Human TLRs 10 and 1 Share Common Mechanisms of Innate Immune Sensing but Not Signaling

This information is current as Yue Guan, Diana Rose E. Ranoa, Song Jiang, Sarita K. of September 26, 2021. Mutha, Xinyan Li, Jerome Baudry and Richard I. Tapping J Immunol 2010; 184:5094-5103; Prepublished online 26 March 2010; doi: 10.4049/jimmunol.0901888

http://www.jimmunol.org/content/184/9/5094 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2010/03/25/jimmunol.090188 Material 8.DC1 http://www.jimmunol.org/ References This article cites 74 articles, 29 of which you can access for free at: http://www.jimmunol.org/content/184/9/5094.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 26, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts 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 © 2010 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Human TLRs 10 and 1 Share Common Mechanisms of Innate Immune Sensing but Not Signaling

Yue Guan,* Diana Rose E. Ranoa,* Song Jiang,* Sarita K. Mutha,† Xinyan Li,* Jerome Baudry,‡ and Richard I. Tapping*,x

TLRs are central receptors of the that drive host inflammation and adaptive immune responses in response to invading microbes. Among human TLRs, TLR10 is the only family member without a defined agonist or function. Phylogenetic analysis reveals that TLR10 is most related to TLR1 and TLR6, both of which mediate immune responses to a variety of microbial and fungal components in cooperation with TLR2. The generation and analysis of chimeric receptors containing the extracellular recognition domain of TLR10 and the intracellular signaling domain of TLR1, revealed that TLR10 senses triacylated lipopep- tides and a wide variety of other microbial-derived agonists shared by TLR1, but not TLR6. TLR10 requires TLR2 for innate immune recognition, and these receptors colocalize in the phagosome and physically interact in an agonist-dependent fashion. Downloaded from Computational modeling and mutational analysis of TLR10 showed preservation of the essential TLR2 dimer interface and lipopeptide-binding channel found in TLR1. Coimmunoprecipitation experiments indicate that, similar to TLR2/1, TLR2/10 complexes recruit the proximal adaptor MyD88 to the activated complex. However, TLR10, alone or in cooperation with TLR2, fails to activate typical TLR-induced signaling, including NF-kB–, IL-8–, or IFN-b–driven reporters. We conclude that human TLR10 cooperates with TLR2 in the sensing of microbes and fungi but possesses a signaling function distinct from that of other TLR2 subfamily members. The Journal of Immunology, 2010, 184: 5094–5103. http://www.jimmunol.org/

he innate immune system is the first line of defense in fense, monocytes/macrophages express most TLRs (6), whereas response to an invading pathogen. Cellular innate immune the expression of individual TLRs in dendritic cells depends upon T defenses are necessary for the direct killing of pathogens, their subtype (9, 10). The microbial agonists for TLRs include and they also mediate the immediate release of proinflammatory a wide variety of structures. For example, TLRs 3, 7, 8, and 9 mediators that enable immune cells to access the site of infection, collectively respond to nucleic acids from bacteria and viruses, as well as the responses of professional APCs essential for gen- and these intracellular family members induce the expression of erating effective adaptive immunity (1, 2). Primary triggers of type-1 IFNs that possess potent antiviral activities (11, 12). In by guest on September 26, 2021 inflammatory and adaptive responses are the TLRs, a family of contrast, TLRs 1, 2, 4, 5, and 6 sense outer membrane components cell surface innate immune sensors that exist in organisms ranging of bacteria, fungi, and protozoan organisms, and these receptors from lower invertebrates to higher mammals. Mammalian TLRs are expressed on the surface of mammalian cells. Although most alert the host to the presence of infection through direct recog- TLRs signal as homodimers, TLR2 requires TLR1 or TLR6 for nition of conserved structural components of viruses, bacteria, activity (13). TLR1/2 and TLR6/2 heterodimers discriminate fungi, and protozoans (3, 4). In addition to microbial and viral among different microbial products that include bacterial lip- products, TLRs sense molecules associated with damage to self- oproteins, bacterial lipoteichoic acids, nonenteric bacterial LPSs, tissues as well as host products of inflammation (5). fungal phospholipomannans, protozoan GPI anchors, and myco- Humans possess 10 TLR family members (1–10), subsets of bacterial lipomannans and phosphatidylinositols (3). The cell which are expressed in leukocytes and the epithelial cells of surface TLRs engage a core signaling pathway, leading to the mucosal surfaces (6–8). In accordance with their role in host de- activation of NF-kB, AP-1, and other transcription factors that drive the production of proinflammatory chemokines, , and cell-adhesion molecules (14). x *Department of Microbiology, †Department of Biochemistry, and College of Med- Despite extensive research on the TLRs, human TLR10 has icine, University of Illinois, Urbana, IL 61801; and ‡Center for Molecular Biophys- ics, Oak Ridge National Laboratory, Oak Ridge, TN 37831 remained an orphan receptor without a known agonist or function. TLR10 was initially cloned in 2001 and shares the greatest ho- Received for publication June 15, 2009. Accepted for publication February 18, 2010. mology with TLR1 and TLR6 (15). In mammals, TLR 10, 1, and 6 This work was supported by the National Institutes of Health, National Institute of Allergy and Infectious Diseases Grant AI052344 (to R.I.T.). are tandemly arranged and seem to have arisen from du- Address correspondence and reprint requests to Dr. Richard Tapping, Department of plication events. Phylogeny supports the idea that TLR10 arose Microbiology, University of Illinois, B103 CLSL MC110, 601 South Goodwin Av- before the duplication that generated TLR1 and TLR6 (16, enue, Urbana, IL 61801. E-mail address: [email protected] 17). Chickens possess two TLRs that share ancestral origins with The online version of this article contains supplemental material. the TLR1-6-10 cluster and two additional TLRs that share origins Abbreviations used in this paper: ECD, extracellular domain; HA, hemagglutinin; with TLR2 (18). The fact that these chicken TLRs mediate re- HKAL, heat-killed Acholeplasma laidlawii; HKSA, heat-killed Staphylococcus sponses to lipopeptides through cooperative interactions indicates aureus; LTA-SA, lipoteichoic acid from S. aureus; LRR, leucine-rich repeat; Malp-2, macrophage-activating lipopeptide-2; MEF, murine embryonic fibroblast; PamCSK4, that this recognition function predates the mammalian expansion S-[2,3-bis(hydroxy)-propyl]-(R)-cysteinyl-(lysyl)3-lysine; Pam3CSK4, N-palmitoyl- of the TLR1-6-10 cluster (19). The TLR2 subfamily is thought to S-[2,3-bis(palmitoyloxy)-propyl]-(R)-cysteinyl-(lysyl)3-lysine; PamCysPamSK4, N- palmitoly-S-[2-hydroxy-3-(palmitoyloxy)propyl]-(R)-cysteinyl-(lysyl)3-lysine; P.G. have evolved under positive and balancing selection and, among LPS, Porphyromonas gingivalis LPS; TIR, Toll/IL-1R; TM, transmembrane. primates, Toll/IL-1R (TIR) domains seem to have undergone www.jimmunol.org/cgi/doi/10.4049/jimmunol.0901888 The Journal of Immunology 5095 purifying selection (20, 21). The independent maintenance of veloped by Jin et al. (25, 26). In short, a truncated portion of the TLR TLR10 and its associated TIR domain suggest a distinct biological ectodomain lacking the C-terminal cap was fused to the highly conserved role for this receptor (16, 19). LRR C-terminal capping module of VLRB.61, a hagfish variable lym- phocyte receptor. With the aid of the computer software DNAWorks TLR10 is an unusual family member in that it is highly expressed (Bethesda, MD) (27), six adjacent sets of ∼60-bp overlapping oligonu- in the B cell lineage, suggesting that it plays a critical role in B cell cleotides were used to synthesize the sequence encoding aa 133–200 of the function. The presence of sequence gaps and retroviral insertions hagfish VLRB.61 clone. Using overlap extension PCR, the VLR coding 9 reveals that mouse TLR10 is a pseudogene, a situation that pre- region was fused to the 3 -end of TLR 1, 2, and 10 ECD, encompassing aa 22–476, 17–508, and 20–474, respectively. The PCR products were cloned cludes the generation and phenotypic assessment of a TLR10 as a BglII/NheI fragment into a modified pDisplay vector containing an knock-out mouse. In this article, we report that human TLR10 HA-tag upstream of the BglII site and an Fc domain of the human IgG1 shares a variety of agonists with TLR1, including a number of cell- downstream of the NheI site (kindly provided by Dr. David M. Kranz, surface components of bacteria and fungi. Similar to TLR1 and Department of Biochemistry, University of Illinois at Urbana Champaign). 9 TLR6, TLR10 requires TLR2 for recognition. However, we found A thrombin cleavage site (LVPRGS) was also added at the 3 -end of the TLRvlr hybrid to allow cleavage of the soluble TLR from the Fc fusion that TLR10 lacks the downstream signaling typically associated . A Flag-tagged TLR2vlr-Fc construct was also engineered. with other TLR2 family members. TLR10 mutants. Site-directed mutagenesis of the TLR10 mutants was performed in TLR10 chimeric receptor TLR10-1. Oligonucleotides Materials and Methods encoding amino acid changes were designed and used to introduce Ethics statement mutations into the TLR10 template through PCR. Veterinary care was provided by the clinical and technical staff of the Divi- TLR retroviral vectors. The retrovirus vector pMX-IRES-GFP (Cell sion of Animal Resources at the University of Illinois at Urbana Champaign, Biolabs, San Diego, CA) was modified by inserting the preprotrypsin Downloaded from which is an American Association for the Accreditation of Laboratory signal sequence, followed by a FLAG linker, into BamHI and NotI site to Animal Care-accredited facility. All animal experiments were approved by generate pMX-preprotrypsin-FLAG plasmid. The coding sequences of the University of Illinois Institutional Animal Care and Use Committee. TLR1, TLR10, and TLR10-1 were directly inserted into the modified retrovirus vector to generate pMX-TLR1, pMX-TLR10, and pMX- Reagents TLR10-1. Synthetic bacterial lipopeptides, N-palmitoyl-S-[2,3-bis(palmitoyloxy)- Cell culture stable cell lines propyl]-(R)-cysteinyl-(lysyl)3-lysine (Pam CSK ), S-[2,3-bis(hydroxy)-pro- http://www.jimmunol.org/ 3 4 The human colonic epithelial cell line SW620 was cultured in RPMI 1640 pyl]-(R)-cysteinyl-(lysyl)3-lysine (PamCSK4), N-palmitoly-S-[2-hydroxy-3- medium containing 10% (v/v) FBS and 2 mM L-glutamine. 293T cells were (palmitoyloxy)propyl]-(R)-cysteinyl-(lysyl)3-lysine (PamCysPamSK ), and 4 cultured in RPMI 1640 medium supplemented with 10% (v/v) FBS and 2 S-[2,3-bis(palmitoyloxy)-propyl]-(R)-cysteinyl-GNNDESNISFKEK (mac- mM L-glutamine. The human TLR10-expressing, stable RAW 264.7 cell rophage-activating lipopeptide-2 [MALP-2]) were purchased from EMC lines were obtained by nucleofection (Lonza, Basel, Swizerland) of HA- Microcollections (Tuebingen, Germany). The microbial-derived agonists, tagged TLR10 plasmid or an empty vector control into RAW 264.7 cells. LPS derived from Porphyromonas gingivalis, heat-killed Acholeplasma lai- After selecting in G418 (0.3 mg/ml) for 3 wk, stable batch cell lines ex- dlawii, lipoteichoic acid from Staphylococcus aureus, and lipomannan from pressing high levels of TLR10 were verified by RT-PCR and flow cytometry Mycobacteria smegmatis, were purchased from Invivogen (San Diego, CA). using anti-HA Ab. Cells were cultured at 37˚C in a humidified environment Zymosan particles and heat-killed S. aureus were purchased from Invitrogen containing 5% CO2.

Life Technologies (Carlsbad, CA). Mycobacteria membrane fractions were by guest on September 26, 2021 Stable HEK 293F cell lines expressing recombinant soluble forms of each received from National Institute of Allergy and Infectious Diseases Contract TLR–ECD–Fc fusion protein were generated by transfection of the rTLR N01 AI-75320 entitled “Tuberculosis Research Materials and Vaccine Test- plasmids using the cationic lipid transfection reagent 293fectin (Invitrogen ing.” The Escherichia coli type II heat-labile enterotoxin LT-IIa B subunits Life Technologies). Cells were then cultured under G418 selection (0.25 were from Dr. T. Connell (University at Buffalo, State University of New mg/ml) for 3–4 wk, and individual stable clones expressing high levels of York, Buffalo, NY). rTLR protein were isolated using limiting dilution cloning and selected by The monoclonal anti-FLAG Ab, HRP-conjugated anti-Flag (M2) mAb, Western blotting. Clones were cultured at 37˚C, 8% CO in serum-free and anti-hemagglutinin (HA) affinity gel were purchased from Sigma- 2 Freestyle 293F expression medium (Invitrogen Life Technologies). Aldrich (St. Louis, MO). HRP-conjugated anti-HA Ab was purchased from Miltenyi Biotec (Auburn, CA). The anti-human TLR1 Ab (clone GD2.F4) Transient transfection and dual-luciferase reporter assay and anti-human TLR2 Ab (clone T2.5) were obtained from eBioscience (San Diego, CA). The secondary Ab, biotin-conjugated donkey anti-mouse IgG SW620 cells or 293T cells were cotransfected with various TLR combi- (H+L chain), and a streptavidin-conjugated fluorophore tertiary Ab were nations along with an experimental promoter-driven firefly luciferase re- obtained from Jackson ImmunoResearch Laboratories (West Grove, PA). porter and a Renilla luciferase transfection control reporter. Transfection was mediated by using Fugene6 (Roche Applied Science, Indianapolis, IN) Plasmid constructs at a lipid/DNA ratio of 4:1. Two days after transfection, cells were stim- ulated with indicated agonists for $6 h, and cell lysates were collected. The primers for construction of the plasmids in this study are listed in Luciferase enzyme activities were measured using the Dual-Luciferase Supplemental Table I. All constructs were verified by complete sequencing reporter assay system (Promega). The values of firefly luciferase were first of both strands of all recombinant insertions. divided by the Renilla luciferase values to normalize the transfection ef- CD4–TLRs. The CD4–TLR4 construct was a kind gift from Dr. Charles ficiency among different wells. Janeway, Jr (Yale University, New Haven, CT). The coding region of Preparation of murine embryonic fibroblasts and retroviral mouse CD4 extracellular domain (ECD) was fused with that of the transmembrane (TM) and intracellular portions of TLR1, TLR2, or transduction TLR10. Both PCR products were cloned together into pCDNA3.1. TLR1-deficient mice were a kind gift of Dr. Shizou Akira (Osaka University, ENA-78 reporter. The promoter region (∼660 bp) of human ENA-78 gene Osaka, Japan). Murine embryonic fibroblasts (MEFs) were prepared from was amplified from human genomic DNA (22). The PCR product was then 13.5-d-old embryos and cultured in DMEM supplemented with 10% FBS. cloned into the pGL3 basic vector (Promega, Madison, WI). Cells at passage 3, 4, or 5 were used for the experiments. For retrovirus infection, the 293T Ampho packaging cell line was transfected with ret- TLR10 chimeric receptors and internal chimeras. TLR1-10, TLR10-1, and roviral vectors using Fugene 6 reagent (Roche Applied Science). The viral TLR10-6 were created by overlap extension PCR (23). TLR10 internal supernatant was harvested 48 h posttransfection and used to infect MEF chimera T10 (6–17)/T1 and T10 (6–17)/T6 were generated using the unique cells by incubating with MEF cells under 10 mg/ml polybrene (Sigma- TLR1 and TLR6 constructs made previously, in which unique restriction Aldrich) for 24 h. After culturing cells in fresh DMEM for another 24 h, sites were engineered at the end of leucine-rich repeat (LRR)5 (24). The the infected MEFs were collected and plated in 96-well plates at a density LRR6–17 region of TLR10 was amplified with appropriate enzyme sites at 4 of 1 3 10 cells per well and cultured for 12 h. For stimulation, cells were both ends and ligated into the digested products of TLR1 or TLR6. treated with increasing concentrations of Pam3CSK4 for 24 h. The con- TLR–ECD–Fc fusions. Pure soluble forms of TLR1, TLR2, and TLR10 centration of mouse IL-6 in the culture supernatant was measured by were produced in large quantities using the hybrid LRR technique de- ELISA using paired Abs (Invitrogen). 5096 INNATE IMMUNE SENSING OF HUMAN TLR10

Confocal microscopy Modeling of the TLR2/10/lipopeptide complex Stable cell lines of RAW 264.7 cells expressing HA-tagged TLR10 were The program MOE version 2006 (Chemical Computing Group, Montreal, grown on chambered microscope slides (Lab-Tek, Nalge Nunc, Rochester, Quebec, Canada) was used for all of the modeling work. The amino acid NY) and incubated with 2 3 106 zymosan particles/ml for 10 min. Cells sequence of the TLR1 ECD was aligned with the sequence of the TLR1 were fixed in 4% paraformaldehyde for 20 min and permeabilized in ace- sequence as in the TLR10/TLR2/lipopetide complex [National Center for tone for 5 min at 220˚C. Nonspecific sites were then blocked by incubating Bioprotein Information Protein Databank entry 2Z7X (26)], using the for 30 min at 4˚C in blocking buffer (PBS/10% rabbit serum/0.03% NaN3). BLOSUM62 scoring matrix as implemented in MOE; 10 models were HA-tagged TLR10 was detected by using mouse anti-HA mAb (Sigma- generated using the HOMOLOGY function provided in the MOE program. Aldrich), biotin-conjugated donkey anti-mouse IgG (Jackson ImmunoR- Each of these 10 models was submitted to an energy minimization using esearch Laboratories), and streptavidin-conjugated Alexa 555 (Molecular the CHARMM force field (28) version 27 as implemented with MOE, with Probes, Eugene, OR). Cells were then costained for mouse TLR2 using solvent effect implicitly described using distance-dependent dielectric and a directly conjugated TLR2.5-Alexa 488 Ab (eBioscience). All images a cutoff for nonbonded interactions between 8 and 10 A˚ . During the ho- were obtained with 403 objective and oil immersion using a Zeiss LSM510 mology modeling, the structure of TLR10 was modeled with the explicit (Carl Zeiss, Thornwood, NY) confocal microscope at the School of Mo- inclusion of TLR2 and lipopeptide structures lecular and Cellular Biology Imaging Facility (University of Illinois, Ur- bana, IL). Vector control cell lines did not exhibit staining for the HA tag. Results Immunoprecipitation and Western blot TLR10 signaling is distinct from that of other TLR2 subfamily HEK 293T cells were transfected in 10-cm tissue culture dishes with members specific TLRs along with MyD88 (0.75 mg HA-tagged TLR2, 2.25 mg AllTLRsarecharacterizedbyanN-terminalECD,composedofLRR FLAG-tagged TLR1 or TLR10, and 1 mg FLAG-tagged MyD88). Twenty-

motifs, followed by a single spanning TM domain and a C-terminal Downloaded from four hours posttransfection, half of the dishes were treated with Pam3CSK4 (200 ng/ml) for 10 min, and all cells were lysed using radio- intracellular TIR signaling domain. To assess the function of TLR10 immunoprecipitation assay buffer. Cell extracts were incubated with anti- withoutknowledgeoftheligand,theECDdomainofthereceptorwas HA affinity gel for 1 h at 4˚C, followed by extensive washes of the beads replaced with that of CD4, an approach that was used to generate with lysis buffer. Samples were separated on 7.5% PAGE and transferred to constitutively active forms of the TLRs (13, 29). Following trans- Hybond-P membrane (GE Healthcare, Piscataway, NJ). Western blotting was performed using HRP-conjugated anti-FLAG Ab (M2) or HRP-con- fection, a CD4–TLR10 fusion failed to induce reporter gene ex- jugated anti-HA Ab. pression from a variety of promoters, including NF-kB, IL-8, and http://www.jimmunol.org/ IFN-b (Fig. 1). In contrast, a CD4–TLR4 positive control activated Protein purification expression of all of these reporter constructs. Because TLR10 is 293F cell lines stably expressing TLR–ECD–Fc fusion were most highly related to TLRs 1 and 6, we hypothesized that, similar to 6 seeded at 0.3 3 10 cells/ml in serum-free medium and incubated with these receptors, it may be a heterodimeric partner for TLR2. As shaking for 5 d. Recombinant protein G Sepharose beads (GE Healthcare; expected, cotransfection of CD4–TLR1 and CD4–TLR2 induced 1 ml 50% slurry) were added to 1 l filtered culture supernatants and stirred k at 4˚C overnight. Protein G beads were recovered by centrifugation at 3000 activation of NF- B and IL-8–driven luciferase reporters. In con- 3 g for 15 min at 4˚C and were subsequently packed in a glass column trast, CD4–TLR10 in combination with CD4–TLR2 failed to acti- connected to the AKTA prime purification system (GE Healthcare). The vate any of the reporters (Fig. 1). Notably, cell-surface expression of beads were washed with 30 ml binding buffer (20 mM sodium phosphate all of the CD4–TLR chimeras was confirmed (Supplemental Fig. 1). by guest on September 26, 2021 buffer [pH 7]) at a flow rate of 1.0 ml/min, and the fusion protein was eluted with 0.1 M glycine-chloride (pH 2.3) in 1.0-ml fractions on tubes Taken together, these results suggest that CD4–TLR10, either alone containing 100 ml neutralizing buffer Tris-HCl (pH 9). The eluted protein or in combination with TLR2, fails to activate reporters commonly was dialyzed overnight against PBS (pH 7.4) at 4˚C. Proteins were con- associated with TLR signaling. centrated using an Amicon Ultra centrifugal filter device with 10,000 nominal m.w. limit (Millipore, Bedford, MA). The protein concentration TLR10 is a partner for TLR2 and shares a variety of agonists was measured using bicinchoninic acid assay. with TLR1 The TLRvlr–Fc fusion protein was then incubated with thrombin (No- vagen, Gibbstown, NJ) at 25˚C for 16–18 h to facilitate the removal of the TLR1 and TLR6 enable TLR2 to discriminate between different Fc tag, at an optimized concentration of 1 U thrombin per 0.25 mg fusion microbial products, including bacterial lipoproteins (30, 31). For protein. After thrombin cleavage, protein G beads were added to the instance, the TLR2/1 dimer mediates responses to the triacylated protein samples to remove Fc fragments, as well as uncut Fc-tagged lipopeptide Pam CSK , whereas the TLR2/6 dimer is required for TLRvlr in solution. After 2 h incubation at 4˚C, the slurry was passed 3 4 through a spin filter (Novagen) to separate the beads containing Fc frag- detection of diacylated lipopeptides, such as MALP-2. Functional ments and uncut proteins from the TLRvlr proteins. The purity of the hybrid proteins was determined by mass spectrometry (Mass Spectrometry Laboratory, School of Chemical Sciences, University of Illinois at Urbana Champaign) using MALDI as the ionization technique and sinapinic acid as a calibration matrix. Final protein concentration after thrombin cleavage was measured using the bicinchoninic acid assay. Receptor-binding assays HA-tagged TLR1vlr and TLR10vlr hybrid proteins (10 mg/ml in PBS [pH 7.4]) were coated in microtiter wells overnight at 4˚C. For all ELISAs, binding steps were performed at room temperature in MES buffer (pH 7.5), and wells were washed with PBS containing 0.05% Tween-20. After blocking with a commercially available blocking buffer (Pierce, Rockford, IL) for 2 h, wells were washed and incubated with an equimolar amount of Flag-tagged TLR2vlr that had been preincubated for 2 h at 25˚C with FIGURE 1. The TLR10 homodimer and TLR2/TLR10 heterodimer do various concentrations of the synthetic ligands Pam3CSK4, MALP-2, and not induce activation of NF-kB, IL-8, and IFN-b promoters. HEK 293 Ac2CSK4 in MES buffer (pH 7.5). One set of preincubation reactions cells were cotransfected with indicated CD4-TLR constructs (400 ng/ml), contained 20 mg/ml blocking anti-TLR2 Ab (T2.5). After washing, wells various luciferase reporters (150 ng/ml), and a Renilla luciferase trans- were incubated with HRP-conjugated anti-Flag (M2) mAb (Sigma-Al- drich). Detection was performed by the addition of an o-phenylenediamine fection control (50 ng/ml). Luciferase activities were measured 48 h tablet (Sigma-Aldrich) dissolved in 0.05 M phosphate-citrate buffer (pH 5) posttransfection. Values represent the level of constitutive reporter acti- containing 0.05% H2O2. The colorimetric signal was stopped with 4 N vation over that of vector alone, whose activity was taken as 1. The error H2SO4, and absorbance was read at 490 nm using an ELISA plate reader. bars represent the SD of three independent values. The Journal of Immunology 5097

analyses demonstrated that the receptor ECD is responsible for Pam3CSK4 supports the idea that signaling from this receptor lipopeptide recognition, and structural studies showed that the complex differs from that of related family members. Pam3CSK4 ligand is coordinately bound by TLR1 and TLR2 (24, To fully confirm the role of TLR10 in sensing microbial lip- 26, 32). Given the phylogenetic relationship among TLRs 1, 6, and opeptides, reconstitution experiments were performed using MEFs 10, we hypothesized that lipopeptides could also be agonists for the derived from TLR1-deficient animals. The TLR1-deficient MEFs latter receptor. To investigate this hypothesis, chimeric receptors were transduced with retroviral vectors expressing human TLR1, were generated in which the ECD and TM domains of human TLR10-1, and wild type TLR10 and then cultured in the presence of TLR10 and TLR1 were swapped (Fig. 2A). After confirming the varying concentrations of Pam3CSK4 (Fig. 2C). As reported pre- expression of the chimeric receptors on the cell surface (Supple- viously, TLR1-deficient cells exhibit very weak responses to mental Fig. 2), we assessed their ability to mediate cellular re- Pam3CSK4 (30). As expected, retroviral expression of human TLR1 sponses to lipopeptides using SW620 cells, a human epithelial cell enabled greater sensitivity and robust responses of MEFs to line in which the activities of TLR2 subfamily members can be Pam3CSK4. Importantly, TLR1-deficient MEFs, reconstituted with reconstituted because of the lack of endogenous expression (24). As chimeric TLR10-1 receptor, also exhibited sensitivity and responses expected, TLR1 enabled cells coexpressing TLR2 to respond to to Pam3CSK4 comparable to that of TLR1-transduced cells. These Pam3CSK4 (Fig. 2B). However, neither TLR10 nor the reciprocal results demonstrate that chimeric TLR10-1 rescues the responses of TLR1-10 chimera, comprising the TLR1 ECD and TM domain TLR1-deficient cells to triacylated lipopeptides. Interestingly, the fused to the TLR10 signaling domain, could reconstitute lipopeptide production of IL-6 in TLR1-deficient MEFs transduced with virus responses when coexpressed with TLR2 (Fig. 2B). Importantly, expressing wild type TLR10 exhibited slightly enhanced dose-de-

TLR10-1, a chimeric receptor in which the ECD and TM domain of pendent responses to Pam3CSK4 compared with an empty virus, Downloaded from TLR1 was replaced with that of TLR10, mediated full responses to suggesting that the cytoplasmic domain of TLR10 may weakly Pam3CSK4 in cooperation with TLR2. Taken together, these results activate IL-6 production in these cells. demonstrate that TLR10 cooperates with TLR2 in the sensing of To better define the agonist specificity of TLR10, cells coex- triacylated lipopeptides. The lack of response of TLR10/2 to pressing TLR10-1 and TLR2 were stimulated with synthetic lip- opeptide compounds and a variety of natural microbial agonists with

differing specificity toward TLR1/2 or TLR2/6. PamCysPamSK4 and http://www.jimmunol.org/ PamCSK4 are TLR2/1-specific agonists that are missing one or both lipids of the diacyl glycerol group of Pam3CSK4, respectively (24). In contrast, MALP-2 is a TLR2/6-specific agonist that possesses the diacyl glycerol group but lacks the third acyl chain on the N- terminal cysteine residue. In combination with TLR2, TLR10-1 by guest on September 26, 2021

FIGURE 2. TLR10 cooperates with TLR2 and senses synthetic lip- opeptides. A, Schematic diagrams of wild type or TLR1 and TLR10 chimeric constructs. B, SW620 cells were cotransfected with the chimeric constructs (180 ng/ml) along with TLR2 (20 ng/ml), an IL-8–driven luciferase reporter, FIGURE 3. TLR10 shares agonist specificity with TLR1. SW620 cells and a Renilla luciferase transfection control, as indicated. Two days after were cotransfected with indicated combinations of TLRs, an IL-8 promoter- transfection, cells were stimulated for 6 h with 20 ng/ml Pam3CSK4, fol- driven luciferase reporter, and Renilla transfection control reporter. Cells lowed by measurement of luciferase activity. Values represent the level of were stimulated for 6 h with 1 mg/ml PamCSK4, 20 ng/ml PamCysPamSK4, reporter activation over that of vector alone, whose activity was taken as 1. C, or 20 ng/ml Malp-2 (A)or107 particles/ml zymosan, 2 mg/ml lipoteichoic MEFs were derived from TLR1 knockout mice and transduced with retro- acid from S. aureus (LTA-SA), 20 ng/ml P. gingivalis LPS (P.G. LPS), 200 viruses encoding the indicated constructs. The cells were subsequently in- mg/ml heat-killed S. aureus (HKSA), or 107 particles/ml heat-killed A. lai- cubated with increasing concentrations of Pam3CSK4 for 12 h, and IL-6 dlawii (HKAL) (B), followed by luciferase activity assays. Values represent concentrations in the culture supernatant were measured by ELISA. Error the level of reporter activation over that of vector alone, whose activity was bars represent the SD of three independent values. taken as 1. Error bars represent the SD of three independent values. 5098 INNATE IMMUNE SENSING OF HUMAN TLR10 mediated responses to TLR1-specific, but not TLR6-specific, syn- thetic lipopeptides (Fig. 3A). Similarly, cellular responses to zy- mosan, atypical LPS, and heat-killed microbes were mediated by TLR2 in combination with TLR1 or TLR10-1. In contrast, cellular responses to lipoteichoic acid were enabled by TLR2 and TLR6, established receptor pairs for this agonist (Fig. 3B). Taken together, the data indicate that TLR10, coupled with TLR2, is able to detect a variety of microbial components and suggest that TLR10 shares agonist specificity with TLR1 but not with TLR6. A TLR10-6 chimeric receptor yielded identical results to that of TLR10-1, supporting the idea that agonist specificity is dictated by the ECDs of the receptors (data not shown). Coexpression of wild type TLR10 with TLR2 did not reconstitute responses to any of the TLR2 ag- onists, further suggesting a unique signaling function for this re- FIGURE 5. TLR10 interacts with TLR2, and the receptor complex recruits ceptor complex. MyD88. HEK 293T cells were cotransfected with indicated TLRs and FLAG- tagged MyD88. One day after transfection, cells were treated or not with TLR2 and TLR10 physically interact and colocalize in Pam3CSK4 at 200 ng/ml for 10 min. Cell lysates were immunoprecipitated phagosomes with anti-HA Ab, and samples were analyzed for the presence of TLR10, TLR1, and MyD88 using anti-FLAG Ab. Whole-cell lysates (WCE) were TLRs 1, 2, and 6 are recruited to phagosomes where they sense and Downloaded from mediate phagocyte responses to incoming microbial cargo (13, 33). analyzed to verify the expression of transfected TLRs and MyD88. To examine TLR10 trafficking during phagocytosis, HA-tagged human TLR10 was stably expressed in the mouse macrophage cell a hybrid LRR technique in which the ECD of each TLR was fused to line RAW264.7. In the absence of the stimulus, TLR10 and TLR2 a C-terminal LRR motif from hagfish (26). The physical association appear to be dispersed on the plasma membrane (Fig. 4). Upon of purified TLR2 with TLR10 ECDs was studied using microtiter incubation with zymosan particles, both receptors are highly en- plate assays: TLR10 was immobilized on the plate followed by the http://www.jimmunol.org/ riched and colocalize in early phagosomes. These results support addition of TLR2 alone or with various ligands. As shown in Fig. 6A, the findings that TLR2 and TLR10 cooperate in mediating re- TLR2 exhibited weak interaction with TLR10, which was greatly sponses to zymosan and show that both receptors localize to the enhanced by the addition of Pam3CSK4 in a dose-dependent fash- phagosome during the ingestion of whole microbes. ion. Addition of an antagonistic anti-TLR2 Ab T2.5 (37) dose-de- To determine whether TLR2 and TLR10 physically interact, pendently inhibited ligand-induced complex formation between coimmunoprecipitation studies were performed using HEK cells TLR2 and TLR10. In contrast, the nonstimulatory control com- transiently expressing affinity-tagged versions of different TLR pound Ac2CSK4, lacking the acyl- and amide-bound long-chain pairs. TLR10 was observed to coimmunoprecipitate with TLR2, and fatty acids of Pam3CSK4, did not enhance the physical association this apparent association between the receptors increased with the between TLR2 and TLR10, showing the requirement of lipid chains by guest on September 26, 2021 addition of Pam3CSK4 ligand (Fig. 5). As expected, a similar ligand- in the formation of a stable heterodimer. Moreover, incubation with enhanced association was observed between TLR2 and TLR1 (34– MALP-2, the diacylated lipopeptide agonist for TLR2/6, had no 36). In addition, the recruitment of the proximal signaling adaptor ability to induce a TLR2/10 complex. Collectively, our results MyD88 to the receptor complexes was assessed in these experi- demonstrated that the ECDs of TLR2 and TLR10 associate with ments. The coimmunoprecipitation of MyD88 with TLR2 alone was each other to form a stable complex in a ligand-dependent manner. barely detectable and was enhanced by the coexpression of TLR1 or As expected, identical results were obtained using TLR1 as a control TLR10 in HEK cells (Fig. 5). The addition of Pam3CSK4 ligand in place of TLR10 (Fig. 6B). greatly enhanced the apparent recruitment of MyD88 to TLR2/1 or TLR 2/10 heterodimers. These results suggest that, similar to TLR2/ Modeling of the TLR2/10/lipopeptide complex reveals structural 1 and TLR2/6 heterodimers, TLR2/10 signaling involves the re- similarity to the TLR2/1/lipopeptide complex cruitment of MyD88 to the activated receptor complex. Similar to TLR1 and TLR6, the ECD of TLR10 is composed of 19 To further investigate ligand-induced physical interactions be- sequentialLRRmotifs.Structurally,LRRsformasolenoidorspringin tween TLRs, soluble forms of TLRs 1, 2, and 10 were produced using which each LRR motif consists of a single turn with the leucines, or other appropriately spaced hydrophobic residues, packed within the interior (38, 39). The solved crystal structure of the TLR2/1/lip- opeptide complex reveals that the TLR2 and TLR1 solenoids form a heterodimer in which the lipid chains of Pam3CSK4 are coordinately bound (26). Two acyl chains of the diacyl glycerol group are ac- commodated by a hydrophobic pocket of TLR2, and the third amide- linked lipid chain occupies a hydrophobic channel in TLR1. The heterodimer is further stabilized by additional interactions at the di- mer interface (26). Because TLR1 and TLR10 share 43% amino acid identity at the ECDs, identical lipopeptide specificity, and TLR2 as a coreceptor, computational modeling was deemed to be a reasonable approach for gaining insights into the structure of the TLR2/10/lip- FIGURE 4. TLR2 and TLR10 colocalize in phagosomes. RAW264.7 cells opeptide complex. The computational model of TLR2/10/lipopeptide stably expressing HA-TLR10 were not treated or were incubated with zy- exhibits the same overall structure as that of the TLR1-containing mosan particles at 2 3 106/ml for 10 min before fixing. Cells were costained complex (Fig. 7A). Similar to TLR1, the model predicts a hydropho- for endogenous TLR2 (Alexa 488) and HA-tagged TLR10 (Alexa 555), bic channel on the convex surface of TLR10 that accommodates the 3 followed by confocal fluorescence microscopy. Original magnification 40. amide-linked lipid chain of the Pam3CSK4 lipopeptide. The Journal of Immunology 5099 Downloaded from

FIGURE 6. Binding of TLR2 to TLR10 or TLR1 is induced by

Pam3CSK4. HA-tagged TLR10 ECD (A) or HA-tagged TLR1 ECD (B) was

immobilized on microtiter wells, followed by incubation with Flag-tagged http://www.jimmunol.org/ FIGURE 7. Modeling of the TLR2/10/Pam CSK complex and func- TLR2 with increasing concentrations of Pam CSK (with or without T2.5 3 4 3 4 tional analyses of receptor chimeras reveal a role for the central LRRs of Ab), MALP-2, or Ac CSK , as indicated. The amount of protein binding was 2 4 TLR10 in Pam CSK -mediated activation. A, A computational model of detected through an HRP-conjugated anti-Flag Ab. Data are representative of 3 4 TLR10 (yellow) in the complex with TLR2 (blue) and Pam CSK (red), a least three independent replicates. 3 4 based upon the crystal structure of the TLR2/1/lipopeptide complex. B, SW620 cells were cotransfected with the indicated combinations of TLRs, Prior to the discovery of the TLR2/1/lipopeptide structure, an IL-8 promoter-driven luciferase reporter, and a Renilla luciferase con- trol. Cells were stimulated with 20 ng/ml Pam CSK or 20 ng/ml Malp-2, domain-swapping experiments between TLR1 and TLR6 suc- 3 4 followed by measurement of luciferase activity. Values represent the level cessfully defined the central LRRs of these receptors as the region

of reporter activation over that of vector alone, whose activity was taken as by guest on September 26, 2021 required for lipoprotein discrimination (24, 40). Because the 1. Error bars represent the SD of three independent values. ligand-binding and receptor dimerization region are predicted by the model to contain the central LRRs of TLR10, the effects of exchanging LRRs 6–17 of TLR1 and TLR6 with those of TLR10 mutants were expressed on the cell surface (Supplemental Fig. 3), were examined as a first approach for validating the model and and their ability to mediate cellular responses to Pam3CSK4 was defining the region of the ECD required for lipopeptide recogni- assessed. To provide a functional readout in the SW620 epithelial tion. As before, the ability of the chimeric receptors, in collabo- cell system, the TLR10 mutants were generated within the ration with TLR2, to mediate responses to different lipopeptides TLR10-1 chimeric receptor. Corresponding amino acid residues in was assessed in SW620 cells (Fig. 7B). As observed previously, in TLR1 were also assessed for comparative purposes. The model conjunction with TLR2, TLR1 and TLR10-1 enabled cellular re- predicts that similar to TLR1, a loop containing aa 311–316 sponses to Pam3CSK4, but not to MALP-2, whose activity was contributes to the dimer interface and the entrance of the hydro- dependent upon TLR6. When LRRs 6–17 of TLR1 were replaced phobic lipid-binding channel (Fig. 8B,8C). The orientation of the with the corresponding region of TLR10, [T10(6–17)/T1], the loop is important, especially because the backbone oxygen be- receptor retained lipopeptide specificity. Additionally, when LRRs tween aa 313 and 314 forms a hydrogen bond with the lipoprotein 6–17 of TLR6 were replaced with those from TLR10, the resulting (Fig. 8B). Phe314 of TLR1 and Ile314 of TLR10 are buried and chimeric receptor [T10(6–17)/T6] exhibited full activity toward make hydrophobic intramolecular contacts that seem to orient the Pam3CSK4 but a complete loss of activity toward MALP-2. These protein backbone of the loop. Substitution with a charged lysine results demonstrate that LRRs 6–17 of TLR10 are responsible for residue at position 314 almost completely abrogated the function defining the lipopeptide specificity of the receptor and support the of both receptors (Fig. 9). In TLR1 and TLR10, Val311 and molecular model of the TLR2/10/lipopeptide complex. Phe312 are positioned at the entrance of the channel; however substitution of either amino acid with a charged residue had little TLR10 and TLR1 interactions with TLR2 are similar but not effect on the activity of either receptor to Pam3CSK4. A nearly identical complete loss of activity was observed following mutation of The residues that contribute to the formation of the binding pocket Gln316, whose side chain in both receptors forms a hydrogen are conserved between TLR1 and TLR10 and preserve the overall bond with the amide oxygen of the Pam3CSK4 lipid chain. The shape of the lipid-binding channel (Fig. 8A,8B). Hydrogen bonds structural requirements for recognition of PamCysPamSK4 by between the peptide portion of Pam3CSK4 and TLR1 are also TLR1 or TLR10 are identical to that of Pam3CSK4, and, as ex- conserved in the TLR10 model (Fig. 8B). To further validate and pected, the various TLR1 and TLR10 mutants exhibited in- explore the structural model, site-directed mutagenesis was per- distinguishable responses to these two ligands (Fig. 9). formed on the residues in the ECD that seem to be crucial for TLR1 and TLR10 possess an extensive TLR2-binding interface ligand binding and TLR2 dimer formation. All of the receptor consisting of a hydrophobic core surrounded by hydrogen-bonding 5100 INNATE IMMUNE SENSING OF HUMAN TLR10

FIGURE 8. The predicted lip- opeptide-binding sites and receptor dimer interface of TLR2/TLR10 complex. A, Sequence alignment of LRRs 9–14 of TLR1 and TLR10. Consensus sequences are shown in red. The mutated sites in TLR1 or TLR10 are labeled with asterisks. B, The modeled lipid-binding channel in TLR10 and the main dimerization interface between TLR10 and TLR2. The side chains of TLR10 residues (gold) that interact with residues in TLR2 (blue) are shown along with carbon (gray), nitrogen (blue), oxy- gen (red), and sulfur (yellow) atoms of the lipopeptide. Hydrogen bonds are depicted as straight pink lines. Lipopeptide elements are shown in gray (carbon), blue (nitrogen), red Downloaded from (oxygen), and yellow (sulfur). All three-dimensional illustrations were produced with UCSF Chimera (41). C, Residues involved in the receptor dimer interface in the TLR2/10/lip- opeptide model or the TLR2/1/lip- http://www.jimmunol.org/ opeptide structure are shown, and interactions are depicted by dashed lines (26).

and ionic-interaction networks (26) (Fig. 8C). In both receptor However, this approach is not feasible for defining the function of complexes, the hydrophobic core overlaps with the entrance of the TLR10 because multiple gaps and insertions have rendered the re- lipid-binding channel; however, the surrounding polar and charged ceptor a pseudogene in mice, as evidenced by sequences from residues that make H-bonds and ionic interactions with TLR2 are a number of inbred strains (43). We found that Mus caroli, which by guest on September 26, 2021 less conserved between TLR1 and TLR10 (Fig. 8C). Hydrophobic emerged ∼4 million years ago and predates the Mus musculus group, interactions with TLR2 contributed by Pro315, Tyr320, Val339, also possesses a TLR10 pseudogene characterized by numerous and Leu359 of TLR1 are contributed by Leu342, Tyr320, Pro339, gaps and insertions (data not shown). The absence of TLR10 in and Ile359 of TLR10, respectively (Fig. 8C). Substitution of mice, coupled with a lack of understanding of TLR10 signaling, has Tyr320, Pro339, or Ile359 with polar or changed amino acids precluded the identification of synthetic or natural agonists and has dramatically reduced TLR10 activity toward Pam3CSK4 (Fig. 9A). rendered this receptor the only remaining orphan human TLR. Similarly, a Val339 mutant of TLR1 also exhibited greatly atten- We performed experiments using chimeric receptors to over- uated activity. In contrast to TLR1, the Tyr320 mutant of TLR10 come these obstacles based upon phylogenetic evidence that retained most of the receptor activity; this residue is predicted by TLR10 is most related to TLR1 and TLR6, both of which in- the model to make more intimate contact than TLR1 with the dependently cooperate with TLR2 in the sensing of a variety of hydrophobic face of TLR2. Pro315 resides at the center of the microbial and fungal components. Through this approach, we hydrophobic core of TLR1 and interacts with Tyr323 of TLR2 report that TLR10 shares a number of microbial agonists with (Fig. 8C). Substitution of Pro315 with leucine, which constitutes TLR1, but not TLR6, and uses TLR2 as a coreceptor. Notably, a naturally occurring polymorphic variant of TLR1, greatly at- expression of a chimeric receptor, which replaces the TLR1 ECD tenuates receptor activity, as reported previously (42). Conversely, with that of TLR10, fully reconstitutes responses of TLR1-deficient substitution of Glu315 of TLR10 with leucine, which is not pre- macrophages to triacylated lipopeptide. We also observed that dicted by the model to contribute to heterodimer formation, re- TLR10 and TLR2 colocalize in the early phagosome and that the tained more than half of the receptor activity (Fig. 9). In ECDs of the two receptors physically interact in a ligand-dependent conclusion, the structural model of the TLR2/10/lipopeptide manner. A computational model of the TLR2/10/lipopeptide complex is consistent with and supported by the mutational complex and mutagenesis studies reveal similarities, as well as studies. Although strikingly similar in overall structure, the TLR1- some subtle differences, in the ligand binding and dimer interface and TLR10-containing complexes seem to possess subtle differ- in comparison with the TLR2/1/lipopeptide crystal structure. ences at the TLR2 dimer interface. The idea that agonist recognition drives receptor dimer formation is supported by the finding that artificial chimeric receptors, which Discussion replace the TLR ECD with CD4 or integrin pairs that form natural TLRs play a central role in host defense by driving appropriate in- dimers, exhibit constitutive activation (13, 29, 44). We found that flammatory and adaptive responses following infection. The early expression of CD4–TLR10, alone or with CD4–TLR2, was unable identification of microbial agonists for the TLRs was greatly facil- to activate NF-kB–, IL-4–, IL-8–, ENA-78–, or IFN-b–driven itated by the systematic generation and phenotypic assessment of promoters in HEK293 or SW620 cells (Fig. 1, data not shown). knockout mice and provided the first insights into TLR function (3). These findings contrast with a report showing that CD4–TLR10 The Journal of Immunology 5101

Despite conservation, there are substantial sequence differences between TIR domains of TLR10 and TLR1. Most notably, TLR1 and TLR6 have identical RNFVPG BB loop sequences, whereas that of TLR10 is composed of SYFDPG and includes a two-amino acid insertionin the precedinga-helix. These and other differences within the TLR10 signaling domain may contribute to a distinct charge distribution in the TLR2/10 surface that confers a unique specificity for adapter binding and downstream signaling (54). We observed that the adaptor MyD88 coimmunoprecipitates with the TLR2/10 heterodimer and that this interaction is enhanced by lipopeptide agonist (Fig. 5). However, TLR2/10 fails to activate signals typically associated with MyD88 activity, including canonical activation of NF-kB, suggesting that recruitment of this adaptor to TLR2/10 is distinct from that of other TLRs, including the closely related TLR2/ 1 and TLR2/6 heterodimers. The idea that recruitment of MyD88 can differ among TLR2 heterodimers is not without precedence, because an I179N point mutation in this adaptor abolishes MyD88-de- pendent signaling of all TLR complexes, including TLR2/1, while

having no effect on MyD88-dependent TLR2/6 signaling (34). Downloaded from In addition to phylogenetic evidence, the unique expression pattern of TLR10 suggests a functional divergence from TLR1 and TLR6. In comparison with these latter two receptors, TLR10 ex- FIGURE 9. Mutation of TLR10 residues essential to TLR2/TLR10/lip- pression seems to be more restricted, with the RNA message found opeptide complex formation affects receptor function. SW620 cells were predominantly in lymphoid tissues, including the spleen, lymph

cotransfected with TLR2, various TLR10 mutants (A), or various TLR1 nodes, thymus, and tonsils (15, 54, 55). An analysis of isolated cell http://www.jimmunol.org/ mutants (B), as indicated, along with an IL-8 promoter-driven luciferase types revealed that high levels of TLR10 expression are restricted reporter and Renilla luciferase control. Cells were stimulated with 20 ng/ml to the B cell lineage, with weaker expression in plasmacytoid Pam3CSK4 or 20 ng/ml PamCysPamSK4 for 6 h, followed by luciferase dendritic cells (42, 56, 57). Although TLR expression levels are activity assays. Results are presented as the percentage of the wild type re- generally low in naive B cells, they are highly induced upon BCR ceptor response; error bars represent the SD of three independent values. stimulation (57, 58). A variety of TLR agonists was shown to act directly on B cells to induce Ab production to T-independent Ags activated NF-kB–, IL-4–, and ENA-78–driven promoters in HEK (59) and to enhance proliferation, isotype switching, and differ- cells (42). Despite our ability to appropriately activate these pro- entiation in response to T-dependent Ags (60–62). TLR7 and moters using CD4–TLR4 or a combination of CD4–TLR1 and TLR9 agonists are generally the focus of many of these studies, by guest on September 26, 2021 CD4–TLR2, we have been unable to detect TLR10-dependent given their potential usefulness as therapeutics (63) and their es- activation of any of the reporter constructs (Fig. 1). In support of tablished role in the autoimmune condition lupus (64). However, our work, an integrin–TLR10 chimera, alone or with an appropriate TLR2 agonists can also provide stimulatory signals to human B integrin–TLR2 partner, does not activate NF-kB, despite the fact cells (65–67). Memory B cells constitutively express a number of that other integrin–TLR combinations are able to do so (44). In TLRs, including TLR2 and TLR10, and can be induced by TLR addition, we found that TLR1, TLR6, or TLR10-1, but not TLR10, agonists to proliferate and differentiate to plasma cells (57, 58). induced NF-kB– or IL-8–driven promoters in response to lip- Concordantly, cell lines derived from mature B cell neoplasias opeptide when coexpressed with TLR2 (Fig. 2). express high levels of TLR10, and a variety of TLR ligands, in- TLR signaling is mediated by an intracellular TIR domain, so cluding lipopeptides, enhances the survival of multiple myeloma named because it shares homology with that of IL-1R (14, 45). The and chronic lymphocytic leukemia cells (57, 68, 69). appropriate association of two intracellular receptor TIR domains In an analogous manner, TLRs were shown to play a direct enables the recruitment of specific TIR domain-containing adaptor costimulatory role during T cell activation (70). Human T cells molecules that ultimately induce the expression of appropriate express high levels of TLR2 following T cell engagement, and immune response genes (14, 45–47). The crystal structure of the these cells produce elevated levels of in response to solved TIR domains of TLR1, TLR2, and TLR10 exhibit a con- lipopeptide (71). TLR2 is constitutively expressed in memory served fold consisting of five parallel b-sheets surrounded by five T cells, which exhibit enhanced responses to lipopeptides. Taken a-helical segments (48, 49). Multiple studies showed that two together with the observations in B cells, the results suggest that exposed segments, designated the BB loop and the DD loop, are TLR10 plays an ancillary role in lymphocyte stimulation, espe- critical for dimer formation and adaptor recruitment that drive cially where the expansion and controlled maintenance of memory downstream signaling (48, 50–52). The TLR10 TIR domain B and T cells are concerned. Further evidence that TLR10 plays crystallizes as a dimer with juxtaposed BB loops (49); however, a role in the control of adaptive immune responses is the finding the idea that this constitutes a physiologically relevant form of the that this receptor is expressed by regulatory T cells through receptor is based upon receptor activity experiments that we transcription factors Foxp3 and NF-AT (72). TLR10 genetic var- cannot reproduce (43). Instead of BB loop interactions, rigid-body iants have been associated with susceptibility to extrapulmonary protein–protein docking studies between the TIR domains of tuberculosis and asthma (73, 74), two diseases in which TLR2- TLR1 and TLR2 revealed a different orientation that involves an mediated adaptive immunity is thought to play a major role. interaction between the BB loop of one monomer and the DD loop The discovery that TLR10 is orthologous to TLR1 in ligand of another (52). In this configuration, His646, Asn700, Gly676, recognition, but not signaling, calls for a re-evaluation of studies in and Tyr737, which are perfectly conserved between TLR1 and which lipoproteins or other TLR2/1 agonists were used in the TLR10, make up the TIR domain dimer interface with TLR2 (53). stimulation of human and other nonmurine mammalian systems. 5102 INNATE IMMUNE SENSING OF HUMAN TLR10

In vivo, the competition of TLR1 and TLR10 for ligand and 23. Horton, R. M., Z. L. Cai, S. N. Ho, and L. R. Pease. 1990. Gene splicing by coreceptors, such as TLR2, is likely to have significant functional overlap extension: tailor-made genes using the polymerase chain reaction. Bio- techniques 8: 528–535. consequences, given the different expression patterns and signaling 24. Omueti, K. O., J. M. Beyer, C. M. Johnson, E. A. Lyle, and R. I. Tapping. 2005. outputs of these two receptor complexes. In addition to providing Domain exchange between human toll-like receptors 1 and 6 reveals a region a means to study TLR10 function, the discovery of lipopeptides and required for lipopeptide discrimination. J. Biol. Chem. 280: 36616–36625. 25. Kim, H. M., B. S. Park, J. I. Kim, S. E. Kim, J. Lee, S. C. Oh, P. Enkhbayar, other microbial agonists for this receptor has important implica- N. Matsushima, H. Lee, O. J. Yoo, and J. O. Lee. 2007. Crystal structure of the tions for the future therapeutic targeting of TLR2. TLR4-MD-2 complex with bound endotoxin antagonist Eritoran. Cell 130: 906– 917. 26. Jin, M. S., S. E. Kim, J. Y. Heo, M. E. Lee, H. M. Kim, S. G. Paik, H. Lee, and Acknowledgments J. O. Lee. 2007. Crystal structure of the TLR1-TLR2 heterodimer induced by We thank the Flow Cytometry Facility as well as the Division of Animal binding of a tri-acylated lipopeptide. Cell 130: 1071–1082. 27. Hoover, D. M., and J. Lubkowski. 2002. DNAWorks: an automated method for Resources of the University of Illinois for excellent service. designing oligonucleotides for PCR-based gene synthesis. Nucleic Acids Res. 30: e43. Disclosures 28. Brooks, B. R., C. L. Brooks 3rd, A. D. Mackerell, Jr., L. Nilsson, R. J. Petrella, B. Roux, Y. Won, G. Archontis, C. Bartels, S. Boresch, et al. 2009. CHARMM: The authors have no financial conflicts of interest. The biomolecular simulation program. J. Comput. Chem. 30: 1545–1614. 29. Medzhitov, R., P. Preston-Hurlburt, and C. A. Janeway, Jr. 1997. A human ho- mologue of the Drosophila Toll protein signals activation of adaptive immunity. References Nature 388: 394–397. 30. Takeuchi, O., S. Sato, T. Horiuchi, K. Hoshino, K. Takeda, Z. Dong, 1. Fearon, D. T. 1997. Seeking wisdom in innate immunity. Nature 388: 323–324. 2. Medzhitov, R., and C. A. Janeway, Jr. 1997. Innate immunity: impact on the R. L. Modlin, and S. Akira. 2002. Cutting edge: role of Toll-like receptor 1 in adaptive immune response. Curr. Opin. Immunol. 9: 4–9. mediating immune response to microbial lipoproteins. J. Immunol. 169: 10–14. Downloaded from 3. Akira, S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innate 31. Takeuchi, O., T. Kawai, P. F. Mu¨hlradt, M. Morr, J. D. Radolf, A. Zychlinsky, immunity. Cell 124: 783–801. K. Takeda, and S. Akira. 2001. Discrimination of bacterial lipoproteins by Toll- 4. Iwasaki, A., and R. Medzhitov. 2004. Toll-like receptor control of the adaptive like receptor 6. Int. Immunol. 13: 933–940. 32. Sandor, F., E. Latz, F. Re, L. Mandell, G. Repik, D. T. Golenbock, T. Espevik, immune responses. Nat. Immunol. 5: 987–995. 5. Zhang, Z., and H. J. Schluesener. 2006. Mammalian toll-like receptors: from E. A. Kurt-Jones, and R. W. Finberg. 2003. Importance of extra- and intracellular endogenous ligands to tissue regeneration. Cell. Mol. Life Sci. 63: 2901–2907. domains of TLR1 and TLR2 in NFkappa B signaling. J. Cell Biol. 162: 1099– 6. Muzio, M., D. Bosisio, N. Polentarutti, G. D’amico, A. Stoppacciaro, 1110. R. Mancinelli, C. van’t Veer, G. Penton-Rol, L. P. Ruco, P. Allavena, and 33. Underhill, D. M., A. Ozinsky, A. M. Hajjar, A. Stevens, C. B. Wilson, http://www.jimmunol.org/ M. Bassetti, and A. Aderem. 1999. The Toll-like receptor 2 is recruited to A. Mantovani. 2000. Differential expression and regulation of toll-like receptors macrophage phagosomes and discriminates between pathogens. Nature 401: (TLR) in human leukocytes: selective expression of TLR3 in dendritic cells. J. 811–815. Immunol. 164: 5998–6004. 34. Jiang, Z., P. Georgel, C. Li, J. Choe, K. Crozat, S. Rutschmann, X. Du, T. Bigby, 7. Zarember, K. A., and P. J. Godowski. 2002. Tissue expression of human Toll-like S. Mudd, S. Sovath, et al. 2006. Details of Toll-like receptor:adapter interaction receptors and differential regulation of Toll-like receptor mRNAs in leukocytes in revealed by germ-line mutagenesis. Proc. Natl. Acad. Sci. USA 103: 10961– response to microbes, their products, and cytokines. J. Immunol. 168: 554–561. 10966. 8. Faure, E., O. Equils, P. A. Sieling, L. Thomas, F. X. Zhang, C. J. Kirschning, 35. Tapping, R. I., and P. S. Tobias. 2003. Mycobacterial lipoarabinomannan me- N. Polentarutti, M. Muzio, and M. Arditi. 2000. Bacterial lipopolysaccharide diates physical interactions between TLR1 and TLR2 to induce signaling. J. activates NF-kappaB through toll-like receptor 4 (TLR-4) in cultured human Endotoxin Res. 9: 264–268. dermal endothelial cells. Differential expression of TLR-4 and TLR-2 in endo- 36. Johnson, C. M., E. A. Lyle, K. O. Omueti, V. A. Stepensky, O. Yegin, E. Alpsoy, thelial cells. J. Biol. Chem. 275: 11058–11063. L. Hamann, R. R. Schumann, and R. I. Tapping. 2007. Cutting edge: A common by guest on September 26, 2021 9. Kadowaki, N., S. Ho, S. Antonenko, R. W. Malefyt, R. A. Kastelein, F. Bazan, polymorphism impairs cell surface trafficking and functional responses of TLR1 and Y. J. Liu. 2001. Subsets of human dendritic cell precursors express different but protects against leprosy. J. Immunol. 178: 7520–7524. toll-like receptors and respond to different microbial antigens. J. Exp. Med. 194: 37. Meng, G., M. Rutz, M. Schiemann, J. Metzger, A. Grabiec, R. Schwandner, 863–869. P. B. Luppa, F. Ebel, D. H. Busch, S. Bauer, et al. 2004. Antagonistic antibody 10. Barton, G. M., and R. Medzhitov. 2002. Control of adaptive immune responses prevents toll-like receptor 2-driven lethal shock-like syndromes. J. Clin. Invest. by Toll-like receptors. Curr. Opin. Immunol. 14: 380–383. 113: 1473–1481. 11. Krieg, A. M., and J. Vollmer. 2007. Toll-like receptors 7, 8, and 9: linking innate 38. Jin, M. S., and J. O. Lee. 2008. Structures of the toll-like receptor family and its immunity to autoimmunity. Immunol. Rev. 220: 251–269. ligand complexes. Immunity 29: 182–191. 12. Uematsu, S., and S. Akira. 2007. Toll-like receptors and Type I interferons. J. 39. Bella, J., K. L. Hindle, P. A. McEwan, and S. C. Lovell. 2008. The leucine-rich Biol. Chem. 282: 15319–15323. repeat structure. Cell. Mol. Life Sci. 65: 2307–2333. 13. Ozinsky, A., D. M. Underhill, J. D. Fontenot, A. M. Hajjar, K. D. Smith, 40. Grabiec, A., G. Meng, S. Fichte, W. Bessler, H. Wagner, and C. J. Kirschning. C. B. Wilson, L. Schroeder, and A. Aderem. 2000. The repertoire for pattern 2004. Human but not murine toll-like receptor 2 discriminates between tri-pal- recognition of pathogens by the innate immune system is defined by cooperation mitoylated and tri-lauroylated peptides. J. Biol. Chem. 279: 48004–48012. between toll-like receptors. Proc. Natl. Acad. Sci. USA 97: 13766–13771. 41. Pettersen, E. F., T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, 14. O’Neill, L. A., and A. G. Bowie. 2007. The family of five: TIR-domain-con- E. C. Meng, and T. E. Ferrin. 2004. UCSF Chimera—a visualization system for taining adaptors in Toll-like receptor signalling. Nat. Rev. Immunol. 7: 353–364. exploratory research and analysis. J. Comput. Chem. 25: 1605–1612. 15. Chuang, T., and R. J. Ulevitch. 2001. Identification of hTLR10: a novel human 42. Omueti, K. O., D. J. Mazur, K. S. Thompson, E. A. Lyle, and R. I. Tapping. Toll-like receptor preferentially expressed in immune cells. Biochim. Biophys. 2007. The polymorphism P315L of human toll-like receptor 1 impairs innate Acta 1518: 157–161. immune sensing of microbial cell wall components. J. Immunol. 178: 6387– 16. Roach, J. C., G. Glusman, L. Rowen, A. Kaur, M. K. Purcell, K. D. Smith, 6394. L. E. Hood, and A. Aderem. 2005. The evolution of vertebrate Toll-like re- 43. Hasan, U., C. Chaffois, C. Gaillard, V. Saulnier, E. Merck, S. Tancredi, C. Guiet, ceptors. Proc. Natl. Acad. Sci. USA 102: 9577–9582. F. Brie`re, J. Vlach, S. Lebecque, et al. 2005. Human TLR10 is a functional 17. Zhou, H., J. Gu, S. J. Lamont, and X. Gu. 2007. Evolutionary analysis for receptor, expressed by B cells and plasmacytoid dendritic cells, which activates functional divergence of the toll-like receptor gene family and altered functional gene transcription through MyD88. J. Immunol. 174: 2942–2950. constraints. J. Mol. Evol. 65: 119–123. 44. Zhang, H., P. N. Tay, W. Cao, W. Li, and J. Lu. 2002. Integrin-nucleated Toll-like 18. Temperley, N. D., S. Berlin, I. R. Paton, D. K. Griffin, and D. W. Burt. 2008. receptor (TLR) dimerization reveals subcellular targeting of TLRs and distinct Evolution of the chicken Toll-like receptor gene family: a story of gene gain and mechanisms of TLR4 activation and signaling. FEBS Lett. 532: 171–176. gene loss. BMC Genomics 9: 62. 45. West, A. P., A. A. Koblansky, and S. Ghosh. 2006. Recognition and signaling by 19. Hughes, A. L., and H. Piontkivska. 2008. Functional diversification of the toll- toll-like receptors. Annu. Rev. Cell Dev. Biol. 22: 409–437. like receptor gene family. Immunogenetics 60: 249–256. 46. Sheedy, F. J., and L. A. O’Neill. 2007. The Troll in Toll: Mal and Tram as 20. Nakajima, T., H. Ohtani, Y. Satta, Y. Uno, H. Akari, T. Ishida, and A. Kimura. bridges for TLR2 and TLR4 signaling. J. Leukoc. Biol. 82: 196–203. 2008. Natural selection in the TLR-related genes in the course of primate evo- 47. Kawai, T., and S. Akira. 2007. Signaling to NF-kappaB by Toll-like receptors. lution. Immunogenetics 60: 727–735. Trends Mol. Med. 13: 460–469. 21. Ferrer-Admetlla, A., E. Bosch, M. Sikora, T. Marque`s-Bonet, A. Ramı´rez- 48. Xu, Y., X. Tao, B. Shen, T. Horng, R. Medzhitov, J. L. Manley, and L. Tong. Soriano, A. Muntasell, A. Navarro, R. Lazarus, F. Calafell, J. Bertranpetit, and 2000. Structural basis for by the Toll/interleukin-1 receptor F. Casals. 2008. Balancing selection is the main force shaping the evolution of domains. Nature 408: 111–115. innate immunity genes. J. Immunol. 181: 1315–1322. 49. Nyman, T., P. Stenmark, S. Flodin, I. Johansson, M. Hammarstro¨m, and 22. Corbett, M. S., I. Schmitt, O. Riess, and A. Walz. 1994. Characterization of the P. Nordlund. 2008. The crystal structure of the human toll-like receptor 10 cy- gene for human neutrophil-activating peptide 78 (ENA-78). Biochem. Biophys. toplasmic domain reveals a putative signaling dimer. J. Biol. Chem. 283: 11861– Res. Commun. 205: 612–617. 11865. The Journal of Immunology 5103

50. Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. Van Huffel, X. Du, D. Birdwell, 62. Ruprecht, C. R., and A. Lanzavecchia. 2006. Toll-like receptor stimulation as a third E. Alejos, M. Silva, C. Galanos, et al. 1998. Defective LPS signaling in C3H/HeJ signal required for activation of human naive B cells. Eur. J. Immunol. 36: 810–816. and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282: 2085–2088. 63. Kanzler, H., F. J. Barrat, E. M. Hessel, and R. L. Coffman. 2007. Therapeutic 51. Toshchakov, V. U., S. Basu, M. J. Fenton, and S. N. Vogel. 2005. Differential targeting of innate immunity with Toll-like receptor agonists and antagonists. involvement of BB loops of toll-IL-1 resistance (TIR) domain-containing Nat. Med. 13: 552–559. adapter proteins in TLR4- versus TLR2-mediated signal transduction. J. Im- 64. Marshak-Rothstein, A. 2006. Toll-like receptors in systemic autoimmune dis- munol. 175: 494–500. ease. Nat. Rev. Immunol. 6: 823–835. 52. Gautam, J. K., L. D. Ashish, J. K. Comeau, Krueger, and M. F. Smith, Jr. 2006. 65. Chiron, D., I. Bekeredjian-Ding, C. Pellat-Deceunynck, R. Bataille, and G. Jego. Structural and functional evidence for the role of the TLR2 DD loop in TLR1/ 2008. Toll-like receptors: lessons to learn from normal and malignant human B TLR2 heterodimerization and signaling. J. Biol. Chem. 281: 30132–30142. cells. Blood 112: 2205–2213. 53. Dunne, A., M. Ejdeback, P. L. Ludidi, L. A. O’Neill, and N. J. Gay. 2003. 66. Ma˚nsson, A., M. Adner, U. Ho¨ckerfelt, and L. O. Cardell. 2006. A distinct Toll- Structural complementarity of Toll/interleukin-1 receptor domains in Toll-like like receptor repertoire in human tonsillar B cells, directly activated by receptors and the adaptors Mal and MyD88. J. Biol. Chem. 278: 41443–41451. PamCSK, R-837 and CpG-2006 stimulation. Immunology 118: 539–548. 54. Shinkai, H., Y. Muneta, K. Suzuki, T. Eguchi-Ogawa, T. Awata, and H. Uenishi. 67. Ganley-Leal, L. M., X. Liu, and L. M. Wetzler. 2006. Toll-like receptor 2-me- 2006. Porcine Toll-like receptor 1, 6, and 10 genes: complete sequencing of diated human B cell differentiation. Clin. Immunol. 120: 272–284. genomic region and expression analysis. Mol. Immunol. 43: 1474–1480. 68. Bohnhorst, J., T. Rasmussen, S. H. Moen, M. Fløttum, L. Knudsen, M. Børset, 55. Opsal, M. A., D. I. Va˚ge, B. Hayes, I. Berget, and S. Lien. 2006. Genomic or- T. Espevik, and A. Sundan. 2006. Toll-like receptors mediate proliferation and ganization and transcript profiling of the bovine toll-like receptor gene cluster survival of multiple myeloma cells. Leukemia 20: 1138–1144. TLR6-TLR1-TLR10. Gene 384: 45–50. 69. Muzio, M., C. Scielzo, M. T. Bertilaccio, M. Frenquelli, P. Ghia, and 56. Hornung, V., S. Rothenfusser, S. Britsch, A. Krug, B. Jahrsdo¨rfer, T. Giese, F. Caligaris-Cappio. 2009. Expression and function of toll like receptors in S. Endres, and G. Hartmann. 2002. Quantitative expression of toll-like receptor chronic lymphocytic leukaemia cells. Br.J.Haematol.144: 507–516. 1-10 mRNA in cellular subsets of human peripheral blood mononuclear cells and 70. Marsland, B. J., and M. Kopf. 2007. Toll-like receptors: paving the path to sensitivity to CpG oligodeoxynucleotides. J. Immunol. 168: 4531–4537. T cell-driven autoimmunity? Curr. Opin. Immunol. 19: 611–614. 57. Bourke, E., D. Bosisio, J. Golay, N. Polentarutti, and A. Mantovani. 2003. The toll- 71. Komai-Koma, M., L. Jones, G. S. Ogg, D. Xu, and F. Y. Liew. 2004. TLR2 is like receptor repertoire of human B lymphocytes: inducible and selective expres- expressed on activated T cells as a costimulatory receptor. Proc. Natl. Acad. Sci. sion of TLR9 and TLR10 in normal and transformed cells. Blood 102: 956–963. USA 101: 3029–3034. Downloaded from 58. Bernasconi, N. L., N. Onai, and A. Lanzavecchia. 2003. A role for Toll-like re- 72. Bell, M. P., P. A. Svingen, M. K. Rahman, Y. Xiong, and W. A. Faubion, Jr. 2007. ceptors in acquired immunity: up-regulation of TLR9 by BCR triggering in naive B FOXP3 regulates TLR10 expression in human T regulatory cells. J. Immunol. cells and constitutive expression in memory B cells. Blood 101: 4500–4504. 179: 1893–1900. 59. Vasilevsky, S., J. Colino, R. Puliaev, D. H. Canaday, and C. M. Snapper. 2008. 73. Ma, X., Y. Liu, B. B. Gowen, E. A. Graviss, A. G. Clark, and J. M. Musser. 2007. Macrophages pulsed with Streptococcus pneumoniae elicit a T cell-dependent Full-exon resequencing reveals toll-like receptor variants contribute to human antibody response upon transfer into naive mice. J. Immunol. 181: 1787–1797. susceptibility to tuberculosis disease. PLoS One 2: e1318. 60. Pasare, C., and R. Medzhitov. 2005. Control of B-cell responses by Toll-like 74. Lazarus, R., B. A. Raby, C. Lange, E. K. Silverman, D. J. Kwiatkowski,

receptors. Nature 438: 364–368. D. Vercelli, W. J. Klimecki, F. D. Martinez, and S. T. Weiss. 2004. TOLL-like http://www.jimmunol.org/ 61. Meyer-Bahlburg, A., S. Khim, and D. J. Rawlings. 2007. B cell intrinsic TLR signals receptor 10 genetic variation is associated with asthma in two independent amplify but are not required for humoral immunity. J. Exp. Med. 204: 3095–3101. samples. Am. J. Respir. Crit. Care Med. 170: 594–600. by guest on September 26, 2021