
. IRF3 is then phosphorylated by
TBK1/IKKε, followed by self-dimerization and translocation into the nucleus
76
.
IRF3 and NF-κB will then bind to their cognate sites to enable the expression of
type I interferons.
1.2.1 RIG-I
Retinoic acid-inducible gene 1(RIG-I) was firstly discovered as a gene induced by
treatment with retinoic acid in acute promyelocytic leukemia cells
77
. It was later
proved to be an IFN-β inducible protein which demonstrates potent antiviral
activities
71
. The essential role of RIG-I to induce IFN-β production was gradually
16

appreciated since then. RIG-I, together with its closely related protein, Melanoma
Differentiation-associated protein 5 (MDA5), are genuine cytoplasmic sensors of
virus-derived double-stranded RNAs (dsRNA), which is derived during viral
replication of most positive-stranded RNA viruses and some DNA viruses
78
.
Negative-stranded RNA viruses, which don’t generate double-stranded RNAs,
activate RIG-I by a 5’ triphosphate short blunt-end dsRNA structure contained in
the panhandle of their viral genomes
79
. RIG-I’s role in antagonizing RNA viruses
including Sendai virus
80
, Newcastle disease viruses
80
, Measles virus
81
, Influenza
virus
82, 83
are wildly studied in both cell lines and KO mice.
The ability of RIG-I to distinguish between self versus non-self RNAs was thought
to depend on its requirement of a triphosphate group at the 5’-terminus of the
RNA (5’ppp-RNA)
84-86
. Recent study showed that RIG-I also recognizes
reoviruses derived disphosphate RNA (5’pp-RNA) and mediates antiviral
responses
87
. Current view of an ideal RIG-I ligand is a 5’blunt, double-stranded
RNA bearing signature phosphate group at its 5’ terminus (either 5’ppp or 5’pp).
Apart from viral infection derived dsRNA by-product, RIG-I was found to be
activated by cellular generated small RNAs to amplify the antiviral response.
Mechanistically, antiviral endoribonucleases RNaseL gets activated during viral
infection and produces small RNA fragments from cellular RNA which contributes
to IFN-β production via RIG-I and MDA5
70
.
1.2.2 RIG-I structure and activation mechanism
RIG-I consists of two N-terminal CARD domains, a central Helicase domain and
a C-terminal domain/Repressor domain
71, 88
.
17

In vitro studies demonstrated that RIG-I activates downstream signaling by
promoting filament formation of its downstream adaptor protein MAVS
89, 90
.
CARD domain is the effector domain which nucleates the formation of MAVS
filament. Structure studies showed that the tetrameric two CARD domains of the
activated RIG-I tetramer adopt a helical structure, acting as a template and
recruit the CARD domain of MAVS to stack along its helical trajectory
91
. This
mechanism is term ‘imprinting’.
RIG-I has two genuine DExD/H RNA Helicase domains which have ATPases
activities
92
. Rather than unwinding annealed double stranded RNA, hydrolysis of
ATP facilitates the movement the RIG-I on the dsRNA which was found to be
required for RIG-I to transduce signaling, as ATPase mutations abolished IFN-β
signaling via RIG-I
71
. However, other studies proposed that binding of ATP is
sufficient for RIG-I to signal whereas ATP hydrolysis promotes RIG-I recycling on
RIG-I/dsRNA complexes, serving as an alternative mechanism to distinguish
between self and non-self RNAs
92
.
In resting cells, RIG-I is kept in a self-inhibitory state. The repressor domain (RD)
at the carboxyl terminus interacts with the two CARD and the central helicase
domains
88
. The C-terminal domain (CTD), which partially overlaps with the
repressor domain, contains a positive cleft which accounts for the binding of RIG-
I to 5’ triphosphate group of dsRNA
93
. Current model of RIG-I activation is: (1)
CTD of RIG-I binds to the 5’ phosphate group (5’ppp or 5’pp) of the viral
infection-derived dsRNA; (2) this binding then triggers the conformational change
of RIG-I which requires binding and hydrolysis of ATP; (3) conformational change
18

drives RIG-I to tetramerize itself on dsRNA while exposing its 2CARD domain; (4)
tetrameric 2CARD domain nucleates the oligomerization of MAVS and induces
its activation
94
.
1.2.3 RIG-I regulation
While RIG-I-IFN signaling plays pivotal role to antagonize viral replication, its
regulation is essential to maintain cellular homeostasis. RIG-I, as the most
upstream sensor of this pathway, undergoes multiple layers of positive/negative
regulations. K63 ubiquitination is one of the most important modifications
required for RIG-I activation. TRIM25, the E3 ligase specifically required for RIG-I
activation, catalyzes the formation of covalent/non-covalent K63 ubiquitin
chains
95
. The K63 ubiquitin chains are essential in stabilizing CARD-CARD
interaction between RIG-I and MAVS. Concomitantly, both cells and some
viruses apply Several de-ubiquitinating enzymes (DUB) to negatively regulate
RIG-I signaling, including A20
96
, CYLD
97
and ORF64 (KSHV)
98
. Cellular E3
ligase RNF125 also negatively regulates RIG-I by K48 ubiquitination and induces
its proteosomal degradation
99
. In resting cells, PKC-α/β phosphorylates RIG-I to
secure its inactive form
100
. Thus, dephosphorylation by PP1 phosphatase is a
prerequisite step of RIG-I activation
101
. Measles virus V protein blocks this action
of the PP1 phosphatase to suppress RIG-I-dependent innate immune
responses
102
. RIG-I is also reported to be negatively regulated by ISG15 induced
ISGylation following IFN induction which serves as a negative feedback of the
IFN signaling
103
. Apart from post-translational modifications, Several cellular
proteins are reported to bind RIG-I and inhibit its activation, including autophagy
19

complex Atg12-Atg5
104
, RIG-I family protein LGP2
105
and a truncated RIG-I
splicing variant (RIG-I SV) lacking the first N-terminal CARD domain
106
.
Besides mimicking host strategy to regulate RIG-I as mentioned above, viruses
applies far more countermeasures to antagonize RIG-I. Some viral proteins
compete with RIG-I to bind dsRNA, including the vaccinia virus E3L
107
, Ebola
virus VP35
108
and so on. Poliovirus, rhinovirus and other picornaviruses directly
cleaves RIG-I with their 3C
pro
protease
109
. The NS1 protein of influenza viruses
binds TRIM25 and blocks TRIM25-mediated K63 ubiquitination on RIG-I
110
. Our
previous studies uncover an unprecedented hijacking mechanism of RIG-I
signaling and a novel modification on RIG-I, deamidation
111
. Murine herpesvirus
68 (γHV68), applies a pseudo-enzyme termed viral glutamine amidotransferase
(vGAT) which recruits and cooperates with one of the cellular GAT enzymes
PFAS, to deamidate RIG-I at Q10, N245 and N445. The triple deamidation of
RIG-I then triggers its activation in the absence of RNA ligand. The weak
activation of RIG-I signaling is then hijacked by γHV68 to promote viral lytic
replication and to escape cytokine production.
1.3 Regulation of NF-κB signaling
1.3.1 Cellular regulation and viral manipulation of NF-κB signaling
Similar to IFN signaling, excess amount of inflammatory responses also dampen
the integrity of cells and tissues, thus activation, termination and fine-tuning of
the NF-κB signaling is of utmost importance for host to maintain homeostasis
when clearing out foreign pathogens. IκB  is one of the NF-κB inducing genes,
which will quickly shuttle the NF-κB dimer back to the cytoplasm post-stimulation.
20

Another example of negative feedback is the induction of A20 deubiquitinase
which can antagonize TNF-α, TLR, RIG-I, etc induced NF-κB responses
112
. This
also emphasizes the strict requirement of ubiquitination during most NF-κB
signaling pathways.
On the other hand, viruses have evolved diversified strategies to modulate NF-κB
signaling. For example, human KSHV and EBV, two oncogenic gamma
herpesviruses, activate NF-κB during latency to exploit its pro-survival
properties
113
, which contributes to oncogenic transformation. Human immune-
deficient virus 1 (HIV-1) and herpes simplex virus 1 (HSV-1) hijacks NF-κB to
activate their own viral gene transcription
114, 115
. Murine gamma herpesvirus 68
(γHV68), a model herpesvirus for human KSHV and EBV, activates RIG-I and
IKK  to promote viral lytic gene expression, while dampening host inflammatory
immune gene expression
116-118
. Other viruses inhibit NF-κB activation to evade
host inflammatory responses. These viruses apply diverse ‘weapons’ that target
key steps of NF-κB signaling pathway, from the upstream PRRs to transcription
factor NF-κB
119
.
1.3.2 Post-translational modification of RelA
Post-translation modification on RelA protein serves as another layer of fine-
tuning NF-κB downstream gene transcriptional activity. RelA is reported to be
phosphorylated on Ser276, which is mediated by protein kinase A (PKAc) and
mitogen- and stress-activated protein kinase-1 (MSK-1)
120-122
. Phosphorylation
on S276 enhances the transcriptional activity of RelA via several mechanisms
including stabilization of RelA-p50 heterodimer, facilitated recruitment of CBP,
21

enhanced binding to CBP/P300
123
and stronger interaction with CDK9/Cyclin T1
complex
124
. Phosphorylation on TAD domain of RelA also plays a role in
regulating its transcriptional activity. Ser536, catalyzed by IKK complex and TBK-
1
125, 126
, is induced by different stimuli such as TNF-α treatment, LPS treatment
and microbial infection
127
. This phosphorylation is reported to result in lower
affinity to IκBα but higher affinity to p300
126, 128
. However, phosphorylation on
S468, catalyzed by IKKε or IKKβ, activates/repress NF-κB in a context-
dependent manner
129, 130
, reflecting the complex nature of this phosphorylation
event. Other phosphorylations, such as S205, S281, S311, S529, T254 and T505
are also reported but less-studied
131
. Concomitantly, several phosphatases,
including protein phosphatase 2A (PP2A) and WIP1 are reported to
dephosphorylate S536 and keep NF-κB at inhibitory state during resting condition
or NF-κB termination
132, 133
. Acetylation of RelA on K221 and K310, catalyzed by
the co-factor CBP/P300, is required for full transcriptional activity
120, 134, 135
.
Mechanistically, K221 acetylation enhances the binding of RelA to κB sites,
whereas K310 acetylation promotes RelA in recruiting CDK9 to phosphorylate
RNA polymerase II. Other sites including K122, K123, K218, K314 and K315 are
also reported to be acetylated by CBP/P300 and possibly play a role in regulating
its transcriptional activity in a context-dependent manner
131
. As such, several
histone deacetylases, including HDAC1, HDAC3 and SIRT1, have been found to
regulate RelA via deacetylation on K218, K221 and K310
136
. Apart from
phosphorylation and acetylation, methylation by Set9
137
and K48 ubiquitination
mediated by SOCS1
138
, PDLIM2
139
also contribute to the precise regulation of
22

RelA induced NF-κB responses. Notably, frequent interplay between these
various post-translational modifications has been discovered. S276 and S536
phosphorylation promotes K310 acetylation
128
. Ubiquitination and acetylation
competes for lysine residues
140
. S468 phosphorylation promotes the
ubiquitination and degradation of RelA
117
. In the thesis, RelA deamidation
represents a novel layer of NF-κB regulation, adding to the complexity of the NF-
κB regulatory network.
1.4 Protein Deamidation and Deamidase
1.4.1 Protein Deamidation
Protein deamidation is the conversion of the amide group of the side chain of
glutamines and asparagines into carboxylic group, resulting in glutamate and
aspartate, respectively. It was a common protein post-translational modification
initially reported more than half a century ago
141
. Classical view of this reaction is
spontaneous/nonenzymatic. After analyzing a large set of proteins, the rate of
asparaginyl deamidation was reported to be determined by primary sequence,
secondary and tertiary structural of the protein and cellular environment
142
. The
ubiquitous distribution of asparagine in proteins prompted the hypothesis that
nonenzymatic protein deamidation serves as an internal clock to time the half-life
of a particular protein
143
. This is supported an inverse correlation between the
total number of asparagine/glutamine and the half-life of the protein. Further
computational analysis and experimental validation of the rate of protein
deamidation offered additional support for the above theory and revealed that the
half-life of nonenzymatic deamidation ranges from days to decades. Thus, the
23

nonenzymatic deamidation is a built-in clock to time the “aging” or functional
decay of the protein.
Until recent years, novel functional consequences of deamidation of a small
subset of proteins were discovered in mammalian cells. Initial studies from host-
bacterial interactions showed that several key proteins in innate immune
signaling and G protein signaling, including ubiquitin, UBC13, Rho and so on, are
being deamidated by bacterial encoded protein deamidases
144
. Deamidation of
ubiquitin
145
and ubc13
146
results in loss of function while deamidated Rho
becomes constitutively active
147
.
In the absence of bacterial manipulations, Bcl-xL, an antiapoptotic protein, is
targeted for PEST guided degradation via deamidation, which is a prerequisite
event of apoptosis in response to DNA damage
148
. During neuron development,
4E-BP2 is deamidated, which promotes its association with the mammalian
Target of Rapamycin (mTOR) and modulates neuronal excitatory synaptic
transmission
149
. It was postulated that the deamidation of Bcl-xL and 4E-BP2 is a
non-enzymatic process. Similarly, nonenzymatic deamidation of proteins are
implicated in other fundamental processes, including cell-matrix interaction and
aging-related diseases (e.g., Alzheimer’s disease). Some of these protein
deamidations share similar patterns as being efficient, regulated and have
important functional consequences other than aging, indicating that certain
protein deamidation is specific, controlled and have significant biological
relevance. However, discovery and characterization of those cellular proteins
responsible for catalyzing/facilitating protein deamidation remain poorly studied.
24

1.4.2 Deamidase
Bacterial effectors are bona fide protein deamidases that adopt a common fold of
cysteine proteases with the signature catalytic triad essential for catalyzing
deamidation
150
. In mammalian cells, transglutaminases and the N-terminal
glutamine amidase (NTQA) can deamidate cellular proteins
151
. Tissue
transglutaminase deamidates glutamine and catalyze the formation of isopeptide
bond between a free amine group (e.g., lysine of a protein or peptide) and the
acyl group of deamidated glutamine residue, known as transamidation
152
. The
transamidation reaction is critical for tissue integrity, skin barrier formation and
blood coagulation (factor XIII). If the acceptor is a water molecule instead of
lysine side chain, transglutaminase then catalyzes glutamine hydrolysis and
release ammonia, a typical deamidation reaction that converts glutamine into
glutamate
152
. In physiological pH, the prototype transglutaminase 2 demonstrates
6-fold higher activity in catalyzing transamidation than deamidation, suggesting
that transamidation is the dominating function of transglutaminase in mammalian
cells. The recently described NQTA catalyzes glutamine deamidation of a model
substrate via N-end rule pathway to facilitate its degradation, though its
physiological substrates remain to be identified. Previous structural study
revealed that the NTQA enzyme shares a similar enzyme fold with the factor VIII
transglutaminase. Our study demonstrated for the first time that glutamine
amidotransferases (GAT) can be potential protein deamidase.
1.4.3 Glutamine Amidotransferases
Glutamine amidotransferases (GAT) are enzymes capable of converting the
amide of the side chain of free glutamines into carboxylic group. They are usually
25

multifunctional proteins catalyzing the synthesis of small molecules by using the
ammonia group donated from glutamines. GATs could be classified into three
classes depending on the mechanism of catalysis
153
. Several enzymes involved
in nucleotide biosynthesis, such as phosphoribosylformylglycinamidine synthase
(PFAS), CTP synthase 1/2 (CTPS1/2), guanine monphosphate synthetase
(GMPS), carbamoyl-phosphate synthetase 2, aspartate transcarbamylase, and
dihydroorotase (CAD), belong to class 1 of GAT which has a typical catalytic triad
consisting of cysteine, histine and glutamate. Interestingly, the class 1 GAT
domain belongs to cysteine protease superfamily, which shares similar fold with
bacterial-encoded protein deamidase. However, whether mammalian GATs can
catalyze protein deamidation to regulate cellular physiology remains poorly
understood. Traditional view on these enzymes stalls at their roles in
biosynthesis with the substrate of deamidation limited to free glutamines. We
previously showed that γHV68 applies viral pseudo-enzyme vGAT to recruit and
activate PFAS, which then catalyzes the protein deamidation of RIG-I
111
. This
provides the first example of protein deamidation catalyzed by GAT enzyme and
demonstrate the substrate promiscuity of PFAS, albeit with the activation by a
pseudo-enzyme. It’s still intriguing as how vGAT manage to alter the substrate
preference of PFAS and whether glutamine amidotransferase and protein
deamidase activity of PFAS correlates with each other. Nevertheless, we believe
that protein deamidation might expand to other members within the cellular
metabolism-related GATs.
26

CAD is an acronym derived from carbamoyl-phosphate synthetase 2 (CPSII),
aspartate transcarbamylase (ATC), and dihydroorotase (DHO). It is a 250kD
protein consisting of three distinct enzymes which catalyze the first three steps of
de novo pyrimidine synthesis. CPSII, the first and rate limiting enzyme of CAD,
catalyzes the formation of carbomyl-phosphate using ammonia, bicarbonate and
ATP
154
. The ammonia group is provided by glutamine via deamidation by
GLNase domain (GAT domain) of CPSII. ATC, as the second step, catalyzes the
formation the carbomyl-aspartate based on aspartate and carbomyl phosphate,
while DHO is responsible for the formation of the pyrimidine ring. The GAT
activity of CAD synchronizes with its CPS activity, as minimal deamidation on
glutamine was observed for the subdomain GLNase itself.
The enzymatic activity of CAD undergoes allosteric regulation. The final product
of the de novo synthesis, Uridine Triphosphate (UTP), has a negative feedback
on CAD. Phosphoribosyl 1-pyrophosphate (PRPP), an intermediate reactant
downstream of CAD, promotes CAD activity
154
. Besides that, CAD is regulated
by post-translational modifications. Phosphorylations on Ser1406 by PKA and
Thr456 by ERK1/2 release CAD from inhibition by UTP binding
155
. More recently,
CAD was found to be phophorylated on Ser1859 by S6K, which promotes its
hexamer formation and enzymatic activity
156, 157
.

