The Journal of Immunology

NLRC5 Deficiency Selectively Impairs MHC Class I- Dependent Killing by Cytotoxic T Cells

Francesco Staehli,* Kristina Ludigs,* Leonhard X. Heinz,† Queralt Seguı´n-Este´vez,‡ Isabel Ferrero,x Marion Braun,x Kate Schroder,{ Manuele Rebsamen,† Aubry Tardivel,* Chantal Mattmann,* H. Robson MacDonald,x Pedro Romero,x Walter Reith,‡ Greta Guarda,*,1 and Ju¨rg Tschopp*,1,2

Nucleotide-binding oligomerization domain-like receptors (NLRs) are intracellular involved in innate-driven inflamma- tory responses. The function of the family member NLR caspase recruitment domain containing 5 (NLRC5) remains a matter of debate, particularly with respect to NF-kB activation, type I IFN, and MHC I expression. To address the role of NLRC5, we generated Nlrc5-deficient mice (Nlrc5D/D). In this article we show that these animals exhibit slightly decreased CD8+ percentages, a phenotype compatible with deregulated MHC I expression. Of interest, NLRC5 ablation only mildly affected MHC I expression on APCs and, accordingly, Nlrc5D/D macrophages efficiently primed CD8+ T cells. In contrast, NLRC5 deficiency dramatically impaired basal expression of MHC I in T, NKT, and NK . NLRC5 was sufficient to induce MHC I expression in a human lymphoid cell line, requiring both caspase recruitment and LRR domains. Moreover, endogenous NLRC5 localized to the nucleus and occupied the proximal promoter region of H-2 . Consistent with downregulated MHC I expression, the elimination of Nlrc5D/D lymphocytes by cytotoxic T cells was markedly reduced and, in addition, we observed low NLRC5 expression in several murine and human lymphoid-derived tumor cell lines. Hence, loss of NLRC5 expression represents an advantage for evading CD8+ T cell-mediated elimination by downmodulation of MHC I levels—a mechanism that may be exploited by transformed cells. Our data show that NLRC5 acts as a key transcriptional regulator of MHC I in lymphocytes and support an essential role for NLRs in directing not only innate but also adaptive immune responses. The Journal of Immunology, 2012, 188: 3820–3828.

he innate is an essential component of 11). Collectively, these data highlight the important contribution our body’s defense strategy against invading pathogens of NLRs to innate-driven inflammation and their broad involve- T and host-derived danger. Prompt detection of hazardous ment in the immune response. by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. agents and early onset of the inflammatory response are guaran- Despite intense investigation, functions for only a few NLRs teed by dedicated receptors, such as nucleotide-binding oligo- have been firmly established (12). NLR caspase recruitment do- merization domain (NOD)-like receptors (NLRs). These are main (CARD) containing protein 5 (NLRC5) (also called NOD4) abundantly expressed by innate immune cells and recognize is a poorly characterized NLR family member with high structural pathogen- and danger-associated molecular patterns (1). Upon homology to NOD1, NOD2, and CIITA (13, 14), and its function activation, a growing number of NLRs are known to assemble in remains a matter of debate. Given the important similarity to inflammasomes, multiprotein platforms serving caspase-1 activa- NOD1 and NOD2, the capacity of NLRC5 to modulate the NF-kB tion, resulting in the maturation of IL-1b (1–3). Other NLRs— and the type I IFN pathways has been assessed. Overexpression namely, NOD1 and NOD2—form NF-kB activating complexes and RNA interference (RNAi)-mediated knockdown studies sug- http://classic.jimmunol.org upon bacterial invasion (4–7) and were also recently shown to gested that NLRC5 negatively regulates both the NF-kB and the promote the induction of type I IFN (8, 9). Unique among NLRs, type I IFN responses (13, 15). However, subsequent studies re- CIITA is the master regulator of MHC class II transcription (10, ported contradictory observations showing that NLRC5 promotes

*Department of Biochemistry, University of Lausanne, 1066 Epalinges, Switzerland; Competence in Research in Molecular Oncology; and work in the laboratory of †Research Center for Molecular Medicine of the Austrian Academy of Science, W.R. was funded by the Swiss National Science Foundation, the National Center Downloaded from 1090 Vienna, Austria; ‡Faculty of Medicine, University of Geneva, 1211 Geneva, of Competence in Research on Neural Plasticity and Repair, the Swiss Multiple Switzerland; xLudwig Center for Cancer Research, University of Lausanne, 1066 Sclerosis Society, and the Geneva Cancer League. Epalinges, Switzerland; and {Institute for Molecular Bioscience, University of Address correspondence and reprint requests to Dr. Greta Guarda, Department of Bio- Queensland, 4067 St. Lucia, Australia chemistry, Chemin des Boveresses 155, 1066 Epalinges, Switzerland. E-mail address: 1G.G. and J.T. share senior authorship. [email protected] 2Deceased. The online version of this article contains supplemental material.

Received for publication September 15, 2011. Accepted for publication February 8, Abbreviations used in this article: B2m, b2 microglobulin; BMDM, bone marrow- 2012. derived macrophage; CARD, caspase recruitment domain; cDC, conventional den- dritic cell; ChIP, chromatin immunoprecipitation; NLR, nucleotide-binding oligomer- This work was supported by grants from the Swiss National Science Foundation, the ization domain-like receptor; NLRC5, NLR CARD containing protein 5; NOD, Institute for Arthritis Research, the Foundation Louis-Jeantet, the European Union nucleotide-binding oligomerization domain; poly(I:C), polyinosinic-polycytidylic Seventh Framework Program Systems Biology Applied to Cancer and acid; RNAi, RNA interference; SP, single positive; TEC, thymic epithelial cell; TEM, AIDS, and the National Center of Competence in Research in Molecular Oncology; thioglycollate-elicited macrophage. K.S. is supported by a C.J. Martin Fellowship from the Australian National Health and Medical Research Council (ID 490993); M.B. and P.R. are supported by the Ó Swiss National Science Foundation; P.R. is supported by the National Center of Copyright 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1102671 The Journal of Immunology 3821

type I IFN production (16, 17). NLRC5 was also proposed to form Enrichment of splenic cell populations, bone marrow-derived an inflammasome, as its knockdown by RNAi led to decreased IL- macrophage and bone marrow-derived dendritic cell 1b secretion in response to a number of pathogens and pathogen- differentiation associated molecular patterns (18). Importantly, it was recently CD4+ T cells were enriched using anti-CD4 magnetic beads (Miltenyi shown that myeloid cells derived from Nlrc5 knockout mice pre- Biotec). CD11b+, CD8+ T, and B cells were enriched using a two-step sented alterations in neither inflammasome activation nor in the procedure. Anti-CD11b (clone M1/70), anti-CD8a (Ly-2), and anti-B220 NF-kB and type I IFN pathways (19). Finally, by analogy to (RA3-6B2) were followed by the appropriate anti-fluorochrome magnetic CIITA, it was suggested that overexpression of NLRC5 in a human beads (Miltenyi Biotec). Bone marrow-derived macrophages (BMDMs) and bone marrow-derived dendritic cells were differentiated as previously T cell line drives the expression of MHC class I and functionally described (27). Thioglycollate-elicited macrophages (TEMs) were isolated related genes, such as b2 microglobulin (B2m) (14). However, 3 d after i.p. injection of a 10% thioglycollate solution. another study showed that RNAi-mediated NLRC5 knockdown in Enrichment of primary human PBMC populations a murine myeloid cell line led to increased MHC I expression (13). Therefore, it is currently unclear whether NLRC5 modulates the We obtained buffy coats from the Lausanne Blood Transfusion Center. Mononuclear cells were isolated from heparinized blood by density cen- MHC I Ag presentation pathway, and with which outcome. + Whereas data on Nlrc5 mRNA tissue distribution consistently trifugation over a Ficoll–Paque gradient (GE Healthcare). CD4 cells, CD8+ T cells, monocytes, B cells, and NK cells were enriched using anti- indicate expression by immune cells (13, 15, 17), the regulation of CD4, -CD8, -CD14, -CD19, and -CD56 magnetic beads, respectively NLRC5 expression is also controversial. In fact, NLRC5 expres- (Miltenyi Biotec). Cellular enrichment was confirmed by FACS and was at sion was shown to be triggered by the TLR3 ligand polyinosinic- least 90%. polycytidylic acid [poly(I:C)], by viral infection, or by IFN-g (13, FACS Abs, staining, and thymocyte gating strategy 16, 17), suggesting a direct role for both type I and II IFNs. This hypothesis was, however, not directly addressed and seems diffi- Cells were preincubated with anti-CD16/32 (2.4G2) to block FcRs, then surface stained using the following Abs raised against CD8a (Ly-2), CD3e cult to reconcile with the proposed MyD88-dependent induction (145-2C11), CD4 (L3T4), CD11b (M1/70), CD11c (N418), CD19 (1D3), of NLRC5 upon LPS triggering of macrophages (15). CD44 (IM7), CD62L (MEL-14), F4/80 (BM8), H-2Db (28-14-8), H-2Kb In this study we show that NLRC5 expression mostly relies on (AF6-88.5.5.3), MHC II (M5/114.15.2), CD1d (1B1), NK1.1 (PK136), the transcription factor STAT1 and is dependent on type I and II HLA-ABC (W6/32), TCRgd (GL3) (all from eBioscience), B220 (RA3- IFNs. Splenocyte analysis of Nlrc5 knockout mice (Nlrc5D/D) 6B2), and CD25 (7D4) (both from BD Biosciences). Allophycocyanin- + Cy7–labeled streptavidin was from eBioscience. Stainings were performed revealed a modest decrease in CD8 T cell percentages and a with an appropriate combination of fluorophores. To gate on thymocytes, tendency to increased NK cell proportion, supporting the hy- the lineage mixture used contained anti-B220, -NK1.1, -F4/80, -CD11c, pothesis of an altered MHC I display. Whereas the expression of and -TCRgd. Propidium iodide (Sigma-Aldrich) was used to exclude dead H-2 genes was only mildly affected in myeloid and intermedi- cells from in vitro cultures. Samples were acquired on a FACSCanto flow cytometer (Becton Dickinson) and analyzed using FlowJo software (Tree ately so in B cells, it was strikingly decreased in T, NKT, and Star). NK lymphocytes. Accordingly, cytotoxic T cell-mediated killing of Nlrc5-deficient T lymphocytes was severely impaired. More- In vitro T cell proliferation and cytotoxic T cell-mediated over, NLRC5 expression was found to be low in several B cell- killing

by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. and T cell-derived tumor lines. Altogether, our results highlight For CD8+ T cell proliferation assays, 1.5 3 105 CFSE (Sigma-Aldrich)- the importance of NLRC5-dependent MHC I expression on lym- labeled, MACS-purified OT-I T cells were incubated together with 5 3 104 phocytes for their cognate elimination by CD8+ T cells. Nlrc5fl/fl and Nlrc5D/D BMDMs and the indicated doses of OVA (Calbio- chem). The third day, cells were harvested and proliferation assessed by FACS. For OT-I–mediated killing of OVA-pulsed target cells, Nlrc5fl/fl and Materials and Methods Nlrc5D/D MACS-purified target CD4+ and CD8+ T cells or T cell-depleted Mice negative fractions were labeled with two levels of CFSE [briefly, CFSE labeling was performed in PBS 8 min at room temperature, at the final Generation of Nlrc5fl/fl mice (Ozgene) is described in Supplemental Fig. 2. D/D fl/fl concentrations of 1.25 (high) or 0.125 (low) mM] and pulsed for 75 min Nlrc5 mice were generated by crossing Nlrc5 to the CMV-Cre deleter with 10 mM SIINFEKL peptide (Anaspec) at 37˚C. For effector T cell strain (The Jackson Laboratory), leading to deletion of the loxP-flanked generation, MACS-purified T cells were stimulated in vitro with 10 mg/ml http://classic.jimmunol.org region also in germ cells, thus allowing selection of Nlrc5-deleted off- plastic-bound functional grade purified anti-CD3 (145-2C11), 2 mg/ml spring and elimination of the CMV-Cre transgene in the following gen- soluble anti-CD28 (37.51) (eBioscience), and human rIL-2 (Glaxo) for eration. C57BL/6 mice were purchased from Harlan; Nlrc5fl/fl, Nlrc5D/D, 2/2 2/2 2/2 2/2 2/2 2 d. Effector cells were then maintained in IL-2 (5 ng/ml)–supplemented Stat1 (20), Ifng (21), Ifnar1 (22), MyD88 (23), Trif (23, medium and used on day 3 or 4. For killing, 5 3 104 differentially labeled 24), and OT-I (25) mice were bred in our animal facility and treated in Nlrc5fl/fl and Nlrc5D/D targets were incubated overnight with the indicated accordance with the Swiss Federal Veterinary Office guidelines. Unless numbers of in vitro activated CD8+ T cells. otherwise specified, 6- to 12-wk-old mice were used. Immunoblot analysis Downloaded from Tumor cell lines Rabbit polyclonal anti-mouse NLRC5 (aa 1–139) and anti-human NLRC5 Human Jurkat and lymphoma cell lines were maintained as previously (aa 1–116) were generated by Adipogen. Other Abs used were monoclonal described (26). Murine tumor cell lines were grown in DMEM supple- 2 anti-tubulin (Sigma-Aldrich), anti-phosphorylated IkBa, polyclonal anti- m b mented with 10% FCS, 50 M -mercaptoethanol, and antibiotics. EL-4 IkBa, anti–caspase-3, anti-phosphorylated IKKa and -b (all from Cell cells are a spontaneous variant derived from EL-4 cells (in this article Signaling), anti-lamin B1, and anti-b-actin (both from Abcam). called EL-4+), characterized by reduced surface MHC I levels (H. Robson MacDonald, unpublished observations). ELISA Media and reagents Cell culture supernatants were assayed for mouse TNF, IL-6 (eBioscience), and IFN-b (PBL IFN Source) according to the manufacturer’s instructions. The medium used for all T cell in vitro experiments was previously de- scribed (27). Ultrapure LPS and poly(I:C) were from Invivogen, IFN-a11 Quantitative PCR and IFN-b from PBL IFN Source, CpG 1826 from Microsynth, and IFN-g from Peprotech. Unless otherwise specified, LPS was used at 100 ng/ml, RNA extraction, retrotranscription to cDNA, and expression analysis were IFN-b at 1000 U/ml, IFN-g at 10 ng/ml, poly(I:C) at 1 mg/ml, CpG at done as previously described (28). Primer sequences are available upon 1 mg/ml, and IFN-a at 1000 U/ml. request. 3822 Nlrc5 DELETION IMPAIRS MHC I-MEDIATED KILLING OF LYMPHOCYTES

NLRC5 constructs, transfection, and nuclear fractionation particular induced the formation of a slower migrating band spe- The full-length human NLRC5 cDNA was assembled from three partial cifically detected by the NLRC5 Ab, suggestive of multilevel cDNA clones (GenBank: BX640842, BC063566, and a Jurkat cell-derived regulation of NLRC5 function (Fig. 1D, Supplemental Fig. 1C–E). cDNA covering the 59 region of human NLRC5 cDNA). Full-length We also investigated the role of IFNs in NLRC5 expression by NLRC5, NLRC5 DCARD (aa 191–1866), and NLRC5 DLRR (aa 1– T cells. Although NLRC5 was detectable in ex vivo splenic CD4+ 664) were cloned into a derivative of pCR3 (Invitrogen) in frame with an and CD8+ T cells, protein levels increased following TCR trig- N-terminal VSV-tag. HEK 293T cells were transfected by the calcium- + phosphate method. Jurkat cells were transfected with expression constructs gering, with more rapid kinetics in CD8 T cells (Fig. 1E, Sup- 2/2 or mock together with plasmid encoding SV40 Large T Ag and plasmid plemental Fig. 1F). As demonstrated by the use of Stat1 , encoding enhanced GFP as previously described for BJAB (26), using Ifnar12/2, and Ifng2/2 T lymphocytes, NLRC5 increase was fully 260 V. Cytoplasmic and nuclear fractions were prepared as previously dependent on autocrine IFN-g signaling in CD4+, and partially in described (29). CD8+ T cells, whereas no role for autocrine type I IFN signaling Chromatin immunoprecipitation was observed. However, we found that Nlrc5 mRNA abundance Chromatin immunoprecipitation (ChIP) experiments were performed as transiently decreased upon TCR triggering, as shown in Supple- previously described (30). Primer sequences are available upon request. mental Fig. 1F, highlighting the existence of important post- transcriptional mechanisms regulating NLRC5 protein levels. Statistical analysis Interestingly, recovery to normal Nlrc5 transcript amounts fol- Statistical differences were calculated with an unpaired Student t test, two- lowing TCR stimulation was dependent on autocrine IFN-g sig- tailed (GraphPad Prism version 5.0). Differences were considered signif- naling (Supplemental Fig. 1F). Accordingly, treatment of CD4+ icant when p # 0.05 (*), very significant when p # 0.01 (**), and ex- and CD8+ T cells with IFN-g was sufficient to increase Nlrc5 tremely significant when p # 0.001 (***). mRNA (Fig. 1F). Altogether, these data show that both type I and Results type II IFNs positively regulate NLRC5 transcription through STAT1, but that additional posttranscriptional mechanisms exist. NLRC5 is highly expressed by lymphocytes D/D mice exhibit minor differences in CD8+ T cell To start, we investigated NLRC5 expression at the protein level and Nlrc5 percentages found it particularly abundant in lymph nodes and spleen (Sup- plemental Fig. 1A). NLRC5 was also clearly detectable in primary To study NLRC5 function, we generated Nlrc5-deficient mice lymphoid organs and tissues highly populated by immune cells, (Supplemental Fig. 2A–E). First, we investigated the role of such as liver and intestine (Supplemental Fig. 1A). A more detailed NLRC5 in NF-kB signaling and type I IFN induction. However, examination of purified splenic cell subsets revealed high NLRC5 NLRC5 ablation did not affect NF-kB activation, TNF, and IL-6 expression in the lymphoid lineage, particularly in CD4+ and CD8+ induction by LPS treatment, or production of IFN-b upon LPS and T cells, whereas it was present to a lesser degree in myeloid cells intracellular poly(I:C) stimulation of BMDMs (Supplemental Fig. (Fig. 1A). A similar pattern was confirmed in human cells, where 2F–I). Furthermore, Nlrc5D/D T cells did not show altered TNF it was also detected in CD56+ NK/NKT cells (Fig. 1B). NLRC5 is secretion upon TCR triggering (Supplemental Fig. 2J). therefore constitutively expressed in immune cells, particularly We next sought to analyze the lymphoid compartment of Nlrc5D/D

by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. by lymphocytes. animals, given the prominent expression of NLRC5 in these cells. Irrespective of their age, Nlrc5D/D mice exhibited only mild dif- NLRC5 is induced by a STAT1-dependent pathway ferences in the proportions of lymphocyte subsets (Fig. 2, Sup- We next investigated the regulation of NLRC5 expression. In plemental Fig. 3A). We consistently observed a slight decrease in BMDMs, Nlrc5 transcript levels were found to be already in- the percentage of CD8+ T cells in spleen, lymph nodes, and blood. creased 4 h after exposure to LPS (Fig. 1C, Supplemental Fig. We therefore analyzed the ability of Nlrc5fl/fl or Nlrc5D/D T cells to 1B). Accordingly, NLRC5 protein augmented following incuba- proliferate and survive upon TCR triggering in vitro, and the ac- tion with LPS in a time-dependent manner in both BMDMs and tivation state ex vivo, and did not observe major differences TEMs (Fig. 1D, Supplemental Fig. 1C, 1D). Of interest, this up- (Supplemental Fig. 3B, 3C). Of interest, NK cells showed a ten- http://classic.jimmunol.org regulation was dependent on the TLR4 adaptor Trif and not on dency to increased percentages in the spleen and the blood of 2/2 2/2 MyD88, as shown by the use of Trif and MyD88 BMDMs young animals (8–12 wk old) (Fig. 2, Supplemental Fig 3A). (Supplemental Fig. 1B–D). This finding suggested a role for Importantly, whereas spleen cellularity in young Nlrc5D/D mice autocrine IFN signaling in the induction of NLRC5, a hypothesis was similar, compared with that in Nlrc5fl/fl controls, older (34- to also supported by the ability of IFNs to induce NLRC5 at both the 36-wk-old) Nlrc5D/D mice often exhibited enlarged spleens (Sup- mRNA and protein level (Fig. 1C, 1D). Accordingly, NLRC5 plemental Fig. 3D). Taken together, these results reveal a mild induction was particularly high when using agonists of TLR3 and +

Downloaded from alteration in the homeostasis of the CD8 lymphoid compartment TLR4, known to strongly induce type I IFN production, and was in Nlrc5D/D animals. abrogated when the autocrine effects of type I IFNs were pre- vented using Ifnar12/2 or Stat12/2 BMDMs (Fig. 1D, Supple- Selected Nlrc5-deficient lymphocyte subsets display markedly mental Fig. 1B–D). Consistent with previous reports, IFN-g also reduced MHC I levels induced NLRC5, an effect preserved in Ifnar12/2 but lost in The differences in CD8+ T cell percentages prompted us to assess Stat12/2 BMDMs (Fig. 1D) (13, 16). However, upon LPS treat- the levels of MHC I, as these cells require MHC I for their thymic ment, only a negligible role for autocrine IFN-g was observed, as selection and peripheral maintenance. Indeed, splenocyte analysis shown by the use of Ifng2/2 BMDMs and TEMs (Fig. 1D, Sup- revealed a dramatic reduction in the display of both H-2D and plemental Fig. 1B–D). H-2K MHC I molecules by Nlrc5D/D CD4+,CD8+ T, NK, NKT, Similar to myeloid cells, B lymphocytes express a range of TLRs and gd T cells (Fig. 3A); an intermediate reduction among B cells; and respond to their agonists. Indeed, LPS-dependent type I IFN and a mild decrease in conventional dendritic cells (cDCs) (Fig. was largely responsible for sustained NLRC5 induction, as dem- 3B). As STAT1 is crucial for NLRC5 expression (Fig. 1E), we onstrated by decreased NLRC5 expression in Ifnar12/2 B cells assessed MHC I levels in T cells derived from Stat1-deficient mice. (Supplemental Fig. 1E). Of interest, exposure to type I IFN in We observed defective MHC I expression by these cells, and the The Journal of Immunology 3823

FIGURE 1. NLRC5 expression is high in lymphocytes, and its induction is STAT1 dependent. (A and B) NLRC5 expression in the indicated splenocyte populations isolated from C57BL/6 animals (A), or among the indicated human PBMC populations (B), was determined by immunoblot analysis. Tubulin was used as loading control. (C) The abundance of Nlrc5 mRNA (relative to Hprt) upon LPS or IFN-b stimulation was determined by quantitative PCR in BMDMs at the indicated times. (D and E) NLRC5 expression was determined by immunoblot analysis in the indicated cell types and stimulatory con- ditions. (D) BMDMs derived from wild-type (wt), Ifnar12/2, Ifng2/2, and Stat12/2 mice were stimulated for 16 h with poly(I:C), LPS, CpG, IFN-a, IFN- b, or IFN-g.(E) Splenic CD4+ and CD8+ T cells from either wt, Ifnar12/2, Ifng2/2,orStat12/2 mice were stimulated for 2 d on plastic-coated anti-CD3 Ab, in the presence of soluble anti-CD28 and IL-2. Cell lysates were collected before stimulation, at day (d) 2 or 5. (F) The abundance of Nlrc5 mRNA following 16 h of IFN-g stimulation was determined by quantitative PCR in CD4+ and CD8+ T cells. One representative experiment of at least two is shown (A–F). **p # 0.01, ***p # 0.001.

by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. defect was more pronounced for H-2K, as in the case of Nlrc5 delineate the specificity of NLRC5 in controlling MHC I ex- deficiency (Fig. 3C). Moreover, the partial dependency of NLRC5 pression, we also investigated the levels of MHC II and the non- expression on STAT1 observed in CD8+ T cells (Fig. 1E) was classical MHC I CD1d on APCs (Fig. 3B). Both these MHC reflected in their milder defect in MHC I expression (Fig. 3C). To molecules were only moderately altered, emphasizing the speci- ficity of NLRC5 in controlling the display of MHC I proteins. We then assessed whether the perturbed levels of MHC I ob- served on splenocytes could also be found in the thymus, the organ where T cell selection proceeds. Indeed, MHC I expression was dramatically decreased on single positive (SP) thymocytes (Fig. http://classic.jimmunol.org 3D). Moderately decreased levels of H-2 molecules were also apparent on cDCs and on cortical and medullary thymic epithelial cells (TECs) (Supplemental Fig. 3E, 3F). Nonetheless, thymocyte development did not present overt abnormalities in Nlrc5D/D mice, except for a slight decrease in CD8+ SP (Supplemental Fig. 3G, 3H).

Downloaded from Altogether, these findings demonstrate that NLRC5 is essential for MHC I display on T, NKT, and NK cells. Of interest, although the display of MHC I in TECs and thymic cDCs is partially reduced in Nlrc5D/D mice, CD8+ T cell numbers and activation state are fairly normal. NLRC5 drives H-2 transcription and requires both CARD and LRR domains FIGURE 2. Nlrc5 deficiency mildly affects lymphocyte homeostasis. To delineate the mechanism by which NLRC5 regulates MHC I Percentages of splenic CD4+ (CD3+, CD4+), CD8+ (CD3+, CD8+), gd + + + + + expression, we assessed the abundance of H-2K and H-2D mRNA (CD3 , TCRgd ) T cells, NKT cells (NK1.1 , CD3 ), NK cells (NK1.1 , fl/fl D/D 2 + fl/fl D/D in Nlrc5 and Nlrc5 T cells (Fig. 4A). Interestingly, we de- CD3 ), and B cells (CD19 ) from Nlrc5 and Nlrc5 mice are shown. D/D Data illustrate mean 6 SEM; results for individual mice are depicted (n = tected decreased amounts of both mRNAs in Nlrc5 T cells. 7) and are representative of at least three independent experiments, al- As MHC I is formed by the association of the H-2 chain with though the increase in gd T cell percentages is not always observed. *p # B2m, we tested the abundance of B2m mRNA and found that the 0.05, ***p # 0.001. latter was affected to a milder extent. We next overexpressed 3824 Nlrc5 DELETION IMPAIRS MHC I-MEDIATED KILLING OF LYMPHOCYTES

FIGURE 3. MHC I expression in lymphocytes is regulated by NLRC5. (A and B) Expression of dif- ferent MHC molecules on the indicated splenic Nlrc5fl/fl and Nlrc5D/D subpopulations was analyzed. Graphs illustrate H-2D and H-2K geometric mean fluorescence intensities (MFIs) (A). Graphs show MFIs of H-2D, H-2K, MHC II, and CD1d expression by B cells and cDCs (CD11chi, CD11bint)(B). (C) Graphs illustrate H-2D and H-2K MFIs of wt and Stat12/2 CD4+ and CD8+ T cells. (D) Graphs depict H-2K and H-2D MFIs on double positive (DP) (line- age2,CD4+,CD8+), CD4+ SP (lineage2, CD3+, CD4+, CD82), and CD8+ SP (lineage2, CD3+, CD42, CD8+) thymocyte subpopulations. Graphs depict mean 6 SEM of MFIs (n = 3), and results are representative of at least two independent experiments (A–D). *p #

by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. 0.05, **p # 0.01, ***p # 0.001. http://classic.jimmunol.org Downloaded from

NLRC5 in the T lymphocyte-derived Jurkat cell line and measured toplasmic and nuclear fractions of splenocytes (Fig. 4C), in both MHC I expression by FACS. We found that whereas constructs the presence and the absence of IFN-b treatment. lacking the CARD or the LRR domain were inactive, full-length To exert its transcriptional regulatory activity, CIITA is recruited NLRC5 strongly increased HLA-AB and -C levels (Fig. 4B). to the promoters of MHC II genes by a DNA-bound protein These data support the hypothesis that NLRC5 acts as a tran- complex called “enhanceosome” (31–33). Enhanceosome factors scriptional regulator; we therefore tested the ability of NLRC5 to bind cooperatively to the SXY sequence, a conserved regulatory shuttle into the nucleus. Indeed, we detected NLRC5 in both cy- module found in the proximal promoter regions of MHC II, MHC The Journal of Immunology 3825

FIGURE 4. NLRC5 occupies the promoters and drives transcription of MHC I genes. (A) Abundance of H-2D, H-2K, and B2m mRNA (relative to Hprt) in Nlrc5D/D normalized to Nlrc5fl/fl CD8+ T cells (mean and SD). **p # 0.01. (B) Jurkat cells were cotransfected with plasmids encoding EGFP and empty vector (–), human full-length NLRC5 (FL), NLRC5 lacking the CARD (DCARD) or lacking the LRR region (DLRR). The graph shows the average MFI of surface HLA-A, -B, and -C expression among EGFPhi cells of two independent experiments. (C) NLRC5 presence in cytoplasmic (Cy) and nuclear (Nu) fractions of splenocytes treated or not with IFN-b for a 16-h period was determined by immunoblot analysis. Lamin B1 and caspase-3 were used as nuclear and cytoplasmic markers, respectively. Results are representative of at least three independent experiments (A–C). (D) Abs specific to mouse NLRC5 were used to immunoprecipitate cross-linked chromatin fragments derived from Nlrc5fl/fl or Nlrc5D/D T cells. Immunoprecipitates were analyzed by quantitative PCR for the abundance of SXY sequences of the H-2D, H-2K, and B2m promoters, and for the promoter sequence of TATA binding protein (TBP). Relative promoter binding is shown. The average and SD of three experiments are depicted.

I, and related genes (34, 35). By analogy, we assessed the ability eliminate infected or transformed cells. We thus investigated the of NLRC5 to occupy the proximal SXY module of the H-2D, H- cytolytic T lymphocyte-dependent killing of Nlrc5D/D T cell tar- 2K, and B2m genes by ChIP. As shown in Fig. 4D, endogenous gets, which present a strong defect in MHC I display. For this NLRC5 was efficiently recruited to the promoter regions of these purpose, Nlrc5fl/fl and Nlrc5D/D T cells were labeled with distin- genes in T cells, whereas binding to the promoter of an unrelated guishable dim and bright CFSE levels, respectively; pulsed with gene (TBP) was not observed. The specificity of our Ab in the the SIINFEKL peptide; and incubated overnight with OT-I CD8+ ChIP assay was ensured by the use of Nlrc5-deficient T cells. effector T cells. Nlrc5D/D target T lymphocytes were much less Altogether, these observations indicate that NLRC5 acts as a efficiently killed by OT-I CD8+ T cells, compared with their transcriptional regulator of MHC I genes and requires both the Nlrc5fl/fl counterparts (Fig. 6A, Supplemental Fig. 4C). Further- CARD and the LRRs for this activity. However, mRNA levels of more, when unpulsed Nlrc5fl/fl and Nlrc5D/D T cells were used as B2m are less affected than those of the H-2D and H-2K genes, targets, or polyclonal effector cells employed as killers, no dif- suggesting the existence of additional mechanisms promoting B2m transcription. by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. NLRC5-dependent MHC I expression is not crucial for CD8+ T cell priming Given the fact that IFNs regulate MHC I expression, we next asked whether IFN-induced MHC I was also affected by Nlrc5 deficiency. To this end, we stimulated Nlrc5fl/fl and Nlrc5D/D T cells with type I or II IFN. Of interest, IFNs were able to induce H-2D and H-2K expression in both Nlrc5fl/fl and Nlrc5D/D CD4+ and CD8+ T cells (Fig. 5A, Supplemental Fig. 4A). As a consequence, although still http://classic.jimmunol.org clearly present, the fold difference in MHC I levels between Nlrc5fl/fl and Nlrc5D/D T cells diminished, in particular upon IFN-b exposure. Similar results were obtained in BMDMs, although the overall difference in H-2K expression between Nlrc5fl/fl and Nlrc5D/D cells was minor, compared with what was observed in T cells (Supplemental Fig. 4B). The mild decrease in MHC I D/D Downloaded from levels found on Nlrc5 APCs led us to attest to the competence of Nlrc5fl/fl and Nlrc5D/D BMDMs to prime transgenic OT-I CD8+ T cells, whose cognate peptide (SIINFEKL) is presented in the context of H-2Kb. Indeed, OT-I T cell proliferation induced by Nlrc5D/D BMDMs was not impaired, compared with that stimu- lated by Nlrc5D/D BMDMs (Fig. 5B). Thus, consistent with the milder defect in MHC I levels in professional APCs, which is further attenuated under inflammatory + conditions, NLRC5 does not play a major role in CD8 T cell FIGURE 5. Nlrc5 deficiency does not prevent MHC I induction by IFN priming. and CD8+ T cell priming by APCs. (A) Histograms show H-2D and H-2K fl/fl D/D + NLRC5 expression by target lymphocytes is essential for their fluorescence by Nlrc5 and Nlrc5 CD4 splenic T cells treated for 16 h B cognate elimination with IFN-b or IFN-g.( ) CFSE profile of OT-I cells cocultured for 3 d with Nlrc5fl/fl and Nlrc5D/D BMDMs, supplied with the indicated doses MHC I expression on nearly all nucleated cells is an essential of OVA (mg/ml) and LPS (10 ng/ml). Data are representative of at least safeguard mechanism allowing CD8+ T cells to identify and three experiments (A and B). 3826 Nlrc5 DELETION IMPAIRS MHC I-MEDIATED KILLING OF LYMPHOCYTES

FIGURE 6. Impaired cytotoxic T cell-mediated killing of Nlrc5D/D target lymphocytes. (A and B) Target T cells isolated from Nlrc5fl/fl and Nlrc5D/D (A), and the T cell-depleted fractions (B), were labeled with dim and bright CFSE concentrations, respectively; pulsed with SIINFEKL peptide; and mixed in a 1:1 ratio. The indicated proportions of effector OT-I cells were then supplied for an overnight period (E:T ratio). Histograms show the relative abundance of surviving [gated on living, propidium iodide (PI)2 cells] Nlrc5fl/fl and Nlrc5D/D target T cells (A) or T cell-depleted splenocytes (B). (C and D) The indicated murine (C) and human (D) tumoral B and T cell lines were screened for the expression levels of NLRC5 by immunoblot analysis. Data are representative of at least two experiments. HD, Healthy donor controls.

ferences in the survival of two target cell populations was ob- count for the remaining NLRC5 expression detectable in STAT1- served (Supplemental Fig. 4C). This finding indicates that the deficient cells. In vitro stimulation of BMDMs highlighted the role preferential killing of Nlrc5fl/fl target T cells is the result of cog- of autocrine type I IFN signaling in NLRC5 induction, whereas nate elimination. We also used T cell-depleted splenocytes, for the NLRC5 expression by TCR-triggered T cells mostly relied on vast majority consisting of B cells, as target cells (Fig. 6B). Al- autocrine IFN-g. though much less pronounced, a survival advantage of T cell- In agreement with Kumar et al. (19), we did not observe defects depleted Nlrc5D/D target cells was also observed, in agreement in NF-kB activation or in type I IFN induction in Nlrc5-deficient with the milder defect in MHC I display detected in B cells (Fig. cells. We found NLRC5 to be crucial for the transcriptional reg- 3B). ulation of MHC I expression in vivo. First, the levels of H-2D and Given that low NLRC5 expression confers an advantage to H-2K proteins and transcripts were markedly reduced in Nlrc5D/D escape CD8+ T cell-mediated elimination, and that downregula- T cells, compared with Nlrc5fl/fl cells. Second, NLRC5 was also tion of surface MHC I molecules is known to mediate immune found to localize in the nucleus and occupy the conserved SXY surveillance evasion (36), we investigated NLRC5 levels in sev- promoter sequence of these genes, in agreement with overex- by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. eral T cell- and B cell-derived tumor lines (Fig. 6C, 6D). Notably, pression data by Meissner et al. (14). These results highlight the the tested murine thymoma (EL-4+, EL-42, EL-4 6.1.10, and ST- analogy between CIITA and NLRC5, and future studies will tell 4), T cell (LSTRA, RMA, RMA-S, and YAC), and B lymphoma whether NLRC5 as well acts as a coactivator of MHC I tran- cell lines (L1210 and M12.4.1) had lower NLRC5 expression than scription together with a specific enhanceosome. did primary splenocytes or B cells (Fig. 6C). Importantly, the We identified a key role for the CARD and LRR regions of tested human cell lines behaved similarly, with Jurkat T cells and NLRC5 in driving MHC I a-chain transcription. The LRR portion half of the B cell lymphomas (U2932, SUDHL-4, SUDHL-6, and is particularly extended and could be involved in the interaction BJAB) expressing negligible NLRC5 levels (Fig. 6D). For the with other transcriptional regulators, similarly to the homologous

http://classic.jimmunol.org C57BL/6 murine and the human cell lines, we also measured region of CIITA (32). H-2D, H-2K, and HLA-A mRNA levels, respectively (Supple- Importantly, we show that the extent of NLRC5-dependent mental Fig. 4D, 4E). We found that abundant MHC I a-chain MHC I expression differs depending on the cell type and on the mRNA always correlated with the presence of NLRC5. However, activation state of the cell. In fact, resting Nlrc5D/D lymphoid T, in some human cell lines the presence of NLRC5 did not lead to NKT, and NK cells were virtually devoid of MHC I expression, substantial HLA-A mRNA, indicating the existence of additional whereas the defect was generally milder in myeloid and B cells, mechanisms negatively regulating transcript levels. reflecting the distribution of NLRC5 expression. The mild MHC I Downloaded from In conclusion, these data demonstrate the importance of reduction observed in APCs could also be explained by the NLRC5-dependent MHC I expression in the cognate elimination presence of CIITA in these cells, as this transcriptional regulator of lymphocytes by cytotoxic T cells and consistently show that participates in MHC I expression as well (37). IFN treatment also NLRC5 expression is low in a large number of lymphoid tumor induced MHC I in Nlrc5-deficient cells, although the expression of cell lines. MHC I in Nlrc5D/D cells remained markedly lower compared with that in control cells. These results confirm the importance of Discussion NLRC5 but indicate that under inflammatory circumstances other In this article, we report the predominant distribution and function factors can concur in MHC I regulation (38). Similar to what we of an NLR in cells of the lymphoid lineage, as NLRC5 is abun- observed for ex vivo cDCs, we found a mild reduction in the levels dantly expressed and required for the regulation of MHC I ex- of MHC I displayed on in vitro-generated BMDMs of Nlrc5D/D pression in lymphocytes. We found that STAT1 plays a crucial role origin, which became even less apparent following inflamma- in NLRC5 induction, consistent with the presence of STAT1 tory challenge. Accordingly, overt differences in the priming of binding sites in the Nlrc5 promoter region, where, in addition, an transgenic OT-I T cells by Nlrc5D/D BMDMs were not observed. NF-kB binding site has been identified (16). The latter may ac- However, as OT-I T cells bear a TCR with high affinity for the The Journal of Immunology 3827

cognate peptide, it is still possible that priming of low- and presentation thus sheds new light on the function of NLRs in intermediate-affinity TCR-bearing T cells is affected by reduced immunity, making them instrumental players not only in the innate MHC I expression in Nlrc5-deficient APCs. response but also in presiding over the adaptive branch of the As MHC molecules play a crucial role in thymocyte selection immune system. processes, we analyzed MHC expression on thymic subpopulations and in thymocyte development. MHC I was strongly reduced Acknowledgments among SP thymocytes, in agreement with the appearance of Nlrc5 We thank T. Decker (Department of Microbiology and Immunobiology, Uni- expression at this differentiation stage (13). Moreover, although versity of Vienna, Vienna, Austria), L.M. Staudt (National Institutes of TECs and thymic cDCs displayed diminished MHC I levels, only Health, Bethesda, MD), M. Aguet (l’Institut Suisse de Recherche Expe´r- a minor decrease in Nlrc5D/D CD8+ T cell proportion was ob- imentale sur le Cancer, Lausanne, Switzerland), and the New York Univer- served. The fairly normal number of peripheral CD8+ T cells was sity School of Medicine (New York, NY) for cell lines and knockout mice; not the consequence of lymphopenia-driven proliferation in the S. Hailfinger, A. Wilson, K. Maslowski, J. Anken, and P. Schneider for periphery, as splenic CD8+ T lymphocytes did not show a signifi- technical help and advice; and M. Thome and H. Acha-Orbea (University cantly altered activation state. On the one hand, the slight decrease of Lausanne, Lausanne, Switzerland) for critical reading of the manuscript. in CD8+ T cell percentage in Nlrc5D/D mice echoes the absence of major defects in priming by APCs. On the other hand, it is Disclosures tempting to speculate that CD8+ T cells may undergo altered The authors have no financial conflicts of interest. thymic selection, skewing the TCR repertoire toward receptors with higher cross-reactivity, compensating for the slightly reduced References MHC I levels encountered in the thymus. 1. Martinon, F., A. Mayor, and J. Tschopp. 2009. The inflammasomes: guardians The presence of normal MHC I levels on some cell types, and of the body. Annu. Rev. Immunol. 27: 229–265. decreased levels on others, leads to a MHC I chimerism in Nlrc5D/D 2. Martinon, F., K. Burns, and J. Tschopp. 2002. The inflammasome: a molecular D/D platform triggering activation of inflammatory caspases and processing of proIL- animals that may perturb NK cell education. Intriguingly, Nlrc5 beta. Mol. Cell 10: 417–426. animals presented a trend toward increased NK cell numbers. 3. Mariathasan, S., K. Newton, D. M. Monack, D. Vucic, D. M. French, W. P. Lee, Although lowered MHC I expression favors NK cell-mediated M. Roose-Girma, S. Erickson, and V. M. Dixit. 2004. Differential activation of the inflammasome by caspase-1 adaptors ASC and Ipaf. Nature 430: 213–218. killing, it also represents an effective way for potentially hazard- 4. Inohara, N., T. Koseki, L. del Peso, Y. Hu, C. Yee, S. Chen, R. Carrio, J. Merino, ous cells to evade cytotoxic T cells. Indeed, lymphocytes derived D. Liu, J. Ni, and G. Nu´n˜ez. 1999. Nod1, an Apaf-1-like activator of caspase-9 D/D + and nuclear factor-kappaB. J. Biol. Chem. 274: 14560–14567. from Nlrc5 animals were less efficiently killed by CD8 T cells, ´ fl/fl 5. Girardin, S. E., I. G. Boneca, L. A. Carneiro, A. Antignac, M. Jehanno, J. Viala, compared with their Nlrc5 counterparts. In agreement with a K. Tedin, M. K. Taha, A. Labigne, U. Za¨hringer, et al. 2003. Nod1 detects physiologic relevance of the reduced recognition and lysis of a unique muropeptide from gram-negative bacterial peptidoglycan. Science 300: 1584–1587. Nlrc5-deficient cells, NLRC5 expression was low in several mu- 6. Ogura, Y., N. Inohara, A. Benito, F. F. Chen, S. Yamaoka, and G. Nunez. 2001. rine transformed cell lines. In addition, several human tumor cell Nod2, a Nod1/Apaf-1 family member that is restricted to monocytes and acti- lines showed decreased NLRC5 expression, particularly Jurkat vates NF-kappaB. J. Biol. Chem. 276: 4812–4818. 7. Girardin, S. E., I. G. Boneca, J. Viala, M. Chamaillard, A. Labigne, G. Thomas, T cells and the tested germinal center B cell-like diffuse large B D. J. Philpott, and P. J. Sansonetti. 2003. Nod2 is a general sensor of peptidoglycan

by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. cell lymphomas. Importantly, reconstitution of NLRC5 expression through muramyl dipeptide (MDP) detection. J. Biol. Chem. 278: 8869–8872. was sufficient to restore MHC I display in Jurkat T cells. More- 8. Sabbah, A., T. H. Chang, R. Harnack, V. Frohlich, K. Tominaga, P. H. Dube, Y. Xiang, and S. Bose. 2009. Activation of innate immune antiviral responses by over, only tumors expressing NLRC5 showed a substantial amount Nod2. Nat. Immunol. 10: 1073–1080. of transcripts encoding MHC I a-chain genes. Nonetheless, the 9. Watanabe, T., N. Asano, S. Fichtner-Feigl, P. L. Gorelick, Y. Tsuji, Y. Matsumoto, T. Chiba, I. J. Fuss, A. Kitani, and W. Strober. 2010. NOD1 levels of MHC I transcripts were not always correlating with contributes to mouse host defense against Helicobacter pylori via induction of surface protein expression, in accordance with the existence of type I IFN and activation of the ISGF3 signaling pathway. J. Clin. Invest. 120: additional posttranscriptional regulatory mechanisms controlling 1645–1662. 10. Steimle, V., L. A. Otten, M. Zufferey, and B. Mach. 1993. Complementation MHC expression (39). Taken together, our results suggest that cloning of an MHC class II transactivator mutated in hereditary MHC class II transformed cells may take advantage of NLRC5 downmod- deficiency (or bare lymphocyte syndrome). Cell 75: 135–146. 11. Chang, C. H., S. Guerder, S. C. Hong, W. van Ewijk, and R. A. Flavell. 1996. http://classic.jimmunol.org ulation, and future studies aimed at evaluating NLRC5 levels in Mice lacking the MHC class II transactivator (CIITA) show tissue-specific im- primary transformed lymphocytes will be crucial to verify this pairment of MHC class II expression. Immunity 4: 167–178. hypothesis. 12. Ting, J. P., J. A. Duncan, and Y. Lei. 2010. How the noninflammasome NLRs function in the . Science 327: 286–290. Several devote part of their restricted genome to encode 13. Benko, S., J. G. Magalhaes, D. J. Philpott, and S. E. Girardin. 2010. NLRC5 proteins interfering with MHC I expression (40, 41). By decreasing limits the activation of inflammatory pathways. J. Immunol. 185: 1681–1691. MHC I display, infected cells evade CD8+ T cell-mediated elim- 14. Meissner, T. B., A. Li, A. Biswas, K. H. Lee, Y. J. Liu, E. Bayir, D. Iliopoulos, P. J. van den Elsen, and K. S. Kobayashi. 2010. NLR family member NLRC5 is

Downloaded from ination. According to our results, the modulation of NLRC5 by a transcriptional regulator of MHC class I genes. Proc. Natl. Acad. Sci. USA 107: viruses having lymphocyte tropism, such as HIV, would represent 13794–13799. an attractive way to escape cytotoxic T cells. It is also predictable 15. Cui, J., L. Zhu, X. Xia, H. Y. Wang, X. Legras, J. Hong, J. Ji, P. Shen, S. Zheng, Z. J. Chen, and R. F. Wang. 2010. NLRC5 negatively regulates the NF-kappaB that NLRC5 plays an important role in transplant rejection. In this and type I signaling pathways. Cell 141: 483–496. context, therapeutic approaches targeting NLRC5 could reduce 16. Kuenzel, S., A. Till, M. Winkler, R. Ha¨sler, S. Lipinski, S. Jung, J. Gro¨tzinger, + H. Fickenscher, S. Schreiber, and P. Rosenstiel. 2010. The nucleotide-binding immune rejection without hampering CD8 T cell priming. oligomerization domain-like receptor NLRC5 is involved in IFN-dependent NLRs are well known for their potent activity in the onset of antiviral immune responses. J. Immunol. 184: 1990–2000. inflammatory responses, particularly in innate immune cells. Our 17. Neerincx, A., K. Lautz, M. Menning, E. Kremmer, P. Zigrino, M. Ho¨sel, H. Bu¨ning, R. Schwarzenbacher, and T. A. Kufer. 2010. A role for the human findings clearly identify NLRC5 as a second NLR family member nucleotide-binding domain, leucine-rich repeat-containing family member regulating MHC expression. These observations raise the fasci- NLRC5 in antiviral responses. J. Biol. Chem. 285: 26223–26232. nating hypothesis that NLRs may control MHC expression in a 18. Davis, B. K., R. A. Roberts, M. T. Huang, S. B. Willingham, B. J. Conti, W. J. Brickey, B. R. Barker, M. Kwan, D. J. Taxman, M. A. Accavitti-Loper, much broader manner than has been recognized so far with the et al. 2011. Cutting edge: NLRC5-dependent activation of the inflammasome. J. well-studied CIITA system. If so, the possibility exists that addi- Immunol. 186: 1333–1337. 19. Kumar, H., S. Pandey, J. Zou, Y. Kumagai, K. Takahashi, S. Akira, and T. Kawai. tional transcriptional regulators belonging to the NLR family may 2011. NLRC5 deficiency does not influence cytokine induction by and participate in this process. The evidence that NLRC5 governs Ag infections. J. Immunol. 186: 994–1000. 3828 Nlrc5 DELETION IMPAIRS MHC I-MEDIATED KILLING OF LYMPHOCYTES

20. Durbin, J. E., R. Hackenmiller, M. C. Simon, and D. E. Levy. 1996. Targeted 30. Masternak, K., N. Peyraud, M. Krawczyk, E. Barras, and W. Reith. 2003. disruption of the mouse Stat1 gene results in compromised innate immunity to Chromatin remodeling and extragenic transcription at the MHC class II locus viral disease. Cell 84: 443–450. control region. Nat. Immunol. 4: 132–137. 21. Dalton, D. K., S. Pitts-Meek, S. Keshav, I. S. Figari, A. Bradley, and 31. Zhu, X. S., M. W. Linhoff, G. Li, K. C. Chin, S. N. Maity, and J. P. Ting. 2000. T. A. Stewart. 1993. Multiple defects of immune cell function in mice with Transcriptional scaffold: CIITA interacts with NF-Y, RFX, and CREB to cause disrupted interferon-gamma genes. Science 259: 1739–1742. stereospecific regulation of the class II major histocompatibility complex pro- 22. Mu¨ller, U., U. Steinhoff, L. F. Reis, S. Hemmi, J. Pavlovic, R. M. Zinkernagel, moter. Mol. Cell. Biol. 20: 6051–6061. and M. Aguet. 1994. Functional role of type I and type II in antiviral 32. Masternak, K., A. Muhlethaler-Mottet, J. Villard, M. Zufferey, V. Steimle, and defense. Science 264: 1918–1921. W. Reith. 2000. CIITA is a transcriptional coactivator that is recruited to MHC 23. Adachi, O., T. Kawai, K. Takeda, M. Matsumoto, H. Tsutsui, M. Sakagami, class II promoters by multiple synergistic interactions with an enhanceosome K. Nakanishi, and S. Akira. 1998. Targeted disruption of the MyD88 gene results complex. Genes Dev. 14: 1156–1166. in loss of IL-1- and IL-18-mediated function. Immunity 9: 143–150. 33. Jabrane-Ferrat, N., N. Nekrep, G. Tosi, L. Esserman, and B. M. Peterlin. 2003. 24. Yamamoto, M., S. Sato, H. Hemmi, K. Hoshino, T. Kaisho, H. Sanjo, MHC class II enhanceosome: how is the class II transactivator recruited to DNA- O. Takeuchi, M. Sugiyama, M. Okabe, K. Takeda, and S. Akira. 2003. Role of bound activators? Int. Immunol. 15: 467–475. adaptor TRIF in the MyD88-independent toll-like receptor signaling pathway. 34. Reith, W., and B. Mach. 2001. The bare lymphocyte syndrome and the regulation Science 301: 640–643. of MHC expression. Annu. Rev. Immunol. 19: 331–373. 25. Clarke, S. R., M. Barnden, C. Kurts, F. R. Carbone, J. F. Miller, and W. R. Heath. 35. van den Elsen, P. J., T. M. Holling, H. F. Kuipers, and N. van der Stoep. 2004. 2000. Characterization of the ovalbumin-specific TCR transgenic line OT-I: Transcriptional regulation of antigen presentation. Curr. Opin. Immunol. 16: 67– MHC elements for positive and negative selection. Immunol. Cell Biol. 78: 75. 110–117. 36. Vesely, M. D., M. H. Kershaw, R. D. Schreiber, and M. J. Smyth. 2011. Natural 26. Hailfinger, S., G. Lenz, V. Ngo, A. