Serine Cluster Phosphorylation Liberates the
C-terminal Helix of IFN Regulatory Factor 7
To Bind Histone Acetyltransferase p300
This information is current as
of June 15, 2017.
Kyoung Jin Lee, Jung Sook Ye, Han Choe, Young Ran
Nam, Nari Kim, Uk Lee and Chul Hyun Joo
J Immunol published online 15 September 2014
http://www.jimmunol.org/content/early/2014/09/13/jimmun
ol.1401290
Permissions
Email Alerts
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2014 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Subscription
Published September 15, 2014, doi:10.4049/jimmunol.1401290
The Journal of Immunology
Serine Cluster Phosphorylation Liberates the C-terminal
Helix of IFN Regulatory Factor 7 To Bind Histone
Acetyltransferase p300
Kyoung Jin Lee,* Jung Sook Ye,* Han Choe,†,‡ Young Ran Nam,* Nari Kim,*
Uk Lee,* and Chul Hyun Joo*,x
I
nterferon regulatory factor 7 (IRF7) is a major transcription
factor that controls expression of type I (a/b) IFN, a cytokine
indispensable for control of viral infection (1). In addition,
a large body of literature describes the involvement of IRF7 in
multiple diseases other than infections. IRF7 was identified as
a protein related with EBV latency (2). IRF7 is abnormally regulated in multiple cancers including hepatocellular carcinoma (3),
astrocytoma (4), lung cancer (5), gastric carcinoma (6). esophageal cancer (7), and breast cancer (8). IRF7 has also been implicated in type 1 diabetes mellitus (9, 10) and systemic lupus
*Department of Microbiology, University of Ulsan College of Medicine, Seoul
138-736, Korea; †Bio-Medical Institute of Technology, University of Ulsan College
of Medicine, Seoul 138-736, Korea; ‡Department of Physiology, University of Ulsan
College of Medicine, Seoul 138-736, Korea; and xCell Dysfunction Research Center,
University of Ulsan College of Medicine, Seoul 138-736, Korea
Received for publication May 19, 2014. Accepted for publication August 18, 2014.
This work was supported by the National Research Foundation of Korea, a Medical
Research Council Grant funded by the Korean government (MSIP 2008-0062286),
and the Asan Institute for Life Sciences (2011-346).
Address correspondence and reprint requests to Prof. Chul Hyun Joo, Department of
Microbiology, University of Ulsan College of Medicine, Asanbyeongwon-gil 86,
Songpa-gu, Seoul 138-736, Korea. E-mail address: [email protected]
Abbreviations used in this article: AD, activation domain; CAD, constitutive activation domain; DA, DNA-binding domain + activation domain; DAR, DNA-binding
domain + activation domain + regulatory domain; DR, DNA-binding domain +
regulatory domain; DBD, DNA-binding domain; H1-4, helix 1-4 in highly conserved
C-terminal region; HCCR, highly conserved C-terminal region; hIRF7, human IFN
regulatory factor 7 isoform; ID, inhibitory domain; IKK, IkB kinase; IRF, IFN
regulatory factor; IRF7 SSAA, IFN regulatory factor S477A/S479A mutant; IRF7
SSDD, IFN regulatory factor S477D/S479D mutant; IRF7wt, IFN regulatory factor 7
wild-type; IRFE, IFN regulatory factor element; ISG, IFN-stimulated gene; MAVS,
mitochondrial antiviral signaling protein; MSA, multiple sequence alignment; NES,
nuclear export signal; NLS, nuclear localization signal; NP-40, Nonidet P-40; PBST,
PBS with 0.1% Tween-20; PRD, positive regulatory domain; PRR, pathogen recognition receptor; PSA, pairwise sequence alignment; RD, regulatory domain; SC,
serine cluster; SRR, serine cluster regulatory region; TBK, TANK binding kinase;
TRIF, Toll/IL-1R domain-containing adapter-inducing IFN-b; VAD, virus activation
domain.
Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.1401290
erythematosus (11–13). However, the detailed transactivation
mechanism of IRF7 remains elusive.
The type I IFN family mainly comprises 13 IFN-a isoforms and
a single IFN-b isoform (14). Because type I IFN limits viral
replication by interfering with cellular homeostasis, its expression
and secretion must be tightly regulated. All cells can secrete
IFN-b immediately in response to viral infections, whereas IFN-a
secretion patterns differ according to cell types. Some cells constitutively express IRF7 in a latent form, allowing them to secrets
IFN-a immediately in response to viral detection. In particular,
plasmacytoid dendritic cell is the major source of IFN-a in the
systemic circulation (15). Other types of cells must first induce
IRF7 expression, resulting in delayed secretion of IFN-a. Although the absolute amount of IFN-a secreted from these latter
cells is much smaller than the amount produced by plasmacytoid
dendritic cell, it is, nonetheless, indispensable for the local primary innate immune response to viral infection.
Delayed IFN-a secretion starts with the recognition of pathogenassociated molecular patterns by pathogen recognition receptors
(PRRs). Viral genomes are recognized as pathogen-associated
molecular patterns by PRRs, extracellular TLRs, or intracellular
retinoic acid-inducible gene I–like helicase receptors (16, 17).
These recognition events trigger signaling cascades mediated by
various adaptor molecules, for which the identities depend on
the type of PRRs involved; these adaptors include mitochondrial
antiviral signaling protein (MAVS), Toll/IL-1R domain-containing
adapter-inducing IFN-b (TRIF), and MyD88 (18). The adaptors
induce formation of signaling complexes that activate IRF3,
and the retinoic acid-inducible gene I–like helicase receptor–
MAVS–IRF3 axis induces primary IFN-b secretion in most cell
types. Activated IRF3 binds to the IFNB1 promoter along with
AP-1 and NF-kB (19). Secreted IFN-b binds with type I IFNs
receptors on the cellular membrane, where it acts in an autocrine
and paracrine manner. Binding of IFN-b to the receptors triggers
JAK-STAT signaling, leading to the formation of IFN-stimulated
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
IFN regulatory factor 7 (IRF7) is a major regulator of type I (ab) IFN secretion. A growing body of evidence shows that IRF7 is
involved in a wide variety of pathologic conditions in addition to infections; however, the detailed mechanism of IRF7 transactivation remains elusive. Our current knowledge of IRF7 transactivation is based on studies of IRF3, another major regulator of
IFN-b secretion. IRF3 and IRF7 are closely related homologs with high sequence similarity in their C-terminal regions, and both
proteins are activated by phosphorylation of a specific serine cluster (SC). Nevertheless, the functional domains of the two proteins
are arranged in an inverted manner. We generated a model structure of the IRF7 C-terminal region using homology modeling and
used it to guide subsequent functional domain studies. The model structure led to the identification of a tripod-helix structure
containing the SC. Based on the model and experimental data, we hypothesized that phosphorylation-mediated IRF7 transactivation is controlled by a tripod-helix structure. Inducible IkB kinase binds a tripod-helix structure. Serial phosphorylation
of the SC by the kinase liberates C-terminal helix from an inhibitory hydrophobic pocket. A histone acetyltransferase P300 binds
the liberated helix. The difference in the P300 binding sites explains why the domain arrangement of IRF7 is inverted relative to
that of IRF3. The Journal of Immunology, 2014, 193: 000–000.
2
TRANSACTIVATION OF IRF7 BY C-TERMINAL TRIPOD-HELIX STRUCTURE
Materials and Methods
Bioinformatics
The amino acid sequences of IRF7 used for multiple sequence alignment
(MSA) were downloaded from GenBank. Generic names and reference
numbers are as follows: Homo sapiens (human; NP_001563.2), Mus musculus (house mouse; NP_001239529.1), Rattus norvegicus (Norway rat;
NP_001028863.1), Sus scrofa (pig; NP_001090897.1), Macaca mulatta
(rhesus monkey; NP_001129572.1), Mustela putorius furo (domestic ferret;
NP_001129572.1), Felis catus (domestic cat; XP_006937768.1), Mesocricetus auratus (golden hamster; XP_005063402.1), Gorilla gorilla
(Western gorilla; XP_004050414.1), Papio anubis (olive baboon;
XP_003909373.1), Pana paniscus (pygmy chimpanzee; XP_003806037.1),
Otolemur garnettii (small-eared galago; XP_003802849.1), Cricetulus
griseus (Chinese hamster; XP_003509823.1), Pongo abelii (Sumatran
orangutan; XP_002821370.1), Chinchilla lanigera (long-tailed chinchilla;
XP_005401881.1), Jaculus jaculus (lesser Egyptian jerboa; XP_004653953.1),
Odobenus rosmarus divergens (Pacific walrus; XP_004403823.1), Saimiri
boliviensis (Bolivian squirrel monkey; XP_003943408.1), Callithrix jacchus
(white-tufted-ear marmoset; XP_002755734.1), Ceratotherium simum simum
(Southern white rhinoceros; XP_004441106.1), Bos taurus (cattle;
NP_001098510.1), Tursiops truncatus (bottlenosed dolphin; XP_004316939.1),
Orcinus orca (killer whale; XP_004278200.1), Cavia porcellus (domestic
guinea pig; XP_004999585.1), Echinops telfairi (small Madagascar hedgehog; XP_004717251.1), Octodon degus(degu; XP_004638094.1), Sorex
araneus (European shrew; XP_004603181.1), Ochotona princeps (American pika; XP_004599294.1), Dasypus novemcinctus (nine-banded armadillo; XP_004459829.1), Trichechus manatus latirostris (Florida manatee;
XP_004389466.1), and Ovis aries (sheep; XP_004019786.1). PSA with
IRF3 (NP_001562.1) was performed using LALIGN version 2.2u (32) with
following parameters; scoring matrix (BLOSUM50), gap open/extension
penalty (214/24). MSA was performed with Clustal V version 1.2.1 (33)
with the following parameters: mBed-like clustering guide tree and iteration,
five iterations, five maximum guide tree iterations, five maximum HMM
iterations. Secondary structure was predicted using Jpred3 version 2.2 (34)
with default parameters. A homology model for IRF7 was built based on the
crystal structures of IRF3 (Protein Data Bank identification 1QWT) using
the homology modeling program MODELER version 9.13 (35). Structural
analyses and figure preparations were performed using PyMol (The PyMOL
Molecular Graphics System, Version 1.5.0.4; Schrödinger).
Cloning
The cloning vectors pCMnGFP, pCMnFlag, pCMnGST, and pCMBB were
constructed for an efficient subcloning by using common multiple cloning
sites (36). The expressing plasmids of IRF7, TBK1, IKKε, and MAVS
were kindly provided by John Hiscott (McGill University, Montreal, QC,
Canada), Tom Maniatis (Columbia University, New York, NY), and
Michaela Gack (Harvard University, Cambridge, MA). The expressing
plasmid of TRIF was acquired from Addgene (plasmid #13644; Cambridge, MA). Luciferase promoter assay construct pGL3-IFNA4 was
described elsewhere (37). The sequence used for primer design was acquired from GenBank (NM_001572, H. sapiens IRF7 transcript variant a).
The following primers were synthesized for the constructs cloning:
DBD_S_NheI (59-CCC GCT AGC GCC TTG GCT CCT GAG AGG
GCA-39), DBD_AS_XhoI (59-CCC CTC GAG CAG CTC CCG GCT
GAG CGC GT-39), AD_S_XhoI (59-CCC CTC GAG TGC TGG CGA
GAA GGC CCA-39), AD_AS_XbaI (59-CCC TCT AGA GGC CTC GCC
TGT CGT TAG T-39), ID_S_XbaI (59-CCC TCT AGA GCG GCC CCA
GAG TC-39), ID_AS_KpnI (59-CCC GGT ACC ACG CTG CGT GCC
CTC TAG GT-39), RD_S_KpnI (59-CCC GGT ACC GAG GGT GTG TCT
TCC CT-39), RD_AS_HindIII (59-CCC AAG CTT GGC GGG CTG CTC
CAG CT-39), SC regulatory region 1 (SRR1)_S_KpnI (59-CCC GGT ACC
TGT GAC ACC CCC ATC TT-39), SRR2_S_KpnI (59-CCC GGT ACC
CGG CAG CGC CGT GGC TCC-39), SRR3_S_KpnI (59-CCC GGT ACC
CTG GAA CCC TGG CTG TGC-39), SRR4_S_KpnI (59-CCC GGT ACC
GAG GGT GTG TCT TCC CT-39), SRR5_S_KpnI (59-CCC GGT ACC
CTC TAT GAC GAC ATC GAG TGC-39), DelH2_S_SpeI (59-CCC ACT
AGT CGG CAG CGC CGT GGC TC-39), DelH2_AS_SpeI (59-CCC ACT
AGT GTC GAA GAT GGG GGT GTC-39), DelH3_S_SpeI (59-CCC ACT
AGT GAG GGT GTG TCT TCC CT-39), DelH3_AS_SpeI (59-CCC ACT
AGT CTT CAC CAG GAC CAG GCT CTT-39), H2_S_XhoI (59-CCC
CTC GAG CCC ATC TTC GAC TTC AGA GTC-39), H2_AS_HindIII (59CCC AAG CTT GCG TGG GGA GCC ACG GCG CT-39), H3_S_NheI
(59-CCC GCT AGC CCC AAG GAG AAG AGC CT-39), H3_AS_HindIII
(59-CCC AAG CTT GCA GAG GCT GAG GCT GC-39), H4_S_NheI (59CCC GCT AGC GTG TCT TCC CTG GAT AG-39), and H4_AS_HindIII
(59-CCC AAG CTT GGC GGG CTG CTC CAG CT-39). Domain fragment
was amplified from pcDNA5/IRF7 wild-type (wt) using the High Fidelity
PCR kit (Stratagene Agilent Technologies, Santa Clara, CA). Amplified
products were purified using a PCR purification kit (Qiagen, Hilden, Germany) and then incubated with indicated restriction enzymes for 1 h.
Digested product was gel-purified using the Gel Extraction Kit (Qiagen). The
pCMBB vector was also digested with corresponding restriction enzymes.
The purified vector and amplified domain fragment were religated using the
Quick Ligation Kit (New England Biolabs, Beverly, MA). Transformation
was performed using the standard heat-shock method. For the helix deletion,
a linear deletion fragment was amplified from pcDNA5/IRF7wt. Amplified
products were purified, digested, and religated as described above. All
constructs were confirmed by restriction enzyme cutting and sequencing.
Cell culture, promoter assay, ELISA, and fluorescence
microscopy
293T cells were grown in DMEM (Life Technologies-BRL, Grand Island, NY)
with 10% FCS and 1% penicillin-streptomycin (Life Technologies-BRL).
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
gene (ISG) factor 3 (20), which activates expression of .300
genes categorized as ISGs (21). IRF7 is an ISG expressed under
the control of ISG factor 3. If a spreading virus is recognized by an
IFN-b–primed cell, induced IRF7 activates immediate secretion of
IFN-a. Via this mechanism of cooperation among neighboring
cells, type I IFN secretion is not only regulated tightly at the initial
phase of infection, but is also amplified robustly at later phases in
proportion to the extent of viral spreading (22).
Among nine mammalian IRFs identified to date, IRF3 and IRF7
are the major transcription factors for type I IFNs (23). IRF proteins
contain a common pentad tryptophan repeat in their N-terminal
DNA-binding domain (DBD) (24). The DBD binds to a motif
called the IFN regulatory factor element (IRFE). Each DBD has
a different binding preference for IRFEs: the lRF3 DBD binds
only to the IFNB1 promoter (22), whereas the IRF7 DBD can bind
to IFNB1 and multiple IFNA promoters (25). After DBD, other
domains are followed. IRF3 and IRF7 both contain an activation
domain (AD), an inhibitory domain (ID), and a regulatory domain
(RD). IRF3 has the arrangement DBD-ID-AD-RD, whereas IRF7
has the inverted arrangement, DBD-AD-ID-RD (Fig. 1C). In
pairwise sequence alignment (PSA), IRF3 ID and IRF7 AD exhibit multiple deletions/insertions with low overall sequence
similarity; however, the IRF3 AD-RD and IRF7 ID-RD are highly
similar. Both IRF3 and IRF7 are activated by phosphorylation of
a unique serine cluster (SC) in the RD; this modification is mediated by TANK binding kinase 1 (TBK1) or inducible IkB kinase
(IKKε) (26, 27). Despite their high similarities at sequence level
and their common mode of activation by SC phosphorylation, the
two proteins are functionally quite different: deletion of IRF7 ID
results in a constitutively active mutant, whereas deletion of corresponding IRF3 AD results in a dominant-negative mutant (28).
Likewise, the RDs of the two proteins play opposite roles: deletion
of the IRF3 RD results in a constitutively active mutant, whereas
deletion of the IRF7 RD results in a dominant-negative mutant. As
a result, IRF3 contains a single AD (aa 134–394) flanked by two
autoinhibitory function domains (ID and RD) (29), whereas IRF7
contains two transactivation domains (AD and RD) separated by
a single ID (aa 247–467) (30).
In the case of IRF3, SC phosphorylation induces the release of
the RD, which covers the binding site for histone acetyltransferase
CBP/P300; binding of CBP/P300 masks a nuclear export signal
(NES), resulting in nuclear translocation of IRF3 (31). Although IRF7
was revealed as a major regulator of IFN-a about a decade ago, the
detailed transactivation mechanism remains elusive, primarily due to
the lack of structural information. In this study, we elucidated the
mechanism of phosphorylation-mediated transactivation using a
model structure, enabling us to answer a longstanding question:
how can IRF3 and IRF7 use a common activation mechanism
despite the inverted arrangements of their functional domains?
The Journal of Immunology
For transient expression assays, plasmid DNA was transfected into
293T cells using Transfectin (Bio-Rad, Richmond, CA) or calcium phosphate (BD Biosciences Clontech, Palo Alto, CA).
293T cells were seeded onto 24-well plates at a density of 105 cells/well 24 h
prior to transfection. Cells were transfected with a Firefly luciferase reporter
plasmid (pGL3-IFNA4 promoter) and a control Renilla luciferase plasmid
(pRL-SV40; Promega, Madison, WI). At 24 h posttransfection, luciferase
activities were measured with a luminometer (Victor 34; PerkinElmer, Waltham, MA) using the Dual-Luciferase Assay Kit (Promega). Firefly luciferase
activity was normalized to the corresponding Renilla luciferase activity, and
relative luciferase units were calculated by dividing the normalized value
by the negative-control normalized value. ELISA was performed using the
VeriKine human IFN-a ELISA kit (PBL InterferonSource, Piscataway, NJ).
293T cells were seeded onto six-well plates at a density of 105 cells/well
24 h before transfection and then transfected with GFP-tagged constructs.
After 24 h, cells were observed using a DM IRB fluorescence microscope
(Leica Microsystems, Tokyo, Japan).
In vivo GST pulldown, immunoprecipitation, and
immunoblotting
EMSA
Cells were washed with PBS and lysed in lysis buffer (10 mM Tris-Cl [pH
8], 60 mM KCl, 1 mM EDTA, 1 mM DTT, 0.5% NP-40, and protease
inhibitor mixture [Roche]) 2 d after transfection. Equivalent amounts of
nuclear extracts were assayed for IRF7 binding by gel-shift analysis using
a biotin-labeled double-stranded oligonucleotide corresponding to the
positive regulatory domain (PRD) III-PRDI region of the IFNB1 promoter
(59-GAA AAC TGA AAG GGA GAA GTG AAA GTG-39). The binding
mixture contained 10 mM Tris-HCl (pH 7.5), 1 mM EDTA, 50 mM NaCl,
2 mM DTT, 5% glycerol, and 0.5% NP-40; 10 mg/ml BSA and 62.5 mg/ml
poly(deoxyinosinic-deoxycytidylic) acid were added to reduce nonspecific
binding. The total volume of the reaction mixture containing whole-cell
extract was 20 ml. After 20 min of incubation with the probe, extracts were
loaded on a DNA Retardation Gel (Invitrogen Life Technologies) prepared
in 0.53 Tris-borate-EDTA. After 1 h running at 100 V, the gel was
transferred to a nylon membrane (Pierce). The membrane was cross-linked
with UV (GS Gene Linker; Bio-Rad), and biotin signals were visualized
using the Chemiluminescent Nucleic Acid Detection Kit (Pierce). To
demonstrate the specificity of protein–DNA complex formation, an Ab
against IRF7 or a 1000-fold molar excess of unlabeled oligonucleotides
was added to the binding mixture before addition of the labeled probe.
Results
Structural model of IRF7 based on bioinformatics analyses
We performed bioinformatics analyses to generate domain constructs for use in subsequent studies. Because functionally critical
residues are often conserved among species, we performed MSA of
IRF7 from 31 mammalian species to identify such residues (Fig. 1C,
conservation frequency). The DBD, the C-terminal part of the ID,
and the RD were highly conserved among species; however, very
little conservation was observed in the AD and the N-terminal
portion of the ID. The MSA underscored the functional importance of the C-terminal portion of the ID and the RD (aa 332–503);
we termed this region the highly conserved C-terminal region
(HCCR). Next, we predicted the secondary structure of IRF7
(Fig. 1C, 2nd structure). Regions with abundant a-helices or
b-sheets correlated with regions of conservation in the MSA; in
particular, the HCCR was predicted to contain highly ordered secondary structures. By contrast, no secondary structure was predicted
between the DBD and HCCR. A comparison of 31 mammalian
HCCRs revealed that the predicted secondary structure in this region is conserved across species. A prominent feature of the HCCR
is the presence of four helices, which we numbered H1 to H4.
To further dissect the HCCR, structure information was necessary;
however, no crystal structure of this region is currently available.
As an alternative, we modeled the HCCR structure by computer
simulation (Fig. 1A). The model, based on the previously reported
IRF3 crystal structure (Research Collaboratory for Structural Bioinformatics Protein Data Bank identification number 1QWT), was
constructed by a homology modeling. The template IRF3 structure
covered the C-terminal region from aa 175–427 (Fig. 1C, IRF3). In
PSA, the HCCRs of IRF3 and IRF7 were sufficiently similar to
allow generation of a reliable structural model of IRF7 HCCR from
the IRF3 crystal structure. The model contained four distinct
a-helices in the corresponding positions revealed in the secondary
structure prediction (Fig. 1A). Notably, H2, H3, and H4 formed
a protruding tripod structure, with the SC positioned between H3
and H4. The RD contained the SC and H4 and sterically occluded
a central hydrophobic core generated by H2 and H3. Based on these
analyses, we generated a set of domain constructs and used them to
explore the transactivation mechanism of IRF7 (Fig. 1B).
The AD and RD are required for transcriptional activation
To investigate the functional roles of the AD and RD, we generated
constructs consisting of various combinations of the DBD, AD, and
RD, as follows: DBD alone; DBD + AD (DA) containing DBD and
AD; DBD + RD (DR) containing DBD and RD; and DBD + AD +
RD (DAR) containing all three domains (Fig. 1B, set a). We
performed IFN-a 4 (IFNA4) promoter assays with increasing
amounts of the constructs (Fig. 2A). All constructs stimulated the
promoter in a concentration-dependent manner. However, the
amplitudes of stimulation were different, the DBD construct barely
stimulated the promoter, whereas DA, DR, and DAR induced
significant stimulations, and transcriptional activation by DAR was
10-fold stronger than activation by DA or DR. These data suggest
that the AD and RD contribute synergistically to activation.
To investigate the influence of upstream signals, we coexpressed
the constructs with various upstream signaling intermediaries
(IKKε, TBK1, MAVS, and TRIF) (Fig. 2B). In cells expressing
IRF7wt, all upstream signals activated the IFNA4 promoter. By
contrast, the DBD, DA, DR, and DAR constructs did not exhibit
responsive stimulations, although once again DAR stimulated the
promoter most strongly regardless of the presence of upstream
signals. Therefore, the ID not only inhibits the AD and RD when
the protein is in a latent status, but also processes upstream signals
involved in transactivation.
Next, to determine the role of domains in activation of endogenous
IFNA genes, we used ELISA to measure secretion of IFN-a in cells
expressing the domain constructs (Fig. 2C). Although DA, DR, and
DAR all stimulated the IFNA4 promoter (Fig. 2B), only DAR induced
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Transfected cells were harvested and resuspended in lysis buffer (50 mM
Tris [pH 8], 150 mM NaCl, 0.5% Nonidet P-40 [NP-40], and protease
inhibitor mixture [Roche Applied Science, Indianapolis, IN]), followed by
incubation at 4˚C for 10 min and centrifugation at 12,000 rpm for 5 min.
The supernatant fractions were incubated with protein A/G-agarose beads
(Pierce, Rockford, IL) for 1 h. Glutathione Sepharose 4B (Amersham,
Piscataway, NJ) was then added to the precleared supernatants and incubated at 4˚C for 3 h. In the case of the immunoprecipitation, the precleared
supernatants were incubated with a specific Ab for 2 h, followed by the
addition of protein A/G agarose beads and incubation for 1 h. The beads
were then washed three times vigorously with washing buffer (50 mM Tris
[pH 8], 300 mM NaCl, 0.5% NP-40, and protease inhibitor mixture
[Roche]). Beads were washed vigorously three times with washing buffer
(50 mM Tris [pH 8], 300 mM NaCl, 0.5% NP-40, and protease inhibitor
mixture). Bead complexes were resuspended and boiled in SDS sample
buffer (Sigma-Aldrich, St. Louis, MO) at 96˚C for 5 min.
Samples were separated by SDS-PAGE and then transferred to a polyvinylidene difluoride membrane (Roche) by semidry transfer (Bio-Rad). The
membrane was blocked for 30 min in PBS with 0.1% Tween-20 (PBST)
containing 5% nonfat milk and then incubated with primary Ab for 1 h. After
three washes with PBST, the membrane was incubated with HRP-conjugated
secondary Ab (Cell Signaling Technology, Beverly, MA) for 1 h and then
washed an additional three times with PBST. Specific signals were visualized
using SuperSignal West Pico Substrate (Pierce) on an ECL machine (LAS-4000;
GE Healthcare, Piscataway, NJ). Primary Abs were purchased from the
following sources: FLAG (M2) and GST Abs from Sigma-Aldrich; IRF7 (F-1),
P300 (NM11), and GFP (T-19) Abs from Santa Cruz Biotechnology (Dallas,
TX); and anti–phospho-IRF7 (Ser471/472) from Cell Signaling Technology.
3
4
TRANSACTIVATION OF IRF7 BY C-TERMINAL TRIPOD-HELIX STRUCTURE
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 1. Bioinformatic analyses of IRF7 domains. (A) Model structure of the highly conserved C-terminal region. Top panel, Cartoon display of
secondary structure. Bottom panel, Positions of four helices based on secondary-structure predictions. (B) Construction sets for domain function studies:
combination of domains (a), separate domains (b), H2 and H3 deletion mutants (c), and helix serial truncations (d). (C) Sequence analyses. Conservation
among mammals: conservation frequency among 31 mammalian sequences at corresponding positions in multiple sequence alignment. 2nd structure:
representative predicted secondary structures; open square, b-sheet; filled square, a-helix. Domains: four major functional domains. IRF3: alignment of
four major functional domains of IRF3. Dotted lines connect corresponding positions in pairwise sequences alignment of IRF3 and IRF7. E, extended
b-sheet.
The Journal of Immunology
5
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 2. Involvement of AD and RD in IRF7 activity. (A) Stimulation of IFNA4 promoter with increasing amounts of construct from set a (Fig. 1B).
293T cells were cotransfected with increasing amounts of constructs (1–32 ng/105 cells with 2-fold increment). Western blots shows protein levels of
b-actin and constructs in lysates of cells used for the promoter assay. Statistical significance was analyzed using one-way ANOVA and Dunnett’s multiple
comparison test. (B) Stimulation by coexpression of upstream signaling genes. 293T cells were transfected with each construct and the indicated upstream
signal protein coding genes. Inset table: fold increment in stimulation (reflecting the data shown in the graph). (A and B) Data represent means 6 SD of
three independent experiments (n = 3). IFN-a ELISA of culture supernatants from 293T cells transfected with the constructs (C) and increasing amounts
of DAR [(D) 4–1000 ng/105 cells with 3-fold increment]. Data represent means 6 SD of three independent experiments (n = 3). *p , 0.05, **p , 0.01,
***p , 0.005, ****p , 0.0001. RLU, relative luciferase unit.
detectable IFN-a secretion, suggesting that both the AD and RD
are required to activate transcription of endogenous IFNA genes.
We next measured IFN-a secretions in cells expressing increasing
amounts of DAR to evaluate the characteristics of transcriptional
activation (Fig. 2D). Secretion was a step function of DAR concentration (i.e., IFN-a was fully off below a threshold concentration [∼100 ng], but essentially fully on above that threshold).
The DBD contains a nuclear localization signal
We compared the DNA-binding abilities of the domain constructs
by EMSA, using a PRDIII-I probe derived from the IRFE in the
IFNB1 promoter (Fig. 3A). The probe exhibited no difference in
shift intensity when incubated with nuclear extracts from 293T
cells overexpressing DBD, DA, DR, or DAR. The IRFE-binding
capacity of the DBD was not enhanced by either AD or RD.
Nuclear translocation is an important mechanism for control of
transcription factor activity. Therefore, we investigated the nuclear
localization of GFP-tagged DBD, DA, DR, and DAR (Fig. 3B). In
the case of DA, DR, and DAR, the majority of GFP signals were
observed in the nucleus with minor cytosolic distribution. DBD
was clearly localized in the nucleus, indicating that nuclear
localization signal (NLS) resides in the DBD. Next, we assessed
6
TRANSACTIVATION OF IRF7 BY C-TERMINAL TRIPOD-HELIX STRUCTURE
the localization of individual domains (Fig. 3C) using GFPtagged DBD, AD, ID, and RD (Fig. 1B, set b). The DBD was
localized in the nucleus; the other domains were evenly distributed throughout the cell. We identified a putative nonclassical bipartite NLS in the DBD, containing abundant arginines,
between aa 67 and 104 (Fig. 3D, bottom panel). To verify that
this sequence functions as an NLS, we generated a point mutant,
GFP-IRF7 K92R, and monitored its nuclear translocation in
response to the upstream signaling factors IKKε, TBK1, MAVS,
and TRIF (Fig. 3D). IRF7wt translocated into the nucleus upon
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 3. NLS in DBD. (A)
EMSA of construct set a (Fig. 1B)
using IFNB PRD III-I probe.
293T cells were transfected with the
indicated constructs. At 48 h posttransfection, nuclear extracts were
subjected to EMSA (right panel).
The same samples were subjected to
immunoblots (IB) with anti-IRF7
and anti–b-actin Abs to determine
the levels of the indicated proteins
(left panel). (B) GFP-tagged constructs (set a) were transfected into
293T cells. (C) GFP-tagged constructs (set b) were transfected into
293T cells. (D) Translocation of
IRF7 K92R mutant in response to
stimulation. GFP-tagged IRF7wt
(GFP-IRF7wt) and K92R mutant
(GFP-IRF7 K92R) were transfected
into 293T cells along with plasmids
expressing inducible IKKε, TBK1,
MAVS, or TRI. Bottom panel, Putative NLS. Red letters represent
hydrophobic residues, and boldface K is the residue mutated to
arginine in the K92R mutant. (B–
D) Localizations were observed
under a fluorescent microscope at
24 h posttransfection. Scale bars,
10 mm. C, cold probe competition;
S, shift; S*, supershift using antiIRF7 Ab.
stimulations, whereas IRF7 K92R did not enter the nucleus,
irrespective of upstream signaling.
A tripod-helix structure plays a critical role in transactivation
In the HCCR structural model, H2, H3, and H4 formed a tripod
complex with intercalating SC (Fig. 1A). We generated H2 and H3
deletion mutants (Fig. 1B, set c) to determine importance of these
helices for the response to stimulation using the IFNA4 promoter
assay (Fig. 4A). H2 or H3 deletion completely abolished activation
by IKKε, TBK1, MAVS, or TRIF. This observation indicates that
The Journal of Immunology
7
the tripod structure with a hydrophobic core is critical for the
transactivation by upstream signals.
To test the role of each helix in transactivation, we generated
a series of constructs consisting of the N-terminal DA fused with
a series of truncated fragments of the C terminus, SRR1 to SRR5
(Fig. 1B, set d). SRR1 contained H2, H3, the SC, and H4; SRR3
contained H3, the SC, and H4; SRR4 contained the SC and H4;
and SRR5 only contained H4. We tested the activities of these
constructs, with or without coexpression of IKKε, in the IFNA4
promoter assay (Fig. 4B). DA-SRR1, DA-SRR2, and DA-SRR3
did not stimulate the promoter, whereas DA-SRR4 and DA-SRR5
both exhibited strong stimulation. Coexpression of IKKε had no
influence over the activities. These data indicate that H3 is sufficient to inhibit H4 transcriptional activity and intact HCCR is
required for proper activation by upstream stimulation. We used
ELISA to measure IFN-a secretion in cells expressing the SRR
constructs (Fig. 4C). Both DA-SRR4 and DA-SRR5 induced
IFN-a secretion; however, DA-SRR5 induced less secretion than
DA-SRR4, even though these constructs activated the IFNA4 promoter to similar extents (Fig. 4B). Therefore, both the SC and H4
are required for efficient activation of endogenous IFNA genes.
We then determined the localization of GFP-tagged SRR constructs (Fig. 4D), in which the N-terminal DA segment was
replaced with GFP to rule out interference from the NLS. GFPSRR1 and GFP-SRR2 exhibited cytosolic localizations, whereas
GFP-SRR3, GFP-SRR4, and GFP-SRR5 were diffused throughout
the cell. These data indicate that the NES resides between H2 and
H3. As noted above, despite its presence in the nucleus, DA-SRR3
neither stimulated the promoter nor induced IFN-a secretion,
suggesting that H3 is sufficient to inhibit H4 activity. Interestingly,
GFP-SRR1 localized in the cytosol whereas GFP-ID diffused
throughout the cells (Fig. 3C), despite both constructs contained
a putative NES region. We checked the nuclear translocation of
GFP-SRR1 and GFP-tagged ID and RD (GFP-IRD) in response to
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 4. A tripod-helix structure plays a critical role in activation. (A and B) IFNA4 promoter assays. IRF7wt, H2 deletion mutant (IRF7 DH2), or H3
deletion mutant (IRF7 DΗ3) was transfected with plasmid expressing inducible IKKε, TBK1, MAVS, or TRIF (A). Stimulation of the IFNA4 promoter by
constructs set d (Fig. 1B, DA/SRR1 to DA/SRR5). Each construct was transfected into 293T cells with or without plasmid expressing IKKε (B). Data are
presented as the means 6 SD of three independent experiments (n = 3). Bottom panel, Immunoblots (IB) showing protein levels of b-actin and constructs in
lysates of cells used for the promoter assay. (C) IFN-a ELISA of culture supernatants from 293T cells transfected with the indicated constructs (set d). Data
are presented as the means 6 SD of three independent experiments (n = 3). (D) GFP-tagged constructs (Fig. 1B, set d) were transfected into 293T cells. (E)
GFP-tagged inhibitory and regulatory domain (IRD) and GFP-SRR1 were transfected into 293T cells along with plasmid expressing IKKε, TBK1, MAVS,
or TRIF as indicated. (D and E) Localizations were observed under a fluorescent microscope at 24 h posttransfection. Scale bars, 10 mm. RLU, relative
luciferase unit.
8
TRANSACTIVATION OF IRF7 BY C-TERMINAL TRIPOD-HELIX STRUCTURE
stimulation (Fig. 4E). When coexpressed with IKKε, TBK1,
MAVS, or TRIF, GFP-IRD diffused into the nucleus. However, the
cytosolic localization of GFP-SRR1 did not change in response to
stimulation. With the promoter assay (Fig. 4B), these data indicate
that intact HCCR (ID and RD) are required for nuclear translocation in response to upstream signals.
H4 binds P300
Intermolecular and intramolecular binding among domains
FIGURE 5. H4 binding with P300. (A–D) In vivo GST pulldown (PD)
assays. GST-IRF7 was transfected into 293T cells with or without plasmid
expressing inducible IKKε (A). GST-tagged construct set a (B), set b (C),
and set d (D) was transfected into 293T cells. At 48 h posttransfection,
whole cell lysates (WCL) were subjected to the GST PD, followed by
immunoblotting (IB) using anti-GST, anti–phospho-IRF7, anti-FLAG, or
anti-P300 Ab.
To investigate the physical associations among domains, we performed GST pulldowns with FLAG-tagged IRF7wt and GSTtagged domain constructs (GST-IRF7wt, -DBD, -AD, -ID, and
-RD) (Fig. 6A). GST-IRF7wt, -DBD, -ID, and -RD bound
IRF7wt, whereas GST-AD did not. The binding of GST-IRF7wt
and FLAG-IRF7wt reflected intermolecular binding (i.e., homodimerization), whereas the binding of the DBD, ID, and RD
constructs with FLAG-IRF7wt might reflect either intra- or intermolecular binding. To rule out the possibility that binding of the
DBD to IRF7wt was mediated by nonspecific DNA fragments
present during the in vivo pulldown procedure, we performed the
assay with FLAG-tagged IRF7 K92R that had no DNA binding
capacity (Fig. 6B). This mutated construct did not bind GST-DBD,
but maintained the ability to bind GST-IRF7wt, -ID, and -RD,
suggesting that homodimerization is mediated by the ID or RD.
Next, we examined the binding of GFP-tagged ID or RD with
other domains (Fig. 6C, 6D). GFP-ID bound GST-ID and GSTRD, whereas GFP-RD bound only GST-ID. Taken together, these
data indicate that the RD-ID association reflects intramolecular
binding, whereas the ID-ID association reflects intermolecular
binding (homodimerization).
ID contained the H2 and H3. In the model structure, hydrophobic
residues in H2 (F407, F411, L414, and A419) are in close contact
with hydrophobic residues in H4 (F495, L496, L499, and A503)
(Fig. 6E). To investigate the role of H2 and H3 in homodimerization, we performed GST pulldowns using GST-IRD with
IRF7wt, IRF7 DH2, and IRF7 DH3, with cotransfection of IKKε
to activate RD (Fig. 6F). IRF7 DH2 and IRF7 DH3 did not bind
GST-IRD, indicating that two helices are required for homodimerization. To evaluate the binding among helices, we cloned
H2 and H3 into a GFP-tagging vector and tested the individual
helix ability to bind GST-RD. The RD bound H2 and H3
(Fig. 6G). These data suggest that the binding between ID and RD
is mediated via the physical interaction among helices.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
The discrepancy between the IFNA4 promoter assays and IFN-a
ELISAs suggested a requirement for a transcriptional coactivator
(Fig. 2). We investigated whether IRF7 could bind with endogenous P300, with or without cotransfection of IKKε (Fig. 5A).
When coexpressed with IKKε, IRF7 was phosphorylated and
bound P300. Next, we assessed the ability of GST-tagged DBD,
DA, DR, and DAR constructs to bind P300 (Fig. 5B). Only GSTDR and GST-DAR bound P300, suggesting that the IRF7 RD
contains a P300-binding site. To confirm that the RD mediates
P300 binding, we subjected individual GST-tagged domains
(GST-DBD, -AD, -ID, and -RD) to pulldown assays (Fig. 5C),
which showed that endogenous P300 bound RD. Based on the
HCCR structural model, in the resting state, H2 and H3 might
prevent H4 from binding P300. Therefore, we investigated
whether endogenous P300 could bind GST-fused SRR constructs
generated by replacing GFP with GST (Fig. 5D). P300 bound
strongly to GST-SRR4, indicating that both the SC and H4 are
required for efficient P300 binding. GST-SRR1 and GST-SRR2
did not bind P300, indicating that H2 and H3 hinder the binding
between H4 and P300.
The Journal of Immunology
9
To determine the IKKε binding site, we performed pulldown
assays with GST-IRF7wt, GST-IRD, and GST-SRR1 (Fig. 7A).
IKKε bound GST-IRD and -SRR1, indicating that the binding site
reside in the HCCR. We then performed immunoprecipitation
with FLAG-tagged IKKε and GFP-tagged helices (GFP-H2, GFPH3, and GFP-H4) (Fig. 7B). GFP-H2 specifically precipitated with
FLAG-tagged IKKε by anti-FLAG Ab. We tested the influence of
GFP-H2 on IRF7 and IKKε in IFNA4 promoter assay, using increasing amounts of expressing plasmids (Fig. 7C). IRF7 activation by IKKε was inhibited by increasing amounts of GFP-H2;
activation could be restored by increasing the level of IKKε, but
not by increasing the level of IRF7. These data indicate that
IRF7wt and GFP-H2 competitively bind IKKε.
V460) are in close contact with hydrophobic residues in the SC
(L478 and L480), and the critical transactivation residues S477
and S479 are intercalated between L478 and L480. We tested
whether IRF7 S477A/S479A (SSAA) or S477D/S479D (SSDD)
mutants could be further activated by IKKε, TBK1, MAVS, or
TRIF (Fig. 8B). The basal activity of IRF7 SSDD was higher than
that of IRF7wt ∼3-fold. In contrast to IRF7 SSAA, which was
nonresponsive, IRF7 SSDD and IRF7wt were activated equivalently by stimulation, supporting total number of charged residues
(phospho-serine and aspartate) in SC is important for the activity.
Based on the HCCR model, phosphorylations of S477 and S479
might break the L457–L478 and L460–L480 hydrophobic interactions. Taken together, these data indicate that H3 binds the SC in
the latent status and that multiple SC phosphorylation releases H4
from the internal hydrophobic core.
SC phosphorylation liberates H4 from the tripod-helix
structure
Discussion
H2 binds IKK«
In the HCCR model, H2, H3, the SC, and H4 form a protruding
tripod structure with an internal hydrophobic core (Figs. 1A, 8A).
The SC contains alternative repeats of serine and hydrophobic
residues. In the model, the hydrophobic residues in H3 (L457 and
We performed a series of functional experiments, based on bioinformatics analyses, aimed at elucidating the mechanism of IRF7
activation by phosphorylation. Our results reveal the specific
functions of the protein’s individual domains, as follows: The
DBD (aa 1–128) contains an NLS and binds with the IRFE; the ID
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 6. Bindings among domains. In vivo GST pulldown (PD) assays. FLAG-tagged IRF7wt (FLAG-IRF7 wt) (A), FLAG-tagged IRF7 K92R point
mutant (FLAG-IRF7K92R) (B), GFP-tagged inhibitory domain (GFP-ID) (C), or GFP-tagged regulatory domain (GFP-RD) (D) was cotransfected into
293T cells with GST-tagged construct set b (Fig. 1B). (E) Hydrophobic residues of H2 and H4 in the model structure. (F and G) IRF7wt, H2 deletion mutant
(IRF7 DH2), or H3 deletion mutant (IRF7 DH3) was transfected with plasmid expressing inducible IKKε and GST-IRD (F). GFP-tagged helix 2 (GFP-H2)
or GFP-tagged helix (GFP-H3) was transfected with GST or FST-RD (G). (A–G) At 48 h posttransfection, whole-cell lysates (WCL) were subjected to the
GST PD, followed by immunoblotting (IB) using anti-GST, anti-FLAG, or anti-GFP Ab.
10
TRANSACTIVATION OF IRF7 BY C-TERMINAL TRIPOD-HELIX STRUCTURE
(aa 259–470), a structural platform with a hydrophobic core, is
involved in IRF7 homodimerization; the NES resides within the
ID; and the RD (aa 471–503) contains the SC and H4, which is
released from the hydrophobic core by serial SC phosphorylation
event; liberated H4 can then bind P300.
Previously studies of IRF7 domain functions used different
isotypes or species of IRF7. To compare these results with our
experimental data obtained using human IRF7 isoform A (hIRF7a),
we aligned the domains by MSA. Lin et al. (28) described the
domains of hIRF7a as the DBD, constitutive AD (CAD), virus AD
(VAD), ID, and serine-rich domain. Their VAD was included in our
ID, which we demonstrated was required for nuclear translocation
of IRF7 (Fig. 4E). We conclude that the VAD is required for the
proper maintenance of the HCCR structure and the mediation of
homodimerization. Yang et al. (38) described the domains of human IRF7 isoform B (hIRF7b) as the DBD, first CAD (CAD1), ID,
and second CAD (CAD2). Their description of the alignment is
identical to ours (CAD1 = AD, CAD2 = RD), except that their
protein contained a deletion of aa 228–256 due to the alternate 59
untranslated region of hIRF7b; the deletion is located in a nonconserved region of our MSA (Fig. 1C) and does not result in
functional impairment. Marié et al. (30) described the domains of
mouse IRF7 isoform A as the DBD, transactivation domain,
autoinhibitory domain, and RD. The mapped regions of human and
mouse IRF7 do not completely match: the transactivation domain
contains our AD and part of the ID, the autoinhibitory domain
contains our H2 and H3, and the RD contains our SC and H4.
Although the AD was required for full activity in the promoter
assay, as well as for IFN-a secretion (Fig. 2), we could not establish a clear functional role for the AD in DNA binding
(Fig. 3A), P300 binding (Fig. 5B, 5C), or control of nuclear
translocation control (Fig. 3B, 3C). The MSA revealed that the
AD is poorly conserved across species (Fig. 1C). Notably, this
domain contains abundant prolines. Due to its bulky side chain,
restricted backbone conformation, and inability to serve as a hydrogen bond donor, the proline residue disrupts a-helices and
b-sheets (39), consistent with the absence of predicted secondary
structure in the AD (Fig. 1C, 2nd structure). We conclude that the
AD might function as a spacer or hinge between the DBD and
HCCR, providing sufficient conformational flexibility for assembly of the basal transcription complex.
Nuclear translocation of a transcription factor translates a cellular signaling event into activation of gene expression. Translocation is controlled by NLS or NES sequences. IRF3 contains an
NLS in the DBD and an NES in the AD (40), and the NES mediates
active export from the nucleus in the resting state (41). Upon
phosphorylation of the SC, CBP/P300 binding masks the NES,
leading to nuclear retention. However, the control of IRF7 translocation is not as clearly defined as that of IRF3. An early study
reported that IRF7 enters the nucleus upon viral stimulation (42,
43), but a later report showed that IRF7 is distributed throughout
the cell, and only the phosphorylated form specifically accumulates in the nucleus (30). We found that IRF7 contains a nonclassical NLS in the DBD that is functionally ablated by the K92R
mutation (Fig. 3C). Notably, K92 is also important for IRFE
binding, and the K92R mutation abolishes the DNA-binding capacity of IRF7 (44). The dual role of this sequence in nuclear
localization and DNA binding underscores the importance of K92
in the DBD structure. A putative NES was reported in the leucinerich region from aa 448–462 previously (28). Our data suggest that
the region between H2 and H3 (aa 420–453) functions as an NES
(Fig. 4D). In the case of IRF3, P300 binding masks the NES;
however, another mechanism of NES hindrance is required for
IRF7, because the binding of P300 to liberated H4 cannot structurally block the NES. One such candidate mechanism is IRF7
homodimerization.
At first, IRF7 RD was thought to exert an inhibitory effect due
to its phosphorylation-mediated activation mechanism, which is
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 7. H2 binding with IKKε. (A) In vivo
GST pulldown (PD) assays. GST-IRF7wt, GSTIRD, or GST-SRR1 was transfected into 293T cells
with FLAG tagged IKKε. At 48 h posttransfection,
whole cell lysates (WCL) were subjected to the
GST PD, followed by immunoblotting (IB) using
anti-GST or anti-FLAG Ab. (B) Immunoprecipitation (IP) assays. GFP, GFP-H2, GFP-H3, or GFPH4 was transfected into 293T cells with FLAG
tagged IKKε. At 48 h posttransfection, WCL were
subjected to the IP using anti-FLAG Ab, followed
by IB using anti-GST or anti-FLAG Ab. (C) IFNA4
promoter assays. IRF7wt was transfected into
293T cells with different concentrations of IKKε
and GFP-H2 (2, 0 ng; +, 20 ng; ++, 40 ng; +++, 60
ng; ++++, 80 ng/105 cells). Data represent the
means 6 SD of three independent experiments (n =
3). Bottom panel, IB showing protein levels of
b-actin and constructs in lysates of cells used for
the promoter assay. RLU, relative luciferase units.
The Journal of Immunology
similar to that of IRF3 (45); however, functional domain analysis
revealed that deletion of the IRF7 RD abolishes transcriptional
activity. Takahasi et al. (46) predicted that the RD was required for
IRF7 homodimerization based on the IRF3 crystal structure. In the
case of IRF3, H2 is the binding site for CBP/P300; the SC and H4
cover the binding site, and phosphorylation of the SC uncovers H4
to unmask the binding site. In this study, we showed that IRF7
H4 is the P300 binding site, which is covered by the hydrophobic
core in the resting state. The difference in the P300 binding sites
explains why the domain arrangement of IRF7 is inverted relative
to that of IRF3, despite the fact that both proteins are activated by
an identical phosphorylation-mediated mechanism.
Based on our HCCR structural model and experimental data,
we hypothesized a mechanism for IRF7 transactivation. Although
the HCCR model is not a crystallographic structure, it provided
sufficient information to guide our functional studies. In the model,
the RD contains the SC and H4 and masks the hydrophobic core
formed by H3 and H4. H2 is in close contact with H4, whereas H3 is
in close contact with the SC, which contains a unique alternating
arrangement of serine and hydrophobic residues. Among these
serine residues, S471/S472 and S477/S479 are critical for transactivation (28, 47). S471 and S472 might be the primary target of
kinases because of their high side-chain solvent accessibilities
(Figs. 1A, 8A), an idea supported by a report that S471A/S472A
mutations abolish a serial phosphorylation of other serine residues
upon viral infection (28). These serial phosphorylation events
separate the SC from the hydrophobic core via an unzipping
mechanism, which is driven by electrostatic repulsion resulting
from the increase in charge. During this process, phosphorylation
of S477/S479 breaks the hydrophobic interaction between H3 and
SC, which is critical for liberation of H4 from the hydrophobic
core. Subsequent phosphorylation of the remaining serine residues
further increases the net charge of the SC. Consequently, the hydrophobic core repulses the highly phosphorylated SC, leading to
liberation of H4, which, in turn can then bind P300.
Compared to the tight regulation of IRF3, IRF7 regulation is
leaky; IRF7 overexpression results in transcriptional activity
even in the absence of upstream stimulation. IRF3 is involved in
the immediate and controlled secretion of IFN-b, whereas IRF7
contributes to robust and amplified type I IFNs secretion during
the late stage of infection (48). These different roles in contexts
enable us to rationalize the differences in the domain arrangements of these two transcription factors: IRF3 is tightly controlled with definite nuclear translocation by upstream signals,
whereas activation of IRF7 is leaky in the activation. Furthermore, IRF3 is expressed at steady state, whereas IRF7 is induced by type I IFNs and has a short t1/2 in the majority of
cells (49). As a result, IRF7 activity is primarily controlled at
the levels of expression and proteasomal degradation. In this
regard, it seems logical that IRF3 has two inhibitory domains,
allowing rigorous control, in contrast to the single inhibitory domain of IRF7. The leaky activation of IRF7 is consistent with this
protein’s role in raising the alarm in situations requiring immediate and robust activation in response to viral spreading.
In addition to phosphorylation, other posttranslational modifications control IRF7 activity by interfering with the protein’s
structure, stability, or localization. TNFR-associated factor 6
modifies IRF7 at C-terminal lysine residues by K63-linked ubiquitination, resulting in activation (50). RTA-associated E3 ubiquitin ligase modifies IRF7 by K48-linked ubiquitination, which
leads to proteasomal degradation (51). Furthermore, small
ubiquitin-related modifier conjugation at lysine 406 inhibits IRF7
activity (52). In addition, IRF7 is acetylated at lysine 92 by the
histone acetyltransferase CBP/P300-associated factor and impairs
its DNA-binding activity (44). Among these modifications,
phosphorylation is the primary mechanism for control of activity. Therefore, a comprehensive knowledge of phosphorylationmediated activation of IRF7 will facilitate better understanding
of other posttranslational modifications.
Disclosures
The authors have no financial conflicts of interest.
References
1. Honda, K., H. Yanai, H. Negishi, M. Asagiri, M. Sato, T. Mizutani, N. Shimada,
Y. Ohba, A. Takaoka, N. Yoshida, and T. Taniguchi. 2005. IRF-7 is the master
regulator of type-I interferon-dependent immune responses. Nature 434: 772–777.
2. Ning, S., A. M. Hahn, L. E. Huye, and J. S. Pagano. 2003. Interferon regulatory
factor 7 regulates expression of Epstein-Barr virus latent membrane protein 1:
a regulatory circuit. J. Virol. 77: 9359–9368.
3. Yu, J., H. Y. Zhang, Z. Z. Ma, W. Lu, Y. F. Wang, and J. D. Zhu. 2003. Methylation profiling of twenty four genes and the concordant methylation behaviours
of nineteen genes that may contribute to hepatocellular carcinogenesis. Cell Res.
13: 319–333.
4. Yu, J., H. Zhang, J. Gu, S. Lin, J. Li, W. Lu, Y. Wang, and J. Zhu. 2004.
Methylation profiles of thirty four promoter-CpG islands and concordant
methylation behaviours of sixteen genes that may contribute to carcinogenesis of
astrocytoma. BMC Cancer 4: 65.
5. Fukasawa, M., M. Kimura, S. Morita, K. Matsubara, S. Yamanaka, C. Endo,
A. Sakurada, M. Sato, T. Kondo, A. Horii, et al. 2006. Microarray analysis of
promoter methylation in lung cancers. J. Hum. Genet. 51: 368–374.
6. Jee, C. D., M. A. Kim, E. J. Jung, J. Kim, and W. H. Kim. 2009. Identification of
genes epigenetically silenced by CpG methylation in human gastric carcinoma.
Eur. J. Cancer 45: 1282–1293.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 8. SC phosphorylation induces liberation of H4. (A) Hydrophobic interactions among helices in the tripod structure. Left panel, Hydrophobic internal core. Right panel, Hydrophobic interaction between H3
and SC. Bottom panel, Alternative repeats of serine and hydrophobic
residues in SC. (B) IFNA4 promoter assays. IRF7wt, IRF7 SSAA, or IRF7
SSDD was transfected along with plasmid expressing inducible IKKε,
TBK1, MAVS, or TRIF. Data represent the means 6 SD of three independent experiments (n = 3). Bottom panel, Immunoblots (IB) showing
protein levels of b-actin and constructs in lysates of cells used for the
promoter assay. RLU, relative luciferase units.
11
12
TRANSACTIVATION OF IRF7 BY C-TERMINAL TRIPOD-HELIX STRUCTURE
30. Marié, I., E. Smith, A. Prakash, and D. E. Levy. 2000. Phosphorylation-induced
dimerization of interferon regulatory factor 7 unmasks DNA binding and a bipartite transactivation domain. Mol. Cell. Biol. 20: 8803–8814.
31. Suhara, W., M. Yoneyama, T. Iwamura, S. Yoshimura, K. Tamura, H. Namiki,
S. Aimoto, and T. Fujita. 2000. Analyses of virus-induced homomeric and
heteromeric protein associations between IRF-3 and coactivator CBP/p300. J.
Biochem. 128: 301–307.
32. Huang, X., and W. Miller. 1991. A time-efficient, linear-space local similarity
algorithm. Adv. Appl. Math. 12: 337–357.
33. Sievers, F., A. Wilm, D. Dineen, and T. J. Gibson. 2011. Fast, scalable generation
of high‐quality protein multiple sequence alignments using Clustal Omega. Mol.
Syst. Biol. 7: 539.
34. Cole, C., J. D. Barber, and G. J. Barton. 2008. The Jpred 3 secondary structure
prediction server. Nucleic Acids Res. 36: W197–W201.
35. Sali, A., and T. L. Blundell. 1993. Comparative protein modelling by satisfaction
of spatial restraints. J. Mol. Biol. 234: 779–815.
36. Joo, C.-H., U. Lee, Y. R. Nam, J. U. Jung, H. Lee, Y. K. Cho, and Y. K. Kim.
2010. Gene therapeutic approach for inhibiting hepatitis C virus replication
using a recombinant protein that controls interferon expression. Antimicrob.
Agents Chemother. 54: 5048–5056.
37. Joo, C.-H., Y. C. Shin, M. Gack, L. Wu, D. Levy, and J. U. Jung. 2007. Inhibition
of interferon regulatory factor 7 (IRF7)-mediated interferon signal transduction
by the Kaposi’s sarcoma-associated herpesvirus viral IRF homolog vIRF3. J.
Virol. 81: 8282–8292.
38. Yang, H., C. H. Lin, G. Ma, M. O. Baffi, and M. G. Wathelet. 2003. Interferon
regulatory factor-7 synergizes with other transcription factors through multiple
interactions with p300/CBP coactivators. J. Biol. Chem. 278: 15495–15504.
39. Williamson, M. P. 1994. The structure and function of proline-rich regions in
proteins. Biochem. J. 297: 249–260.
40. Kumar, K. P., K. M. McBride, B. K. Weaver, C. Dingwall, and N. C. Reich.
2000. Regulated nuclear-cytoplasmic localization of interferon regulatory factor
3, a subunit of double-stranded RNA-activated factor 1. Mol. Cell. Biol. 20:
4159–4168.
41. Yoneyama, M., W. Suhara, Y. Fukuhara, M. Fukuda, E. Nishida, and T. Fujita.
1998. Direct triggering of the type I interferon system by virus infection: activation of a transcription factor complex containing IRF-3 and CBP/p300. EMBO
J. 17: 1087–1095.
42. Au, W. C., P. A. Moore, D. W. LaFleur, B. Tombal, and P. M. Pitha. 1998.
Characterization of the interferon regulatory factor-7 and its potential role in the
transcription activation of interferon A genes. J. Biol. Chem. 273: 29210–29217.
43. Sato, M., N. Hata, M. Asagiri, T. Nakaya, T. Taniguchi, and N. Tanaka. 1998.
Positive feedback regulation of type I IFN genes by the IFN-inducible transcription factor IRF-7. FEBS Lett. 441: 106–110.
44. Caillaud, A., A. Prakash, E. Smith, A. Masumi, A. G. Hovanessian, D. E. Levy,
and I. Marié. 2002. Acetylation of interferon regulatory factor-7 by p300/CREBbinding protein (CBP)-associated factor (PCAF) impairs its DNA binding. J.
Biol. Chem. 277: 49417–49421.
45. Qin, B. Y., C. Liu, S. S. Lam, H. Srinath, R. Delston, J. J. Correia, R. Derynck,
and K. Lin. 2003. Crystal structure of IRF-3 reveals mechanism of autoinhibition
and virus-induced phosphoactivation. Nat. Struct. Biol. 10: 913–921.
46. Takahasi, K., N. N. Suzuki, M. Horiuchi, M. Mori, W. Suhara, Y. Okabe,
Y. Fukuhara, H. Terasawa, S. Akira, T. Fujita, and F. Inagaki. 2003. X-ray crystal
structure of IRF-3 and its functional implications. Nat. Struct. Biol. 10: 922–927.
47. Caillaud, A., A. G. Hovanessian, D. E. Levy, and I. J. Marié. 2005. Regulatory
serine residues mediate phosphorylation-dependent and phosphorylationindependent activation of interferon regulatory factor 7. J. Biol. Chem. 280:
17671–17677.
48. Levy, D. E., I. Marié, E. Smith, and A. Prakash. 2002. Enhancement and diversification of IFN induction by IRF-7-mediated positive feedback. J. Interferon
Cytokine Res. 22: 87–93.
49. Prakash, A., and D. E. Levy. 2006. Regulation of IRF7 through cell type-specific
protein stability. Biochem. Biophys. Res. Commun. 342: 50–56.
50. Ning, S., A. D. Campos, B. G. Darnay, G. L. Bentz, and J. S. Pagano. 2008.
TRAF6 and the three C-terminal lysine sites on IRF7 are required for its
ubiquitination-mediated activation by the tumor necrosis factor receptor family
member latent membrane protein 1. Mol. Cell. Biol. 28: 6536–6546.
51. Yu, Y., and G. S. Hayward. 2010. The ubiquitin E3 ligase RAUL negatively
regulates type i interferon through ubiquitination of the transcription factors
IRF7 and IRF3. Immunity 33: 863–877.
52. Kubota, T., M. Matsuoka, T. H. Chang, P. Tailor, T. Sasaki, M. Tashiro, A. Kato,
and K. Ozato. 2008. Virus infection triggers SUMOylation of IRF3 and IRF7,
leading to the negative regulation of type I interferon gene expression. J. Biol.
Chem. 283: 25660–25670.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
7. Chen, J., L. Guo, D. A. Peiffer, L. Zhou, O. T. M. Chan, M. Bibikova,
E. Wickham-Garcia, S.-H. Lu, Q. Zhan, J. Wang-Rodriguez, et al. 2008. Genomic profiling of 766 cancer-related genes in archived esophageal normal and
carcinoma tissues. Int. J. Cancer 122: 2249–2254.
8. Bidwell, B. N., C. Y. Slaney, N. P. Withana, S. Forster, Y. Cao, S. Loi,
D. Andrews, T. Mikeska, N. E. Mangan, S. A. Samarajiwa, et al. 2012. Silencing
of Irf7 pathways in breast cancer cells promotes bone metastasis through immune escape. Nat. Med. 18: 1224–1231.
9. Li, Q., B. Xu, S. A. Michie, K. H. Rubins, R. D. Schreriber, and H. O. McDevitt.
2008. Interferon-alpha initiates type 1 diabetes in nonobese diabetic mice. Proc.
Natl. Acad. Sci. USA 105: 12439–12444.
10. Xu, W.-D., Y.-J. Zhang, K. Xu, Y. Zhai, B.-Z. Li, H.-F. Pan, and D.-Q. Ye. 2012. IRF7,
a functional factor associates with systemic lupus erythematosus. Cytokine 58: 317–320.
11. Fu, Q., J. Zhao, X. Qian, J. L. H. Wong, K. M. Kaufman, C. Y. Yu, M. Y. Mok,
J. B. Harley, J. M. Guthridge, Y. W. Song, et al; Hwee Siew Howe.; Tan Tock
Seng Hospital Lupus Study Group. 2011. Association of a functional IRF7
variant with systemic lupus erythematosus. Arthritis Rheum. 63: 749–754.
12. Lin, L.-H., P. Ling, and M.-F. Liu. 2011. The potential role of interferonregulatory factor 7 among Taiwanese patients with systemic lupus erythematosus. J. Rheumatol. 38: 1914–1919.
13. Salloum, R., B. S. Franek, S. N. Kariuki, L. Rhee, R. A. Mikolaitis, M. Jolly,
T. O. Utset, and T. B. Niewold. 2010. Genetic variation at the IRF7/PHRF1 locus
is associated with autoantibody profile and serum interferon-alpha activity in
lupus patients. Arthritis Rheum. 62: 553–561.
14. Levy, D. E., I. J. Marié, and J. E. Durbin. 2011. Induction and function of type I
and III interferon in response to viral infection. Curr. Opin. Virol. 1: 476–486.
15. Siegal, F. P., N. Kadowaki, M. Shodell, P. A. Fitzgerald-Bocarsly, K. Shah,
S. Ho, S. Antonenko, and Y. J. Liu. 1999. The nature of the principal type 1
interferon-producing cells in human blood. Science 284: 1835–1837.
16. Yoneyama, M., M. Kikuchi, T. Natsukawa, N. Shinobu, T. Imaizumi,
M. Miyagishi, K. Taira, S. Akira, and T. Fujita. 2004. The RNA helicase RIG-I
has an essential function in double-stranded RNA-induced innate antiviral
responses. Nat. Immunol. 5: 730–737.
17. Seth, R. B., L. Sun, C.-K. Ea, and Z. J. Chen. 2005. Identification and characterization of MAVS, a mitochondrial antiviral signaling protein that activates
NF-kappaB and IRF 3. Cell 122: 669–682.
18. Kawai, T., and S. Akira. 2010. The role of pattern-recognition receptors in innate
immunity: update on Toll-like receptors. Nat. Immunol. 11: 373–384.
19. Lin, R., C. Heylbroeck, P. M. Pitha, and J. Hiscott. 1998. Virus-dependent
phosphorylation of the IRF-3 transcription factor regulates nuclear translocation, transactivation potential, and proteasome-mediated degradation. Mol. Cell.
Biol. 18: 2986–2996.
20. Stark, G. R., and J. E. Darnell, Jr. 2012. The JAK-STAT pathway at twenty.
Immunity 36: 503–514.
21. Ivashkiv, L. B., and L. T. Donlin. 2014. Regulation of type I interferon responses.
Nat. Rev. Immunol. 14: 36–49.
22. Marié, I., J. E. Durbin, and D. E. Levy. 1998. Differential viral induction of
distinct interferon-alpha genes by positive feedback through interferon regulatory factor-7. EMBO J. 17: 6660–6669.
23. Sato, M., H. Suemori, N. Hata, M. Asagiri, K. Ogasawara, K. Nakao, T. Nakaya,
M. Katsuki, S. Noguchi, N. Tanaka, and T. Taniguchi. 2000. Distinct and essential roles of transcription factors IRF-3 and IRF-7 in response to viruses for
IFN-alpha/beta gene induction. Immunity 13: 539–548.
24. Veals, S. A., C. Schindler, D. Leonard, X. Y. Fu, R. Aebersold, J. E. Darnell, Jr.,
and D. E. Levy. 1992. Subunit of an alpha-interferon-responsive transcription
factor is related to interferon regulatory factor and Myb families of DNA-binding
proteins. Mol. Cell. Biol. 12: 3315–3324.
25. Lin, R., P. Génin, Y. Mamane, and J. Hiscott. 2000. Selective DNA binding and
association with the CREB binding protein coactivator contribute to differential
activation of alpha/beta interferon genes by interferon regulatory factors 3 and 7.
Mol. Cell. Biol. 20: 6342–6353.
26. tenOever, B. R., S. Sharma, W. Zou, Q. Sun, N. Grandvaux, I. Julkunen, H. Hemmi,
M. Yamamoto, S. Akira, W.-C. Yeh, et al. 2004. Activation of TBK1 and IKKvarepsilon kinases by vesicular stomatitis virus infection and the role of viral ribonucleoprotein in the development of interferon antiviral immunity. J. Virol. 78: 10636–10649.
27. Sharma, S., B. R. tenOever, N. Grandvaux, G. P. Zhou, R. Lin, and J. Hiscott.
2003. Triggering the interferon antiviral response through an IKK-related
pathway. Science 300: 1148–1151.
28. Lin, R., Y. Mamane, and J. Hiscott. 2000. Multiple regulatory domains control IRF7 activity in response to virus infection. J. Biol. Chem. 275: 34320–34327.
29. Lin, R., Y. Mamane, and J. Hiscott. 1999. Structural and functional analysis of
interferon regulatory factor 3: localization of the transactivation and autoinhibitory domains. Mol. Cell. Biol. 19: 2465–2474.