27

2. Deamidation of RIG-I by Herpes Simplex Virus 1 (HSV-1) UL37
to evade dsRNA induced IFN-β production
2.1 Introduction
In response to viral infection, the retinoic acid-inducible gene I (RIG-I) senses
viral RNA and triggers antiviral immune signaling, leading to the production of
type I interferons and pro-inflammatory cytokines
71
. We recently reported that
gamma-herpesviruses activate RIG-I via deamidation to evade antiviral cytokine
production
111
. The gamma herpesvirus-specific homologues of glutamine
amidotransferase (vGAT) interact with cellular GAT PFAS and induce RIG-I
deamidation. Interestingly, infection of herpes simplex virus 1 also induced RIG-I
deamidation (Figure 1A), despite that the HSV-1 genome encodes no apparent
sequence homologue of vGAT. We then discovered that tegument protein UL37,
a functional homologue of vGAT, bound and induced RIG-I deamidation to inhibit
its activation by dsRNA derived from HSV-1 infection. Mass spectrometry
analysis on purified RIG-I identified two asparigines of the central helicase
domain (Hel2i) that were deamidated by UL37. Deamidation significantly
diminished the RNA-binding affinity and the ATPase activity of RIG-I which is
required for it to transduce signaling. Mutation of the two asparagines into
glutamines conferred RIG-I resistance to deamidation which results in restoration
of RIG-I activation and upregulated IFN induction during HSV-1 infection. On the
other hand, HSV-1 carrying UL37 mutants, which fail to deamidate RIG-I,
induced RIG-I activation and elevated IFN induction. Altogether, these results
suggest that deamidation by UL37 is a mechanism for HSV-1 to evade dsRNA-
induced activation of RIG-I.
28

2.2 Results
2.2.1 HSV-1 infection inhibits RIG-I activation.
Positive-RNA and DNA viruses are known to generate double-stranded RNAs
during active replication
78
. The function of these dsRNA intermediate remains
elusive, as they are regarded as replication byproduct due to symmetric RNA
transcripts
158
. By using an antibody specific to dsRNA, dsRNA species were
readily detected during HSV-1 replication in Human Foreskin Fibroblast (HFF)
(Figure 1B). DsRNA was extracted, annealed and enriched in vitro based on a
previously-described method
158
, which potently activated RIG-I downstream
signaling in an interferon promoter luciferase assay (Figure 1C). Knocking down
of RIG-I significantly impaired the dsRNA-induced interferon signaling (Figure
2A), demonstrating that HSV-1 infection produced dsRNA intermediates that
were able to activate RIG-I to induce interferon response. This is supported by a
previous study that PKR, a dsRNA-dependent ISG, gets activated during
ICP34.5 (PKR signaling inhibitor) null HSV-1 infection. Surprisingly, knocking
down of RIG-I failed to impact type I interferon gene expression induced by HSV-
1 infection both in 293T cells and HeLa cells (Figure 1D, 1E, 2C, 2D). In
comparison, RIG-I knockdown abolished Sendai virus-induced interferon gene
expression (Figure 2B). RIG-I’s lack of response in HSV-1 infection, compared
with its activation by dsRNA extracted from HSV-1 infected cells, implies that it
might be inhibited by HSV-1. Indeed, while Sendai virus triggered robust RIG-I
oligomerization, an indication of its activation, HSV-1 infection failed to do so. In
the presence of HSV-1, however, Sendai virus was unable to activate RIG-I
(Figure 1F). This observation demonstrated that HSV-1 inhibits RIG-I activation
29

which is induced by either Sendai virus or HSV-1 dsRNA. In agree with the
inhibitory effect, HSV-1 failed to trigger significant interferon gene expression
while strongly inhibiting Sendai virus triggered responses (Figure 1G, 2F, 2G).
On the contrary, HSV-1 weakly induced NF-κB gene expression and failed to
inhibit Sendai virus induced responses (Figure 2E). Notably, HSV-1 ICP0
antagonizes IRF3 which serves as RIG-I downstream transcription factors,
making the inhibition on RIG-I-dependent interferon signaling an integrated
outcome
159
.
2.2.2 UL37 interacts with RIG-I.
We next sought to identify the HSV-1 protein responsible for RIG-I inhibition. We
screened the interactions between RIG-I and HSV-1 tegument proteins and
immediate early proteins using co-immunoprecipitation. One of the tegument
proteins, UL37, showed the strongest binding to RIG-I (Figure 3A, 4A and data
not shown). UL37 interacted with RIG-I when over-expressed in 293T cells
(Figure 4B). More importantly, UL37 could be readily detected when endogenous
RIG-I was immunoprecipitated during HSV-1 infection (Figure 3B). When we
fractionized HSV-1 infected cell lysate by size exclusion chromatography, we
found that part of UL37 overlapped with the most abundant RIG-I fraction, an
indication of potential interactions between the two during HSV-1 infection
(Figure 3C). TRAF6, a known UL37 binding protein, was found within the same
fraction of UL37 and RIG-I, which means it could potentially mediate the
interaction between RIG-I and UL37. This was supported by its role in mediating
UL37’s interaction with other innate immune signaling proteins. The majority of
UL37 was eluted in a bigger-sized fraction with no TRAF6 involved. This portion
30

of UL37 could be involved in other functions such as mediating virion trafficking
while binding to UL36, however, further evidence will be needed to support this
hypothesis.
Co-immunoprecipitation assay with truncated RIG-I mutants indicated that UL37
preferentially bound to the RIG-I 2CARD and RNA helicase domain (Figure 4C).
Interestingly, RIG-I also bound to MDA5, a closely related RIG-I sibling sensor
(Figure 4D).
2.2.3 UL37 inhibits RIG-I activation and RIG-I-dependent IFN signaling.
To test if UL37 accounts for RIG-I inhibition during HSV-1 infection, 293T cells
stably expressing UL37 or control were generated (Figure 5A). Upon Sendai
virus infection, the induction of the interferon genes was significantly impaired in
the presence of UL37 (Figure 5B). In contrary, UL37 overexpression induced NF-
κB gene expression regardless of Sendai virus infection (Figure 5B, 6A). UL37
was reported to activate NF-κB via binding to TRAF6. In agreement with that,
UL37 itself induced NF-κB activation but not type I interferon as shown in the
luciferase reporter assay (Figure 6B, 6C). UL37 induced similar NF-κB responses
in wild-type or knockout RIG-I MEF cells (Figure 6D), indicating that RIG-I is not
involved in the activation of NF-κB by UL37. These observations demonstrated a
distinct mechanism and the opposite consequences of UL37’s manipulation on
IRF and NF-κB signaling, both of which ends up facilitating HSV-1 replication.
UL37’s inhibition on RIG-I signaling was further confirmed by reporter assay
using different IFN reporter plasmid and another RIG-I ligand (low molecular
weight Poly[I:C]) (Figure 5C,5D). The inhibition on RIG-I signaling was specific,
31

as UL37 failed to block MDA5 or CGAS-induced interferon induction in the
reporter assays (Figure 6E, 6F). IRF3 dimerization and nuclear translocation are
two downstream signaling events of RIG-I activation which serve as prerequisite
steps of type I interferon gene induction. UL37 significantly impaired the
dimerization of IRF3 induced by Sendai virus (Figure 5E). Furthermore, IRF3
nuclear translocation was blocked by UL37 (Figure 6G). Based on these data, we
sought to test whether UL37 directly inhibits RIG-I or any of the RIG-I
downstream signaling events. Transient transfection of RIG-I downstream protein
(e.g. MAVS, TBK1, constitutively active IRF3) bypasses the requirement of
upstream signaling event and activates IFN signaling. We reasoned that if UL37
inhibits RIG-I signaling at any downstream step, at least one of these proteins-
induced IFN responses will be inhibited by UL37. However, rather than inhibiting,
UL37 significantly promoted MAVS, TBK1 or IRF-5D induced IFN-β reporter
expression (Figure 5F). This was reasonable since IFN-β reporter consists of
both IRF and κB binding sites. It has been reported that NF-κB activation
facilitates the production of type I interferons. As such, UL37 expression
saturated NF-κB, resulting in elevated reporter expression (Figure 6G). As UL37
inhibited rather than promoting RIG-I signaling induced by dsRNA, we thought
that it might directly antagonize RIG-I activation. Indeed, when compared to
control, more RIG-I adopted an inactive monomeric form with ectopically
expressed UL37, which was in perfect agreement with HSV-1 infection (Figure
5H). Furthermore, Sendai virus induced oligomerization of RIG-I was abolished
by UL37. We strengthened this finding by RIG-I dimerization assay in a two
32

differently tagged RIG-I expressing cell lines. UL37 abolished RIG-I dimerization
induced by Sendai virus (Figure 5I). All these data demonstrated that UL37 is
responsible for directly inhibiting RIG-I activation during HSV-1 infection.
2.2.4 UL37 deamidates RIG-I at N495 and N549.
We previously discovered that γHV68 pseudo-deamidase vGAT, though lacking
intrinsic catalytic ability, recruits and activates cellular PFAS to deamidate and
activate RIG-I, resulting in the viral hijack of the RIG-I signaling pathway. UL37
and vGAT have Several functional similarities: (1) both are large tegument
proteins; (2) activate NF-κB, albeit via different mechanism; (3) are involved in
viral trafficking and early gene transcription; and (4) manipulate RIG-I. Although
no sequence homology was found between the two, these functional similarities
lend support to the deduction that UL37 might regulate RIG-I via deamidation.
Indeed, overexpression of UL37 induced the shift of RIG-I to the anode (positive
site) when analyzed by 2D gel electrophoresis, indicating a net gain of negative
charge (Figure 7A). In comparison, neither did UL37 alter the charge status of
MDA5, nor did it induce the shift of TRAF6 which was reported to be the UL37
major interacting protein (Figure 8A, 8B). 6-Diazo-5-oxo-L-norleucine (DON), an
irreversible competitive inhibitor for deamidation, potently block the shift of RIG-I
induced by UL37 (Figure 8C), supporting the conclusion that UL37, indeed,
deamidates RIG-I. To identify the deamidation sites, we purified RIG-I from either
HSV-1 infected or UL37 overexpressed cells (Figure 8D) and performed tandem
mass spectrometry to identify a 1Da gain of molecular weight on peptides
containing glutamines or asparagines. Mass spectrometry analyses identified two
asparagines undergoing deamidation followed by both HSV-1 infection and UL37
33

deamidation, N495 and N594 (Figure 7B, 7C). We then mutated the two
asparagines (N) to the aspartic acids (D) to generate the deamidated RIG-I (RIG-
I-DD). UL37 was able to induce the deamidation of wild-type RIG-I, however, it
failed to do so for the deamidated RIG-I (RIG-I-DD) (Figure 7D). This further
indicated that UL37 deamidates RIG-I at N495 and N549. Next we sought to
characterize the mechanism of the deamidation. We reasoned that UL37 might
act like another pseudo-enzyme to recruit and activate some cellular GAT
enzyme, as it is functionally similar to vGAT. However, the knockdown of all
known GAT enzymes failed to restore UL37’s inhibition on RIG-I signaling
(Figure 8E). This suggested that UL37 probably deamidates RIG-I via a different
mechanism, which could be that UL37 recruits unknown enzyme/factors, or UL37
itself is a genuine viral deamidase. Purified UL37 from 293T cells deamidated
RIG-I in vitro. The in vitro deamidation of RIG-I could be blocked by pre-treating
UL37 with inhibitor DON (Figure 7E). This data strongly suggested that UL37
might be a genuine viral deamidase. Further experiments, including solving the
structure of UL37 and identification of its catalytic triad, are being carried out and
will help to unveil the mechanism of the deamidation.
2.2.5 Deamidation desensitizes RIG-I to dsRNA and impairs its ability to
activate IFN signaling
Since RIG-I is deamidated by UL37 during HSV-1 infection, we turned to test
whether deamidation results in the loss of function of RIG-I. Overexpression of
wild-type RIG-I, single deamidated RIG-I and double deamidated RIG-I (RIG-I-
DD) all fail to activate IFN-β or NF-κB (Figure 10A), in stark contrast with the
constitutively active RIG-I N-terminal 2CARD domain. The structure of RIG-I is
34

available. N495 and N549 belongs to the specialized helicase domain (Hel2i) of
RIG-I. They locate at two neighboring α-helix, in close proximity to the RNA
binding region (Figure 9A). According to structural predictions, deamidation of
these two asparagines will disrupt local hydrogen bonding and impair the integrity
of the α-helix. As RNA helicase domain contributes to stable RNA binding and
RIG-I signaling, we hypothesized that deamidation of these two asparagines
impairs the binding of RIG-I to dsRNA. Indeed, in the dsRNA-gel shift assay,
deamidated RIG-I showed a significantly reduced dsRNA-binding activity when
compared to wild-type RIG-I which readily shifted dsRNA species to form a
stable RIG-I-dsRNA complex (Figure 9B). Deamidated RIG-I formed a higher
order of complex along the dsRNA. Interestingly, higher order of RIG-I
oligomerization status didn’t result in stronger activation, demonstrating that only
the tetrameric form of RIG-I could lead to an optimally-folded 2CARD tetramer,
which serves as the basis for nucleating MAVS oligomerization. ATPases assay
further revealed that deamidated RIG-I showed a dramatically reduced ATP
hydrolysis activity (Figure 9C, 9D, 10C, 10D). Both dsRNA binding and ATP
hydrolysis are required for dsRNA-induced RIG-I activation, as a consequence,
deamidated RIG-I failed to respond to Sendai virus induced oligomerization
(Figure 9E). In the context of wild-type and deamidated RIG-I ‘reconstituted’
MEFs, deamidation of RIG-I significantly impaired the type I interferon gene
induction by either Vescular Stomatitis Virus (VSV) or Sendai virus infection
(Figure 9F, 10E). Consequently, the ability of RIG-I to antagonize VSV replication
35

was largely reduced upon deamidation (Figure 10F). In summary, deamidation
inhibits RIG-I activation by desensitizing RIG-I to dsRNA.
2.2.6 HSV-1 carrying deamidation-deficient UL37 mutations induce stronger
IFN responses.
To further assess the contribution of RIG-I deamidation in HSV-1 infection, we
sought to uncouple RIG-I deamidation in HSV-1 infection by introducing
deamidation-deficient UL37 into the virus. We screened for UL37 mutants which
are defective in deamidating RIG-I. Truncations and mutations on UL37 were
limited to a minimum degree in order to keep the integrity and other functions of
UL37. As TRAF6 is required for RIG-I to interact with innate immune adaptors,
TRAF6 binding mutant UL37 (PVEDDE PVADDE) which we termed UL37-
1101A, indeed, showed defect in binding to RIG-I (Figure 12E). By truncation
analysis (data not shown), we found that UL37-d19, a UL37 mutant generated
from the deletion of 19 C-terminal residues while keeping the TRAF6 binding
motif, failed to inhibit RIG-I, which was similar to UL37-1101A and a TRAF6 motif
deletion mutant [UL37(1-1063)] (Figure 11A). Notably, failure to inhibit Sendai
virus-induced RIG-I signaling was not due to UL37’s defect in NF-κB activation,
as UL37-d19 potently induced NF-κB activation similar to the wild-type UL37
(Figure 12A). Neither UL37-1101A nor UL37-d19 deamidated RIG-I (Figure 11B,
12B), demonstrating the strong correlation between deamidation and RIG-I
inhibition. As UL37 is essential for HSV-1’s replication in cell culture, HSV-1
carrying UL37 mutations (1101A and d19) was produced by homologous
recombination of UL37 wild-type/mutant and UL37 KO HSV-1 Bacmid DNA (data
not shown). We then quantify the genome copy of these recombinant viruses
36

and infect RIG-I stable cell line at same MOI. Gel filtration analysis showed that
both HSV-1 carrying deamidation-deficient UL37 robustly induced RIG-I
oligomerization, while wild-type UL37 rescued HSV-1 failed to induce RIG-I
activation as expected (Figure 11C). These two recombinant viruses also
induced significantly stronger TBK-1 phosphorylation, another indication of IFN
signaling activation (Figure 11D). Consequently, both UL37 mutant HSV-1
induced significantly higher interferon gene expression compared to UL37 wild-
type HSV-1 in 293T and HeLa cells (Figure 11E, 12C). Lastly, these mutant
viruses showed defects in viral replication in HFF cells, partially indicating that
the elevated IFN signaling antagonized HSV-1 replication (Figure 12D).
Nevertheless, we cannot rule out the possibility that these mutations impaired the
integrity of UL37 which resulted in viral trafficking or envelopment defects and
reduced viral yield. Further experiment will be conducted to assess the
contribution of RIG-I induced interferons in inhibiting HSV-1 replication.
2.2.7 Deamidation-resistant RIG-I-QQ is activated by HSV-1 and induces
stronger IFN responses.
To further confirm that N495 and N594 indeed accounts for the inhibition of RIG-I
signaling upon HSV-1 infection, we sought to create a deamidation-resistant
RIG-I and try to restore its activity upon HSV-1 infection. Under structural
guidance, we found that N594 forms hydrogen bonding with the backbone of
T506 in the neighboring α-Helix, indicating that residual mutations at neighboring
α-Helix to restore the structural impact brought by deamidation at N594 would be
implausible. Thus, we decided to mutate N594 to H594 or Q594 to maintain the
hydrogen bonding, albeit with slightly different bond length. Unfortunately, N495
37

locates in a comparatively flexible loop where the information of precise local
environment is missing. By using 2D gel electrophoresis, we found that mutating
asparagines to glutamines (RIG-I-QQ) (Figure 13A) presented RIG-I the ability to
resist deamidation induced by UL37 (Figure 13B). More importantly, RIG-I-QQ
oligomerization could be readily induced by both Sendai virus and HSV-1 (Figure
13C), indicating that RIG-I-QQ serves as a functionally restored mutant.
Luciferase assay further showed that while Sendai virus induced comparable
interferon reporter expression in RIG-I-QQ and RIG-I wild-type cells, HSV-1
infection induced significantly higher interferon expression in RIG-I-QQ cells
compared to RIG-I wild-type (Figure 13D). This was further confirmed by real-
time analysis of interferon gene induction where RIG-I-QQ induced stronger IFN
responses compared with RIG-I wild-type (Figure 13E). Finally, HSV-1 replication
was significantly reduced in RIG-I-QQ stably expressing HFF cell when
compared to RIG-I wild-type (Figure 13F). All these data demonstrated the
necessity of deamidating N495 and N549 to evade RIG-I signaling for the
purpose of HSV-1 lytic replication.
2.3 Discussion
RNA transcripts arising from symmetric transcription was discovered in cells
infected by a number of DNA viruses, including poxyviruses, adenoviruses,
papovaviruses and HSV-1. A portion of these transcripts (~15%), which are
complementary and self-annealing, forms dsRNA and can serve as potential
ligands to activate dsRNA sensors
158
. Although initially believed to be located in
the nucleus, these dsRNA were also found in the cytoplasm during HSV-1 and
KSHV infection. The functions of these dsRNA are poorly understood, and
38

current view regards them as by-products due to symmetric viral transcription.
Previous research indicated that dsRNA sensor PKR can be activated by HSV-1
in the absence of its viral inhibitor ICP34.5
160
. RIG-I was also reported to be
required for KSHV to induce IFN-β responses
161
. In spite of these findings, there
is no evidence linking the dsRNA to these sensors. We here provide the first
evidence that dsRNA extracted from HSV-1 can indeed activate RIG-I, which
also brings interesting questions: (1) what is the identity of these dsRNAs; (2) do
these dsRNAs regulate viral transcription; and (3) how are these dsRNA
transported between cytoplasm and nucleus. In order to address these questions,
the pulldown of deamidation-resistant RIG-I (RIG-I-QQ) in HSV-1 infected cells
will be performed to allow the sequencing of these dsRNAs. The sequenced
transcripts could then be mapped on the viral genome for further characterization.
Although RIG-I is not responsive to HSV-1 in several human cells we tested,
such as 293T and HeLa, knockdown of it significantly impairs HSV-1 induced
IFN-β responses in mouse cell lines (MEF, RAW, etc)
162
. In addition to that, HSV-
1 infection in murine cell lines induces much stronger IFN-β when compared to
human cells. This strongly demonstrates that the manipulation of IFN signaling
by HSV-1 is host specific, which is evolved during long-term virus-host arms
races. Murine RIG-I, though similar to human RIG-I, has several regions of
distinct local peptide sequences. In fact, N495, which is deamidated by HSV-1, is
not conserved in mRIG-I. It would be very interesting to test if the binding or
deamidation of mRIG-I by UL37 is impaired, which leaves the mRIG-I functional
during HSV-1 infection in murine cell lines. More importantly, this observation
39

