Mitogen Derived − Mycoplasma Arthritidis Activities of HLA-DR Potentiates the Superantigenic Coexpression of TLR2 Or TLR4 With

Home , MHC class II

Coexpression of TLR2 or TLR4 with HLA-DR Potentiates the Superantigenic Activities of Mycoplasma arthritidis−Derived Mitogen This information is current as of September 29, 2021. Marina T. Shio, Ghada S. Hassan, Waris A. Shah, Amal Nadiri, Youssef El Fakhry, Hogmin Li and Walid Mourad J Immunol published online 3 February 2014 http://www.jimmunol.org/content/early/2014/01/31/jimmun ol.1300591 Downloaded from

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 29, 2021 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 © 2014 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published February 3, 2014, doi:10.4049/jimmunol.1300591 The Journal of Immunology

Coexpression of TLR2 or TLR4 with HLA-DR Potentiates the Superantigenic Activities of Mycoplasma arthritidis–Derived Mitogen

Marina T. Shio,* Ghada S. Hassan,* Waris A. Shah,* Amal Nadiri,* Youssef El Fakhry,* Hogmin Li,†,‡ and Walid Mourad*

Mycoplasma arthritidis–derived mitogen (MAM) is a member of the superantigen family that structurally differs from other members while still capable of initiating cognate APC/T cell interaction. In addition to the critical role of MHC class II molecules, it has been suggested that TLR2 and TLR4 may cooperate with MHC class II during MAM-induced responses. In this study, we investigated the direct involvement of TLR2 and TLR4 in MAM binding and presentation to T cells. Our results showed that

MAM fails to bind to TLR2- and TLR4-transfected cells. However, coexpression of TLR2 or TLR4 with HLA-DR signiﬁcantly Downloaded from increases MAM binding and the subsequent T cell activation compared with cells expressing HLA-DR alone. The upregulated MAM binding and activity in HLA-DR/TLR–transfected cells is abrogated by an anti–HLA-DR Ab. Interestingly, we also found that MAM complexed with soluble HLA-DR is capable of binding to both TLR2 and TLR4. The enhancing effect of TLR2 or TLR4 on MAM-induced T cell proliferation was not due to TLR ligand contamination in the MAM preparation. Taken together, these results strongly suggest that binding of MAM to HLA-DR leads to a conformational change in MAM structure allowing its

interaction with TLR2 and TLR4 and a better recognition by T cells. The Journal of Immunology, 2014, 192: 000–000. http://www.jimmunol.org/

uperantigens (Sags) are a family of bacterial and viral ability to trigger and/or exacerbate the collagen-induced arthritis products that activate a large number of T cells by a pathway in genetically susceptible rodent strains (10), and its nonrelated- S that is distinct from that of conventional Ags. Upon direct ness to other Sags (13). Our group has previously reported that binding of Sags to MHC class II molecules, MHC class II/Sag MAM has a novel fold that is distinct from all other Sags, pro- complexes are formed without Ag processing (1) or MHC restric- vided the structure/interaction of MAM/MHC class II by cocrystal tion (2) and are recognized by the variable regions of the b-chain of analysis, and determined the crystal structure of the first ternary the TCR (TCR Vb) leading to T cell activation (3, 4). Mycoplasma complex (TCR/MAM/MHC class II) (14, 15). Moreover, we dem- by guest on September 29, 2021 arthritidis–derived mitogen (MAM) produced by M. arthritidis is onstrated that MAM not only binds TCR Vb but also interacts with a member of this family (5, 6). Similar to other Sag group members, TCR Va (16) suggesting that TCRs recognize MAM in a substantially MAM mimics nominal Ag in initiating cognate T/B cells interac- different way than conventional peptide Ags and other known Sags. tion that leads to B cell proliferation and differentiation (7, 8), can In addition to the pivotal role of MHC class II in MAM-induced be detected in dimeric form in solution and its dimerization status is responses, it has been proposed that other immune receptors, necessary for its binding to MHC class II (6), is capable of inducing namely TLR2 and TLR4, may mediate MAM-induced responses inflammatory cytokines in human monocytes and fibroblast-like (17, 18). These TLRs are highly expressed on dendritic cells and synoviocytes (9), and of triggering T cell activation (6). macrophages, which are the most efficient in Sag-induced T cell In addition to the above described similarities, MAM has unique activation (19). TLR2 recognizes a large entity of microbial prod- features that differentiate it from other bacterial Sags, namely its ucts originating from a variety of species such as Gram-negative strong preference for class II H2-E or HLA-DR (10, 11), its usage and -positive bacteria, mycoplasma, and yeasts. TLR2 ligands include of the CDR3 region of the TCR as well as the Vb-chain (12), its peptidoglycans and lipoproteins from Gram-positive bacteria and lipopeptides such as macrophage-activating lipopeptide-2 from Mycoplasma fermentans (20, 21). TLR4, together with MD-2 and *Laboratoire d’immunologie cellulaire et molećulaire, Centre Hospitalier de l’Uni- CD14 (22), senses a certain group of ligands in viruses and en- versite´ de Montreál, Montreal, Quebec H2X 0A9, Canada; †Department of Biomed- ical Sciences, State University of New York, Albany, NY 12201; and ‡Wadsworth dogenous pathogen-associated molecular patterns. MAM was shown Center, New York State Department of Health, Albany, NY 12208 to elicit diverse cytokine profiles in mice that differ in their ex- Received for publication March 1, 2013. Accepted for publication December 27, pression of TLR2 and TLR4 (23). It induces a type 2 cytokine 2013. profile in mice that express both TLR2 and TLR4 and a type 1 This work was supported by Canadian Institutes of Health Research Grant MOP- cytokine profile in mice that express only TLR2. More recent data 102750. outlined a possible interaction of MAM with these TLRs by showing Address correspondence and reprint requests to Dr. Walid Mourad, Centre de Re- that expression of TLR4 on MHC class II–positive cells negatively cherche du Centre Hospitalier de l’Universite´ de Montreál, Viger Tower, 900 Saint Denis Street, Montreal, Quebec H2X 0A9, Canada. E-mail address: mw.mourad@ influenced TLR2 expression as well as inflammatory response in- umontreal.ca duced upon MAM stimulation (18). Taken together, these obser- Abbreviations used in this article: HEK, human embryonic kidney; MAM, Myco- vations suggest the cooperation between MHC class II and TLRs plasma arthritidis–derived mitogen; Sag, superantigen; XTT, 2,3-bis(2-methoxy-4- during MAM-induced responses. nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide. In the current study, we further characterize this cooperation Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00 between MHC class II and TLRs in MAM binding and presentation.

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1300591 2 HLA-DR/TLR COOPERATION FOR MAM BINDING AND ACTIVITY

Our results indicate that ligation of MAM with HLA-DR triggers 96-well microtiter plate in phenol red-free RPMI 1640 complete medium. a conformational change that promotes MAM binding to TLR2 The coculture were either stimulated or not with different concentrations of and TLR4 leading to a significant enhancement of T cell activation. MAM. In some experiments, the cells were pretreated with different concentrations of anti–HLA-DR for 1 h before MAM addition. After 48 h Given the present findings, MAM could be considered as a pertinent of incubation at 37˚C, the supernatants were collected; diluted (final vol- tool to study the cross-talk between innate and adaptive immunity. ume, 100 ml in phenol red-free RPMI 1640 complete medium), and added to the IL-2–dependent T cell clone (CTLL-2) already seeded (6 3 103/ well) in a 96-well microtiter plate. After the 48 h, 50 ml XTT/PMS mix (5 Materials and Methods ml of XTT at 0.9 mg/ml, and 100 ml of PMS at 0.3 mg/ml) were added to Abs and reagents CTLL2 cells, and the plate was further incubated for 2–4 h. The absor- bance was measured at 450/650-nm filter in a Dynatech microplate reader, Anti-hTLR4 (HTA125 an IgG2a mAb) (24) was provided by Dr. K. Miyake and the concentration of IL-2 was determined in relation to a standard (Saga Medical School, Saga, Japan); anti-hTLR2 (TL2.1, an IgG2a mAb) curve made with serial 2-fold dilutions of the recombinant cytokine. The was purchased from eBioscience (San Diego, CA); and anti–HLA-DR results were expressed as units per milliliter. mAb (L243, IgG2a), which recognizes a conformational epitope on the DRa-chain, was obtained from the American Type Culture Collection Statistical analysis (Manassas, VA). Anti-staphylococcal enterotoxin B mAb (8C12, an IgG2a mAb) was generated in our laboratory and was used as isotype control. Student t test was used when comparing two groups and ANOVA/ Recombinant MAM was generated and purified as we described previously Bonferroni test when comparing more than two groups. The differences (5). MAM also was biotinylated by the Pierce EZ-linked NHS-biotin kit were considered significant at a 5% level. (Pierce), according to the manufacturer’s protocol. FBS and hygromycin were purchased from Wisent. Zeocin purchased from InvivoGen, and

penicillin (P), streptomycin (S), and L-glutamine (G) were procured from Results Downloaded from Life Technologies (Burlington, ON, Canada). LPS–Porphyromonas gin- Coexpression of TLR2 or TLR4 with HLA-DR enhances MAM givalis, DMEM, RPMI 1640 medium, and Alexa 488–conjugated goat binding anti-mouse IgG was from Invitrogen. XTT (2,3-bis(2-methoxy-4-nitro-5- sulfophenyl)-2H-tetrazolium-5-carboxanilide), phenazine methosulfate It has been shown that MAM modulates cytokine profile in TLR4- (PMS), and LPS–Escherichia coli were obtained from Sigma-Aldrich deficient mice, because tlr4def mice exhibit a Th1 cytokine profile, (Oakville, ON, Canada). PE-conjugated streptavidin was procured from whereas the wild-type counterpart mice present a Th2 profile upon BD Biosciences (Mississauga, ON, Canada).

stimulation with MAM (23). Abs directed against TLR2 or TLR4 http://www.jimmunol.org/ Plasmid constructs prevent MAM-induced cytokine production by murine macrophages and human monocyte cell line, suggesting a possible in- Human TLR2/pEFBOS and TLR4/pCMV-flag were a gift from Dr. A. Ahmad (St. Justine Hospital, Montreal, QC, Canada) and Dr. K. Fitzgerald (Uni- teraction between MAM and TLR2 or TLR4 (18). However, direct versity of Massachusetts, Worcester, MA), respectively. hTLR2 was subcl- evidence showing the interaction of MAM and TLRs is still oned into pcDNA3.1/hygromycin at XhoI/KpnI restriction sites. HLA-DRa missing. To determine whether MAM can bind to TLR2 or TLR4, and DRb-chains were cloned into the pBudCE4-A-Zeocine vector (25). we used HEK-293 cells that were transfected with DNA coding Empty pBudCE4-A-Zeocine vector and pcDNA3.1 A-Hygromycin were for HLA-DR, TLR2, or TLR4 or with empty vector. HEK-293 a gift from Drs. J. Thibodeau and R. P. Sekaly, respectively (University of Montreal, Montreal, QC, Canada). cells were used along this study, because they do not initially express these MHC class II, TLR molecules or any accessory by guest on September 29, 2021 Cell lines and transfections molecules that can affect MAM-induced T cell activation. Stably The murine T cell clone 3DT (Vb8.2) was maintained in RPMI 1640 transfected HEK-293 cells were sorted and first analyzed for medium supplemented with 10% FBS, 2 mM L-glutamine, 100 U/ml HLA-DR, TLR2, or TLR4 expression by cytometry (Fig. 1A). The penicillin, 100 mg/ml streptomycin, and 5 mM 2-ME. The IL-2–depen- transfected cells were then used to study the binding capacity of dent CTLL-2 T cell clone was cultured in RPMI 1640 medium supple- 2 2 2 biotinylated MAM. As shown in Fig. 1B, although there is a sig- mented with 5 U/ml IL-2. MHC class II , TLR2 , and TLR4 human embryonic kidney (HEK-293) cells were maintained in DMEM supple- nificant binding of MAM to HEK-293 cells expressing HLA-DR, mented with 5% FBS, 2 mM L-glutamine, 100 U/ml penicillin, and 100 MAM was unable to bind to TLR2- or TLR4-transfected cells. mg/ml streptomycin. HEK-293 stably transfected with HLA-DR and/or Because there is no detectable MAM binding to TLR2- and TLR4- TLR2 or TLR4 were performed by the calcium phosphate coprecipita- transfected cells, we then investigated whether coexpression of tion technique as previously described (26) and cloned with appropriate selective agents. hTLR4 stable clones were obtained by cotransfection with HLA-DR molecules could influence MAM binding. For this purpose, pcDNA3.1/hygromycin (10:1 ratio). Selected clones were maintained with HEK-293 cells stably expressing HLA-DR were cotransfected complete media containing 1.6 mg/ml Zeocine (HLA-DR) and/or 0.2 mg/ with TLR2, TLR4, or empty vector. Stably cotransfected HEK-293 ml hygromycin (TLR2 or TLR4). cells were sorted and analyzed for HLA-DR, TLR2, and TLR4 MAM binding and flow cytometry coexpression by cytometry (Fig. 2A). The cotransfected cells were then used to study the binding capacity of biotinynaled MAM. 5 HEK-293 cells were trypsinized, washed, adjusted to 3 3 10 cells, and Results presented in Fig. 2B and 2C show that MAM binding incubated with different concentrations (0.01–0.5 mg) of biotin-conjugated MAM in 100 ml RPMI 1640 medium/10% FBS for 2 h on ice. In some was significantly upregulated in cells cotransfected with both experiments, the cells were preincubated for 15 min with different con- HLA-DRandTLR2orTLR4ascomparedwithcellsexpress- centrations of anti–HLA-DR mAb L243 before incubating them with ingonlyHLA-DR. MAM. After washing, cells were incubated with PE-conjugated strepta- Because we have previously shown that anti–HLA-DR Ab 3 5 vidin for 30 min on ice. For surface staining, cells (3 10 cells) were (L243), but not anti-MHC class I or anti-CD40, completely inhibited incubated with anti-TLR2 mAb, anti-TLR4, anti–HLA-DR, or an isotype- matched control mAb (8C12, mouse IgG2a) in 100 ml RPMI 1640 medium the MAM binding (28), the L243 Ab was used to evaluate the co- for 30 min on ice and followed by Alexa 488–conjugated goat anti-mouse operation between HLA-DR and TLR2 or TLR4 in MAM binding. IgG for 30 min on ice. After secondary staining, cells were washed, and the As shown in Fig. 3A and 3B, treatment with L243 inhibited MAM fluorescence was acquired on a LSR I cytometer (BD Biosciences) and binding to HLA-DR– as well as to HLA-DR/TLR2– and HLA-DR/ analyzed using CellQuest software (BD Biosciences). TLR4–transfected cells in a dose-dependent manner (Fig. 3B). The T cell proliferation and IL-2 measurement major difference between the observed inhibitions is that lower The assay was performed as previously described (27) with some mod- doses of L243 (1 mg/ml) was required to completely inhibit MAM ifications. Briefly, 2 3 104 APCs (HEK-293–transfected cells) were binding to HLA-DR–transfected cells as compared with 5 mg/ml cocultured with 8 3 104 T cell clones in a final volume of 200 ml/well in for HLA-DR/TLR2– and HLA-DR/TLR4–cotransfected cells. The Journal of Immunology 3 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021 FIGURE 1. No detectable binding of MAM to TLR2- or TLR4- expressing cells. (A) Expression levels of HLA-DR, TLR2, and TLR4 in stably transfected HEK-293 cells. Cells were first incubated with L243, TL2.1, or HTA125 Abs against HLA-DR, TLR2, and TLR4, respectively, followed by Alexa 488–conjugated goat anti-mouse IgG. IgG2a Ab was FIGURE 2. Expression of TLR2 and TLR4 enhances MAM binding to used as an isotype match control. Data are representative of three inde- HLA-DR/HEK-293–transfected cells. (A) Expression levels of HLA-DR, pendent experiments. (B) Transfected cells (0.3 3 106/100 ml) were incu- TLR2, and TLR4 in stably transfected HEK-293 cells. Cells were stained bated without MAM-biot. (negative control, empty plot) or with MAM-biot. using L243, TL.2, or HTA125 Abs against HLA-DR, TLR2, and TLR4, (20 nM, gray plot), followed by streptavidin–PE. Cells were analyzed by respectively. IgG2a Ab was used as an isotype match control. Data are flow cytometry for MAM binding. Data are representative of three inde- representative of three independent experiments. (B) Transfected cells pendent experiments. (0.3 3 106/100 ml) were incubated without MAM-biot. (negative control, empty plot) or with MAM-biot. (20 nM, gray plot), followed by streptavidin–PE. Cells were analyzed by flow cytometry. Data are representative of three independent experiments. (C) Histograms represent the mean 6 These results indicate that HLA-DR is the main MAM receptor SEM of three independent MAM-binding experiments. *p , 0.05. and strongly support the cooperation between HLA-DR and TLR2 or TLR4 in MAM binding. transfected cells (Fig. 4B), confirming the cooperation of HLA-DR MAM ligation with soluble HLA-DR influences its binding to in MAM binding to TLR molecules. TLR2 and TLR4 molecules It was proposed earlier that Sag binding to MHC class II molecules Coexpression of TLR2 or TLR4 with HLA-DR enhances leads to a conformational change that allow Sag/MHC class II MAM-induced T cell activation complexes to interact with TCR on T cells and induces subsequent The cooperation demonstrated between the MHC class II and the T cell activation (29). This hypothesis was supported by our co- TLRs could have biological implications. Indeed, it has previously crystal analysis (MAM/HLA-DR) (15), and the ternary complex, been shown that TLR2 and TLR4 could affect cytokine production TCR/MAM/HLA-DR (16). To verify whether the conformational by macrophages and monocyte possibly via MHC class II sig- change induced in MAM upon its binding to MHC class II could naling. In our study, to characterize the role of TLR in MAM- influence the MAM binding to TLRs, we generated soluble HLA-DR induced cell activation, transfected and cotransfected HEK-293 and used it for the binding experiments. Fig. 4A shows that there cells were cocultured with MAM-responsive murine 3DT cell was no detectable binding of MAM to empty vector–, TLR2- or clones that express Vb8.2 (6). Cocultured cells were left unsti- TLR4-transfected cells. However, the addition of soluble HLA-DR mulated or stimulated with different concentrations of MAM. allows significant binding of MAM to both TLR2- and TLR4- After 48 h, levels of IL-2 were assessed in cell supernatants as 4 HLA-DR/TLR COOPERATION FOR MAM BINDING AND ACTIVITY

FIGURE 3. HLA-DR mAb completely blocks MAM binding. (A) Transfected cells (0.3 3 106/100 ml) expressing HLA- DR or HLA-DR in combination with TLR-2 or TLR-4 were incubated without MAM-biot. (negative control, empty plot) or with MAM-biot. (20 nM) in the absence (black plot) or presence (gray plot) of L243 anti–HLA-DR Abs (2 mg/ml) for 2 h on ice, followed by streptavidin–PE. Cells were analyzed by ﬂow cytometry for MAM binding. (B) Histograms represent mean 6 SEM of three independent experiments in which different concentrations of anti–HLA-DR were used. *p , 0.05. Downloaded from http://www.jimmunol.org/ a parameter of T cell activation. Similarly to our previously re- T cells upon stimulation with MAM in a dose-dependent manner. ported data (6), MAM induced T cell activation of 3DT cells when MAM presented by HEK-293 cells transfected with empty vector presented by HLA-DR. Interestingly, the coexpression of TLR2 or cells transfected with only TLR2 or TLR4 failed to induce any (Fig. 5A) or TLR4 (Fig. 5B) together with HLA-DR molecules on detectable IL-2 production. It is worth noting that, in contrast to HEK-293 cells signiﬁcantly increased the activation of cocultured other cell types such as human monocytes, monocytic cell line, by guest on September 29, 2021

FIGURE 4. Soluble HLA-DR allows MAM binding to TLR2 and TLR4: MAM-biot. was added to HEK-293 cells transfected with empty vector, with TLR2 or TLR4, alone (A) or in combination with soluble HLA-DR (B) and incubated for 1 h at 37˚C. Washed cells were then incubated with streptavidin–PE. After 1 h, cells were washed and analyzed by ﬂow cytometry. The Journal of Immunology 5

the Fig. 6, L243 completely inhibited T cell activation when MAM was presented by HLA-DR alone. In addition, when MAM was presented by HEK-293 expressing HLA-DR in combination with TLR2 (Fig. 6A) or TLR4 (Fig. 6B), L243 Ab was able to abrogate, but to a lesser extent, the MAM-induced T cell activation. To rule out the possibility that the observed T cell activation was due to contaminants in the MAM preparation that could ligate TLRs, namely the LPSs, T cell clones cocultured with transfected HEK-293 cells were treated with MAM and/or LPS from the Porphyromonas gingivali, ligand of TLR2 (32) or from the E. coli, ligand of TLR4 (33). As shown in Fig. 6, both LPSs were not able to induce the proliferation of T cells in coculture with HEK-293– transfected with TLR2 (Fig. 6C) or TLR4 (Fig. 6D). Moreover, both LPSs could not synergize with MAM in activating T cells cocultured with HEK-293 expressing HLA-DR and TLR2 (Fig. 6C) or TLR4 (Fig. 6D). Indeed, IL-2 levels detected upon MAM stimulation were not altered when the stimulation was undertaken

with a combination of MAM and both types of LPSs. These Downloaded from results also were conﬁrmed using polymyxin B (10 mg/ml) to inhibit LPS-induced effects (5267 6 383 U/ml by MAM versus 4349 6 338 U/ml IL-2 by MAM + polymyxin B). Furthermore, on the basis of the heat sensitivity of MAM and LPS, MAM being heat-sensitive (34) and LPS being heat-resistant (35), heating

MAM at 95˚C for 15 min (36) abrogated its capacity to induce http://www.jimmunol.org/ T cell activation in the coculture with HLA-DR/TLR2– or HLA- DR/TLR4–transfected HEK-293 cells (data not shown). These results indicate that the enhanced MAM-mediated T cell activation in the presence of both HLA-DR and TLRs (TLR2 and TLR4) was not due to a contamination of MAM preparation with TLR FIGURE 5. Coexpression of TLR2 or TLR4 with HLA-DR enhanced ligands, outlining as such the biological interaction between MAM MAM-induced T cell activation: HEK-293 cells stably transfected with and TLRs themselves in the presence of HLA-DR molecules. Given empty vector, HLA-DR, TLR2, HLA-DR/TLR2 (A), TLR4, or HLA-DR/ our data, we suggest a model outlining the cooperation between 4 4 TLR4 (B)(23 10 cells/200 ml) were cocultured with 8 3 10 cells/200 MHC class II and TLR2 or TLR4 molecules for MAM binding and by guest on September 29, 2021 ml of a murine Vb8.2 (3DT). Cells were stimulated in the presence or activity (Fig. 7). absence of MAM at the indicated concentration. After 48 h of incubation, the supernatants were collected, and the IL-2 release was assessed with bioassay using CTLL-2 cells, followed by XTT metabolization assay. IL-2 Discussion levels are represented by arbitrary units per milliliter. Results represent the The present study highlights the role of TLRs 2 and 4 as important mean 6 SEM of four independent experiment. *p , 0.05 comparing TLR coreceptors for MAM. Along with HLA-DR, TLR2 and TLR4 are positive with negative cells. able to increase MAM binding and its presentation to T cells. Indeed, data presented in the current investigation indicate that MAM exhibited an increased binding to cells coexpressing both and fibroblast-like synoviocytes (30, 31), we could not perceive HLA-DR and TLR2 or TLR4 when compared with cells expressing any signaling event in our transfected and cotransfected HEK-293 HLA-DR alone. Moreover, MAM was able to bind cells expressing cells following addition of MAM. Therefore, the IL-2 response TLR2 or TLR4 only in the presence of soluble HLA-DR molecules. observed in the above experiments was not due to IL-1 or IL-6 The enhancement of MAM binding was completely abrogated by production by HEK-293 because no such cytokines were detected pretreatment with anti–HLD-DR mAb. On the biological level, in the cell supernatants (data not shown). Along the same line of cells expressing both HLA-DR and TLR molecules and presenting evidence, the enhancement of MAM activity in HLA-DR/TLR2– MAM to T cells in cocultures induced a more pronounced acti- or HLA-DR/TLR4–cotransfected cells was not due to a MAM- vation of T cells as compared with cells expressing only HLA-DR induced increase in HLA-DR expression in these cells, because on their surface. This enhanced response was proven specific to a stimulation with MAM failed to affect HLA-DR expression (data MAM because TLR ligands failed to mount any response when not shown). Taken together, our results demonstrate a greater IL-2 used as stimulants. production in T cells upon their stimulation with MAM presented It has been previously demonstrated that MAM was capable of + + by TLR2 or TLR4 and HLA-DR, correlating with our binding data activating THP1, a human monocytic cell line (HLA-DR /TLR2 / + that outlined an increased binding of MAM in the presence of TLRs TLR4 ), and that such activation was inhibited by Abs directed and MHC class II molecules. against both TLRs. In addition, macrophages derived from C3H/ HeJ mice that exhibit a hyporeactive form of TLR4 (nonrespon- Enhanced MAM-induced T cell activation is dependent on sive to LPS) and from C3H/HeN TLR2 knockout mice (TLR22/2/ HLA-DR but totally independent of TLR2 or TLR4 ligands TLR4+/+) were shown to produce inflammatory cytokines upon To further evaluate the cooperation between TLR2 or TLR4 and their stimulation with MAM, whereas C3H/HeJ TLR2KO cells HLA-DR molecules for T cell activation induced by MAM, the lost such ability (18). Even though these interesting findings cocultures were treated with different doses of L243. As shown in outline a cross-talk between MAM and TLRs, they did not dem- 6 HLA-DR/TLR COOPERATION FOR MAM BINDING AND ACTIVITY Downloaded from http://www.jimmunol.org/

FIGURE 6. Enhanced MAM-induced T cell activation is HLA-DR-dependent but totally independent of TLR2 or TLR4 ligands. (A and B) Cells stably transfected with HLA-DR, HLA-DR/TLR2 (A), or HLA-DR/TLR4 (B)(23 104 cells/200 ml) were cocultured with 8 3 104 cells/200 ml cells of a murine Vb8.2 (3DT)–positive T cell line. Cells were stimulated with MAM in the presence or absence of different doses of anti–HLA-DR (L243). After 48 h of incubation, the supernatants were collected, and the IL-2 releases were assessed with bioassay using CTLL-2 cells, followed by XTT metabolization assay. IL-2 levels are represented by arbitrary units per milliliter. Results represent the mean of triplicate of one experiment. (C and D) Cocultured cells were stimulated with MAM (250 ng/ml) in the presence or absence of 100 ng/ml LPS from P. gingivalis (ligand of TLR2) (C) or from E. coli (ligand of TLR4) by guest on September 29, 2021 (D). After 48 h of incubation, the supernatants were collected and the IL-2 release was assessed with bioassay using CTLL-2 cells, followed by XTT metabolization assay. IL-2 levels are represented by arbitrary units per milliliter. Results represent the mean of triplicate of a representative experiment of two independent experiments. onstrate the direct binding between both types of molecules. Data compared with cell expressing only HLA-DR, a ﬁnding strongly from our current study show that MAM is not capable of directly suggesting the cooperation between TLRs and HLA-DR mole- binding TLR2 or TLR4 using HEK-293 cells expressing either of cules for MAM binding. To outline the signiﬁcance of the in- these TLRs and immunostaining techniques. However, our study creased MAM binding to cells cotransfected with HLA-DR and demonstrate that MAM exhibit an increased binding to HEK-293 TLR2 or TLR4 compared with cells expressing HLA-DR alone, cells expressing both HLA-DR and TLR2 or TLR4 molecules as anti–HLA-DR mAb was added and shown to completely inhibit

FIGURE 7. Schematic model of MAM interactions with MHC class II, TLR2 or TLR4, and TCR. The Journal of Immunology 7 the binding of MAM to HLA-DR cells and HLA-DR/TLR2 or larly TLR3, 4, and 9. Indeed, Liu et al. (47) have demonstrated HLA-DR/TLR4 cells. that mice knocked out for their MHC class II molecules exhibited Therefore, our data suggest that MAM upon its binding to HLA- a reduced TLR-mediated immune response, namely TNF, IL-6, DR undertake a conformational change allowing its interaction and IFN-b, when challenged with TLR4, TLR3, or TLR9 ligands. with TLR2 and TLR4. These findings are in accordance with results The authors further outlined the importance of MHC class II in described above, where MAM was capable of interacting with full activation of the TLR system in vitro using peritoneal mac- monocytic cells expressing TLR2 and TLR4 (18). Indeed, we rophages and dendritic cells derived from MHC2/2 mice. These consider that the cross-talk that was previously reported between cells produced a decreased immune response when stimulated with MAM and TLRs in monocytes and macrophages and leading to TLR ligands as compared with WT cells, highlighting as such the cell activation (18) is really due to the well-known presence of specific role of MHC class II in TLR signaling (47). In addition to MHC class II molecules in these cells, in particular the HLA-DR the role of MHC class II molecules in TLR-mediated immune (37). To further confirm our interpretation of these data, macro- responses, TLRs themselves could be also implicated in MHC phages derived from SLJ mice that lack MAM-reactive MHC class II signaling. Activation of TLR4 was shown to upregulate class II alleles as well as Chinese hamster ovary cells, which do the autophagy-dependent presentation of Ag by MHC class II not express any MHC molecules and that were transfected with molecules (48). Our data in this study outline another role of TLR CD14/TLR2 or CD14/TLR4, were shown to be irresponsive to the in MHC class II signaling by cooperating with the latter molecules stimulation by MAM. However, these cells could be activated by for MAM binding and upregulating as such the MAM-induced LPS or macrophage-activating lipopeptide-2, the ligands of TLR4 activation of T cells. and TLR2, respectively (18). Therefore, the previously described In view of data presented in this study, we can suggest a model Downloaded from observations outlining the interaction of MAM with TLRs in of interaction between MHC class II molecules, TLRs and MAM. MHC class II–positive cells support our findings that MHC class II Binding of MHC class II to MAM induces a conformational molecules themselves positively influence MAM binding to TLR2 change in this latter allowing its interaction with TLR2 and TLR4 and TLR4. and subsequently a better presentation to T cells (Fig. 7). Confor- Interestingly, the cooperation between TLRs and HLA-DR mole- mational changes occurring upon ligand/receptor interactions have

cules was not restricted to the binding to MAM but included also long been reported in literature. For instance, binding of Ag/MHC http://www.jimmunol.org/ the biological response upon stimulation with MAM. It is well complex to TCR induces structural changes in this latter allowing established that MAM, like other Sags, bind to MHC class II the initiation of the activation cascade (49). We previously re- molecules on APCs and is presented as such to T cells by inter- ported that MAM uniquely dimerizes HLA-DR molecule and acting with their TCR, and more specifically with the region Vb itself adopts novel fold with homodimer formation (15). MAM’s 3.1, 13.1, 14, and 17 in humans and Vb 6, 8.1, and 5.1 in mice unique binding could be causing some unexplored conforma- (38). MAM induces different Vb repertoires depending on the tional changes to the complex of MAM/HLA-DR. The additional HLA-DR haplotype, which indicates that the Vb specificity of binding of MAM/HLA-DR to TLR2/4 provides more complexity MAM is determined by the MAM/HLA-DR complex (39, 40). In to the conformation of the tricomplex MAM/HLA-DR/TLR2/4 human T cells, MAM was shown to induce an increased in- that could be providing more binding sites to MAM, allowing as by guest on September 29, 2021 tracellular Ca2+ concentration (41). In addition, a wide array of such a better interaction with TCR on T cells. However, the exact cytokines was triggered by MAM in human cells derived from nature of this association needs to be explored by crystallographic peripheral blood or spleen (42, 43). MAM was shown to induce study of the trimer complex. In addition, many questions remain to a type 2 cytokine profile in mice that express both TLR2 and be resolved concerning the intracellular signaling implicated in TLR4 and a type 1 cytokine profile in mice that express only the TLR/MHC class II cross-talk, its effect on APCs themselves, TLR2 (23, 44). In our study, we wanted to investigate the bio- and the possible association with other adaptor molecules. logical outcome of MAM binding to TLR2 and TLR4 cooperating with HLA-DR molecules on surface of APCs. Indeed, MAM Acknowledgments presented by cells expressing only TLR2 or TLR4 failed to induce We thank Drs. Kensuke Miyake, Ali Ahmad, K. Fitzgerald, J. Thibodeau, any detectable T cell response, whereas it triggered slight T cell and R.P. Sekaly for providing the molecular biology materials. activation reflected by IL-2 production when presented by cells transfected with HLA-DR. Most importantly, a significant T cell Disclosures activation was observed when MAM was presented by cells The authors have no financial conflicts of interest. coexpressing HLA-DR/TLR2 or HLA-DR/TLR4. This enhancement is not due to an increased HLA-DR expression as similar levels of HLA-DR were observed in the HEK-293 cells after References MAM stimulation (data not shown). To confirm the specificity for 1. Dellabona, P., J. Peccoud, J. Kappler, P. Marrack, C. Benoist, and D. Mathis. MAM in this T cell response, we stimulated the system above 1990. Superantigens interact with MHC class II molecules outside of the antigen groove. Cell 62: 1115–1121. using LPS. Our data showed that adding LPS to the coculture 2. Fleischer, B. 1991. [Stimulation of the immune system by microbial “super- system has no effect on T cell activation. These results strongly antigens”]. Immun. Infekt. 19: 8–11. 3. Cole, B. C., D. R. Kartchner, and D. J. Wells. 1989. Stimulation of suggest a biological synergism between HLA-DR and TLR mol- mouse lymphocytes by a mitogen derived from Mycoplasma arthritidis. VII. ecules with respect to stimulation with MAM. The cooperation Responsiveness is associated with expression of a product(s) of the Vb8 gene between TLRs and MHC class II molecules was first outlined in family present on the T cell receptor a/b for antigen. J. Immunol. 142: 4131– 4137. 2000, where macrophages exhibiting low or no MHC class II 4. White, J., A. Herman, A. M. Pullen, R. Kubo, J. W. Kappler, and P. Marrack. expression failed to respond to LPS stimulation (45). More re- 1989. The Vb-specific superantigen staphylococcal enterotoxin B: stimulation of cently, Frei et al. (46) have shown that MHC class II molecules mature T cells and clonal deletion in neonatal mice. Cell 56: 27–35. 5. Langlois, M. A., P. Etongue´-Mayer, M. Ouellette, and W. Mourad. 2000. Binding colocalize with TLR2 and enhance TLR2- as well as TLR4- of Mycoplasma arthritidis‑derived mitogen to human MHC class II molecules induced responses upon stimulation with bacterial lipoprotein or via its N terminus is modulated by invariant chain expression and its C terminus is required for T cell activation. Eur. J. Immunol. 30: 1748–1756. LPS, respectively. An interesting study in 2011 reported a peculiar 6. Langlois, M.-A., Y. El Fakhry, and W. Mourad. 2003. Zinc-binding sites in the N cross-talk between MHC class II molecules and TLRs, particu- terminus of Mycoplasma arthritidis‑derived mitogen permit the dimer formation 8 HLA-DR/TLR COOPERATION FOR MAM BINDING AND ACTIVITY

required for high affinity binding to HLA-DR and for T cell activation. J. Biol. 28. Etongue´-Mayer, P., M. A. Langlois, M. Ouellette, H. Li, S. Younes, R. Al- Chem. 278: 22309–22315. Daccak, and W. Mourad. 2002. Involvement of zinc in the binding of Myco- 7. Mourad, W., R. S. Geha, and T. Chatila. 1990. Engagement of major histo- plasma arthritidis-derived mitogen to the proximity of the HLA-DR binding compatibility complex class II molecules induces sustained, lymphocyte groove regardless of histidine 81 of the b chain. Eur. J. Immunol. 32: 50–58. function-associated molecule 1-dependent cell adhesion. J. Exp. Med. 172: 29. Li, H., A. Llera, E. L. Malchiodi, and R. A. Mariuzza. 1999. The structural basis 1513–1516. of T cell activation by superantigens. Annu. Rev. Immunol. 17: 435–466. 8. Tumang, J. R., D. N. Posnett, B. C. Cole, M. K. Crow, and S. M. Friedman. 1990. 30. Mourad, W., K. Mehindate, T. J. Schall, and S. R. McColl. 1992. Engagement of Helper T cell‑dependent human B cell differentiation mediated by a mycoplas- major histocompatibility complex class II molecules by superantigen induces mal superantigen bridge. J. Exp. Med. 171: 2153–2158. inflammatory cytokine gene expression in human rheumatoid fibroblast-like 9. Mehindate, K., J. Thibodeau, M. Dohlsten, T. Kalland, R. P. Se´kaly, and synoviocytes. J. Exp. Med. 175: 613–616. W. Mourad. 1995. Cross-linking of major histocompatibility complex class II 31. Mehindate, K., R. al-Daccak, L. Rink, S. Mecheri, J. He´bert, and W. Mourad. molecules by staphylococcal enterotoxin A superantigen is a requirement for 1994. Modulation of Mycoplasma arthritidis‑derived superantigen-induced cy- inflammatory cytokine gene expression. J. Exp. Med. 182: 1573–1577. tokine gene expression by dexamethasone and interleukin-4. Infect. Immun. 62: 10. Cole, B. C., and C. L. Atkin. 1991. The Mycoplasma arthritidis T-cell mitogen, 4716–4721. MAM: a model superantigen. Immunol. Today 12: 271–276. 32. Zhou, Q., T. Desta, M. Fenton, D. T. Graves, and S. Amar. 2005. Cytokine 11. Cole, B. C., A. D. Sawitzke, E. A. Ahmed, C. L. Atkin, and C. S. David. 1997. profiling of macrophages exposed to Porphyromonas gingivalis, its lipopoly- Allelic polymorphisms at the H-2A and HLA-DQ loci influence the response of saccharide, or its FimA protein. Infect. Immun. 73: 935–943. murine lymphocytes to the Mycoplasma arthritidis superantigen MAM. [Pub- 33. Hoshino, K., O. Takeuchi, T. Kawai, H. Sanjo, T. Ogawa, Y. Takeda, K. Takeda, lished erratum appears in 1998 Infect. Immun. 66: 2399] Infect. Immun. 65: and S. Akira. 1999. Cutting edge: Toll-like receptor 4 (TLR4)‑deficient mice are 4190–4198. hyporesponsive to lipopolysaccharide: evidence for TLR4 as the Lps gene 12. Hodtsev, A. S., Y. Choi, E. Spanopoulou, and D. N. Posnett. 1998. Mycoplasma product. J. Immunol. 162: 3749–3752. superantigen is a CDR3-dependent ligand for the T cell antigen receptor. J. Exp. 34. Atkin, C. L., S. Wei, and B. C. Cole. 1994. The Mycoplasma arthritidis super- Med. 187: 319–327. antigen MAM: purification and identification of an active peptide. Infect. Immun. 13. Cole, B. C., K. L. Knudtson, A. Oliphant, A. D. Sawitzke, A. Pole, M. Manohar, 62: 5367–5375. L. S. Benson, E. Ahmed, and C. L. Atkin. 1996. The sequence of the Myco- 35. Gao, B., and M. F. Tsan. 2003. Endotoxin contamination in recombinant human plasma arthritidis superantigen, MAM: identification of functional domains and heat shock protein 70 (Hsp70) preparation is responsible for the induction of Downloaded from comparison with microbial superantigens and plant lectin mitogens. J. Exp. Med. tumor necrosis factor a release by murine macrophages. J. Biol. Chem. 278: 183: 1105–1110. 174–179. 14. Li, H., Y. Zhao, Y. Guo, S. J. Vanvranken, Z. Li, L. Eisele, and W. Mourad. 2007. 36. Ribeiro-Dias, F., M. T. Shio, J. Timenetsky, A. P. Oliane, C. C. Metran, Mutagenesis, biochemical, and biophysical characterization of Mycoplasma F. B. Pessoa, and S. Jancar. 2003. Mycoplasma arthritidis superantigen arthritidis‑derived mitogen. Mol. Immunol. 44: 763–773. (MAM)‑induced macrophage nitric oxide release is MHC class II restricted, 15. Zhao, Y., Z. Li, S. J. Drozd, Y. Guo, W. Mourad, and H. Li. 2004. Crystal interferon g dependent, and Toll-like receptor 4 independent. Exp. Cell Res. 286: structure of Mycoplasma arthritidis mitogen complexed with HLA-DR1 reveals 345–354.

a novel superantigen fold and a dimerized superantigen-MHC complex. Struc- 37. Yunis, J. J., H. Band, F. Bonneville, and E. J. Yunis. 1989. Differential expression http://www.jimmunol.org/ ture 12: 277–288. of MHC class II antigens in myelomonocytic leukemia cell lines. Blood 73: 931– 16. Wang, L., Y. Zhao, Z. Li, Y. Guo, L. L. Jones, D. M. Kranz, W. Mourad, and 937. H. Li. 2007. Crystal structure of a complete ternary complex of TCR, super- 38. Baccala, R., L. R. Smith, M. Vestberg, P. A. Peterson, B. C. Cole, and antigen and peptide-MHC. Nat. Struct. Mol. Biol. 14: 169–171. A. N. Theoﬁlopoulos. 1992. Mycoplasma arthritidis mitogen. Vb engaged in 17. Mu, H. H., J. Humphreys, F. V. Chan, and B. C. Cole. 2006. TLR2 and TLR4 mice, rats, and humans, and requirement of HLA-DRa for presentation. Arthritis differentially regulate B7-1 resulting in distinct cytokine responses to the my- Rheum. 35: 434–442. coplasma superantigen MAM as well as to disease induced by Mycoplasma 39. Posnett, D. N., A. S. Hodtsev, S. Kabak, S. M. Friedman, B. C. Cole, and arthritidis. Cell. Microbiol. 8: 414–426. N. Bhardwaj. 1993. Interaction of Mycoplasma arthritidis superantigen with 18. Mu, H. H., N. D. Pennock, J. Humphreys, C. J. Kirschning, and B. C. Cole. 2005. human T cells. Clin. Infect. Dis. 17(Suppl. 1): S170–S175. Engagement of Toll-like receptors by mycoplasmal superantigen: downregula- 40. Alvarez-Ossorio, L., M. Johannsen, M. Russlies, H. Kirchner, and L. Rink. 1997. tion of TLR2 by MAM/TLR4 interaction. Cell. Microbiol. 7: 789–797. HLA-dependent heterogeneous T cell response to Mycoplasma arthritidis‑der-

19. Nestle, F. O., C. Thompson, Y. Shimizu, L. A. Turka, and B. J. Nickoloff. 1994. ived superantigen (MAS). Med. Microbiol. Immunol. (Berl.) 185: 245–251. by guest on September 29, 2021 Costimulation of superantigen-activated T lymphocytes by autologous dendritic 41. Matthes, M., H. Schrezenmeier, J. Homfeld, S. Fleischer, B. Malissen, cells is dependent on B7. Cell. Immunol. 156: 220–229. H. Kirchner, and B. Fleischer. 1988. Clonal analysis of human T cell activation 20. Borrello, S., C. Nicolo`, G. Delogu, F. Pandolfi, and F. Ria. 2011. TLR2: by the Mycoplasma arthritidis mitogen (MAS). Eur. J. Immunol. 18: 1733–1737. a crossroads between infections and autoimmunity? Int. J. Immunopathol. 42. Rink, L., A. Kruse, W. Nicklas, J. Hoyer, and H. Kirchner. 1992. Induction of Pharmacol. 24: 549–556. cytokines in human peripheral blood and spleen cells by the Mycoplasma 21. Takeuchi, O., A. Kaufmann, K. Grote, T. Kawai, K. Hoshino, M. Morr, arthritidis‑derived superantigen. Lymphokine Cytokine Res. 11: 105–108. P. F.Muhlradt,€ and S. Akira. 2000. Cutting edge: preferentially the R- 43. Rink, L., W. Nicklas, L. Alvarez-Ossorio, M. Koester, and H. Kirchner. 1994. stereoisomer of the mycoplasmal lipopeptide macrophage-activating lipopeptide-2 Differential induction of tumor necrosis factor a in murine and human leuko- activates immune cells through a Toll-like receptor 2‑ and MyD88-dependent cytes by Mycoplasma arthritidis‑derived superantigen. Infect. Immun. 62: 462– signaling pathway. J. Immunol. 164: 554–557. 467. 22. Miller, S. I., R. K. Ernst, and M. W. Bader. 2005. LPS, TLR4 and infectious 44. Cole, B. C., H. H. Mu, and A. D. Sawitzke. 2000. The mycoplasma superantigen disease diversity. Nat. Rev. Microbiol. 3: 36–46. MAM: role in arthritis and immune-mediated disease. Int. J. Med. Microbiol. 23. Mu, H. H., A. D. Sawitzke, and B. C. Cole. 2001. Presence of Lps(d) mutation 290: 489–490. influences cytokine regulation in vivo by the Mycoplasma arthritidis mitogen 45. Piani, A., J. P. Hossle, T. Birchler, C. A. Siegrist, D. Heumann, G. Davies, superantigen and lethal toxicity in mice infected with M. arthritidis. Infect. S. Loeliger, R. Seger, and R. P. Lauener. 2000. Expression of MHC class II Immun. 69: 3837–3844. molecules contributes to lipopolysaccharide responsiveness. Eur. J. Immunol. 30: 24. Shimazu, R., S. Akashi, H. Ogata, Y. Nagai, K. Fukudome, K. Miyake, and 3140–3146. M. Kimoto. 1999. MD-2, a molecule that confers lipopolysaccharide respon- 46. Frei, R., J. Steinle, T. Birchler, S. Loeliger, C. Roduit, D. Steinhoff, R. Seibl, siveness on Toll-like receptor 4. J. Exp. Med. 189: 1777–1782. K. Buchner,€ R. Seger, W. Reith, and R. P. Lauener. 2010. MHC class II mole- 25. Finzi, A., A. Brunet, Y. Xiao, J. Thibodeau, and E. A. Cohen. 2006. Major cules enhance Toll-like receptor mediated innate immune responses. PLoS ONE histocompatibility complex class II molecules promote human immunodefi- 5: e8808. ciency virus type 1 assembly and budding to late endosomal/multivesicular body 47. Liu, X., Z. Zhan, D. Li, L. Xu, F. Ma, P. Zhang, H. Yao, and X. Cao. 2011. compartments. J. Virol. 80: 9789–9797. Intracellular MHC class II molecules promote TLR-triggered innate immune 26. Brunet, A., A. Samaan, F. Deshaies, T. J. Kindt, and J. Thibodeau. 2000. responses by maintaining activation of the kinase Btk. Nat. Immunol. 12: 416– Functional characterization of a lysosomal sorting motif in the cytoplasmic tail 424. of HLA-DOb. J. Biol. Chem. 275: 37062–37071. 48. Watts, C., M. A. West, and R. Zaru. 2010. TLR signalling regulated antigen 27. Thibodeau, J., C. De Souza, M. Smorawinska, and J. P. Thirion. 1992. Selection presentation in dendritic cells. Curr. Opin. Immunol. 22: 124–130. of mouse cells with amplified metallothionein genes retaining their glucocorti- 49. Ma, Z., P. A. Janmey, and T. H. Finkel. 2008. The receptor deformation model of coid inducibility. FEBS Lett. 310: 75–78. TCR triggering. FASEB J. 22: 1002–1008.