The Journal of Immunology

Human Cytomegalovirus Envelope Glycoproteins B and H Are Necessary for TLR2 Activation in Permissive Cells1

Karl W. Boehme,* Mario Guerrero,*† and Teresa Compton2*†

Human CMV (HCMV) is a ubiquitous member of the Herpesviridae family and an opportunistic pathogen that poses significant health risks for immunocompromised patients. HCMV pathogenesis is intimately tied to the immune status of the host, thus characterization of the innate immune response to HCMV infection is critical for understanding disease progression. Previously, we identified TLR2 as a host factor that detects and initiates inflammatory cytokine secretion in response to HCMV independent of viral replication. In this study, we show that two entry-mediating envelope gp, gp B (gB) and gp H (gH), display determinants recognized by TLR2. Neutralizing Abs against TLR2, gB and gH inhibit inflammatory cytokine responses to HCMV infection, suggesting that inflammatory cytokine stimulation by HCMV is mediated by interactions between these envelope gp and TLR2. Furthermore, both gB and gH coimmunoprecipitate with TLR2 and TLR1, indicating that these envelope gp directly interact with TLR2 and that a TLR2/TLR1 heterodimer is a functional sensor for HCMV. Because our previous studies were conducted in model cell lines, we also show that TLR2 is expressed by HCMV permissive human fibroblast cell strains, and that TLR2 is a functional sensor in these cells. This study further elucidates the importance and potency of envelope gp as a class of molecules displaying pathogen-associated molecular patterns that are recognized with immediate kinetics by TLRs in permissive cells. The Journal of Immunology, 2006, 177: 7094–7102.

uman CMV (HCMV)3 is a ubiquitous member of the We recently identified TLR2 as a host factor that activates in- Herpesviridae family that causes significant morbidity flammatory cytokine secretion in response to HCMV (17). The H and mortality in immune compromised patients (1). TLRs are a family of pathogen-recognition receptors that initiate Similar to other herpesviruses, HCMV establishes a lifelong rela- innate immune responses to a myriad of invading microbes, in- tionship with its host as a latent infection, and disease can result cluding viruses (18, 19). Eleven mammalian TLRs have been iden- from either primary infection or reactivation from latency (2, 3). tified, and they are predominantly expressed on phagocytic cells HCMV has an extremely broad tissue tropism that allows it to such as dendritic cells and macrophages; however, most cells ex- infect nearly every organ system in the body (4, 5). Consequently, press at least a subset of TLRs (19). The primary consequences of HCMV disease presents itself in a variety of clinical sequelae (1). TLR activation include NF-␬B activation, inflammatory cytokine

by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. It is a major cause of postoperative disease in chemically immu- secretion, dendritic cell maturation, up-regulation of immune co- nosuppressed transplant recipients and greatly increases the risk of stimulatory molecules, and for a subset of TLRs, the production of graft rejection (6–8). HCMV is also a leading cause of congenital type I IFN (19–22). TLRs detect microorganisms on the basis of birth defects, and infection during the first trimester of pregnancy unique molecular structures termed pathogen-associated molecular often results in neurological and cognitive disorders in the devel- patterns (PAMPs). Analysis of the innate response to bacterial oping child (9, 10). Furthermore, HCMV has been implicated as a PAMPs such as LPS, peptidoglycan, and unmethylated CpG DNA factor in coronary artery disease (11–14). There is currently no are a cornerstone of TLR research, and great strides have been vaccine for HCMV, and existing therapeutics exhibit toxicity pre- made in our understanding of the relationship between bacteria and cluding long-term use (15, 16). Thus, an understanding of the viral the innate immune system (23–28). In contrast, the mechanisms by http://classic.jimmunol.org and cellular determinants of the immune response to HCMV is critical for the development of new vaccines and therapies. which the TLR system recognizes and responds to viruses have only begun to be explored. Viral genomic nucleic acids are one major class of PAMP. TLR3 (dsRNA), TLR7 (ssRNA), TLR8 *McArdle Laboratory for Cancer Research, University of Wisconsin Medical School, (ssRNA), and TLR9 (CpG DNA) (29–33) signal from the endo- Madison, WI 53706; and †Department of Biomolecular Chemistry, University of some (34–38) where degradation of virus particles exposes the Wisconsin, Madison, WI 53706 viral genome for detection by this panel of TLRs (29, 31, 32). Downloaded from Received for publication April 17, 2006. Accepted for publication August 29, 2006. Although significantly less well studied, envelope gps that deco- The costs of publication of this article were defrayed in part by the payment of page rate the exterior of the virion are an emerging class of TLR activators charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. (18). To date, three envelope gp have been identified as TLR agonists. 1 This work was supported by National Institutes of Health Grants RO1AI34998 The fusion protein from respiratory syncytial virus and the mouse and R21A154915 (to T.C.) and National Institutes of Health Training Grant mammary tumor virus envelope protein activate TLR4, while the T32GM07215 (to K.W.B. and M.G.). hemagglutinin protein from measles virus activates TLR2 (39–42). 2 Address correspondence and reprint requests to Dr. Teresa Compton, 100 Tech- Interestingly, a shared feature of these gp is that they play critical roles nology Square, Novartis Institute for Biomedical Research, Cambridge, MA 02139. E-mail address: [email protected] in the entry of their respective viruses, and this shared feature suggests 3 Abbreviations used in this paper: HCMV, human CMV; PAMP, pathogen-associ- that the molecular machinery used by viruses for entry is also targeted ated molecular pattern; gB, gp B; gH, gp H; gL, gp L; gO, gp O; NHDF, normal by the innate immune system (43, 44). human dermal fibroblast; HEK, human embryonic kidney; MOI, multiplicity of in- fection; eGFP, enhanced GFP; VSV-G, vesicular stomatitis virus G; HSV-1, herpes Although we demonstrated previously that TLR2 is activated by simplex virus type 1; CHO, Chinese hamster ovary. HCMV, the molecular trigger for TLR2 has not been determined

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00 The Journal of Immunology 7095

(17). In contrast with the RNA viruses listed above, HCMV dis- centration of 1 mg/ml. The cells were subsequently adapted to suspension plays as many as 12 envelope gp, four of which are required for in chemically defined CHO medium (Invitrogen Life Technologies) sup- entry. gp B (gB) works in concert with a tripartite complex com- plemented with 1% PSF and geneticin at a concentration of 1 mg/ml. For protein isolation, cells were pelleted and lysed by sonication in lysis buffer prised of gp H (gH), gp L (gL), and gp O (gO) to mediate the (50 mM NaPO4, 300 mM NaCl, 0.5% Tween 20, 10 mM imidazole (pH binding and entry of HCMV virons into host cells (45–48). In 8.0)). Cell debris was removed by centrifugation at 27,000 ϫ g for 30 min. addition to their roles in entry, there is a growing body of evidence The supernatants were incubated for2hat4°Cunder rotation with Ni- that gB and gH elicit responses from cells that are reminiscent of NTA-agarose beads (Qiagen). The beads were transferred to a chromatog- raphy column and washed with 10-column volumes of lysis buffer, fol- TLR activation. Abs against gB and gH block the induction of lowed by 10-column volumes of wash buffer (50 mM NaPO4, 300 mM various innate markers, including NF-␬B (49, 50), and cells ex- NaCl, 20 mM imidazole (pH 8.0)). gBs-GFP was eluted in 4 ml of elution ␬ posed to soluble forms of gB activate NF- B and the type I IFN buffer (50 mM NaPO4, 300 mM NaCl, 300 mM imidazole (pH 8.0)) and (49–53). Additionally, an anti-Id bearing the image of gH activates dialyzed overnight in PBS (50 mM NaPO4, 150 mM NaCl (pH 8.0)) at 4°C. NF-␬B (50). Based on these observations, we hypothesized that gB Dialyzed protein was separated from low m.w. contaminants by size-ex- clusion chromatography. Samples were loaded onto a 50-ml column con- and gH are the target of innate sensing by the host cell. In this taining Sephacryl S-200 substrate (Amersham Biosciences) in 1ϫ PBS study, we show that HCMV gB and gH activate TLR2 and asso- (Invitrogen Life Technologies) and run through by gravity flow at 4°C. ciate with TLR1 and TLR2. Abs against gB and gH, but not gL, Collected fractions were stored at Ϫ80°C. inhibit the inflammatory cytokine response to HCMV, and both gB Construction and generation of TLR2⌬C and TLR4⌬C-encoding and gH coimmunoprecipitate with TLR2 and TLR1, indicating that retroviruses the functional sensor for HCMV is a TLR2/TLR1 heterodimer. We also extend our initial studies to HCMV permissive human fibro- The mutants were constructed using full-length FLAG epitope-tagged blast cells and show that TLR2 mediates NF-␬B activation and TLR2 and TLR4 provided by B. Williams (Cleveland Clinic Foundation, Cleveland, OH). The TLR2 and TLR4 cytoplasmic tails were deleted by inflammatory cytokine responses in cells that support productive PCR mutagenesis using a common upstream primer (5Ј-TAA TAT ACC HCMV infection. GGT GCC ACC ATG TCT GCA CTT CTG ATC C-3Ј) incorporating an AgeI restriction site and TLR2-specific (5Ј-TTA AAT GCG GCC GCT Materials and Methods TAT GTA TTT CAT ATA CCA CAG GCC-3Ј) and TLR4-specific (5Ј- TTA AAT GCG GCC GCT TAT GTA GCA GCC AGC AAG AAG C-3Ј) Cell lines, reagents, and virus downstream primers incorporating NotI restriction sites. The fragments Human embryonic kidney (HEK) 293T cells (American Type Culture Col- were digested and cloned into the retroviral transfer vector pCMMP.MCS. lection) and normal human dermal fibroblast (NHDF) (Cambrex) cells IRES-GFP (a gift from B. Sugden, University of Wisconsin, Madison, WI). The constructs were confirmed by sequencing (University of Wisconsin were grown in 5% CO2 in DMEM (Invitrogen Life Technologies) supple- mented with 10% FBS (HyClone) and 1% penicillin-streptomycin-ampho- Biotechnology Center), and recombinant retroviruses were generated as tericin B-fungizone (PSF; BioWhittaker). Monomac-6 cells were main- described previously (59). NHDF cells were transduced with retroviruses tained in Ham’s F12 medium supplemented with 10% FBS and 1% PSF in encoding TLR2⌬C, TLR4⌬C, or an empty vector control in a minimal volume for1hinthepresence of 5 ␮g/ml polybrene. At 96 h posttrans- a5%CO2 environment. LPS (from Escherichia coli 0111:B4) was ob- tained from Sigma-Aldrich and repurified by phenol extraction as de- duction, GFP-positive cells were collected by FACS and used as indicated. scribed previously (54). Recombinant human IL-1␤ was obtained from Cytokine ELISAs R&D Systems, Pam3CSK4 was obtained from EMC Microcollections, and soluble CD14 (sCD14) was from Biometec. The AD169 strain of HCMV by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. Ninety-six-well plates were seeded with cells at a density of 5000 cells per was propagated in NHDF cells. Virion particles were purified from infected well. At 24 h postplating, the growth medium was removed and replaced supernatants by density-gradient centrifugation (55–57), and titers were with serum-free DMEM. After 24 h serum starvation, the cells were chal- determined as described previously on NHDF cell monolayers (58). lenged as indicated. At 18 h postchallenge, the supernatants were harvested RT-PCR and IL-6 or IL-8 levels were determined by ELISA. OptEIA IL-6 or IL-8 dual Ab detection assay (BD Pharmingen) was used according to the man- Total RNA was harvested from NHDF, HEK/CD14, or Monomac 6 cells ufacturer’s instructions. For blocking Ab studies, virions were preincu- using RNA-STAT 60 (Tel-Test B) according to the manufacturer’s instruc- bated for 15 min with isotype control (eBioscience), anti-gB 27-78 or 9-3 tions. RNA was quantitated, and 1 ␮g of RNA was used for RT-PCR (60), or anti-gH 14-4b (61) mouse mAbs at 100 ␮g/ml. For gL, a rabbit analysis with TLR2 and GAPDH-specific primers. RNA was quantitated, polyclonal anti-gL 6394 (62) or rabbit IgG (Sigma-Aldrich) were used at and 1 ␮g of RNA was used for RT-PCR analysis using the rTth DNA 100 ␮g/ml. For TLR2 blocking, Ab studies the cells were preincubated for http://classic.jimmunol.org polymerase (Applied Biosystems) with TLR2 (5Ј-GCC AAA GTC TTG 30 min with isotype control or anti-TLR2 Abs (eBioscience) at 1 ␮g/ml. ATT GAT TGG-3Ј and 5Ј-TTG AAG TTC TCC AGC TCC TG-3Ј) and GAPDH (5Ј-GAG CCA AAA GGG TCA TC-3Ј and 5Ј-GTG GTC ATG Coimmunoprecipitations and immunoblotting AGT CCT TC-3Ј)-specific primers. The gB, gH, and gL coding sequences from HCMV strain AD169 were Cloning and purification of recombinant gBs-GFP cloned previously into the pCAGGS expression vector (63). The vector pCVSVG encoding vesicular stomatitis virus G (VSV-G) was a gift from gBs-GFP was constructed by fusing the ectodomain of gB (strain AD169) Y. Kawaoka (University of Wisconsin, Madison, WI). pFLAG-TLR1,

Downloaded from ending at aa 750 to the enhanced GFP (eGFP). The ectodomain of gB was pFLAG-TLR2, and pFLAG-TLR6 plasmids were donated by B. Williams amplified with the upstream primer (5Ј-CTC GAG CTC GAG ATG GAA (Cleveland Clinic Foundation). For coimmunoprecipitation experiments, TCC AGG ATC-3Ј) incorporating a XhoI site and the downstream primer 293T cells were cotransfected with plasmids encoding gB, gH, gL, VSV-G (5Ј-TCT AGA TCT AGA GGG GTT TTT GAG GAA-3Ј) incorporating an pFLAG-TLR1, pFLAG-TLR2, or pFLAG-TLR6 as indicated. Transfec- XbaI site. The eGFP sequence was amplified using the upstream primer tions were performed using Lipofectamine 2000 (Invitrogen Life Technol- (5Ј-TCT AGA TCT AGA ATG GTG AGC AAG-3Ј) incorporating an XbaI ogies) according to the manufacturer’s directions. Dose-response precipi- site and the downstream primer (5Ј-CGC GGC CGC GGC TCA CTT GTA tations were performed in 6-well plates, with total DNA concentrations CAG CTC-3Ј) incorporating a NotI site. The gB fragment was inserted into ranging from 0 to 2 ␮g/ml. For radiolabeled immunoprecipitation assays, the pCI-Neo vector (Promega) using XhoI/XbaI. The eGFP fragment was cells were incubated for 24 h in DMEM supplemented with 150 ␮Ci/ml subsequently inserted using XbaI/NotI.A6ϫ histidine tag was added to the 35S-express label (NEN-DuPont). Cells were harvested at 48 h posttrans- 3Ј end of eGFP upstream of the stop codon using the downstream primer fection in lysis buffer (TBS plus 2% TX-100 (pH 8.8)). Lysates were clar- (5Ј-TCA GTG GTG GTG GTG GTG GTG CTT GTA CAG CTC-3Ј). ified twice by high-speed microcentrifugation (13,000 rpm, 5 min, 4°C), Oligonucleotide primers were synthesized at the University of Wisconsin diluted 2-fold in lysis buffer, and incubated with 30 ␮l of anti-FLAG- Biotechnology Center. The construct was transfected into Chinese hamster conjugated bead slurry (M2; Sigma-Aldrich) for 12 h with continuous ovary (CHO) pgsD 677 cells for the generation of a stable gBs-GFP-ex- rocking at 4°C. The beads were pelleted under low-speed microcentrifu- pressing cell line. At 48 h posttransfection, GFP-positive cells were col- gation (2500 rpm, 2 min, 4°C) and washed five times in lysis buffer. Prod- lected by FACS (University of Wisconsin Flow Cytometry Facility) and ucts were eluted by competition with 3ϫ FLAG peptide (Sigma-Aldrich) subjected to geneticin (Invitrogen Life Technologies) selection at a con- and analyzed by 10% SDS-PAGE gel followed by immunoblotting using 7096 HCMV ENVELOPE gp ACTIVATE TLR2

the following Abs: anti-gB 27-78 (60), anti-gH 6824 (62), anti-gL 6394 (62), anti-VSV-G 15F9 (Sigma-Aldrich), or anti-FLAG M2 (Sigma- Aldrich). For the radiolabeled immunoprecipitation assay, the cells were transfected with gB and FLAG-TLR2 expression constructs individually or in combination. At 24 h posttransfection, the medium was replaced with DMEM supplemented 10% FBS, 1% PSF, and 150 ␮Ci/ml 35S-express label (NEN-DuPont). The cells were harvested at 48 h posttransfection and processed as described above. The immunoprecipitation products were re- solved by 10% SDS-PAGE, the gel was dried to Whatman paper, and exposed to film. Images were collected using a Typhoon phosphor imager. I␬B␣ degradation assay Cells were serum starved for 24 h before infection and pretreated with cycloheximide (100 ␮g/ml) for 1 h before infection. The cells were treated ␤ ␮ ␮ with IL-1 (100 pg/ml), Pam3CSK4 (20 g/ml), LPS (1 g/ml) plus 4 (1 ␮g/ml), or infected with UV-HCMV (multiplicity of infection (MOI) ϭ 10) as measured before UV treatment. At 3 h posttreatment,the cells were harvested by scraping in Nonidet P-40 lysis buffer (50 mM Tris (pH 7.4), 150 mM NaCl, 30 mM NaF, 5 mM EDTA, 10% glycerol, 40 mM 2-glyc-

erophosphate; Sigma-Aldrich), 1 mM Na3VO4 (Sigma-Aldrich), 0.1 mm of PMSF, protease inhibitor mixture (56), and 1% Nonidet P-40). The cells were subjected to two freeze-thaw cycles, and insoluble material was re- moved by microcentrifugation (13,000 rpm, 5 min, 4°C). The total protein content of each sample was quantitated using the Bio-Rad protein assay reagent. Equivalent amounts of total protein for each sample were sepa- rated by 10% SDS-PAGE. I␬B␣ and actin levels were analyzed by immu- noblotting as described previously (56) using anti-I␬B␣ (sc-371; Santa Cruz Biotechnologies) and anti-actin Abs. Densitometry was performed using the ImageQuant software system (Amersham Biosciences). Statistical analysis The means of triplicate samples were compared using an unpaired Stu- dent’s t test with GraphPad Prism software (version 4.00; GraphPad). FIGURE 1. gB and gH elicit cytokine responses to HCMV. A, UV- Results inactivated HCMV virions (MOI ϭ 1) were incubated with the indicated gB and gH elicit inflammatory cytokine responses from cells Abs (100 ␮g/ml) for 15 min before infection of NHDF cells. At 18 h To assess the ability of envelope gp to elicit inflammatory cytokine postinfection, the supernatants were harvested and IL-6 levels were deter- Ͻ ء responses we used a panel of neutralizing Abs to block interactions mined by ELISA. Error bars indicate SD. , p 0.05 in comparison to untreated control. B, HEK cells expressing CD14 alone or in combination between gB and gH and receptors on the surface of the cell (Fig. with TLR2 or TLR4 were mock treated, treated with Pam3CSK4 (100 ng/ by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. 1A). Transcriptionally inert UV-inactivated HCMV virions (UV- ml) or LPS (20 ␮g/ml), infected with HCMV (MOI ϭ 0.01), treated with HCMV) were incubated with Abs for 15 min before infection, and gB-sGFP (250 nM), or treated with soluble eGFP alone (2 ␮M). At 18 h IL-6 levels were measured by ELISA at 18 h postinfection as a posttreatment, the supernatants were harvested and IL-8 levels were deter- marker of inflammatory cytokine activation. IL-6 levels were di- mined by ELISA. Error bars indicate SD. minished by pretreatment of virions with gB (27-78 and 9-3) and gH-specific (14-4b) Abs, whereas the isotype control Ab had a modest effect on the IL-6 response. Furthermore, a rabbit poly- interactions (64). Based on the preceding data, we hypothesized clonal Ab against gL did not affect the IL-6 response (data not that gB and gH directly interact with TLR2. To test this hypoth- shown). These data suggest that gB and gH, but not gL, are inter- esis, we performed coimmunoprecipitation experiments from 293T http://classic.jimmunol.org acting with cell surface receptors that elicit inflammatory cytokine cells cotransfected with gB and FLAG-TLR2 expression con- secretion. structs. FLAG-TLR2 was immunoprecipitated with anti-FLAG Additionally, the capacity of a soluble form of gB to activate Ab-conjugated agarose beads, the proteins resolved by SDS- TLR2 was assessed. The ectodomain of gB (HCMV strain AD169) PAGE, and products detected by immunoblotting for FLAG-TLR2 was fused to the eGFP (gB-GFP), the resulting protein purified and or gB (Fig. 2A). A dose-dependent pulldown of gB was observed used to challenge HEK cells expressing CD14 alone (HEK/CD14) from cells cotransfected with a constant amount of gB-expression

Downloaded from or in combination with TLR2 (HEK/CD14/TLR2) or TLR4 (HEK/ plasmid and increasing levels of FLAG-TLR2-expression plasmid. CD14/TLR4) and IL-8 levels were measured by ELISA (Fig. 1B). The amount of gB precipitated increased in proportion to the level To date, we have been unable to generate a soluble form of the of FLAG-TLR2 input. Immunoprecipitations from 35S-labeled

gH/gL complex. The TLR2- and TLR4-specific controls, Pam3CSK4 cells confirmed that coexpression of both gB and FLAG-TLR2 is and LPS, elicited IL-8 responses from HEK/CD14/TLR2 and HEK/ required for this interaction (Fig. 2B). gH also coprecipitated with CD14/TLR4, respectively; and, consistent with our previous study, TLR2 indicating that both gB and gH physically interact with HCMV virions only activated cells expressing TLR2 (17). Simi- TLR2 (Fig. 2E). The envelope gp from VSV-G was included as a larly, gB-GFP induced IL-8 secretion in a TLR2-dependent man- specificity control and did not coprecipitate with any of the TLRs ner. A soluble eGFP control did not elicit cytokine responses. tested (Fig. 2G). These data further support the hypothesis that HCMV envelope gp In vivo, TLR2 functions as a heterodimer in combination with serve as agonists for TLR2. either TLR1 or TLR6 (65, 66). To determine the interacting part- ner for TLR2, gB and gH coimmunoprecipitation experiments gB and gH physically associate with TLR2 and TLR1 were performed with TLR1 and TLR6. Both gB and gH copre- TLR ectodomains are composed of varying numbers of leucine- cipitated with TLR1, but not TLR6, suggesting that the operative rich repeats, a motif that is commonly involved in protein–protein sensor for HCMV gB and gH is a TLR2/TLR1 heterodimer (Fig. The Journal of Immunology 7097

FIGURE 2. HCMV envelope glycoproteins coimmunoprecipitate with TLR2 and TLR1. 293T cells were cotransfected with a constant amount of gB expression plasmid and increasing amounts of plasmids encoding FLAG-TLR2 (A), FLAG-TLR1 (C), or FLAG-TLR6 (D). At 48 h posttransfection, immunoprecipitations were per- formed with anti-FLAG agarose beads. The pre- cipitation products were separated by SDS-PAGE and immunoblotted with anti-gB and anti-FLAG Abs as indicated. B, 293T cells were transfected with gB or FLAG-TLR2 expression plasmids as indicated. At 24 h posttransfection, the cells were metabolically labeled with [35S]methionine. At 48 h posttransfection, immunoprecipitations were performed as described above, the products were separated by SDS-PAGE, and labeled proteins were detected by autoradiography. Bands corre- sponding to gB and TLR2 are indicated. 293T cells were transfected with gH (E), gL (F), or VSV-G (G) expression plasmids, along with FLAG-TLR1, FLAG-TLR2, or FLAG-TLR6 as indicated. At 48 h posttransfection, immunoprecipitations were performed as described above. The precipitation products were analyzed with anti-gH, anti-gL, anti- VSV-G, or anti-FLAG Abs as indicated. by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. http://classic.jimmunol.org

2, C–E). In contrast, gL did not coprecipitate with TLR2, TLR1, or known whether these cells express TLR2. RT-PCR analysis of TLR6 (Fig. 2F). These results are consistent with inability of anti- total RNA from NHDF cells revealed the presence of TLR2 (Fig.

Downloaded from gL Abs to inhibit the inflammatory cytokine response to HCMV. 3), as well as TLR1, TLR6, and TLR4 (data not shown). RNA Together, these results indicate that gB and gH directly interact harvested from Monomac-6 and 293T cells were included as pos- with a TLR2/TLR1 heterodimer to activate inflammatory cytokine itive and negative controls for TLR2 expression, respectively (Fig. responses from cells. 3). Efforts to detect TLR2 protein expression in NHDF cells have been unsuccessful; however, the presence of the TLR2 transcript, HCMV activates TLR2 in permissive fibroblasts coupled with the ability of NHDF cells to respond to the synthetic

To appreciate the relationship between TLR2 and HCMV in a TLR2 ligand Pam3CSK4 (Fig. 4), indicates that these cells express context that is physiologically relevant to the life cycle of HCMV, TLR2. Additionally, TLR1, TLR6, and TLR4 transcripts were de- it is critical to use cells that are fully permissive for HCMV in- tected in NHDF cells by RT-PCR (data not shown). NHDF cells fection. The initial studies that identified TLR2 as a host factor secrete IL-6 in response to zymosan (data not shown). mediating innate responses to HCMV used cell types that do not To determine whether TLR2 mediates innate responses to support HCMV replication (17). Thus, we endeavored to translate HCMV in NHDF cells, we tested the effect of an anti-TLR2 Ab on our findings into HCMV permissive human fibroblast cells, the the cytokine response to HCMV (Fig. 4). Compared with medium best-characterized cell culture system for the study of HCMV. The alone (No Ab), an isotype control Ab had no effect on the IL-6 TLR repertoire of NHDF cells has not been reported, and it is not response to UV-HCMV infection, the TLR-independent control 7098 HCMV ENVELOPE gp ACTIVATE TLR2

FIGURE 5. Dominant-negative TLR constructs. A, Dominant-negative TLR2 and TLR4 constructs were generated that lack the cytoplasmic TLR1 ⌬ ⌬ FIGURE 3. TLR2 is expressed in NHDF cells. RT-PCR analysis of to- domain (TLR2 C and TLR4 C, respectively). The dominant-negative tal RNA harvested from NHDF, HEK, and Monomac-6 cells using TLR2 constructs contain an aminoterminal FLAG epitope tag for detection pur- (left) and GAPDH-specific primers (right). The resulting PCR products poses. B, Expression of the constructs was confirmed by immunoprecipi- were analyzed by 1% agarose gel electrophoresis and visualized by tation-immunoblot analysis. ethidium bromide staining. in combination with eGFP (59). A control population expressing a GFP vector was also generated. GFP-positive cells were collected IL-1␤, or the synthetic TLR2 ligand Pam CSK . In contrast, the 3 4 by FACS and expression of the dominant-negative constructs was levels of IL-6 secreted in response to UV-HCMV and the TLR2 confirmed by immunoprecipitation and immunoblotting (Fig. 5B). control Pam CSK were dramatically reduced by pretreatment 3 4 IL-6 secretion was used as a marker of inflammatory cytokine with the TLR2 blocking Ab, but no effect was observed on the activation after challenge of the dominant-negative cell panel with response to IL-1␤. These results indicate that inflammatory cyto- UV-HCMV (Fig. 6). The GFP vector control cells responded nor- kine responses to HCMV in permissive NHDF cells are mediated mally to all stimuli, including UV-HCMV. TLR2⌬C-expressing by TLR2. cells responded normally to IL-1␤ and LPS ϩ sCD14, the TLR4 To further address the role of TLR2 in NHDF cells, we con- ligand. However, these cells displayed a reduced response to UV- structed dominant-negative versions of TLR2 and TLR4 by re- HCMV and the TLR2 control ligand Pam CSK , confirming that moving their cytoplasmic tails (TLR2⌬C and TLR4⌬C, respec- 3 4 TLR2 mediates inflammatory cytokine responses to HCMV in per- tively) (Fig. 5A). These signaling-defective constructs lack the missive NHDF cells. The IL-6 response from TLR4⌬C-expressing TLR1 domain common to all TLRs and cannot recruit the cyto- ␤

by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. cells to IL-1 and Pam CSK were unaffected, and as predicted, plasmic adaptor molecules that propagate downstream signaling 3 4 the response to LPS ϩ sCD14 was completely eliminated. Inter- events (67). NHDF cells were transduced with recombinant retro- estingly, the response to UV-HCMV was partially diminished in viruses that coexpress FLAG epitope-tagged TLR2⌬C or TLR4⌬C these cells. No role for TLR4 was found in previous studies (17); however, these experiments were performed in nonpermissive http://classic.jimmunol.org Downloaded from

FIGURE 4. Anti-TLR2 Abs block inflammatory cytokine responses to FIGURE 6. Dominant-negative TLR2 inhibits inflammatory cytokine HCMV. NHDF cells were incubated without Ab (No Ab) or with 1 ␮g/ml responses to HCMV in permissive fibroblasts. NHDF cells expressing a isotype control Ab or anti-TLR2 Ab for 30 min before HCMV challenge. GFP control vector, TLR2⌬C, or TLR4⌬C were mock infected, treated ␤ ␮ ␤ ␮ ␮ ϩ The cells were treated with IL-1 (1 pg/ml), Pam3CSK4 (20 g/ml), or with IL-1 (1 pg/ml), Pam3CSK4 (40 g/ml), LPS (1 g/ml) sCD14 (1 infected with UV-inactivated HCMV at a MOI of 1. The supernatants were ␮g/ml), or infected with UV-inactivated HCMV at a MOI of 10. The harvested at 18 h postinfection, and IL-6 levels were determined by supernatants were harvested at 18 h postinfection, and IL-6 levels were p Ͻ 0.05, compared with ,ء .p Ͻ 0.001 in comparison to No Ab determined by ELISA. Error bars indicate SD ,ء .ELISA. Error bars indicate SD control. GFP control. The Journal of Immunology 7099 by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd.

FIGURE 7. HCMV activates NF-␬B via TLR2 in permissive fibroblasts. NHDF cells expressing a GFP control vector (A), TLR2⌬C(B), or TLR4⌬C ␤ ␮ (C) were treated with cycloheximide for 30 min before challenge. The cells were mock infected, treated with IL-1 (1 pg/ml), Pam3CSK4 (40 g/ml), LPS (1 ␮g/ml) ϩ sCD14 (1 ␮g/ml), or infected with UV-inactivated HCMV at a MOI of 10. At 3 h postchallenge, whole-cell lysates were prepared. I␬B␣ and actin levels were determined by SDS-PAGE analysis followed by immunoblotting (left panel). Right panel, The immunoblot shown in the upper panel was subjected to densitometric analysis using ImageQuant software. The intensity of the I␬B␣ bands for each sample was normalized to the intensity of the corresponding actin band. The resulting I␬B␣ intensity was plotted as a percentage of the I␬B␣ signal in the mock-infected cells for each transductant. The results are indicative of two independent experiments. http://classic.jimmunol.org

cells. It is possible that TLR4 is involved in the innate response to trol cells IL-1␤ caused complete I␬B␣ degradation, whereas ϩ HCMV in permissive cells and this possibility is currently under Pam3CSK4 and LPS sCD14 induced a lesser degree of degra- consideration. dation. Furthermore, I␬B␣ degradation is blocked in response to

Downloaded from Pam CSK and LPS ϩ sCD14 in TLR2⌬C and TLR4⌬C-express- ␬ 3 4 TLR2 mediates NF- B activation in response to HCMV infection ing cells, respectively. Similar results were observed in a second Another signature TLR response is activation of the pleiotropic experiment. In response to UV-HCMV near-complete degradation transcription factor NF-␬B (19). Previous studies have shown that of I␬B␣ is observed in GFP vector control and TLR4⌬C-express- HCMV activates NF-␬B within minutes after infection, kinetics ing cells. However, in cells expressing TLR2⌬C, the level of I␬B␣ that are suggestive of receptor-induced signaling (50, 68, 69). To degradation is reduced. Densitometric analysis of the blots indi- determine whether TLR2 mediates NF-␬B activation upon HCMV cates that, although dominant-negative TLR2 completely prevents infection, we used our dominant-negative TLR cell panel to assess I␬B␣ degradation in response to the TLR2 control ligand ␬ ␣ ␬ ␬ ␣ the degradation of I B as a marker of NF- B activation (Fig. 7). Pam3CSK4,I B degradation in response to HCMV infection is I␬B␣ binds and sequesters NF-␬B in the cytoplasm as part of a not completely blocked (Fig. 7, lower panel). This observation transcriptionally inactive complex (70). Many stimuli, including suggests that TLR2 is not the only mechanism by which HCMV TLRs, induce I␬B␣ degradation thereby releasing NF-␬B to trans- can activate NF-␬B. Together, these observations indicate that locate to the nucleus where it complexes with numerous other TLR2 mediates a portion of NF-␬B activation in response to factors to modulate transcription. Thus, the loss of I␬B␣ though HCMV infection and further support the hypothesis that TLR2 is degradation correlates with the activation of NF-␬B. In GFP con- a key cellular factor for the innate immune response to HCMV. 7100 HCMV ENVELOPE gp ACTIVATE TLR2

Discussion In addition to HCMV, several other members of the Herpesviri- The goal of this study is to further elucidate the relationship be- dae activate innate immunity through TLRs. Herpes simplex virus tween viruses and the host innate immune response. We previously type 1 (HSV-1), HSV-2 and mouse CMV harbor CpG-rich ge- identified TLR2 as a cellular factor that mediates innate immune nomes that activate TLR9 (32, 33, 84, 85), and HSV-1 and vari- responses to HCMV infection (17). However, many questions re- cella-zoster virus activate TLR2 (86, 87). An emerging possibility main with respect to the mechanism by which HCMV activates is that herpesviruses are subject to innate detection by multiple TLR2, as well as the effect of TLR2 activation on the virus. In this TLRs, with each TLR providing a distinct contribution to the over- study, we show that two HCMV envelope gp, gB and gH, activate all response. For instance, TLR2 is associated with inflammatory TLR2. mAbs against both gB and gH inhibit cytokine responses to cytokine responses, whereas TLR9 elicits the secretion of type I HCMV, and both gB and gH physically associate with TLR2 in IFNs. Using multiple TLRs would allow the host to tailor its re- coimmunoprecipitation experiments. gB and gH also coprecipitate sponse to fit the pathogen through the combined actions of each with TLR1, but not TLR6, indicating that the functional sensor for TLR. In addition, HCMV infects a variety of cell types in vivo, HCMV is a TLR2/TLR1 heteromeric complex. Our previous stud- including fibroblasts, endothelial cells, epithelial cells, monocytes/ ies used cell types that do not support HCMV infection, such as macrophages, smooth muscle cells, stromal cells, neuronal cells, HEK and CHO indicator lines and murine fibroblasts. In this study, and hepatocytes (4, 5), and each of these cell types may express a we extend our studies into HCMV permissive human fibroblasts, unique subset of TLRs and respond differently to HCMV infection. which allows for a greater appreciation for the role that TLR2 Thus, it is possible that the cell type infected and the different combinations of TLRs activated may have a profound influence on plays in HCMV biology. Although human fibroblasts have been the outcome of infection. Experiments addressing the role of mul- shown previously to respond to TLR2 ligands, we demonstrate in tiple TLRs simultaneously will provide valuable insights into how this study that these cells express TLR2 mRNA (71, 72). Using each TLR influences the global immune response to herpesviruses. TLR2 function-blocking Abs and dominant-negative TLR con- HCMV, like all herpesviruses, establishes a lifelong association structs, we show that TLR2 mediates NF-␬B activation and in- with the host as a latent infection. To accomplish this goal, HCMV flammatory cytokine responses to HCMV in these cells. Together, maintains a particularly close relationship with the host immune these data add to the growing body of evidence suggesting that system and employs multiple immune modulation strategies that HCMV can activate innate immunity during binding and entry into allow it to avoid detection by the host and persist in the face of a host cells. We propose that envelope gp gB and gH, already well potent immune response (3). Because of this close relationship, it appreciated for their roles as mediators of virus entry, also interact is tempting to speculate that HCMV may have adapted to use TLR directly with TLR2 and TLR1 during entry to initiate a signaling responses to its advantage. HCMV disseminates in neutrophils and cascade that results in the activation of NF-␬B and secretion of monocytes, and CD14-positive cells are hypothesized as a reser- inflammatory cytokines. voir for latent virus (88). Each of these cell types are either acti- HCMV gB and gH brings the number of viral envelope gp that vated by TLRs or are subject to recruitment by the mixture of are detected by TLRs to five (39–42). Notably, a shared feature of cytokines and chemokines that result from TLR activation. Thus, all of these gp is that they play critical roles in the binding and HCMV may have evolved to use TLR responses as a means of entry of their respective viruses (43–47, 60, 73–79). Viral enve- by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. recruit its dissemination and latency vehicles to the site of infec- lope proteins are a compelling target for the TLR system, as they tion, where these cells could then become infected. Further exam- are the first component of the virus to come into contact with the ination of the role that TLRs play at both the cellular and organ- cell. Consequently, detection of viral envelope gp would allow the ismal levels may provide further clues toward understanding the cell to set the innate response in motion at the earliest stages of complex relationship between herpesviruses and their hosts. infection, perhaps even before the virus entering the cell. The rapid recognition and response could provide a temporal advantage for Acknowledgments the host immune response, which would be extremely beneficial We thank Bryan Williams (Cleveland Clinic Foundation) for the TLR1, for combating a viral infection. TLR2, TLR4, and TLR6 plasmids; and Yoshi Kawaoka and Bill Sugden

http://classic.jimmunol.org Activation of TLRs by envelope gp also suggests that the pro- (both from the University of Wisconsin, Madison, WI) for the VSV-G cesses of virus entry and innate immune activation are coordinated. construct and retroviral vectors, respectively. We also thank members of The viral envelope is studded with numerous copies of each gp, the Compton laboratory for critical review of the manuscript. and each copy is able to interact with one or more cellular recep- Disclosures tors. Multiple interactions between viral envelope proteins and dif- The authors have no financial conflict of interest. ferent types of cellular receptors may induce the formation of an

Downloaded from organized structure reminiscent of the immunological synapse References (80). This type of receptor clustering would allow the cell to syn- 1. Pass, R. F. 2001. Cytomegalovirus. In Fields Virology. D. M. Knipe chronize innate immune activation with the process of viral entry. and P. M. Howley, eds. Lippincott-Raven, Philadelphia, pp. 2675–2706. 2. Sissons, J. G., M. Bain, and M. R. Wills. 2002. Latency and reactivation of Cellular integrins have been identified as receptors for HCMV gB human cytomegalovirus. J. Infect. 44: 73–77. and gH (46, 47) and have also been linked to TLRs (81, 82). It is 3. Soderberg-Naucler, C., and J. Y. Nelson. 1999. Human cytomegalovirus latency possible that HCMV binding to integrins could facilitate interac- and reactivation: a delicate balance between the virus and its host’s immune system. Intervirology 42: 314–321. tion with TLR2/TLR1 heterodimers. Furthermore, receptor clus- 4. Ibanez, C. E., R. Schrier, P. Ghazal, C. Wiley, and J. A. Nelson. 1991. Human tering may provide a mechanism by with integrin and TLR sig- cytomegalovirus productively infects primary differentiated macrophages. J. Vi- rol. 65: 6581–6588. naling can be coordinated. Fig. 7 indicates that HCMV-mediated 5. Sinzger, C., M. Kahl, K. Laib, K. Klingel, P. Rieger, B. Plachter, and G. Jahn. NF-␬B activation is only partially attributable to TLR2. As NF-␬B 2000. Tropism of human cytomegalovirus for endothelial cells is determined by activation is also a downstream consequence of integrin usage, it a post-entry step dependent on efficient translocation to the nucleus. J. Gen. Virol. 81: 3021–3035. is possible that TLR2 and integrins both contribute to the activa- 6. Alford, C. A., and W. Britt. 1996. Cytomegalovirus. In Fields Virology. tion of NF-␬B upon HCMV infection (83). However, it remains to B. N. Fields, D. M. Knipe, and P. M. Howley, eds. Lippincott-Raven, Philadel- ␬ phia, pp. 2493–2523. be determined whether NF- B activation by TLR2 and integrins is 7. Ljungman, P. 1996. Cytomegalovirus infections in transplant patients. Scand coordinated or coincidental. J. Infect. Dis. 100(Suppl.): 59–63. The Journal of Immunology 7101

8. Razonable, R. R., and C. V. Paya. 2003. Herpesvirus infections in transplant TLR9 signals after translocating from the ER to CpG DNA in the lysosome. Nat. recipients: current challenges in the clinical management of cytomegalovirus and Immunol. 5: 190–198. Epstein-Barr virus infections. Herpes 10: 60–65. 38. Nishiya, T., and A. L. DeFranco. 2004. Ligand-regulated chimeric receptor ap- 9. Ramsay, M. E., E. Miller, and C. S. Peckham. 1991. Outcome of confirmed proach reveals distinctive subcellular localization and signaling properties of the symptomatic congenital cytomegalovirus infection. Arch. Dis. Child. 66: Toll-like receptors. J. Biol. Chem. 279: 19008–19017. 1068–1069. 39. Kurt-Jones, E. A., L. Popova, L. Kwinn, L. M. Haynes, L. P. Jones, R. A. Tripp, 10. Pass, R. F., S. Stagno, G. J. Myers, and C. A. Alford. 1980. Outcome of symp- E. E. Walsh, M. W. Freeman, D. T. Golenbock, L. J. Anderson, and tomatic congenital cytomegalovirus infection: results of long-term longitudinal R. W. Finberg. 2000. Pattern recognition receptors TLR4 and CD14 mediate follow-up. Pediatrics 66: 758–762. response to respiratory syncytial virus. Nat. Immunol. 1: 398–401. 11. Streblow, D. N., S. L. Orloff, and J. A. Nelson. 2001. Do pathogens accelerate 40. Rassa, J. C., J. L. Meyers, Y. Zhang, R. Kudaravalli, and S. R. Ross. 2002. atherosclerosis? J. Nutr. 131: 2798S–2804S. Murine retroviruses activate B cells via interaction with Toll-like receptor 4. 12. Melnick, J. L., C. Hu, J. Burek, E. Adam, and M. E. DeBakey. 1994. Cytomeg- Proc. Natl. Acad. Sci. USA 99: 2281–2286. alovirus DNA in arterial walls of patients with atherosclerosis. J. Med. Virol. 42: 41. Bieback, K., E. Lien, I. M. Klagge, E. Avota, J. Schneider-Schaulies, W. P. 170–174. Duprex, H. Wagner, C. J. Kirschning, V. Ter Meulen, and S. Schneider-Schaulies. 13. Melnick, J. L., E. Adam, and M. E. DeBakey. 1995. Cytomegalovirus and ath- 2002. Hemagglutinin protein of wild-type measles virus activates Toll-like erosclerosis. BioEssays 17: 899–903. receptor 2 signaling. J. Virol. 76: 8729–8736. 14. Hendrix, M. G., M. M. Salimans, C. P. van Boven, and C. A. Bruggeman. 1990. 42. Jude, B. A., Y. Pobezinskaya, J. Bishop, S. Parke, R. M. Medzhitov, High prevalence of latently present cytomegalovirus in arterial walls of patients A. V. Chervonsky, and T. V. Golovkina. 2003. Subversion of the innate immune suffering from grade III atherosclerosis. Am. J. Pathol. 136: 23–28. system by a retrovirus. Nat. Immunol. 4: 573–578. 15. Faulds, D., and R. C. Heel. 1990. Ganciclovir: a review of its antiviral activity, 43. Hernandez, L. D., L. R. Hoffman, T. G. Wolfsberg, and J. M. White. 1996. pharmacokinetic properties and therapeutic efficacy in cytomegalovirus infec- Virus-cell and cell-cell fusion. Annu. Rev. Cell Dev. Biol. 12: 627–661. tions. Drugs 39: 597–638. 44. McClure, M. O., M. A. Sommerfelt, M. Marsh, and R. A. Weiss. 1990. The pH 16. Chrisp, P., and S. P. Clissold. 1991. Foscarnet: a review of its antiviral activity, independence of mammalian retrovirus infection. J. Gen. Virol. 71: 767–773. pharmacokinetic properties, and therapeutic use in immunocompromised patients 45. Boyle, K. A., and T. Compton. 1998. Receptor-binding properties of a soluble with cytomegalovirus retinitis. Drugs 41: 104–129. form of human cytomegalovirus glycoprotein B. J. Virol. 72: 1826–1833. 17. Compton, T., E. A. Kurt-Jones, K. W. Boehme, J. Belko, E. Latz, D. T. 46. Feire, A. L., H. Koss, and T. Compton. 2004. Cellular integrins function as entry Golenbock, and R. W. Finberg. 2003. Human cytomegalovirus activates inflam- receptors for human cytomegalovirus via a highly conserved disintegrin-like do- matory cytokine responses via CD14 and Toll-Like receptor 2. J. Virol. 77: main. Proc. Natl. Acad. Sci. USA 101: 15470–15475. ␣ ␤ 4588–4596. 47. Wang, X., D. Y. Huang, S. M. Huong, and E. S. Huang. 2005. Integrin v 3 is 18. Boehme, K. W., and T. Compton. 2004. Innate sensing of viruses by Toll-like a coreceptor for human cytomegalovirus. Nat. Med. 11: 515–521. receptors. J. Virol. 78: 7867–7873. 48. Spear, P. G., and R. Longnecker. 2003. Herpesvirus entry: an update. J. Virol. 77: 19. Takeda, K., T. Kaisho, and S. Akira. 2003. Toll-like receptors. Annu. Rev. Im- 10179–10185. munol. 21: 335–376. 49. Yurochko, A. D., and E. S. Huang. 1999. Human cytomegalovirus binding to 20. Akira, S., and K. Takeda. 2004. Toll-like receptor signalling. Nat. Rev. Immunol. human monocytes induces immunoregulatory gene expression. J. Immunol. 162: 4: 499–511. 4806–4816. 50. Yurochko, A. D., E. S. Hwang, L. Rasmussen, S. Keay, L. Pereira, and 21. Hertzog, P. J., L. A. O’Neill, and J. A. Hamilton. 2003. The interferon in TLR E. S. Huang. 1997. The human cytomegalovirus UL55 (gB) and UL75 (gH) signaling: more than just antiviral. Trends Immunol. 24: 534–539. glycoprotein ligands initiate the rapid activation of Sp1 and NF-␬B during in- 22. Wagner, H. 2004. The immunobiology of the TLR9 subfamily. Trends Immunol. fection. J. Virol. 71: 5051–5059. 25: 381–386. 51. Boyle, K. A., R. L. Pietropaolo, and T. Compton. 1999. Engagement of the 23. Schwandner, R., R. Dziarski, H. Wesche, M. Rothe, and C. J. Kirschning. 1999. cellular receptor for glycoprotein B of human cytomegalovirus activates the in- Peptidoglycan- and lipoteichoic acid-induced cell activation is mediated by Toll- terferon-responsive pathway. Mol. Cell. Biol. 19: 3607–3613. like receptor 2. J. Biol. Chem. 274: 17406–17409. 52. Boehme, K. W., J. Singh, S. T. Perry, and T. Compton. 2004. Human cytomeg- 24. Yoshimura, A., E. Lien, R. R. Ingalls, E. Tuomanen, R. Dziarski, and alovirus elicits a coordinated cellular antiviral response via envelope glycoprotein D. Golenbock. 1999. Cutting edge: recognition of Gram-positive bacterial cell B. J. Virol. 78: 1202–1211. wall components by the innate immune system occurs via Toll-like receptor 2. 53. Simmen, K. A., J. Singh, B. G. Luukkonen, M. Lopper, A. Bittner, N. E. Miller, J. Immunol. 163: 1–5. M. R. Jackson, T. Compton, and K. Fruh. 2001. Global modulation of cellular by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. 25. Qureshi, S. T., L. Lariviere, G. Leveque, S. Clermont, K. J. Moore, P. Gros, and transcription by human cytomegalovirus is initiated by viral glycoprotein B. D. Malo. 1999. Endotoxin-tolerant mice have mutations in Toll-like receptor 4 Proc. Natl. Acad. Sci. USA 98: 7140–7145. (Tlr4). J. Exp. Med. 189: 615–625. 54. Hirschfeld, M., Y. Ma, J. H. Weis, S. N. Vogel, and J. J. Weis. 2000. Cutting 26. Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. Van Huffel, X. Du, D. Birdwell, edge: repurification of lipopolysaccharide eliminates signaling through both hu- E. Alejos, M. Silva, C. Galanos, et al. 1998. Defective LPS signaling in C3H/HeJ man and murine Toll-like receptor 2. J. Immunol. 165: 618–622. and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282: 2085–2088. 55. Compton, T., R. R. Nepomuceno, and D. M. Nowlin. 1992. Human cytomega- 27. Hemmi, H., O. Takeuchi, T. Kawai, T. Kaisho, S. Sato, H. Sanjo, M. Matsumoto, lovirus penetrates host cells by pH-independent fusion at the cell surface. Virol- K. Hoshino, H. Wagner, K. Takeda, and S. Akira. 2000. A Toll-like receptor ogy 191: 387–395. recognizes bacterial DNA. Nature 408: 740–745. 56. Huber, M. T., and T. Compton. 1997. Characterization of a novel third member 28. Bauer, S., C. J. Kirschning, H. Hacker, V. Redecke, S. Hausmann, S. Akira, of the human cytomegalovirus glycoprotein H-glycoprotein L complex. J. Virol. H. Wagner, and G. B. Lipford. 2001. Human TLR9 confers responsiveness to 71: 5391–5398. bacterial DNA via species-specific CpG motif recognition. Proc. Natl. Acad. Sci. 57. Theiler, R. N., and T. Compton. 2002. Distinct glycoprotein O complexes arise http://classic.jimmunol.org USA 98: 9237–9242. in a post-Golgi compartment of cytomegalovirus-infected cells. J. Virol. 76: 29. Diebold, S. S., T. Kaisho, H. Hemmi, S. Akira, and E. S. C. Reis. 2004. Innate 2890–2898. antiviral responses by means of TLR7-mediated recognition of single-stranded 58. Compton, T. 1993. An immortalized human fibroblast cell line is permissive for RNA. Science 303: 1529–1531. human cytomegalovirus infection. J. Virol. 67: 3644–3648. 30. Heil, F., H. Hemmi, H. Hochrein, F. Ampenberger, C. Kirschning, S. Akira, 59. Kinzler, E. R., and T. Compton. 2005. Characterization of human cytomegalo- G. Lipford, H. Wagner, and S. Bauer. 2004. Species-specific recognition of sin- virus glycoprotein-induced cell-cell fusion. J. Virol. 79: 7827–7837. gle-stranded RNA via Toll-like receptor 7 and 8. Science 303: 1526–1529. 60. Britt, W. J. 1984. Neutralizing antibodies detect a disulfide-linked glycoprotein 31. Lund, J. M., L. Alexopoulou, A. Sato, M. Karow, N. C. Adams, N. W. Gale, complex within the envelope of human cytomegalovirus. Virology 135: 369–378.

Downloaded from A. Iwasaki, and R. A. Flavell. 2004. Recognition of single-stranded RNA viruses 61. Urban, M., M. Klein, W. J. Britt, E. Hassfurther, and M. Mach. 1996. Glyco- by Toll-like receptor 7. Proc. Natl. Acad. Sci. USA 101: 5598–5603. protein H of human cytomegalovirus is a major antigen for the neutralizing hu- 32. Lund, J., A. Sato, S. Akira, R. Medzhitov, and A. Iwasaki. 2003. Toll-like re- moral immune response. J. Gen. Virol. 77: 1537–1547. ceptor 9-mediated recognition of herpes simplex virus-2 by plasmacytoid den- 62. Kinzler, E. R., R. N. Theiler, and T. Compton. 2002. Expression and reconsti- dritic cells. J. Exp. Med. 198: 513–520. tution of the gH/gL/gO complex of human cytomegalovirus. J. Clin. Virol. 33. Tabeta, K., P. Georgel, E. Janssen, X. Du, K. Hoebe, K. Crozat, S. Mudd, 25(Suppl. 2): S87–S95. L. Shamel, S. Sovath, J. Goode, et al. 2004. Toll-like receptors 9 and 3 as es- 63. Theiler, R. N., and T. Compton. 2001. Characterization of the signal peptide sential components of innate immune defense against mouse cytomegalovirus processing and membrane association of human cytomegalovirus glycoprotein O. infection. Proc. Natl. Acad. Sci. USA 101: 3516–3521. J. Biol. Chem. 276: 39226–39231. 34. Ahmad-Nejad, P., H. Hacker, M. Rutz, S. Bauer, R. M. Vabulas, and H. Wagner. 64. Bell, J. K., G. E. Mullen, C. A. Leifer, A. Mazzoni, D. R. Davies, and 2002. Bacterial CpG-DNA and lipopolysaccharides activate Toll-like receptors at D. M. Segal. 2003. Leucine-rich repeats and pathogen recognition in Toll-like distinct cellular compartments. Eur. J. Immunol. 32: 1958–1968. receptors. Trends Immunol. 24: 528–533. 35. Jurk, M., F. Heil, J. Vollmer, C. Schetter, A. M. Krieg, H. Wagner, G. Lipford, 65. Lien, E., T. J. Sellati, A. Yoshimura, T. H. Flo, G. Rawadi, R. W. Finberg, and S. Bauer. 2002. Human TLR7 or TLR8 independently confer responsiveness J. D. Carroll, T. Espevik, R. R. Ingalls, J. D. Radolf, and D. T. Golenbock. 1999. to the antiviral compound R-848. Nat Immunol. 3: 499. Toll-like receptor 2 functions as a pattern recognition receptor for diverse bac- 36. Lee, H. K., S. Dunzendorfer, and P. S. Tobias. 2004. Cytoplasmic domain-me- terial products. J. Biol. Chem. 274: 33419–33425. diated dimerizations of Toll-like receptor 4 observed by ␤-lactamase enzyme 66. Hajjar, A. M., D. S. O’Mahony, A. Ozinsky, D. M. Underhill, A. Aderem, fragment complementation. J. Biol. Chem. 279: 10564–10574. S. J. Klebanoff, and C. B. Wilson. 2001. Cutting edge: functional interactions 37. Latz, E., A. Schoenemeyer, A. Visintin, K. A. Fitzgerald, B. G. Monks, between Toll-like receptor (TLR) 2 and TLR1 or TLR6 in response to phenol- C. F. Knetter, E. Lien, N. J. Nilsen, T. Espevik, and D. T. Golenbock. 2004. soluble modulin. J. Immunol. 166: 15–19. 7102 HCMV ENVELOPE gp ACTIVATE TLR2

67. Akira, S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innate 78. Keay, S., and B. Baldwin. 1991. Anti-idiotype antibodies that mimic gp86 of immunity. Cell 124: 783–801. human cytomegalovirus inhibit viral fusion but not attachment. J. Virol. 65: 68. Yurochko, A. D., M. W. Mayo, E. E. Poma, A. S. Baldwin, Jr., and E. S. Huang. 5124–5128. 1997. Induction of the transcription factor Sp1 during human cytomegalovirus 79. Bold, S., M. Ohlin, W. Garten, and K. Radsak. 1996. Structural domains involved infection mediates upregulation of the p65 and p105/p50 NF-␬B promoters. J. Vi- in human cytomegalovirus glycoprotein B-mediated cell-cell fusion. J. Gen. Vi- rol. 71: 4638–4648. rol. 77: 2297–2302. 69. Yurochko, A. D., T. F. Kowalik, S. M. Huong, and E. S. Huang. 1995. Human 80. Kinashi, T. 2005. Intracellular signalling controlling integrin activation in lym- cytomegalovirus upregulates NF-␬B activity by transactivating the NF-␬B p105/ phocytes. Nat. Rev. Immunol. 5: 546–559. p50 and p65 promoters. J. Virol. 69: 5391–5400. 81. Ogawa, T., Y. Asai, M. Hashimoto, and H. Uchida. 2002. Bacterial fimbriae ␬ ␬ 70. Baldwin, A. S., Jr. 1996. The NF- B and I B proteins: new discoveries and activate human peripheral blood monocytes utilizing TLR2, CD14, and CD11a/ insights. Annu. Rev. Immunol. 14: 649–683. CD18 as cellular receptors. Eur. J. Immunol. 32: 2543–2550. 71. Hatakeyama, J., R. Tamai, A. Sugiyama, S. Akashi, S. Sugawara, and H. Takada. 82. Triantafilou, M., and K. Triantafilou. 2002. Lipopolysaccharide recognition: 2003. Contrasting responses of human gingival and periodontal ligament fibro- CD14, TLRs, and the LPS-activation cluster. Trends Immunol. 23: 301–304. blasts to bacterial cell-surface components through the CD14/Toll-like receptor 83. Guo, W., and F. G. Giancotti. 2004. Integrin signalling during tumor progression. system. Oral Microbiol. Immunol. 18: 14–23. Nat. Rev. Mol. Cell Biol. 5: 816–826. 72. Tabeta, K., K. Yamazaki, S. Akashi, K. Miyake, H. Kumada, T. Umemoto, and H. Yoshie. 2000. Toll-like receptors confer responsiveness to lipopolysaccharide 84. Krug, A., G. D. Luker, W. Barchet, D. A. Leib, S. Akira, and M. Colonna. 2004. from Porphyromonas gingivalis in human gingival fibroblasts. Infect. Immun. 68: Herpes simplex virus type 1 activates murine natural interferon-producing cells 3731–3735. through Toll-like receptor 9. Blood 103: 1433–1437. 73. Kari, B., R. Radeke, and R. Gehrz. 1992. Processing of human cytomegalovirus 85. Krug, A., A. R. French, W. Barchet, J. A. Fischer, A. Dzionek, J. T. Pingel, envelope glycoproteins in and egress of cytomegalovirus from human astrocy- M. M. Orihuela, S. Akira, W. M. Yokoyama, and M. Colonna. 2004. TLR9- toma cells. J. Gen. Virol. 73: 253–260. dependent recognition of MCMV by IPC and DC generates coordinated cytokine 74. Lopper, M., and T. Compton. 2004. Coiled-coil domains in glycoproteins B and responses that activate antiviral NK cell function. Immunity 21: 107–119. H are involved in human cytomegalovirus membrane fusion. J. Virol. 78: 86. Kurt-Jones, E. A., M. Chan, S. Zhou, J. Wang, G. Reed, R. Bronson, 8333–8341. M. M. Arnold, D. M. Knipe, and R. W. Finberg. 2004. Herpes simplex virus 1 75. Tugizov, S., D. Navarro, P. Paz, Y. Wang, I. Qadri, and L. Pereira. 1994. Func- interaction with Toll-like receptor 2 contributes to lethal encephalitis. Proc. Natl. tion of human cytomegalovirus glycoprotein B: syncytium formation in cells Acad. Sci. USA 101: 1315–1320. constitutively expressing gB is blocked by virus-neutralizing antibodies. Virology 87. Wang, J. P., E. A. Kurt-Jones, O. S. Shin, M. D. Manchak, M. J. Levin, and 201: 263–276. R. W. Finberg. 2005. Varicella-zoster virus activates inflammatory cytokines in 76. Utz, U., W. Britt, L. Vugler, and M. Mach. 1989. Identification of a neutralizing human monocytes and macrophages via Toll-like receptor 2. J. Virol. 79: epitope on glycoprotein gp58 of human cytomegalovirus. J. Virol. 63: 12658–12666. 1995–2001. 88. Gerna, G., E. Percivalle, F. Baldanti, S. Sozzani, P. Lanzarini, E. Genini, 77. Wang, X., S. M. Huong, M. L. Chiu, N. Raab-Traub, and E. S. Huang. 2003. D. Lilleri, and M. G. Revello. 2000. Human cytomegalovirus replicates abor- Epidermal growth factor receptor is a cellular receptor for human cytomegalo- tively in polymorphonuclear leukocytes after transfer from infected endothelial virus. Nature 424: 456–461. cells via transient microfusion events. J. Virol. 74: 5629–5638. by guest on October 1, 2021. Copyright 2006 Pageant Media Ltd. http://classic.jimmunol.org Downloaded from