raises potential concern for the use of murine cell lines and mice to characterize
HSV-1 infection in vitro and in vivo, as the innate immune system might be
reprogrammed in a totally different way.
Despite the fact that IRF and NF-κB coordinate the production of type I
interferons, lots of viruses exploit NF-κB activation to either transactivate viral
genes or antagonize cell apoptosis. For gamma herpesviruses, EBV and KSHV
activate NF-κB during latency to maintain cell viability which contributes to the
malignant transformation. Lytic viruses like HSV-1 and γHV68 either activates or
hijacks NF-κB pathway to promote viral gene transcription. However, none of
these viruses favor IRF activation. KSHV encodes protein like vFLIP which
directly activates NF-κB without interfering IRF
163
. Interestingly, γHV68 and HSV-
1 apply tegument proteins to co-regulate the two pathways. vGAT activates IRF
and NF-κB upstream signaling via RIG-I deamidation while inhibits IRF activation
downstream of MAVS. UL37 inhibits IRF signaling through RIG-I deamidation
and activates NF-κB directly via its TRAF6 binding motif. Studying these intricate
mechanisms which differentially manipulate these two closely related signaling
pathways will allow us to manipulate IRF and NF-κB in a separated, modularized
manner. That is to say, we could learn from viruses to control innate immune
responses systematically.
The TRAF6 binding motif (PVEDDE) of UL37 is necessary for it to induce NF-κB
response and mediates the interaction of UL37 with many innate immune
molecules
164
. We find that mutation of this motif would disrupt the binding of
UL37 to RIG-I. This could be explained by three possibilities: (1) TRAF6
40

mediates all the interactions of UL37; (2) the motif (PVEDDE) is not solely a
TRAF6 interaction motif; (3) TRAF6 binds and induces K63 ubiquitination of
UL37, while K63 ubiquitin chain mediates or stabilizes the binding of UL37 to
other proteins. The frequently detected smearing pattern of UL37 in SDS gel
electrophoresis is likely due to its ubiquitination by TRAF6, which support the
third hypothesis. In fact, the activation of RIG-I requires the covalent/non-
covalent association with K63 ubiquitin chains. Besides RIG-I, K63 ubiquitin
chains are reported to be actively involved in RLR, TLR induced IRF and NF-κB
signaling. They are required for multiple steps of the signaling event, from
upstream sensor signaling to downstream kinase activation. Hence, K63
ubiquitination might be guiding UL37 to specifically target important molecules to
manipulate host innate immune signaling. Further studies are needed to
characterize the ubiquitination on UL37.
UL37 induces RIG-I deamidation at N495 and N549, which locate at a
specialized insertion domain (HEL2i) within the second helical domain (HEL2).
vGAT induces RIG-I deamidation at Q10, N245 and N445, with the later two
residues locating in first helical domain. It may seem intriguing that RIG-I
undergoes different deamidation upon HSV-1 and γHV68 which result in
inhibition and activation, respectively. However, these facts could help to explain
this difference. Firstly, vGAT shares no homology to UL37, while PFAS is not
involved in the deamidation of RIG-I by UL37, indicating that the binding interface
and the mechanism might be different. Secondly, the N245, N445 double
deamidated RIG-I induced by vGAT is not responsive to dsRNA, similar to the
41

functional consequence of UL37 induced RIG-I deamidation. In fact, destabilizing
RIG-I/MDA5 by binding to its helical domain is a common evasion strategy
applied by RNA viruses such as Measles Virus. From an evolutionary
perspective, we think the deamidation in helical domain is shared by these two
viruses to evade dsRNA-induced IFN responses. Deamidation of Q10 (induced
by vGAT) in the 2CARD domain, which accounts for the activation of RIG-I when
combining with deamidation in helical domain, might be evolved later for gamma
herpesviruses 68 to establish a unique hijack mechanism. While it might be
challenging to prove the hypothesis, identification of the sites induced by KSHV
and other herpesviruses will help to provide explanations.
It’s been reported that UL37 binds to other innate immune signaling molecules,
including MyD88, IRAK1, etc. It would be interesting to probe whether these
binding proteins undergo deamidation by UL37, provided that tegument proteins
of Herpesvirus usually manipulate multiple innate immune signaling pathways.
On the other hand, discovery and characterization of more substrates will help to
delineate the conserved motif of UL37 deamidation. Like phosphorylation, we
believe deamidases-catalyzed protein deamidation is specific and requires
particular primary/secondary local peptide information. In support to this claim,
UL37 binds to TRAF6 and MDA5 but fails to induce deamidation of either protein.
Finding such a deamidation motif will significantly facilitate the discovery of more
deamidated substrates in diverse cellular pathways.
Last but not least, we are focusing on delineate the mechanism of RIG-I
deamidation by UL37. None of the annotated GATs is required for the
42

deamidation, while immunoprecipitated UL37 deamidates RIG-I in the test tube,
leading to hypothesis that UL37 might be a genuine viral deamidase.
Nevertheless, purification from mammalian cells might carry residual amounts of
UL37 binding proteins which are invisible by silver staining. We couldn’t rule out,
though not likely, the existence of an uncharacterized deamidase co-purified with
UL37 which compromises the conclusion of the in vitro deamidation. In answer to
these concerns, truncated UL37 (581-1123) has been generated which shows
inhibition on RIG-I signaling in a reporter assay (data not shown). This fragment
could be readily expressed in e.coli, satisfying the requirement of an ‘in vitro
reconstituted’ system. Whether it could deamidate RIG-I in vitro will be studied.
Moreover, we reason that solving structure of the UL37 is still the golden
standard to reveal its true identity. We are collaborating with structural biologist to
obtain the crystals of UL37. All deamidases discovered so far catalyze the
deamidation via cysteine-mediated hydrolysis. There are 14 cysteines in UL37,
suggesting that mutation individually is a plausible strategy to screening for the
catalytic residue. In fact, we have got two cysteine mutations which result in
failure of UL37 to inhibit RIG-I (data not shown). More experiment will have to be
carried out to characterize their ability to deamidate RIG-I. Notably, the pitfall of
this strategy might be that mutation of cysteines could influence the structural
integrity of UL37, more than achieving the ‘catalytic-dead’ effect. Nevertheless, it
will provide important information, together with the structural study, to facilitate
our understanding of UL37.

43

Figure 1

44

Figure 1. HSV-1 infection inhibits RIG-I activation.
(A) 293/RIG-I cells were infected with HSV-1 (MOI = 5) and Sendai virus (100
[HAU]/ml) for 10hrs. WCLs were analyzed by 2D gel electrophoresis and
immunoblotting.
(B) Human Foreskin Fibroblast cells were infected with GFP HSV-1 (MOI = 5)
for 6hrs. Immunofluorescence assay for dsRNA was performed using J2
antibody.
(C) Total RNA was extracted from mock and infected HeLa cells and dsRNA
was prepared as indicated. 293T cells were transfected with PRDIII (ISRE)
reporter cocktail and subsequently transfected with dsRNA for 15hrs.
PRDIII fold induction was determined by luciferase assay.
(D) 293T cells stably expressing control or RIG-I shRNAs were prepared by
lentiviral transduction. WCLs were analyzed by SDS page and
immunoblotting.
(E) Control and RIG-I knockdown 293T cells were infected with HSV-1 (MOI =
5) for the indicated hours. RNA was extracted and cDNA was prepared to
determine IFN-β and ISG56 mRNA expression by real-time PCR analysis.
(F) 293T/RIG-I cells were infected by Sendai virus (100 [HAU]/ml, 10hrs),
HSV-1 (MOI = 5, 12hrs) and HSV-1(MOI = 5, 12hrs) superinfected with
Sendai virus (100 [HAU]/ml, 10hrs). Purified RIG-I was analyzed by gel
filtration and immunoblotting.
(G) 293T cells were infected by Sendai virus, HSV-1 and HSV-1 superinfected
with Sendai virus as mentioned above. RNA was extracted and cDNA
was prepared to determine IFN-β and ISG56 mRNA expression by real-
time PCR analysis.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)
45

Figure 2
B E
F G
A
C D

46

Figure 2. HSV-1 infection inhibits RIG-I activation.
(A) 293T cells stably expressing control or RIG-I shRNA was prepared and
analyzed by immunoblotting. The stable cell lines were then transfected
with PRDIII (ISRE) reporter cocktail and subsequently transfected with
dsRNA extracted from mock or HSV-1 infected HeLa for 15hrs. PRDIII fold
induction was determined by luciferase assay.
(B) Control and RIG-I knockdown 293T cells were infected with Sendai virus
(100 [HAU]/ml) for 10hrs. RNA was extracted and cDNA was prepared to
determine IFN-β and ISG56 mRNA expression by real-time PCR analysis.
(C) HeLa cells stably expressing control or RIG-I shRNA were prepared by
lentiviral transduction. RNA was extracted and cDNA was prepared to
determine RIG-I mRNA expression by real-time PCR analysis.
(D) Control and RIG-I knockdown HeLa cells were infected with HSV-1 (MOI =
5) for indicated hours. RNA was extracted and cDNA was prepared to
determine IFN-β and ISG56 mRNA expression by real-time PCR analysis.
(E) 293T cells were infected by Sendai virus (100 [HAU]/ml, 10hrs), HSV-1
(MOI = 5, 12hrs) and HSV-1(MOI = 5, 12hrs) superinfected with Sendai
virus (100 [HAU]/ml, 10hrs). RNA was extracted and cDNA was prepared
to determine IL8 and CXCL2 mRNA expression by real-time PCR analysis.
(F) 293T cells were transfected with PRDIII reporter cocktail and
subsequently infected by Sendai virus, HSV-1 and HSV-1 superinfected
with Sendai virus as mentioned above. PRDIII reporter expression was
determined by luciferase assay.
(G) 293T cells were transfected with PRDIII reporter cocktail and
subsequently infected by Sendai virus, HSV-1 and HSV-1 superinfected
with Sendai virus as mentioned above. WCLs were analyzed by SDS
page and immunoblotting with indicated antibodies.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

47

Figure 3

48

Figure 3. UL37 interacts with RIG-I.
(A) The indicated mix of HSV-1 ORFs and RIG-I were ectopically expressed
in 293T cells. WCLs were harvested and RIG-I was immunoprecipitated,
followed by SDS page analysis and immunoblotting with the indicated
antibodies.
(B) 293T cells were infected by HSV-1 (MOI = 5) for the indicated hrs.
Endogenous RIG-I was immunoprecipitated, analyzed by SDS page and
immunblotting with the indicated antibodies.
(C) 293T cells were infected by HSV-1 (MOI = 5) for 15hrs. WCLs were
analyzed by gel filtration and immunoblotting with the indicated antibodies.

49

Figure 4

50

Figure 4. UL37 interacts with RIG-I.
(A) The indicated HSV-1 ORFs and RIG-I were ectopically expressed in 293T
cells. WCLs were harvested and ORFs were immunoprecipitated, followed
by SDS page analysis and immunoblotting with the indicated antibodies.
(B) UL37 and RIG-I UL37 were ectopically expressed in 293T cells. WCLs
were harvested and UL37 was immunoprecipitated, followed by SDS page
analysis and immunoblotting with the indicated antibodies.
(C) RIG-I truncated mutants and UL37 were ectopically expressed in 293T
cells. RIG-I mutants were immunoprecipitated and immunoblotted for
associated UL37 with the indicated antibody.
(D) UL37 and RIG-I/MDA5 were ectopically expressed in 293T cells. RIG-
I/MDA5 was immunoprecipitated and immunoblotted for associated UL37
with the indicated antibody.
51

Figure 5

52

Figure 5. UL37 inhibits RIG-I activation and RIG-I-dependent IFN signaling.
(A) Control or UL37 stably expressing 293T cells were prepared by lentiviral
transduction. WCLs were analyzed by SDS page and immunoblotting.
(B) Control and 293T/UL37 cells were infected with Sendai virus (100
[HAU]/ml) for 10hrs. RNA was extracted and cDNA was prepared to
determine IFN-β, ISG56 and IL8 mRNA expression by real-time PCR
analysis.
(C) Control and 293T/UL37 cells were transfected with PRDIII (ISRE) reporter
cocktail and subsequently transfected with Poly [I:C] for 15hrs. PRDIII
reporter expression was determined by luciferase assay.
(D) 293T cells were transfected with IFN-β or PRDIII (ISRE) reporter cocktail
and increasing amounts of UL37. Transfected cells were subsequently
infected with Sendai virus (100 [HAU]/ml) for 15hrs. PRDIII reporter
expression was determined by luciferase assay.
(E) Control and 293T/UL37 cells were infected with Sendai virus (100
[HAU]/ml) for 8hrs. WCLs were harvested, analyzed by native page and
immunoblotting using IRF3 antibody.
(F) 293T cells were transfected with IFN-β reporter cocktail and UL37 with
MAVS, TBK1 or IRF3-5D. IFN-β reporter expression was determined by
luciferase assay.
(G) UL37 targets RIG-I to inhibit IRF signaling pathway. UL37 activates NF-κB
via TRAF6, which explains the upregulation of IFN-β when co-transfected
with RIG-I downstream proteins as shown in Figure 3(F).
(H) 293T/RIG-I cells stably expressing control or UL37 were prepared by
lentiviral transduction. 293T/RIG-I/Control and 293T/RIG-I/UL37 were then
infected with Sendai virus (100 [HAU]/ml) for 8hrs. RIG-I was purified by
WCLs and analyzed by gel filtration and immunoblotting.
(I) 293T/RIG-I-V5/RIG-I-Flag and 293T/RIG-I-V5/RIG-I-Flag/UL37 cells were
prepared by lentiviral transduction. Cells were infected with Sendai virus
(100 [HAU]/ml) for 8hrs. RIG-I-Flag was immunoprecipitated and the
associated RIG-I-V5 was immunoblotted with V5 antibody.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)
53

Figure 6

54

Figure 6. UL37 inhibits RIG-I activation and RIG-I-dependent IFN signaling.
(A) Control and 293T/UL37 cells were infected with Sendai virus (100
[HAU]/ml) for 10hrs. RNA was extracted and cDNA was prepared to
determine IL8 and CXCL2 mRNA expression by real-time PCR analysis.
(B) 293T cells were transfected with NF-κB reporter cocktail and increasing
amounts of UL37. NF-κB reporter expression was determined by
luciferase assay.
(C) 293T cells were transfected with PRDIII reporter cocktail and increasing
amounts of UL37. PRDIII reporter expression was determined by
luciferase assay.
(D) Rig-i
+/+
and Rig-i
-/-
were transfected with NF-κB reporter cocktail and UL37
via NEON transfection system. NF-κB reporter expression was determined
by luciferase assay.
(E) 293T cells were transfected with PRDIII reporter cocktail and MDA5 with
increasing amounts of UL37. PRDIII reporter expression was determined
by luciferase assay.
(F) 293T/STING cells were transfected with PRDIII reporter cocktail and
CGAS with UL37 or UL37-1101A. PRDIII reporter expression was
determined by luciferase assay.
(G) HeLa was tranfected with control or UL37, and then infected with Sendai
virus (100 [HAU]/ml) for 8hrs. IRF3 nuclear translocation was visualized by
immunofluorescence assay with IRF3 antibody
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

55

Figure 7

56

Figure 7. UL37 deamidates RIG-I at N495 and N549.
(A) 293/Flag-RIG-I cells were transfected with a plasmid containing either a
vector or UL37. WCLs were analyzed by 2D gel electrophoresis and
immunoblotting.
(B) 293T/RIG-I cells were infected with HSV-1 (MOI = 5) for 10hrs or
transfected with a plasmid containing UL37. RIG-I was purified and
analyzed by tandem mass spectrometry. Two deamidated residues were
identified. N495 and N549 were converted into D (in red) due to
deamidation.
(C) Quantification of deamidated N495 and N549 of RIG-I from mock, HSV-1
infected and UL37 transfected cells.
(D) 293T cells were transfected with plasmids containing RIG-I or the RIG-I
double deamidation mutant (RIG-I-DD) without or with a plasmid
containing UL37. WCLs were analyzed by 2D gel electrophoresis and
immunoblotting.
(E) GST-RIG-I and UL37 were purified from 293T cells to homogeneity and
analyzed by silver staining (right). In vitro deamidation of RIG-I by UL37
and DON (500µ M) treated UL37 was analyzed by 2D gel electrophoresis
and immunoblotting.
57

Figure 8

58

Figure 8. UL37 deamidates RIG-I at N495 and N549.
(A) 293T cells were transfected with MDA5 and a plasmid containing either a
vector or UL37. WCLs were analyzed by 2D gel electrophoresis and
immunoblotting.
(B) 293T cells were transfected with TRAF6 and a plasmid containing either a
vector or UL37. WCLs were analyzed by 2D gel electrophoresis and
immunoblotting.
(C) 293T/RIG-I cells were transfected with UL37 and subsequently treated
with/without DON (10µ M). WCLs were analyzed by 2D gel electrophoresis
and immunoblotting. β-actin served as an internal control.
(D) 293T/RIG-I cells were treated with/without HSV-1 infection (MOI = 5) or
UL37 overexpression. RIG-I was purified and analyzed by SDS page and
coommassie staining.
(E) 293T cells were infected with the lentiviruses containing shRNA against
GAT family members. 48hrs later, cells were transfected with PRDIII
(ISRE) reporter cocktail and increasing amount of UL37. Transfected cells
were subsequently infected with Sendai virus (100 [HAU]/ml) for 15hrs.
PRDIII reporter expression was determined by luciferase assay.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

59

Figure 9

60

Figure 9. Deamidation desensitizes RIG-I to dsRNA and impairs its ability to
activate IFN signaling.
(A) N495 and N549 locate in the Helical2i domain, in close proximity to
dsRNA binding region.
(B) Purified RIG-I and RIG-I-DD (200 nM) were incubated with
32
P-labeled 5′-
triphosphate dsRNA (4 nM), with and without a 200-fold excess of cold 5′-
triphosphate dsRNA. RNA-RIG-I complex was analyzed by PAGE and
autoradiography.
(C) Purified RIG-I and RIG-I-DD (20 nM) were incubated with increasing
concentrations of ATP in the presence of 5′-triphosphate dsRNA (80 nM)
and analyzed by ATP hydrolysis assay.
(D) Purified RIG-I and RIG-I-DD (20 nM) were incubated with increasing
concentrations of dsRNA in the presence of ATP (1000 µ M) and analyzed
by ATP hydrolysis assay.
(E) 293T/RIG-I and 293T/RIG-I-DD cells were prepared by lentiviral
transduction and infected with Sendai virus (100 [HAU]/ml) for 8hrs. RIG-I
and RIG-I-DD were purified from WCLs and analyzed by gel filtration and
immunoblotting.
(F) Rig-i
−/−
MEFs were ‘reconstituted’ with RIG-I or RIG-I-DD and analyzed by
immunoblotting (top). ‘Reconstituted’ MEFs were infected with VSV (MOI
= 1) for 10hrs. RNA was extracted and cDNA was prepared to determine
IFN-β and ISG56 mRNA expression by real-time PCR analysis.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

61

Figure 10

62

Figure 10. Deamidation desensitizes RIG-I to dsRNA and impairs its ability
to activate IFN signaling.
(A) 293T cells were transfected with PRDIII reporter cocktail and RIG-I, RIG-I-
N495D, RIG-I-N549D, RIG-I-DD and RIG-I-N respectively. PRDIII reporter
expression was determined by luciferase assay.
(B) 293T cells were transfected with NF-κB reporter cocktail and RIG-I, RIG-I-
N495D, RIG-I-N549D, RIG-I-DD and RIG-I-N, respectively. NF-κB reporter
expression was determined by luciferase assay.
(C) Purified RIG-I, RIG-I-N495D, RIG-I-N549D and RIG-I-DD (20 nM) were
incubated with increasing concentrations of ATP in the presence of 5′-
triphosphate dsRNA (80 nM) and analyzed by ATP hydrolysis assay.
(D) Purified RIG-I, RIG-I-N495D, RIG-I-N549D and RIG-I-DD (20 nM) were
incubated with increasing concentrations of dsRNA in the presence of
ATP (1000 µ M) and analyzed by ATP hydrolysis assay.
(E) ‘Reconstituted’ MEFs as shown in Figure 5(F) were infected with Sendai
virus (100 [HAU]/ml) for 10hrs. RNA was extracted and cDNA was
prepared to determine IFN-β and ISG56 mRNA expression by real-time
PCR analysis.
(F) Rig-i
+/+
, Rig-i
−/−
and ‘reconstituted’ MEFs were infected with VSV (MOI =
0.1) for the indicated hrs. Supernatant was harvested at indicated time
points and viral titer was measured by plaque assay in VERO cells.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

63

Figure 11

64

Figure 11. HSV-1 carrying deamidation-deficient UL37 mutations induce
stronger IFN responses.
(A) 293T cells were transfected with PRDIII reporter cocktail and increasing
amount of wild-type UL37, UL37-1101A, UL37 (1-1063) and UL37-d19,
respectively. Transfected cells were subsequently infected with Sendai
virus (100 [HAU]/ml) for 15hrs. PRDIII reporter expression was determined
by luciferase assay.
(B) 293T/RIG-I cells were transfected with wild-type UL37, UL37-1101A or
UL37-d19. WCLs were analyzed by 2D gel electrophoresis and
immunoblotting.
(C) 293T/RIG-I cells were infected with HSV-1 carrying wild-type UL37, UL37-
1101A or UL37-d19 (MOI = 1) for 15hrs. RIG-I was immunoprecipitated
and analyzed by gel filtration and immunoblotting.
(D) 293T cells were infected with HSV-1 carrying wild-type UL37, UL37-1101A
or UL37-d19 (MOI = 5) for the indicated hrs. WCLs were analyzed by SDS
page and immunoblotting with TBK1 Ser172 phospho-specifc antibodies.
(E) 293T cells were infected with HSV-1 carrying wild-type UL37, UL37-1101A
or UL37-d19 (MOI = 5) for the indicated hrs. RNA was extracted and
cDNA was prepared to determine IFN-β and ISG56 mRNA expression by
real-time PCR analysis.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

65

Figure 12

66

Figure 12. HSV-1 carrying UL37 mutations induce stronger IFN responses.
(A) 293T cell was transfected with NF-κB reporter cocktail and increasing
amounts of wild-type UL37, UL37-1101A, UL37 (1-1063) and UL37-d19
respectively. NF-κB reporter expression was determined by luciferase
assay.
(B) 293T/RIG-I cells were infected with HSV-1 carrying wild-type UL37, UL37-
1101A or UL37-d19 (MOI = 1) for 15hrs. WCLs were analyzed by 2D gel
electrophoresis and immunoblotting. β-actin served as an internal control.
(C) HeLa cells were infected with HSV-1 carrying wild-type UL37, UL37-
1101A or UL37-d19 (MOI = 5) for the indicated hrs. RNA was extracted
and cDNA was prepared to determine IFN-β and ISG56 mRNA expression
by real-time PCR analysis.
(D) HFF cells were infected with HSV-1 carrying wild-type UL37, UL37-1101A
or UL37-d19 (MOI = 1) for the indicated hrs. Supernantant was harvested,
and the viral titer was measured by plaque assay in VERO cells.
(E) RIG-I and UL37 mutants were ectopically expressed in 293T cells. UL37
mutants were immunoprecipitated and immunoblotted for associated RIG-I
with indicated antibody.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

67

Figure 13

68

Figure 13. Deamidation-resistant RIG-I-QQ is activated by HSV-1 and
induces stronger IFN responses.
(A) 293T stably expressing wild-type RIG-I, RIG-I-DD and RIG-I-QQ were
prepared by lentiviral transduction. WCLs were analyzed by SDS page
and immunoblotting.
(B) 293T/RIG-I and 293T/RIG-I-QQ cells were infected with HSV-1 (MOI = 5)
for 10hrs. WCLs were analyzed by 2D gel electrophoresis and
immunoblotting.
(C) 293T/RIG-I-QQ cells were infected with HSV-1 (MOI = 5) for 10hrs. RIG-I-
QQ was purified and analyzed by gel filtration and immunoblotting.
(D) 293T/RIG-I, 293T/RIG-I-DD and 293T/RIG-I-QQ cells were transfected
with PRDIII reporter cocktail and subsequently infected with Sendai virus
(100 [HAU]/ml) for 15hrs or HSV-1 (MOI = 5) for the indicated hrs. PRDIII
reporter expression was determined by luciferase assay.
(E) 293T, 293T/RIG-I, 293T/RIG-I-DD and 293T/RIG-I-QQ cells were infected
with HSV-1 (MOI = 5) for the indicated hrs. RNA was extracted and cDNA
was prepared to determine IFN-β and ISG56 mRNA expression by Real-
time PCR analysis.
(F) HFF stably expressing RIG-I or RIG-I-QQ cells were prepared by lentiviral
transduction. Cells were then infected by HSV-1 (MOI = 1) for 24hrs.
Supernatant was harvested and viral titer was measure by plaque assay in
VERO cells.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

69

Figure 14

Figure 14. Model: HSV-1 UL37 deamidates RIG-I to evade dsRNA induced
IFN responses.

70

3. Deamidation of RelA by CAD to regulate NF-κB response
3.1 Introduction
The gamma herpesvirus-specific homologues of glutamine amidotransferase
(vGAT), though lacking intrinsic catalytic activity, recruits and activates cellular
GAT PFAS to deamidate RIG-I
111
. This provides the first example of protein
deamidation catalyzed by the GAT enzyme, demonstrating the substrate
promiscuity of PFAS. In this study, we wish to explore this promiscuity in the
entire GAT family. We aim to investigate whether protein deamidation, catalyzed
by GAT enzymes, regulates NF-κB signaling. By applying an unbiased screening
within the GAT family, we found CAD negatively regulated NF-κB signaling
induced by different stimuli. Mechanistically, CAD directly interacted with and
deamidated RelA at N64 and N139 in vitro and in vivo. Deamidation of RelA
inversely correlated with NF-κB activation, implying its negative role on NF-κB
signaling. Indeed, the deamidated RelA showed impaired responses to induce
NF-κB gene expression in response to Sendai virus infection. Altogether, our
data discover a new protein deamidase CAD and unveil a novel post-
translational modification on NF-κB regulation, RelA deamidation.
3.2 Results
3.2.1 CAD negatively regulates RelA induced NF-κB responses.
To date, there are 11 annotated GATs in mammalian cells
153
, which allow us to
establish an unbiased screening by knockdown one each and monitor functional
output. In this screening, we focus on RIG-I-NF-κB signaling axis and study
whether protein deamiation, catalyzed by any GAT family members, plays a role
in regulating this pathway. Knockdown of CAD and PPAT significantly
71

upregulated NF-κB reporter expression induced by Sendai virus (Figure 15A). As
mentioned earlier, CAD is a large (~250kDa) multi-functional enzyme involved in
the first three steps of de novo pyrimidine synthesis. We decided to focus on
CAD in this project as it showed the most robust phenotype. To further validate
the screening, CAD stably knockdown cells were prepared and treated with two
different stimuli, Sendai virus and TNF-α, which activate NF-κB via distinct
upstream signaling events but merge at the activation of the IKK kinase complex.
In both case, NF-κB downstream gene expression were significantly increased in
the absence of CAD (Figure 15B, 16A). We observed similar upregulation in CAD
knockdown HFF cells (Figure 16B, S1C). Altogether, these data indicated that
CAD negatively regulates NF-κB signaling, which is cell-type independent and
likely acts downstream of IKK kinase complex. To probe the target(s) along the
NF-κB axis being regulated by CAD, NF-κB responses were induced by RIG-I
2CARD, MAVS, IKKβ or RelA. Knockdown of CAD, in all these induced groups,
upregulated NF-κB reporter expression (Figure 15C, 15D). This observation
demonstrated that the negative regulation by CAD acts on the most downstream
component of NF-κB signaling, the transcription factors. Notably, the negative
regulation of CAD on NF-κB was specific as knockdown of CAD failed to regulate
IFN-β reporter expression (Figure 16D). Feeding Uridine to the cell failed to
restore the effect of CAD knockdown, indicating that the upregulated NF-κB
expression was not due to the deficiency of final product of de novo pyrimidine
synthesis (Figure 16E). Serving as another control, knockdown of CAD
downstream enzyme UMPS, which mediates the final step of uridine synthesis,
72

failed to regulate NF-κB responses in a similar setting (Figure 16F). In contrast to
knockdown, overexpression of CAD inhibited RIG-I 2CARD, IKKβ and RelA
induced NF-κB reporter expression (Figure 15E. 15F). Among them, the IKKβ
induced NF-κB was less inhibited by CAD. One of the possible explanations were
that overexpression of IKKβ could potently phosphrylate CAD, leading to
inhibition of its activity on RelA. The CPSII enzyme, which is the first enzyme of
CAD containing the GAT domain, failed to inhibit RelA induced NF-κB response
as compared with full length CAD (Figure 16G).
3.2.2 CAD interacts with RelA.
Since CAD directly inhibited RelA induced NF-κB response, we went on to test if
CAD interacts with and regulates RelA. A previous research on mapping of RelA
binding proteins by mass spectrometry has already identified CAD as a potential
binding protein. Indeed, overexpressed CAD could readily interact with
endogenous RelA by co-immunoprecipitation assay (Figure 17A). As mentioned
earlier, CAD consists of three enzymes, carbamoyl-phosphate synthetase 2
(CPSII), aspartate transcarbamylase (ATC), and dihydroorotase (DHO). CPSII
has two functional enzymatic domains, GAT and CPS domains termed CPSab.
Co-immunoprecipitation assay indicated that GAT, CPSab and DHO interacted
with RelA, but not ATC (Figure 17B). As ATC locates at the very carboxylic
terminus, the interaction between CAD and RelA was mediated by the N-terminal
part of CAD. RelA has an N-terminal RHD domain and C-terminal TAD domain,
while CAD interacted with the RHD domain of RelA (Figure 17C). The domain
interactions between CAD and RelA were depicted in Figure 17D.
73

3.2.3 CAD deamidates RelA.
Since CAD contains the GAT domain which could serve as a potential protein
deamidase, we wish to know whether protein deamidation occurs on RelA which
accounts for its inhibition by CAD. A previous study supported the hypothesis
that CAD might be a genuine protein deamidase, as it inhibits NOD2 signaling
and GAT activity is required for this inhibition. To probe for this hypothesis, we
applied a competitive inhibitor PALA to specifically antagonize CAD enzyme
activity. Although targeting CPSab, PALA can inhibit GAT activity indirectly as the
activity of GAT parallels with CPSab in order to balance substrate (ammonia)
channeling. PALA treatment upregulated basal level NF-κB which was similar to
knockdown of CAD (Figure 15C, 18A), demonstrating that CAD enzymatic
activity is required for the inhibition on RelA. Mutation of catalytic residue
cysteine of the GAT domain abolished CAD’s inhibition on NF-κB (Figure 18B),
which further narrows down the enzymatic activity to the GAT domain. We then
performed 2D gel electrophoresis to assess the charge status of RelA. In resting
cells, RelA had two major species. Upon knockdown of CAD, however, majority
of RelA shifted to the more positively charged species (Figure 18C). In 293T cell,
RelA adopted three species and overexpression of CAD drove the shift of RelA
to its most negatively charged species (Figure 18D). As a control, NF-κB1 (p50),
which was found abundantly in the RelA heterodimer, failed to present any
changes upon CAD expression (Figure 18D). Immunoprecipiated CAD (but not
its catalytic-dead mutant) induced RelA shift in vitro (Figure 18E). The in vitro
deamdation was blocked by GAT inhibitor DON. These data indicated that CAD
indeed induces RelA deamidation. Interestingly, Sendai virus or TNF-α induced
74

NF-κB activation resulted in loss of RelA deamidation (Figure 18F), which was in
stark contrast to elevated RelA phosphorylation upon TNF-α treatment (Figure
18G).
3.2.4 Deamidation of RelA at N64 and N139 impairs its ability to
transactivate NF-κB downstream gene expression.
Tandem mass spectrometry of purified RelA identified N64 and N139 as the sites
undergoing deamidation by CAD overexpression (Figure 19A, 19B). Both N64
and N139 are found within the RHD domain, which was characterized as the
binding domain of RelA to CAD. Rather than in the DNA binding interface, these
two asparagines locate in the unshielded region of RelA, making it accessible for
them to interact with CAD regardless of DNA binding (Figure 19C). Both N64
deamidation and N139 deamidation of RelA resulted in its significantly impaired
induction of NF-κB response (Figure 19D, 19E). Moreover, double deamidation
of the two residues (RelA-DD) almost abolished NF-κB activation, demonstrating
the dramatic phenotype alteration of RelA deamidation (Figure 19E). Mutation of
N64 to alanine (RelA-N64A) results in a gain-of-function RelA which induced
significantly stronger NF-κB and showed resistance to inhibition by CAD (Figure
19D, 19F). When reconstituted into Rela
-/-
MEF (Figure 19G), wild-type RelA
MEF showed significantly higher basal expression of NF-κB dependent genes
than deamidated RelA MEF (Figure 19H). Moreover, deamidated RelA showed a
much weaker response to Sendai virus induced NF-κB gene expression (Figure
19H). Taken together, CAD induces RelA deamidation at N64 and N139 to inhibit
its ability to transactivate the transcription of NF-κB downstream genes.
75

3.3 Discussion
Emerging studies have demonstrated that deamidation of key proteins (RIG-I,
UBC13, Ubiquitin) in innate immune signaling are important strategy to
manipulate host defense by microbes. Compared to bacterial which encodes
deamidases, viruses often rely on host where they exploit existing regulatory
mechanisms to benefit its own life cycle. This implies that deamidase-catalyzed
protein deamidation might be an intrinsic yet less appreciated post-translational
modification in cells regulating specific signaling events. The studies of Bcl-xl and
4E-BP2 deamidations demonstrated that protein deamidations play essential
roles during apoptosis and neuronal development. However, these studies
stalled at the classical view of the spontanenous deamidation, concluding that
these events were results from abrupt changing of the local environment.
We believe that there are deamidases within the cell which catalyze certain
protein deamidations. As mentioned earlier, the discovery of the first cellular GAT
PFAS as a deamidase, albeit with the requirement of a viral pseudoenzyme,
leads us to expand the study to GATs-catalyzed protein deamidations. There are
two strategies to discover potential GAT substrates: (1) perform proteome mass
spectrometry to identify substrates undergoing deamidation upon a specific
stimulus; (2) screen GATs for deamidases and search for their substrates via
interaction and functional verification. In this study, we applied the second
strategy and focused on NF-κB which is a prototype innate immune signaling.
The knockdown screening, as compared to overexpression system, has the
advantage of achieving genuine negative regulator, due to the fact that
knockdown of these biosynthetic genes might result in poor cell condition and
76

reduced luciferase expression. As such, the outcome of an upregulated reporter
expression would be more persuasive. Notably, knockdown of either CAD or
PPAT didn’t alter cell condition or growth rate, indicating that salvage pathways
or nucleotides in the medium have compensated for the deficiencies in de novo
purine and pyrimidine synthesis.
Several previous findings suggested that CAD is a potential deamidase and a
negative NF-κB regulator. Firstly, CAD negatively regulates NOD2 activation
which is induced by its ligand MDP in a GAT-activity dependent manner
165
.
Secondly, CAD interacts with important proteins in NF-κB signaling axis,
including NEMO and RelA in the NF-κB interactome study
166
. We here show that
indeed CAD interacts with RelA and negatively inhibits RelA-induced NF-κB
activation. Intriguingly, inhibition on RelA requires full-length of CAD but not its
first enzyme CPSII containing the GAT domain. As substrate channeling effect is
important to maintain full activity of CAD, we think that the CPSII alone, though
with intact GAT domain, might have minimal enzymatic activity which explains its
defect in deamidating and inhibiting RelA. CAD is shown to be both in the
cytoplasm and nucleus with significantly higher catalytic activity for the nuclear
portion. Immunostaining showed that CAD co-localized with RelA in the nucleus
(data not shown). Thus we postulated that the deamidation on RelA is taking
place mainly in the nucleus.
RelA adopted distinct deamidation status in a cell-dependent manner. The wild-
type RelA and single deamidated RelA in HCT116 and 293T could match to each
other, whereas significantly more double deamidated RelA was discovered in
77

293T. We think that the protein level and catalytic activity of CAD depict the
outcome of basal level deamidation of RelA, and CAD expression is significantly
higher in 293T compared with HCT116 (data not shown). RelA failed to achieve a
complete double deamidation regardless of the activity of CAD. In both
overexpression of CAD or in vitro incubation with purified CAD, RelA resulted in a
mixture of single and double deamidated species. Notably, purified RelA from
293T already carries single deamidated species, which might due to its
deamidation by endogenous CAD. We have already purified RelA RHD domain
from E.coli and will use it in the in vitro deamidation assay to avoid the
deamidation caused by endogenous CAD.
Interestingly, similar to knockdown of CAD, NF-κB activation by Sendai virus and
TNF-α shifted RelA to a wild-type dominant pattern, yet CAD expression wasn’t
significantly altered during the stimuli. This leads to the postulation that RelA
deamidation is a dynamic and regulated post-translational modification during
NF-κB activation. Such dynamic process might contribute to fine-tuning the NF-
κB responses. In order to study it, it is necessary to monitor RelA deamidation
during difference phases of activation. As 2D gel electrophoresis is labor-
intensive, we have generated specific antibody against N64 and N139
deamidated RelA which will enable us to visualize RelA deamidation kinetics post
NF-κB activation. Meanwhile, what governs the transition of RelA into wild-type
during NF-κB activation remains unclear. One possibility is that there is an
unknown mechanism which degrades deamidated RelA upon stimulation. If this
is true, the cell needs to (1) degrade deamidated RelA within short period of time
78

(in the case of TNF-α treatment, less than 30 min); (2) be able to distinguish the
deamidated species from wild-type ones. In fact, γHV68 RTA, which serves as
E3 ligase to target RelA for degradation, fail to induce the degradation of
deamidated RelA (unpublished data). It is reasonable for the cell to have similar
mechanism to distinguish between the two. Another possibility would be the
inhibition of CAD upon stimulation. CAD is known to interact with IKK complex.
Upon stimulation, it is likely to be targeted by one of the IKK kinases (e.g. IKKβ)
which phosphorylates CAD and inhibits its activity. As a consequence of CAD
inhibition, the continuous turnover of RelA will finally result in a wild-type
dominant manner. More experiments will be conducted to test these hypotheses.
Deamidated RelA clearly showed a defect in inducing classical NF-κB responses.
However, wild-type and deamidated RelA bound to NF-κB cognate sequence
with comparable affinity in vitro (data not shown). In collaboration with structural
biologists, we solved the crystal structure of the RHD domain of deamidated
RelA. Agreeing with the in vitro binding, deamidation of N64 and N139 showed
no difference in DNA binding, but significantly changed the local environment of
RHD interface reported to be involved in transcriptional cofactor binding.
Therefore our current hypothesis is that deamidation of RelA impairs or alters the
recruitment of essential transcriptional coactivators. We are screening for these
coactivators by mass spectrometry and co-immunoprecipitation assay. There is
also a possibility that deamidated RelA retains its transcriptional activity, albeit
with altered DNA targets in cell. Since all the current data are based on cognate
NF-κB promoter sequence and several classical NF-κB downstream genes, our
79

conclusions are limited. We are conducting a total mRNA sequencing to overlook
the global gene expression alterations upon introducing deamidated RelA.
Finally, deamidation of CAD on RelA is not simply another add-on to RelA post-
translational modifications. Since CAD functions as the rate-limiting step in the
nucleotide biosynthesis while RelA is the central hub of NF-κB initiation, this
regulation represents a novel crosstalk between innate immunity and nucleotide
metabolism. Our data so far indicated that deamidation of RelA by CAD parallels
with CAD enzymatic activity in de novo nucleotide synthesis. This leads to an
interesting hypothesis that cells in robust de novo nucleotide biosynthesis would
have a dampened NF-κB response. Upon viral infection, however, biosynthesis
could be stalled purposefully to limit the replication of virus and promote innate
immune responses synergistically. One reasonable explanation of introducing
such a balance between the two is to allow the optimal usage of resources like
ATP as both systems, once activated, require a ton of it. Future studies will help
to establish a bigger picture of innate immunity and nucleotide metabolism
beyond CAD and RelA.

80

Figure 15

81

Figure 15. CAD negatively regulates RelA induced NF-κB responses.
(A) 293T cells were infected with lentivirus containing shRNA against GAT
family members. 48hrs later, cells were transfected with NF-κB reporter
cocktail were subsequently infected with Sendai virus (100 [HAU]/ml) for
15hrs. NF-κB reporter expression was determined by luciferase assay.
(B) 293T stably expressing control of shRNA against CAD was prepared by
lentiviral transduction and analyzed by SDS page and immunoblotting
(top). Control/CAD knockdown 293T cells were then infected with Sendai
virus (100 [HAU]/ml) for 10hrs or treated with TNF-α (10 ng/ml) for 8hrs.
RNA was extracted and cDNA was prepared to determine IL8 mRNA
expression by real-time PCR analysis. The knowndown cells were
supplemented with Uridine (10 µ g/ml).
(C) Control/CAD knockdown 293T cells were transfected with NF-κB reporter
cocktail and RIG-I-N, MAVS or IKKβ. NF-κB reporter expression was
determined by luciferase assay.
(D) Control/CAD knockdown 293T cells were transfected with NF-κB reporter
cocktail and RelA. NF-κB reporter expression was determined by
luciferase assay.
(E) 293T cells were transfected with NF-κB reporter cocktail and RIG-I-N or
IKKβ with increasing amount of CAD. NF-κB reporter expression was
determined by luciferase assay.
(F) 293T cells were transfected with NF-κB reporter cocktail and RelA with
increasing amount of CAD. NF-κB reporter expression was determined by
luciferase assay.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

82

Figure 16

83

Figure 16. CAD negatively regulates RelA induced NF-κB activation.
(A) Control/CAD knockdown 293T cells were then infected with Sendai virus
(100 [HAU]/ml) for 10hrs or treated with TNF-α (10 ng/ml) for 8hrs. RNA
was extracted and cDNA was prepared to determine TNF-α mRNA
expression by real-time PCR analysis. The knowndown cells were
supplemented with Uridine (10 µ g/ml). N.D. = Not Detected.
(B) BJ5 cells stably expressing control of shRNA against CAD was prepared
by lentiviral transduction. RNA was extracted from stable cell lines and
cDNA was prepared to determine CAD mRNA expression by real-time
PCR analysis.
(C) Control/CAD knockdown BJ5 cells were infected with Sendai virus (100
[HAU]/ml) for 10hrs. RNA was extracted and cDNA was prepared to
determine IL8 mRNA expression by real-time PCR analysis. The
knowndown cells were supplemented with Uridine (10 µ g/ml).
(D) Control/CAD knockdown 293T cells were transfected with IFN-β reporter
cocktail and TBK1 or RIG-I-N. IFN-β reporter expression was determined
by luciferase assay.
(E) Control/CAD knockdown 293T cells were supplemented with/without
Uridine (10 µ g/ml). Cells were then transfected with NF-κB reporter
cocktail and RelA. NF-κB reporter expression was determined by
luciferase assay.
(F) 293T cells stably expressing shRNA against UMPS were prepared by
lentiviral transduction. Control/CAD knockdown/UMPS knockdown 293T
cells were transfected with NF-κB reporter cocktail and subsequently
infected with Sendai virus (100 [HAU]/ml) for 15hrs. NF-κB reporter
expression was determined by luciferase assay.
(G) 293T cells were transfected with NF-κB reporter cocktail and RelA with
increasing amount of CPSII. NF-κB reporter expression was determined
by luciferase assay.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)

84

Figure 17

85

Figure 17. CAD interacts with RelA.
(A) CAD was ectopically expressed in 293T cells. WCLs were harvested and
CAD was immunoprecipitated. The immunoprecipitates were analyzed by
SDS page and immunoblotting with indicated antibody.
(B) CAD truncation mutants were ectopically expressed in 293T cells. WCLs
were harvested and truncation mutants were immunoprecipitated, followed
by SDS page analysis and immunoblotting with indicated antibody.
(C) RelA truncation mutants and CAD were ectopically expressed in 293T
cells. WCLs were harvested and CAD was immunoprecipitated, followed
by SDS page analysis and immunoblotting with indicated antibody.
(D) The domain interactions of CAD and RelA were depicted.

86

Figure 18

87

Figure 18. CAD deamidates RelA.
(A) 293T cells were transfected with NF-κB reporter cocktail and subsequently
treated with increasing concentration of PALA. NF-κB reporter expression
was determined by luciferase assay.
(B) 293T cells were transfected with NF-κB reporter cocktail and RelA with
increasing amount of CAD or its catalytic-dead mutant (CAD-ED). NF-κB
reporter expression was determined by luciferase assay
(C) HCT116 cells stably expressing control or shRNA against CAD were
prepared by lentiviral transduction. WCLs were then analyzed by 2D gel
electrophoresis and immunoblotting.
(D) 293T cells were transfected with vector or CAD. WCLs were then
analyzed by 2D gel electrophoresis and immunoblotting with indicated
antibodies.
(E) CAD and CAD-ED were purified from 293T cells to homogeneity and
analyzed by silver staining (right). In vitro deamidation of RelA by CAD,
DON (50µ M) treated CAD or CAD-ED was analyzed by 2D gel
electrophoresis and immunoblotting.
(F) HCT116 cells were infected with Sendai virus (100 [HAU]/ml) for 6hrs or
treated with TNF-α (10 ng/ml) for 30 min. WCLs were analyzed by 2D gel
electrophoresis and immunoblotting.
(G) HCT116 cells were treated with TNF-α (10 ng/ml) for 30 min. WCLs were
analyzed by SDS page and immunoblotting with RelA Ser536 phospho-
specific antibody.

88

Figure 19

89

Figure 19. Deamidation of RelA at N64 and N139 impairs its ability to
transactivate NF-κB downstream gene expression.
(A) 293T cells were ectopically transfected with RelA and a plasmid
containing vector or CAD. RelA was purified and analyzed by tandem
mass spectrometry for deamidation. Two deamidated residues were
identified. N64 and N139 were converted into D (in red) due to
deamidation.
(B) N64 and N139 were located at RHD domain of RelA, away from DNA
binding groove.
(C) 293T cells were transfected with NF-κB reporter cocktail increasing
amount of wild-type RelA, RelA N64 to alanine mutant (RelA-N64A) or
N64 deamidated RelA (RelA-N64D). NF-κB reporter expression was
determined by luciferase assay.
(D) 293T cells were transfected with NF-κB reporter cocktail increasing
amount of wild-type RelA, N64 deamidated RelA (RelA-N64D), N139
deamidated RelA (RelA-N139D) or double deamidated RelA (RelA-DD).
NF-κB reporter expression was determined by luciferase assay.
(E) 293T cells were transfected with NF-κB reporter cocktail and RelA, RelA-
N64D or RelA-N64A with increasing amount of CAD. NF-κB reporter
expression was determined by luciferase assay.
(F) Rela
-/-
MEFs were ‘reconstituted’ with wild-type RelA or deamidated RelA
(RelA-DD). WCLs were harvested and analyzed by SDS page and
immunoblotting.
(H) ‘Reconstituted’ MEFs were infected with Sendai virus (100 [HAU]/ml) for
10hrs. RNA was extracted and cDNA was prepared to determine MIP2
and CCL5 mRNA expression by real-time PCR analysis.
**p<0.01, ***p<0.001, Error bars denote SD (n = 3)
90

Figure 20

Figure 5. Model of RelA deamidation by CAD

91

4. Materials and Methods
4.1 Cell lines and Viruses
HEK293T, HeLa, VERO, HCT116, BHK21, mouse embryonic fibroblasts (MEFs)
and human foreskin fibroblasts (HFF) were cultured in Dulbecco’s modified
Eagle’s medium (DMEM, Corning) supplemented with 10% heat-inactivated fetal
bovine serum (FBS; HyClone), penicillin (100 U/mL) and streptomycin (100
μg/mL). Wild-type, Rig-i
-/-
MEFs were described previously. Rela
-/-
MEFs was a
gift from Dr. Lin-feng Chen. 293T-STING was a gift from Dr. Jae U Jung. Wild-
type HSV-1 (KOS strain), HSV-1 GFP and other HSV-1 mutant viruses were
amplified in VERO cells. HSV-1 UL37-1101A was a gift from Dr. David Knipe.
HSV-1 UL37-WT rescued virus and HSV-1 UL37-d19 rescued virus were
generated by homologous recombination via transfecting 293T cells with HSV-1
ΔUL37 (KOS) Bacmid and respective UL37 mutant sequences. VSV was
amplified in BHK21 cells. Sendai virus was purchased from Charles River
Laboratories.
4.2 Constructs
Luciferase reporter plasmids for NF-κB, IFN-β promoter, PRDIII (ISRE) promoter,
mammalian expression plasmids for RIG-I and their truncated mutants, MDA5,
MAVS, IKKβ, TBK1,IRF3-5D, RelA were described previously. Human and
Mouse CAD expression plasmid was purchased from Origene. The non-silencing
(control) shRNA plasmid and shRNA plasmids for human RIG-I, human GAT
family, human UMPS were purchased from Thermo Scientific. HSV-1 expression
library was a gift from Dr. Weiming Yuan. Mammalian expression plasmids for
truncated mutants of UL37, truncated mutants of CAD and lentiviral expression
92

plamid for UL37 were generated by standard molecular biology techniques. All
point mutants, including RIG-I, RelA and UL37, were generated by site-directed
mutagenesis and confirm by sequencing. HSV-1 ΔUL37 (KOS) Bacmid was a gift
from Dr. Weiming Yuan and Dr. Thomas C. Mettenleiter.
4.3 Antibodies and reagents
Antibody against UL37 was a gift from Dr. Weiming Yuan. Antibodies against
GST (Z-5), IRF3 (FL-425), RelA (C-20), TRAF6 (D10) and RIG-I (H-300) were
purchased from Santa Cruz Biotechnology. The antibodies against FLAG (M2)
(Sigma), V5 (Bethyl Group), CAD (Bethyl Lab), RIG-I (Enzo Life Sciences), PFAS
(Abcam), dsRNA-J2 (SCICONS), P-S172 TBK-1 (Cell Signaling), P-S536 RelA
(Bethyl Lab) and β-actin (Abcam) were purchased from the indicated suppliers.
Human and mouse TNF-α were purchased from R&D systems. N-
(phosphonacetyl)-L-aspartate (PALA) was obtained from NCI. 6-Diazo-5-oxo-L-
norleucine (DON) was purchased from Sigma. Low molecular weight Poly [I:C]
was purchased from Invivogen. Lipo2000 was purchased from Life Technologies.
4.4 Lentivirus-mediated Stable Cell Line Construction
Lentiviruses were generated following second generation lentiviral packaging
protocol. Briefly, HEK293T cells were transfected with the packaging plasmids
VSV-G and DR8.91 and the pCDH lentiviral expression vector or lentiviral shRNA
plasmids. At 48hrs post transfection, supernatant was harvested and filtered
(concentrated if necessary). HEK293T cells, MEFs, HeLa, HCT116 or HFF cells
were infected with the supernatant in the presence of polybrene (8 μg/ml). Cells
93

were selected 48hrs post infection and maintained in 10%FBS DMEM
supplemented with puromycin (1~2 μg/ml).
4.5 Dual-Luciferase Reporter Assay
HEK293T cells, seeded in 24-well plates (~50% cell density), were transfected
with IFN-β, PRDIII (ISRE) or NF-κB reporter plasmid cocktail (50 ng of luciferase
reporter plasmid and 5 ng of pRL Renilla luciferase control vector) and
expression plasmid (empty plasmid, one or multiple plasmids depend on the
experiment) by calcium phosphate precipitation. Cells were infected with Sendai
virus (100 HA/ml) or HSV-1 for 16hrs, transfected with Poly [I:C] for 24 hrs post
transfection for 16hrs or directly harvested 30-36 hrs post transfection. Whole cell
lysates were subject to measurement of firefly luciferase activity and renilla
luciferase activity by microplate reader (FLUOstar Omega).
4.6 Protein expression and purification
HEK293T cells were transfected with expression vector containing Flag-tag
fusion protein. Cells were harvested and lysed with Triton X-100 buffer (20 mM
Tris, pH=7.5, 150 mM NaCl, 1.5 mM MgCl
2
, 20 mM β-glycerophosphate, 1 mM
sodium orthovanadate, 10% glycerol, 0.5 mM EGTA, 0.5% Triton X-100)
supplemented with a protease inhibitor cocktail (Roche). Whole cell lysates were
sonicated and centrifuged at 12,000 rpm for 15 min. Supernatant was harvested,
filtered, pre-cleared with protein A/G agarose beads at 4° C for 1hr and then
incubated with anti-Flag agarose beads at 4° C for 4hrs. The agarose beads were
washed extensively and eluted with 0.2 mg/ml 3xFlag peptide. The eluted
proteins were analyzed by SDS gel electrophoresis and silver staining.
94

4.7 RNA mobility shift assay
RNA shift assays were performed as previously described. Radiolabeled γ-
32
P-5’-
ppp-dsRNA was purchased from Invivogen. Purified RIG-I and RIG-I mutant was
incubated with dsRNA at room temperature for 15 min. Binding buffer contains
20 mM Tris-HCl (pH=8.0), 1.5 mM MgCl
2
and 1.5mM DTT. Unlabeled ppp-dsRNA
was used as competitor. The reaction mixtures were run on 5 % native
polyacrylamide gels at a constant voltage of 200V. Gels were dried and
subjected to phosphorimaging.
4.8 Co-immunoprecipitation (Co-IP) and Immunoblotting
For exogenous protein Co-IP, HEK293T cells were transfected with indicated
expression plasmids for 48hrs. For endogenous Co-IP, cells were directly
harvested and lysed with NP40 buffer (50 mM Tris-HCl, pH = 7.4, 150 mM NaCl,
1% NP-40, 5 mM EDTA) supplemented with 20 mM β-glycerophosphate and 1
mM sodium orthovanadate. Whole cell lysates were sonicated, centrifuged and
pre-cleared with protein A/G agarose for 1hr. Pre-cleared samples were then
incubated with indicated antibodies overnight and protein A/G agarose for 1hr at
4° C, or with antibody/glutathione-conjugated agarose for 4hrs at 4° C. The
agarose beads were washed extensively and samples were eluted by boiling at
95 ° C for 10 min. Co-immunoprecipitated proteins were analyzed by SDS gel
electrophoresis and immunoblot.
All the immunoblotting were performed using the indicated primary antibodies
and IRDye800-conjugated secondary antibodies (Licor). Proteins were visualized
by Odyssey infrared imaging system (Licor).
95

4.9 IRF3 Dimerization assay
IRF3 dimerization assay was performed as previously described. Briefly,
HEK293T-UL37 and control stable cell lines were infected with Sendai virus (100
HA/ml) for 8hrs. Cells were harvested and lyzed with Lysis buffer (20 mM Tris-
HCl, pH = 7.4, 150 mM NaCl, 10% Glycerol, 0.5% NP-40, 1mM sodium
orthovanadate) and centrifuged at 12,000g for 5 min. Supernatant was mixed
with loading buffer (40% glycerol, 0.005% Bromophenol Blue) and DOC (0.5%
final concentration). The mixture was subjected to 9% native page
electrophoresis for 60 min at 200V, 4° C. IRF3 was visualized by immunoblotting.
4.10 RIG-I Purification and Gel Filtration
Virus-infected 293T(HeLa)/RIG-I-Flag stable cells or transfected 293T cells were
harvested and lysed in cold Triton X-100 buffer (20 mM Tris, pH = 7.5, 150 mM
NaCl, 1.5 mM MgCl
2
, 20 mM β-glycerophosphate, 1 mM sodium orthovanadate,
10% glycerol, 0.5 mM EGTA, 0.5% Triton X-100, 1 mM PMSF and 10 µ g/ml
leupeptin). Centrifuged supernatant was filtered and subjected to incubation with
anti-Flag-conjugated agarose beads for 2 hours at 4° C. Beads were then
extensively washed and proteins were eluted with 3xFlag peptide at 0.2 mg/ml.
Gel filtration with superose 6 was performed as described previously. Briefly,
purified proteins (200-300 µ l) were loaded to superose 6 column and subjected to
gel filtration analysis with Buffer B (20 mM Tris-HCl, pH = 7.6, 150 mM NaCl, 1
mM EDTA, 0.5 mM EGTA, 0.5% Triton X-100, 20 mM NaF, 20 mM β-
glycerophosphate, 1 mM Na
3
VO
4
, 2.5 mM metabisulphite [sodium salt], 5 mM
benzamidine). Elution was collected in 0.5 ml fractions and aliquots of fractions
were analyzed by immunoblotting.
96

4.11 Whole cell lysate Gel Filtration
Mock/HSV-1 infected cells (2 10
7
) were harvested and lysed in 300 µ l cold Triton
X-100 buffer. Samples were sonicated briefly and centrifuged. Supernatant was
filtered and loaded to superose 6 column and subjected to gel filtration analysis
with Buffer B. Elution was collected in 0.5 ml fractions and aliquots of fractions
were analyzed by immunoblotting.
4.12 Two-dimensional Gel Electrophoresis
Cells (1 10
6
) were lysed in 150 µ l rehydration buffer (8 M Urea, 2% CHAPS, 0.5%
IPG Buffer, 0.002% bromophenol blue). Whole cell lysates were treated by one
pulse of sonication and centrifuged at 20,000 g for 15 min. Supernatants were
loaded to IEF strips for focusing with a program comprising: 0 V, 10 h
(rehydration); 300 V, 1 h; 1000 V, 1 h; 1000 V- 5000 V, 1 h; 5000 V, 4 h. After
IEF, strips were incubated with SDS equilibration buffer (50 mM Tris-HCl pH =
8.8, 6 M urea, 30% glycerol, 2% SDS, 0.001% Bromophenol Blue) containing 10
mg/ml DTT for 15 min and same buffer containing 2-iodoacetamide for 15 min.
Strips were washed with SDS-PAGE buffer, resolved by SDS-PAGE, and
analyzed by immunoblotting.
4.13 Quantitative Real-time PCR (qRT-PCR)
Quantitative Real-time PCR was performed as previously described. Cells were
infected or treated with indicated Viruses/Ligands for indicated time period. Total
RNA was extracted using TRIzol reagent (Invitrogen). Complementary cDNA was
synthesized from DNase I-treated total RNA using reverse transcriptase
(Invitrogen). cDNA was diluted and qRT-PCR was performed using SYBR Green
Master Mix (Applied Biosystems) by real-time PCR instrument (Applied
97

Biosystems). Relative mRNA expression for each target gene was calculated by
the 2
-ΔΔCt
method using β-Actin as an internal control. The sequences of qRT-
PCR primers are as follows:
Human β-actin forward 5’-CTGGCACCCAGCACAATG-3’
reverse 5’-GCCGATCCACACGGAGTACT-3’
Human IFN- β forward 5’-AGGACAGGATGAACTTTGAC-3’
reverse 5’-TGATAGACATTAGCCAGGAG-3’
Human ISG56 forward 5’-TCTCAGAGGAGCCTGGCTAA-3’
reverse 5’-TGACATCTCAATTGCTCCAG-3’
Human IL8 forward 5'-GGCACAAACTTTCAGAGACAG-3'
reverse 5'-ACACAGAGCTGCAGAAATCAGG-3'
Human CXCL2 forward 5’-GGGCAGAAAGCTTGTCTCAA-3’
reverse 5’-GCTTCCTCCTTCCTTCTGGT-3’
Human CAD forward 5’-TGCTCACCTATCCTCTGATCG-3’
reverse 5’-GCTGGGAGTAGGACAGCAC-3’
Mouse β-actin forward 5’- ACGGCCAGGTCATCACTATTG-3’
reverse 5’-CAAGAAGGAAGGCTGGAAAAGA-3’
Mouse CCL-5 forward 5’-CCTGCTGCTTTGCCTACCTCTC-3’
reverse 5’-ACACACTTGGCGGTTCCTTCGA-3’
Mouse MIP2 forward 5’-CTCTCAAGGGCGGTCAAAAAGTT-3’
reverse 5’-TCAGACAGCGAGGCACATCAGGTA-3’
Mouse IL8 forward 5ꞌ-CACCTCAAGAACATCCAGAGCT-3ꞌ
98

reverse 5ꞌ-CAAGCAG AACTGAACTACCATCG-3ꞌ
Mouse TNF- α forward 5′-ACTGAACTTCGGGGTGATCGGTCC-3’
reverse 5′-GTGGGTGAGGAGCACGTAGTCG-3’

4.14 In vitro Deamidation Assay
GST-RIG-I was purified from transfected 293T cells to homogeneity as
determined by silver staining. In vitro on-column deamidation of RIG-I was
performed as previously reported. Briefly, ~0.2 μg of UL37 (with or without
treatment by 500μM DON for 15min), and 0.6 μg of GST-RIG-I (bound to
glutathione-conjugated agarose) were added to a total volume of 30 μl. The
reaction was carried out at 30° C for 45 min in deamidation buffer (50 mM Tris-
HCl, pH=7.5, 100 mM NaCl, 5mM MgCl
2
). The GST beads were then washed
with deamidation buffer and GST-RIG-I was eluted with rehydration buffer (8 M
Urea, 2% CHAPS, 0.5% IPG Buffer, 0.002% bromophenol blue) at room
temperature. Samples were then analyzed by two-dimensional gel
electrophoresis and immunoblotting.
GST-RelA was purified from transfected from transfected 293T cells to
homogeneity as determined by silver staining. In vitro on-column deamidation of
RelA was performed similar to RIG-I. 0.2 μg of CAD, CAD-ED or CAD pre-
treated with DON were mixed with 0.6 μg agarose-bound GST-RelA. The
reaction was carried out at 30° C for 45 min in CAD deamidation buffer (100 mM
Tris–HCl, pH = 8.0, 7.5% dimethylsulfoxide, 2.5% glycerol, 100 mM KCl, 1 mM
DTT, 20.2 mM aspartate, 1.5 mM ATP, 200 mM phosphoribosyl 5’-
99

pyrophosphate, 3.5 mM MgCl
2
, and 5 mM NaHCO
3
). The GST beads were then
washed with deamidation buffer and GST-RIG-I was eluted with rehydration
buffer (8 M Urea, 2% CHAPS, 0.5% IPG Buffer, 0.002% bromophenol blue) at
room temperature. Samples were then analyzed by two-dimensional gel
electrophoresis and immunoblotting.
4.15 In Vitro ATPase Activity Assay
The ATPase activity assays for RIG-I and RIG-I-DD were performed as follows.
Purified RIG-I or RIG-I-DD was incubated with 5’-ppp-dsRNA (Invivogen) at 37° C
for 20 min in ATPase reaction buffer (50 mM Tris-HCl, pH = 7.5, 2.5 mM MgCl
2
,
and ATP. Released phosphates were measured using a PiColorLockTM
phosphate detection reagent (Innova Biosciences). For reactions with varying
concentrations of ATP, the concentrations of the proteins and RNA were 20 nM
and 80 nM, respectively. For reactions with varying concentrations of the RNA,
the concentrations of the proteins and ATP were 20 nM and 500 μM, respectively.
4.16 Mass Spectrometric Analysis
Samples were denatured, reduced with DTT, alkylated with Iodacetamide, and
digested with trypsin as described previously (Zhou et al., 2011). Samples were
analyzed using an LC/MS system consisting of an Eksigent NanoLC Ultra 2D
(Dublin, CA) and Thermo Fisher Scientific LTQ Orbitrap XL (San Jose, CA).
Briefly, peptides were separated in a 10 cm column (75 μm inner diameter)
packed in-house with 5 μm C18 beads on a Eksigent NanoLC Ultra 2D system
using a binary gradient of buffer A (0.1% formic acid) and buffer B (0.1% formic
acid and 80% ACN). The peptides were loaded directly without any trapping
column with buffer A at a flow rate of 300 nL/min. Elution was carried out at a
100

flow rate of 250 nL/min, with a linear gradient from 10% to 35% buffer B in 95 min
followed by 50% buffer B for 15 min. At the end of the gradient, the column was
washed with 90% buffer B and equilibrated with 5% buffer B for 10 min. The
eluted peptides were sprayed into the LTQ Orbitrap XL. The sourcewas operated
at 2.1-2.25 kV, with no sheath gas flow, with theion transfer tube at 250° C. MS
spectra in the range of m/z 350–2000 were acquired in the orbitrap at a FWHM
resolution of 30,000 after accumulation to an AGC target value of 500,000 in the
linear ion trap with 1 microscan.
For peptide sequencing and modification site localization, the same precursors
selected for fragmentation by CID, and fragment ions were analyzed in the liner
ion trap. The five most abundant precursor ions were selected for fragmentation
by CID. The instrument was operated in data-dependent acquisition mode,
whereby five CID data-dependent MS/MS scans succeeded the high resolution
MS scan. For all sequencing events, dynamic exclusion was enabled to minimize
repeated sequencing. Peaks selected for fragmentation more than once within 60
s were excluded from selection (10 ppm window).
Proteome Discoverer 1.4 (Thermo Fisher Scientific) was used for protein
identificationusing Sequest algorithms. The following criteria were followed. For
MS/MS spectra, variable modifications were selected to include N,Q deamination,
M oxidation and C carbamidomethylation with a maximum of four modifications.
Searches were conducted against Uniprot or in-house customer database. Up to
two missed cleavages were allowed for protease digestion and peptide had to be
fully tryptic.MS1 tolerance was 10 ppm and MS2 tolerance was set at 0.8 Da.
101

Peptides reported via search engine were accepted only if they met the false
discovery rate of 1%. There is no fixed cutoff score threshold, but instead spectra
are accepted until the 1% FDR rate is reached. Only peptides with a minimum of
six amino acid lengths were considered for identification. We also validated the
identifications by manual inspection of the mass spectra.
4.17 Statistical Analysis
Statistical analysis was performed by unpaired two-tailed Student’s t-test. P-
value less than 0.05 is considered statistically significant.

102

5. References
1. Akira S, Takeda K: Toll-like receptor signalling, Nature Reviews
Immunology 2004, 4:499-511
2. Kawai T, Akira S: The role of pattern-recognition receptors in innate
immunity: update on Toll-like receptors, Nature immunology 2010, 11:373-384
3. Takeuchi O, Akira S: Pattern recognition receptors and inflammation, Cell
2010, 140:805-820
4. Cai X, Chiu YH, Chen ZJ: The cGAS-cGAMP-STING pathway of cytosolic
DNA sensing and signaling, Molecular cell 2014, 54:289-296
5. Gross O, Poeck H, Bscheider M, Dostert C, Hannesschlager N, Endres S,
Hartmann G, Tardivel A, Schweighoffer E, Tybulewicz V, Mocsai A, Tschopp J,
Ruland J: Syk kinase signalling couples to the Nlrp3 inflammasome for anti-
fungal host defence, Nature 2009, 459:433-436
6. Shi J, Zhao Y, Wang Y, Gao W, Ding J, Li P, Hu L, Shao F: Inflammatory
caspases are innate immune receptors for intracellular LPS, Nature 2014,
514:187-192
7. Besch R, Poeck H, Hohenauer T, Senft D, Hacker G, Berking C, Hornung
V, Endres S, Ruzicka T, Rothenfusser S, Hartmann G: Proapoptotic signaling
induced by RIG-I and MDA-5 results in type I interferon-independent apoptosis in
human melanoma cells, The Journal of clinical investigation 2009, 119:2399-
2411
8. Taniguchi T, Ogasawara K, Takaoka A, Tanaka N: IRF family of
transcription factors as regulators of host defense, Annual review of immunology
2001, 19:623-655
9. Miyamoto M, Fujita T, Kimura Y, Maruyama M, Harada H, Sudo Y, Miyata
T, Taniguchi T: Regulated expression of a gene encoding a nuclear factor, IRF-1,
that specifically binds to IFN-beta gene regulatory elements, Cell 1988, 54:903-
913
10. Sharma S, tenOever BR, Grandvaux N, Zhou GP, Lin R, Hiscott J:
Triggering the interferon antiviral response through an IKK-related pathway,
Science 2003, 300:1148-1151
11. Fitzgerald KA, McWhirter SM, Faia KL, Rowe DC, Latz E, Golenbock DT,
Coyle AJ, Liao SM, Maniatis T: IKKepsilon and TBK1 are essential components
of the IRF3 signaling pathway, Nature immunology 2003, 4:491-496
12. Lin R, Heylbroeck C, Pitha PM, Hiscott J: Virus-dependent
phosphorylation of the IRF-3 transcription factor regulates nuclear translocation,
transactivation potential, and proteasome-mediated degradation, Molecular and
cellular biology 1998, 18:2986-2996
13. Yoneyama M, Suhara W, Fukuhara Y, Fukuda M, Nishida E, Fujita T:
Direct triggering of the type I interferon system by virus infection: activation of a
transcription factor complex containing IRF-3 and CBP/p300, The EMBO journal
1998, 17:1087-1095
14. Marie I, Durbin JE, Levy DE: Differential viral induction of distinct
interferon-alpha genes by positive feedback through interferon regulatory factor-7,
The EMBO journal 1998, 17:6660-6669
103

15. Sato M, Hata N, Asagiri M, Nakaya T, Taniguchi T, Tanaka N: Positive
feedback regulation of type I IFN genes by the IFN-inducible transcription factor
IRF-7, FEBS letters 1998, 441:106-110
16. Sato M, Suemori H, Hata N, Asagiri M, Ogasawara K, Nakao K, Nakaya T,
Katsuki M, Noguchi S, Tanaka N, Taniguchi T: Distinct and essential roles of
transcription factors IRF-3 and IRF-7 in response to viruses for IFN-alpha/beta
gene induction, Immunity 2000, 13:539-548
17. Honda K, Taniguchi T: IRFs: master regulators of signalling by Toll-like
receptors and cytosolic pattern-recognition receptors, Nature reviews
Immunology 2006, 6:644-658
18. Sen R, Baltimore D: Multiple nuclear factors interact with the
immunoglobulin enhancer sequences, Cell 1986, 46:705-716
19. Ghosh S, Hayden MS: New regulators of NF-kappaB in inflammation,
Nature Reviews Immunology 2008, 8:837-848
20. Hacker H, Karin M: Regulation and function of IKK and IKK-related
kinases, Sci STKE 2006, 2006:re13
21. Hayden MS, Ghosh S: Shared principles in NF-kappaB signaling, Cell
2008, 132:344-362
22. Baker RG, Hayden MS, Ghosh S: NF-kappaB, inflammation, and
metabolic disease, Cell metabolism 2011, 13:11-22
23. Sun SC: Non-canonical NF-kappaB signaling pathway, Cell research 2011,
21:71-85
24. Ghosh S, Karin M: Missing pieces in the NF-kappaB puzzle, Cell 2002,
109 Suppl:S81-96
25. Fujita T, Nolan GP, Ghosh S, Baltimore D: Independent modes of
transcriptional activation by the p50 and p65 subunits of NF-kappa B, Genes &
development 1992, 6:775-787
26. Jacobs MD, Harrison SC: Structure of an IkappaBalpha/NF-kappaB
complex, Cell 1998, 95:749-758
27. Schreck R, Rieber P, Baeuerle PA: Reactive oxygen intermediates as
apparently widely used messengers in the activation of the NF-kappa B
transcription factor and HIV-1, The EMBO journal 1991, 10:2247-2258
28. Beg AA, Finco TS, Nantermet PV, Baldwin AS, Jr.: Tumor necrosis factor
and interleukin-1 lead to phosphorylation and loss of I kappa B alpha: a
mechanism for NF-kappa B activation, Molecular and cellular biology 1993,
13:3301-3310
29. O'Neill LA: Signal transduction pathways activated by the IL-1
receptor/toll-like receptor superfamily, Current topics in microbiology and
immunology 2002, 270:47-61
30. Chow JC, Young DW, Golenbock DT, Christ WJ, Gusovsky F: Toll-like
receptor-4 mediates lipopolysaccharide-induced signal transduction, The Journal
of biological chemistry 1999, 274:10689-10692
31. Meylan E, Curran J, Hofmann K, Moradpour D, Binder M, Bartenschlager
R, Tschopp J: Cardif is an adaptor protein in the RIG-I antiviral pathway and is
targeted by hepatitis C virus, Nature 2005, 437:1167-1172
104

32. Seth RB, Sun L, Ea CK, Chen ZJ: Identification and characterization of
MAVS, a mitochondrial antiviral signaling protein that activates NF-kappaB and
IRF 3, Cell 2005, 122:669-682
33. Kawai T, Takahashi K, Sato S, Coban C, Kumar H, Kato H, Ishii KJ,
Takeuchi O, Akira S: IPS-1, an adaptor triggering RIG-I- and Mda5-mediated
type I interferon induction, Nature immunology 2005, 6:981-988
34. Xu LG, Wang YY, Han KJ, Li LY, Zhai Z, Shu HB: VISA is an adapter
protein required for virus-triggered IFN-beta signaling, Molecular cell 2005,
19:727-740
35. Sun L, Wu J, Du F, Chen X, Chen ZJ: Cyclic GMP-AMP synthase is a
cytosolic DNA sensor that activates the type I interferon pathway, Science 2013,
339:786-791
36. Unterholzner L, Keating SE, Baran M, Horan KA, Jensen SB, Sharma S,
Sirois CM, Jin T, Latz E, Xiao TS, Fitzgerald KA, Paludan SR, Bowie AG: IFI16 is
an innate immune sensor for intracellular DNA, Nature immunology 2010,
11:997-1004
37. Zhang Z, Yuan B, Bao M, Lu N, Kim T, Liu YJ: The helicase DDX41
senses intracellular DNA mediated by the adaptor STING in dendritic cells,
Nature immunology 2011, 12:959-965
38. Takaoka A, Wang Z, Choi MK, Yanai H, Negishi H, Ban T, Lu Y, Miyagishi
M, Kodama T, Honda K, Ohba Y, Taniguchi T: DAI (DLM-1/ZBP1) is a cytosolic
DNA sensor and an activator of innate immune response, Nature 2007, 448:501-
505
39. Deng L, Wang C, Spencer E, Yang L, Braun A, You J, Slaughter C,
Pickart C, Chen ZJ: Activation of the IkappaB kinase complex by TRAF6 requires
a dimeric ubiquitin-conjugating enzyme complex and a unique polyubiquitin chain,
Cell 2000, 103:351-361
40. Zandi E, Rothwarf DM, Delhase M, Hayakawa M, Karin M: The IkappaB
kinase complex (IKK) contains two kinase subunits, IKKalpha and IKKbeta,
necessary for IkappaB phosphorylation and NF-kappaB activation, Cell 1997,
91:243-252
41. De Andrea M, Ravera R, Gioia D, Gariglio M, Landolfo S: The interferon
system: an overview, European journal of paediatric neurology : EJPN : official
journal of the European Paediatric Neurology Society 2002, 6 Suppl A:A41-46;
discussion A55-48
42. Pestka S, Krause CD, Walter MR: Interferons, interferon-like cytokines,
and their receptors, Immunological reviews 2004, 202:8-32
43. Der SD, Zhou A, Williams BR, Silverman RH: Identification of genes
differentially regulated by interferon alpha, beta, or gamma using oligonucleotide
arrays, Proceedings of the National Academy of Sciences of the United States of
America 1998, 95:15623-15628
44. Darnell JE, Kerr IM, Stark GR: Jak-Stat Pathways and Transcriptional
Activation in Response to Ifns and Other Extracellular Signaling Proteins,
Science 1994, 264:1415-1421
45. Levy DE, Darnell JE: STATs: Transcriptional control and biological impact,
Nat Rev Mol Cell Bio 2002, 3:651-662
105

46. Stark GR, Darnell JE: The JAK-STAT Pathway at Twenty, Immunity 2012,
36:503-514
47. Darnell JE: STATs and gene regulation, Science 1997, 277:1630-1635
48. Aaronson DS, Horvath CM: A road map for those who don't know JAK-
STAT, Science 2002, 296:1653-1655
49. Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P,
Rice CM: A diverse range of gene products are effectors of the type I interferon
antiviral response, Nature 2011, 472:481-485
50. Kerr IM, Brown RE, Hovanessian AG: Nature of Inhibitor of Cell-Free
Protein-Synthesis Formed in Response to Interferon and Double-Stranded-Rna,
Nature 1977, 268:540-542
51. Meurs E, Chong K, Galabru J, Thomas NSB, Kerr IM, Williams BRG,
Hovanessian AG: Molecular-Cloning and Characterization of the Human Double-
Stranded-Rna Activated Protein-Kinase Induced by Interferon, Cell 1990, 62:379-
390
52. Haller O, Kochs G: Human MxA Protein: An Interferon-Induced Dynamin-
Like GTPase with Broad Antiviral Activity, J Interf Cytok Res 2011, 31:79-87
53. Hovanessian AG, Justesen J: The human 2 '-5 ' oligoadenylate synthetase
family: Unique interferon-inducible enzymes catalyzing 2 '-5 ' instead of 3 '-5 '
phosphodiester bond formation, Biochimie 2007, 89:779-788
54. Sheehy AM, Gaddis NC, Choi JD, Malim MH: Isolation of a human gene
that inhibits HIV-1 infection and is suppressed by the viral Vif protein, Nature
2002, 418:646-650
55. Stremlau M, Owens CM, Perron MJ, Kiessling M, Autissier P, Sodroski J:
The cytoplasmic body component TRIM5 alpha restricts HIV-1 infection in Old
World monkeys, Nature 2004, 427:848-853
56. Zhao C, Denison C, Huibregtse JM, Gygi S, Krug RM: Human ISG15
conjugation targets both IFN-induced and constitutively expressed proteins
functioning in diverse cellular pathways, Proceedings of the National Academy of
Sciences of the United States of America 2005, 102:10200-10205
57. Diamond MS, Farzan M: The broad-spectrum antiviral functions of IFIT
and IFITM proteins, Nature Reviews Immunology 2013, 13:46-57
58. Sadler AJ, Williams BRG: Interferon-inducible antiviral effectors, Nature
Reviews Immunology 2008, 8:559-568
59. Nanduri S, Rahman F, Williams BRG, Qin J: A dynamically tuned double-
stranded RNA binding mechanism for the activation of antiviral kinase PKR,
Embo Journal 2000, 19:5567-5574
60. Nallagatla SR, Hwang J, Toroney R, Zheng XF, Cameron CE, Bevilacqua
PC: 5 '-triphosphate-dependent activation of PKR by RNAs with short stem-loops,
Science 2007, 318:1455-1458
61. Dey M, Cao C, Dar AC, Tamura T, Ozato K, Sicheri F, Dever TE:
Mechanistic link between PKR dimerization, autophosphorylation, and elF2 alpha
substrate recognition, Cell 2005, 122:901-913
62. Romano PR, Garcia-Barrio MT, Zhang XL, Wang QZ, Taylor DR, Zhang F,
Herring C, Mathews MB, Qin J, Hinnebusch AG: Autophosphorylation in the
activation loop is required for full kinase activity in vivo of human and yeast
106

eukaryotic initiation factor 2 alpha kinases PKR and GCN2, Molecular and
cellular biology 1998, 18:2282-2297
63. Taylor DR, Lee SB, Romano PR, Marshak DR, Hinnebusch AG, Esteban
M, Mathews MB: Autophosphorylation sites participate in the activation of the
double-stranded-RNA-activated protein kinase PKR, Molecular and cellular
biology 1996, 16:6295-6302
64. Roberts WK, Hovanessian A, Brown RE, Clemens MJ, Kerr IM: Interferon-
Mediated Protein-Kinase and Low-Molecular-Weight Inhibitor of Protein-
Synthesis, Nature 1976, 264:477-480
65. Liu ZL, Pan QH, Ding SL, Qian J, Xu FW, Zhou JM, Cen S, Guo F, Liang
C: The Interferon-Inducible MxB Protein Inhibits HIV-1 Infection, Cell Host
Microbe 2013, 14:398-410
66. Kane M, Yadav SS, Bitzegeio J, Kutluay SB, Zang T, Wilson SJ,
Schoggins JW, Rice CM, Yamashita M, Hatziioannou T, Bieniasz PD: MX2 is an
interferon-induced inhibitor of HIV-1 infection, Nature 2013, 502:563-+
67. Goujon C, Moncorge O, Bauby H, Doyle T, Ward CC, Schaller T, Hue S,
Barclay WS, Schulz R, Malim MH: Human MX2 is an interferon-induced post-
entry inhibitor of HIV-1 infection, Nature 2013, 502:559-+
68. Rebouillat D, Hovanessian AG: The human 2 ',5 '-oligoadenylate
synthetase family: Interferon-induced proteins with unique enzymatic properties,
J Interf Cytok Res 1999, 19:295-308
69. Clemens MJ, Williams BRG: Inhibition of Cell-Free Protein-Synthesis by
Pppa2'p5'a2'p5'a - Novel Oligonucleotide Synthesized by Interferon-Treated L-
Cell Extracts, Cell 1978, 13:565-572
70. Malathi K, Dong BH, Gale M, Silverman RH: Small self-RNA generated by
RNase L amplifies antiviral innate immunity, Nature 2007, 448:816-U819
71. Yoneyama M, Kikuchi M, Natsukawa T, Shinobu N, Imaizumi T, Miyagishi
M, Taira K, Akira S, Fujita T: The RNA helicase RIG-I has an essential function in
double-stranded RNA-induced innate antiviral responses, Nature immunology
2004, 5:730-737
72. Kang DC, Gopalkrishnan RV, Wu QP, Jankowsky E, Pyle AM, Fisher PB:
mda-5: An interferon-inducible putative RNA helicase with double-stranded RNA-
dependent ATPase activity and melanoma growth-suppressive properties,
Proceedings of the National Academy of Sciences of the United States of
America 2002, 99:637-642
73. Adhikari A, Xu M, Chen ZJ: Ubiquitin-mediated activation of TAK1 and IKK,
Oncogene 2007, 26:3214-3226
74. Oganesyan G, Saha SK, Guo BC, He JQ, Shahangian A, Zarnegar B,
Perry A, Cheng GH: Critical role of TRAF3 in the Toll-like receptor-dependent
and -independent antiviral response, Nature 2006, 439:208-211
75. Liu SQ, Cai X, Wu JX, Cong Q, Chen X, Li T, Du FH, Ren JY, Wu YT,
Grishin NV, Chen ZJJ: Phosphorylation of innate immune adaptor proteins MAVS,
STING, and TRIF induces IRF3 activation, Science 2015, 347:1217-U1217
76. Yoneyama M, Suhara W, Fukuhara Y, Fukuda M, Nishida E, Fujita T:
Direct triggering of the type I interferon system by virus infection: activation of a
107

transcription factor complex containing IRF-3 and CBP/p300, Embo Journal 1998,
17:1087-1095
77. Barral PM, Sarkar D, Su ZZ, Barber GN, DeSalle R, Racaniello VR, Fisher
PB: Functions of the cytoplasmic RNA sensors RIG-I and MDA-5: Key regulators
of innate immunity, Pharmacol Therapeut 2009, 124:219-234
78. Weber F, Wagner V, Rasmussen SB, Hartmann R, Paludan SR: Double-
stranded RNA is produced by positive-strand RNA viruses and DNA viruses but
not in detectable amounts by negative-strand RNA viruses, Journal of virology
2006, 80:5059-5064
79. Schlee M, Roth A, Hornung V, Hagmann CA, Wimmenauer V, Barchet W,
Coch C, Janke M, Mihailovic A, Wardle G, Juranek S, Kato H, Kawai T, Poeck H,
Fitzgerald KA, Takeuchi O, Akira S, Tuschl T, Latz E, Ludwig J, Hartmann G:
Recognition of 5 ' Triphosphate by RIG-I Helicase Requires Short Blunt Double-
Stranded RNA as Contained in Panhandle of Negative-Strand Virus, Immunity
2009, 31:25-34
80. Kato H, Sato S, Yoneyama M, Yamamoto M, Uematsu S, Matsui K,
Tsujimura T, Takeda K, Fujita T, Takeuchi O, Akira S: Cell type-specific
involvement of RIG-I in antiviral response, Immunity 2005, 23:19-28
81. Plumet S, Herschke F, Bourhis JM, Valentin H, Longhi S, Gerlier D:
Cytosolic 5 '-Triphosphate Ended Viral Leader Transcript of Measles Virus as
Activator of the RIG I-Mediated Interferon Response, Plos One 2007, 2:
82. Loo YM, Fornek J, Crochet N, Bajwa G, Perwitasari O, Martinez-Sobrido L,
Akira S, Gill MA, Garcia-Sastre A, Katze MG, Gale M: Distinct RIG-I and MDA5
signaling by RNA viruses in innate immunity, Journal of virology 2008, 82:335-
345
83. Kato H, Takeuchi O, Sato S, Yoneyama M, Yamamoto M, Matsui K,
Uematsu S, Jung A, Kawai T, Ishii KJ, Yamaguchi O, Otsu K, Tsujimura T, Koh
CS, Sousa CRE, Matsuura Y, Fujita T, Akira S: Differential roles of MDA5 and
RIG-I helicases in the recognition of RNA viruses, Nature 2006, 441:101-105
84. Hornung V, Ellegast J, Kim S, Brzozka K, Jung A, Kato H, Poeck H, Akira
S, Conzelmann KK, Schlee M, Endres S, Hartmann G: 5 '-triphosphate RNA is
the ligand for RIG-I, Science 2006, 314:994-997
85. Saito T, Owen DM, Jiang FG, Marcotrigiano J, Gale M: Innate immunity
induced by composition-dependent RIG-I recognition of hepatitis C virus RNA,
Nature 2008, 454:523-527
86. Pichlmair A, Schulz O, Tan CP, Naslund TI, Liljestrom P, Weber F, Sousa
CRE: RIG-I-mediated antiviral responses to single-stranded RNA bearing 5 '-
phosphates, Science 2006, 314:997-1001
87. Goubau D, Schlee M, Deddouche S, Pruijssers AJ, Zillinger T, Goldeck M,
Schuberth C, Van der Veen AG, Fujimura T, Rehwinkel J, Iskarpatyoti JA,
Barchet W, Ludwig J, Dermody TS, Hartmann G, Sousa CRE: Antiviral immunity
via RIG-I-mediated recognition of RNA bearing 5 '-diphosphates, Nature 2014,
514:372-+
88. Saito T, Hirai R, Loo YM, Owen D, Johnson CL, Sinha SC, Akira S, Fujita
T, Gale M: Regulation of innate antiviral defenses through a shared repressor
108

domain in RIG-I and LGP2, Proceedings of the National Academy of Sciences of
the United States of America 2007, 104:582-587
89. Xu H, He XJ, Zheng H, Huang LJ, Hou FJ, Yu ZH, de la Cruz MJ,
Borkowski B, Zhang XW, Chen ZJJ, Jiang QX: Structural basis for the prion-like
MAVS filaments in antiviral innate immunity, Elife 2014, 3:
90. Hou FJ, Sun LJ, Zheng H, Skaug B, Jiang QX, Chen ZJ: MAVS Forms
Functional Prion-like Aggregates to Activate and Propagate Antiviral Innate
Immune Response (vol 146, pg 448, 2011), Cell 2011, 146:841-841
91. Wu B, Peisley A, Tetrault D, Li ZL, Egelman EH, Magor KE, Walz T,
Penczek PA, Hur S: Molecular Imprinting as a Signal-Activation Mechanism of
the Viral RNA Sensor RIG-I, Molecular cell 2014, 55:511-523
92. Anchisi S, Guerra J, Garcin D: RIG-I ATPase Activity and Discrimination of
Self-RNA versus Non-Self-RNA, Mbio 2015, 6:
93. Cui S, Eisenacher K, Kirchhofer A, Brzozka K, Lammens A, Lammens K,
Fujita T, Conzelmann KK, Krug A, Hopfner KP: The C-terminal regulatory domain
is the RNA 5 '-triphosphate sensor of RIG-I, Molecular cell 2008, 29:169-179
94. Loo YM, Gale M: Immune Signaling by RIG-I-like Receptors, Immunity
2011, 34:680-692
95. Gack MU, Shin YC, Joo CH, Urano T, Liang C, Sun LJ, Takeuchi O, Akira
S, Chen ZJ, Inoue SS, Jung JU: TRIM25 RING-finger E3 ubiquitin ligase is
essential for RIG-I-mediated antiviral activity, Nature 2007, 446:916-U912
96. Lin RT, Yang L, Nakhaei P, Sun Q, Sharif-Askari E, Julkunen I, Hiscott J:
Negative regulation of the retinoic acid-inducible gene I-induced antiviral state by
the ubiquitin-editing protein A20, Journal of Biological Chemistry 2006, 281:2095-
2103
97. Friedman CS, O'Donnell MA, Legarda-Addison D, Ng A, Cardenas WB,
Yount JS, Moran TM, Basler CF, Komuro A, Horvath CM, Xavier R, Ting AT: The
tumour suppressor CYLD is a negative regulator of RIG-I-mediated antiviral
response, Embo Rep 2008, 9:930-936
98. Inn KS, Lee SH, Rathbun JY, Wong LY, Toth Z, Machida K, Ou JHJ, Jung
JU: Inhibition of RIG-I-Mediated Signaling by Kaposi's Sarcoma-Associated
Herpesvirus-Encoded Deubiquitinase ORF64, Journal of virology 2011,
85:10899-10904
99. Arimoto KI, Takahashi H, Hishiki T, Konishi H, Fujita T, Shimotohno K:
Negative regulation of the RIG-I signaling by the ubiquitin ligase RNF125,
Proceedings of the National Academy of Sciences of the United States of
America 2007, 104:7500-7505
100. Gack MU, Nistal-Villan E, Inn KS, Garcia-Sastre A, Jung JU:
Phosphorylation-mediated negative regulation of RIG-I antiviral activity, Journal
of virology 2010, 84:3220-3229
101. Wies E, Wang MK, Maharaj NP, Chen K, Zhou S, Finberg RW, Gack MU:
Dephosphorylation of the RNA Sensors RIG-I and MDA5 by the Phosphatase
PP1 Is Essential for Innate Immune Signaling, Immunity 2013, 38:437-449
102. Mesman AW, Zijlstra-Willems EM, Kaptein TM, de Swart RL, Davis ME,
Ludlow M, Duprex WP, Gack MU, Gringhuis SI, Geijtenbeek TB: Measles virus
109

suppresses RIG-I-like receptor activation in dendritic cells via DC-SIGN-mediated
inhibition of PP1 phosphatases, Cell Host Microbe 2014, 16:31-42
103. Kim MJ, Hwang SY, Imaizumi T, Yoo JY: Negative feedback regulation of
RIG-1-mediated antiviral signaling by interferon-induced ISG15 conjugation,
Journal of virology 2008, 82:1474-1483
104. Jounai N, Takeshita F, Kobiyama K, Sawano A, Miyawaki A, Xin KQ, Ishii
KJ, Kawaii T, Akira S, Suzuki K, Okuda K: The Atg5-Atg12 conjugate associates
with innate antiviral immune responses, Proceedings of the National Academy of
Sciences of the United States of America 2007, 104:14050-14055
105. Yoneyama M, Kikuchi M, Matsumoto K, Imaizumi T, Miyagishi M, Taira K,
Foy E, Loo YM, Gale M, Akira S, Yonehara S, Kato A, Fujita T: Shared and
unique functions of the DExD/H-box helicases RIG-I, MDA5, and LGP2 in
antiviral innate immunity, J Immunol 2005, 175:2851-2858
106. Gack MU, Kirchhofer A, Shin YC, Inn KS, Liang CY, Cui S, Myong S, Ha T,
Hopfner KP, Jung JU: Roles of RIG-1 N-terminal tandem CARD and splice
variant in TRIM25-mediated antiviral signal transduction, Proceedings of the
National Academy of Sciences of the United States of America 2008, 105:16743-
16748
107. Marq JB, Hausmann S, Luban J, Kolakofsky D, Garcin D: The Double-
stranded RNA Binding Domain of the Vaccinia Virus E3L Protein Inhibits Both
RNA- and DNA-induced Activation of Interferon beta, Journal of Biological
Chemistry 2009, 284:25471-25478
108. Cardenas WB, Loo YM, Gale M, Hartman AL, Kimberlin CR, Martinez-
Sobrido L, Saphire EO, Basler CF: Ebola virus VP35 protein binds double-
stranded RNA and inhibits alpha/beta interferon production induced by RIG-I
signaling, Journal of virology 2006, 80:5168-5178
109. Barral PM, Sarkar D, Fisher PB, Racaniello VR: RIG-I is cleaved during
picornavirus infection, Virology 2009, 391:171-176
110. Gack MU, Albrecht RA, Urano T, Inn KS, Huang IC, Carnero E, Farzan M,
Inoue S, Jung JU, Garcia-Sastre A: Influenza A Virus NS1 Targets the Ubiquitin
Ligase TRIM25 to Evade Recognition by the Host Viral RNA Sensor RIG-I, Cell
Host Microbe 2009, 5:439-449
111. He S, Zhao J, Song S, He X, Minassian A, Zhou Y, Zhang J, Brulois K,
Wang Y, Cabo J, Zandi E, Liang C, Jung JU, Zhang X, Feng P: Viral Pseudo-
Enzymes Activate RIG-I via Deamidation to Evade Cytokine Production,
Molecular cell 2015, 58:134-146
112. Shembade N, Harhaj EW: Regulation of NF-kappa B signaling by the A20
deubiquitinase, Cell Mol Immunol 2012, 9:123-130
113. de Oliveira DE, Ballon G, Cesarman E: NF-kappaB signaling modulation
by EBV and KSHV, Trends Microbiol 2010, 18:248-257
114. Varin A, Manna SK, Quivy V, Decrion AZ, Van Lint C, Herbein G,
Aggarwal BB: Exogenous Nef protein activates NF-kappa B, AP-1, and c-Jun N-
terminal kinase and stimulates HIV transcription in promonocytic cells. Role in
AIDS pathogenesis, The Journal of biological chemistry 2003, 278:2219-2227
115. Patel A, Hanson J, McLean TI, Olgiate J, Hilton M, Miller WE,
Bachenheimer SL: Herpes simplex type 1 induction of persistent NF-kappa B
110

nuclear translocation increases the efficiency of virus replication, Virology 1998,
247:212-222
116. Dong X, Feng P: Murine gamma herpesvirus 68 hijacks MAVS and
IKKbeta to abrogate NFkappaB activation and antiviral cytokine production, PLoS
pathogens 2011, 7:e1002336
117. Dong X, He Z, Durakoglugil D, Arneson L, Shen Y, Feng P: Murine
gammaherpesvirus 68 evades host cytokine production via replication
transactivator-induced RelA degradation, Journal of virology 2012, 86:1930-1941
118. Dong X, Feng H, Sun Q, Li H, Wu TT, Sun R, Tibbetts SA, Chen ZJ, Feng
P: Murine gamma-herpesvirus 68 hijacks MAVS and IKKbeta to initiate lytic
replication, PLoS pathogens 2010, 6:e1001001
119. Zhao J, He S, Minassian A, Li J, Feng P: Recent advances on viral
manipulation of NF-kappaB signaling pathway, Current opinion in virology 2015,
15:103-111
120. Ishinaga H, Jono H, Lim JH, Kweon SM, Xu HD, Ha UH, Xu HD, Koga T,
Yan C, Feng XH, Chen LF, Li JD: TGF-beta induces p65 acetylation to enhance
bacteria-induced NF-kappa B activation, Embo Journal 2007, 26:1150-1162
121. Jamaluddin M, Wang SF, Boldogh I, Tian B, Brasier AR: TNF-alpha-
induced NF-kappa B/Re1A Ser(276) phosphorylation and enhanceosome
formation is mediated by an ROS-dependent PKAc pathway, Cell Signal 2007,
19:1419-1433
122. Vermeulen L, De Wilde G, Van Damme P, Vanden Berghe W, Haegeman
G: Transcriptional activation of the NF-kappa B p65 subunit by mitogen- and
stress-activated protein kinase-1 (MSK1), Embo Journal 2003, 22:1313-1324
123. Zhong HH, Voll RE, Ghosh S: Phosphorylation of NF-kappa B p65 by PKA
stimulates transcriptional activity by promoting a novel bivalent interaction with
the coactivator CBP/p300, Molecular cell 1998, 1:661-671
124. Nowak DE, Tian B, Jamaluddin M, Boldogh I, Vergara LA, Choudhary S,
Brasier AR: RelA Ser(276) phosphorylation is required for activation of a subset
of NF-kappa B-dependent genes by recruiting cyclin-dependent kinase 9/cyclin
T1 complexes, Molecular and cellular biology 2008, 28:3623-3638
125. Sakurai H, Chiba H, Miyoshi H, Sugita T, Toriumi W: I kappa B kinases
phosphorylate NF-kappa B p65 subunit on serine 536 in the transactivation
domain, Journal of Biological Chemistry 1999, 274:30353-30356
126. Buss H, Dorrie A, Schmitz ML, Hoffmann E, Resch K, Kracht M:
Constitutive and interleukin-1-inducible phosphorylation of p65 NF-kappa B at
serine 536 is mediated by multiple protein kinases including I kappa B kinase
(IKK)-alpha, IKK beta, IKK epsilon, TRAF family member-associated (TANK)-
binding kinase 1 (TBK1), and an unknown kinase and couples p65 to TATA-
binding protein-associated factor II31-mediated interleukin-8 transcription,
Journal of Biological Chemistry 2004, 279:55633-55643
127. Lamb A, Yang XD, Tsang YHN, Li JD, Higashi H, Hatakeyama M, Peek
RM, Blanke SR, Chen LF: Helicobacter pylori CagA activates NF-kappa B by
targeting TAK1 for TRAF6-mediated Lys 63 ubiquitination, Embo Rep 2009,
10:1242-1249
111

128. Chen LF, Williams SA, Mu YJ, Nakano H, Duerr JM, Buckbinder L,
Greene WC: NF-kappa B RelA phosphorylation regulates RelA acetylation,
Molecular and cellular biology 2005, 25:7966-7975
129. Mattioli I, Geng H, Sebald A, Hodel M, Bucher C, Kracht M, Schmitz ML:
Inducible phosphorylation of NF-kappa B p65 at serine 468 by T cell
costimulation is mediated by IKK epsilon, Journal of Biological Chemistry 2006,
281:6175-6183
130. Schwabe RF, Sakurai H: IKK beta phosphorylates p65 at S468 in
transactivaton domain 2, Faseb J 2005, 19:1758-+
131. Huang B, Yang XD, Lamb A, Chen LF: Posttranslational modifications of
NF-kappaB: another layer of regulation for NF-kappaB signaling pathway, Cell
Signal 2010, 22:1282-1290
132. Yang JM, Fan GH, Wadzinski BE, Sakurai H, Richmond A: Protein
phosphatase 2A interacts with and directly dephosphorylates Re1A., Journal of
Biological Chemistry 2001, 276:47828-47833
133. Chew J, Biswas S, Shreeram S, Humaidi M, Wong ET, Dhillion MK, Teo H,
Hazra A, Fang CC, Lopez-Collazo E, Bulavin DV, Tergaonkar V: WIP1
phosphatase is a negative regulator of NF-kappa B signalling, Nat Cell Biol 2009,
11:659-U493
134. Chen LF, Mu YJ, Greene WC: Acetylation of ReIA at discrete sites
regulates distinct nuclear functions of NF-kappa B, Embo Journal 2002, 21:6539-
6548
135. Gringhuis SI, den Dunnen J, Litjens M, Hof BV, van Kooyk Y, Geijtenbeek
TBH: C-type lectin DC-SIGN modulates toll-like receptor signaling via Raf-1
kinase-dependent acetylation of transcription factor NF-kappa B, Immunity 2007,
26:605-616
136. Chen LF, Fischle W, Verdin E, Greene WC: Duration of nuclear NF-kappa
B action regulated by reversible acetylation, Science 2001, 293:1653-1657
137. Yang XD, Huang B, Li MX, Lamb A, Kelleher NL, Chen LF: Negative
regulation of NF-kappa B action by Set9-mediated lysine methylation of the RelA
subunit, Embo Journal 2009, 28:1055-1066
138. Ryo A, Suizu F, Yoshida Y, Perrem K, Liou YC, Wulf G, Rottapel R,
Yamaoka S, Lu KP: Regulation of NF-kappa B signaling by Pin1-dependent
prolyl isomerization and ubiquitin-mediated proteolysis of p65/RelA, Molecular
cell 2003, 12:1413-1426
139. Tanaka T, Grusby MJ, Kaisho T: PDLIM2-mediated termination of
transcription factor NF-kappa B activation by intranuclear sequestration and
degradation of the p65 subunit, Nature immunology 2007, 8:584-591
140. Li H, Wittwer T, Weber A, Schneider H, Moreno R, Maine GN, Kracht M,
Schmitz ML, Burstein E: Regulation of NF-kappa B activity by competition
between RelA acetylation and ubiquitination, Oncogene 2012, 31:611-623
141. Mycek MJ, Waelsch H: Enzymatic Deamidation of Proteins, Journal of
Biological Chemistry 1960, 235:3513-3517
142. Robinson NE, Robinson AB: Prediction of protein deamidation rates from
primary and three-dimensional structure, Proceedings of the National Academy
of Sciences of the United States of America 2001, 98:4367-4372
112

143. Robinson NE, Robinson AB: Molecular clocks, Proceedings of the
National Academy of Sciences of the United States of America 2001, 98:944-949
144. Washington EJ, Banfield MJ, Dangi JL: What a Difference a Dalton Makes:
Bacterial Virulence Factors Modulate Eukaryotic Host Cell Signaling Systems via
Deamidation, Microbiol Mol Biol R 2013, 77:527-539
145. Cui JX, Yao Q, Li S, Ding XJ, Lu QH, Mao HB, Liu LP, Zheng N, Chen S,
Shao F: Glutamine Deamidation and Dysfunction of Ubiquitin/NEDD8 Induced by
a Bacterial Effector Family, Science 2010, 329:1215-1218
146. Sanada T, Kim M, Mimuro H, Suzuki M, Ogawa M, Oyama A, Ashida H,
Kobayashi T, Koyama T, Nagai S, Shibata Y, Gohda J, Inoue J, Mizushima T,
Sasakawa C: The Shigella flexneri effector OspI deamidates UBC13 to dampen
the inflammatory response, Nature 2012, 483:623-U149
147. Flatau G, Lemichez E, Gauthier M, Chardin P, Paris S, Fiorentini C,
Boquet P: Toxin-induced activation of the G protein p21 Rho by deamidation of
glutamine, Nature 1997, 387:729-733
148. Deverman BE, Cook BL, Manson SR, Niederhoff RA, Langer EM, Rosova
I, Kulans LA, Fu XY, Weinberg JS, Heinecke JW, Roth KA, Weintraub SJ: Bcl-X-
L deamidation is a critical switch in the regulation of the response to DNA
damage, Cell 2002, 111:51-62
149. Bidinosti M, Ran I, Sanchez-Carbente MR, Martineau Y, Gingras AC,
Gkogkas C, Raught B, Bramham CR, Sossin WS, Costa-Mattioli M,
DesGroseillers L, Lacaille JC, Sonenberg N: Postnatal Deamidation of 4E-BP2 in
Brain Enhances Its Association with Raptor and Alters Kinetics of Excitatory
Synaptic Transmission, Molecular cell 2010, 37:797-808
150. Buetow L, Flatau G, Chiu K, Boquet P, Ghosh P: Structure of the Rho-
activating domain of Escherichia coli cytotoxic necrotizing factor 1, Nat Struct Biol
2001, 8:584-588
151. Wang HQ, Piatkov KI, Brower CS, Varshavsky A: Glutamine-Specific N-
Terminal Amidase, a Component of the N-End Rule Pathway, Molecular cell
2009, 34:686-695
152. Molberg O, Mcadam SN, Korner R, Quarsten H, Kristiansen C, Madsen L,
Fugger L, Scott H, Noren O, Roepstorff P, Lundin KEA, Sjostrom H, Sollid LM:
Tissue transglutaminase selectively modifies gliadin peptides that are recognized
by gut-derived T cells in celiac disease (vol 4, pg 713, 1998), Nat Med 1998,
4:974-974
153. Massiere F, Badet-Denisot MA: The mechanism of glutamine-dependent
amidotransferases, Cell Mol Life Sci 1998, 54:205-222
154. Nakashima A, Kawanishi I, Eguchi S, Yu EH, Eguchi S, Oshiro N, Yoshino
K, Kikkawa U, Yonezawa K: Association of CAD, a multifunctional protein
involved in pyrimidine synthesis, with mLST8, a component of the mTOR
complexes, J Biomed Sci 2013, 20:
155. Graves LM, Guy HI, Kozlowski P, Huang M, Lazarowski E, Pope RM,
Collins MA, Dahlstrand EN, Earp HS, Evans DR: Regulation of carbamoyl
phosphate synthetase by MAP kinase, Nature 2000, 403:328-332
156. Robitaille AM, Christen S, Shimobayashi M, Cornu M, Fava LL, Moes S,
Prescianotto-Baschong C, Sauer U, Jenoe P, Hall MN: Quantitative
113

Phosphoproteomics Reveal mTORC1 Activates de Novo Pyrimidine Synthesis,
Science 2013, 339:1320-1323
157. Ben-Sahra I, Howell JJ, Asara JM, Manning BD: Stimulation of de Novo
Pyrimidine Synthesis by Growth Signaling Through mTOR and S6K1, Science
2013, 339:1323-1328
158. Jacquemont B, Roizman B: Rna-Synthesis in Cells Infected with Herpes-
Simplex Virus .10. Properties of Viral Symmetric Transcripts and of Double-
Stranded-Rna Prepared from Them, Journal of virology 1975, 15:707-713
159. Lin R, Noyce RS, Collins SE, Everett RD, Mossman KL: The herpes
simplex virus ICP0 RING finger domain inhibits IRF3- and IRF7-mediated
activation of interferon-stimulated genes, Journal of virology 2004, 78:1675-1684
160. Chou J, Chen JJ, Gross M, Roizman B: Association of a M(R)-90,000
Phosphoprotein with Protein-Kinase Pkr in Cells Exhibiting Enhanced
Phosphorylation of Translation Initiation-Factor Eif-2-Alpha and Premature
Shutoff of Protein-Synthesis after Infection with Gamma(1)34.5(-) Mutants of
Herpes-Simplex-Virus-1, Proceedings of the National Academy of Sciences of
the United States of America 1995, 92:10516-10520
161. West JA, Wicks M, Gregory SM, Chugh P, Jacobs SR, Zhang ZG, Host
KM, Dittmer DP, Damania B: An Important Role for Mitochondrial Antiviral
Signaling Protein in the Kaposi's Sarcoma-Associated Herpesvirus Life Cycle,
Journal of virology 2014, 88:5778-5787
162. Rasmussen SB, Jensen SB, Nielsen C, Quartin E, Kato H, Chen ZJ,
Silverman RH, Akira S, Paludan SR: Herpes simplex virus infection is sensed by
both Toll-like receptors and retinoic acid-inducible gene-like receptors, which
synergize to induce type I interferon production, J Gen Virol 2009, 90:74-78
163. Guasparri I, Keller SA, Cesarman E: KSHV vFLIP is essential for the
survival of infected lymphoma cells, J Exp Med 2004, 199:993-1003
164. Liu XQ, Fitzgerald K, Kurt-Jones E, Finberg R, Knipe DM: Herpesvirus
tegument protein activates NF-kappa B signaling through the TRAF6 adaptor
protein, Proceedings of the National Academy of Sciences of the United States of
America 2008, 105:11335-11339
165. Richmond AL, Kabi A, Homer CR, Marina-Garcia N, Nickerson KP,
Nesvizhskii AI, Sreekumar A, Chinnaiyan AM, Nunez G, McDonald C: The
Nucleotide Synthesis Enzyme CAD Inhibits NOD2 Antibacterial Function in
Human Intestinal Epithelial Cells, Gastroenterology 2012, 142:1483-+
166. Bouwmeester T: A physical and functional map of the human TNF-alpha,
NF-kappa, B signal transduction pathway (vol 6, pg 97, 2004), Nat Cell Biol 2004,
6:

Abstract (if available)

Abstract Innate immunity is the first line of defense against foreign pathogens. In response to viral infection, the retinoic acid-inducible gene I (RIG-I) senses viral RNA and activates the interferon regulatory factor (IRF) and nuclear factor κ of B cell (NF-κB), leading to the production of type I interferons and pro-inflammatory cytokines. Protein deamidation is a poorly characterized post-translational modification. Emerging studies implicate protein deamidation in regulating fundamental biological processes. We recently reported that gamma-herpesviruses activate RIG-I via deamidation to evade antiviral cytokine production. The gamma herpesvirus homologues of glutamine amidotransferase (vGAT) interact with cellular phosphoribosylformylglycinamidine synthetase (PFAS) to deamidate RIG-I, indicating that deamidase-catalyzed protein deamidation plays essential roles in regulating innate immune signaling. In this thesis, we characterized the regulatory roles of deamidation of the RIG-I receptor and the NF-κB transcription factor RelA (also known as p65) that are mediated by herpes simplex virus 1 (HSV-1) UL37 and cellular CAD deamidase, respectively. ❧ We demonstrated that infection of HSV-1 induced RIG-I deamidation and prevented RIG-I activation by Sendai virus, a prototype RNA virus known to activate RIG-I. A focused screen identified the tegument protein UL37 as the major RIG-I-interacting protein and UL37 was sufficient to induce RIG-I deamidation and block RIG-I activation. Biochemical analyses identified two asparigines, N495 and N549, of the central helicase domain (Hel2i) that were deamidated by UL37 expression and HSV-1 infection. Deamidation significantly diminished the RNA-sensing activity, ATPase activity and antiviral activity of RIG-I. Conversely, uncoupling RIG-I deamidation via mutations that generated deamidation-resistant RIG-I or deamidation-deficient UL37 resulted in more robust antiviral immune response and reduced HSV-1 replication. Altogether, this study demonstrates that deamidation is a novel mechanism to evade RIG-I-mediated antiviral immunity by HSV-1, and suggests that UL37 may function as a protein deamidase. ❧ Cellular PFAS, when complexed with gamma herpesvirus vGAT, deamidates RIG-I in vitro, demonstrating that cellular metabolic glutamine amidotransferases (GATs) possess intrinsic protein-deamidating activity. Employing an unbiased screening targeting the GAT-containing enzymes, we discovered that CAD (carbamoyl-phosphate synthetase 2, aspartate transcarbamylase, and dihydroorotase) potently deamidated RelA and diminished NF-κB activation. Mass spectrometry analysis identified two asparagines within the insertion region of RelA that were deamidated by CAD. Biochemical analyses indicated that CAD is a bona fide deamidase of RelA. Interestingly, deamidated RelA demonstrated normal affinity for classic NF-κB-responsive element in vitro, while failed to induce pro-inflammatory responses in cells upon stimulation. This work collectively defined CAD as a novel deamidase that deamidates RelA to dampen NF-κB activation. ❧ In summary, I have discovered a cellular GAT and a potential viral deamidase that regulate innate immune signaling via deamidating key components. This study also defines two novel protein deamidations in mammalian cells and their fundamental roles in host innate defense and viral immune evasion.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Characterization of three novel variants of the MAVS adaptor

PDF

CAD-mediated metabolic reprogramming is regulated by innate immune kinases TBK1 and IKKβ

PDF

Selective innate immune activation by murine gamma-herpesvirus 68 (ɣHV68)

PDF

Protein deamidation mediated metabolic reprogramming during KSHV lytic replication

PDF

The effects of hepatitis C virus infection on host immune response and signaling pathways

PDF

Paneth cell α-defensins in mouse enteric innate immunity

PDF

Molecular mechanism for the immune evasion of CD1d antigen presentation by herpes simplex virus-1 UL56 protein

Asset Metadata

Creator Zhao, Jun (author)

Core Title Study of protein deamidation in innate immune signaling

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 10/29/2017

Defense Date 11/24/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag CAD,deamidase,deamidation,HSV-1,innate immune signaling,OAI-PMH Harvest,RelA,RIG-I,UL37

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Gao, Shou-Jiang (committee chair), Feng, Pinghui (committee member), Jung, Jae U. (committee member), Zandi, Ebrahim (committee member)

Creator Email junz@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-246999

Unique identifier UC11278960

Identifier etd-ZhaoJun-4309.pdf (filename),usctheses-c40-246999 (legacy record id)

Legacy Identifier etd-ZhaoJun-4309.pdf

Dmrecord 246999

Document Type Dissertation

Format application/pdf (imt)

Rights Zhao, Jun

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA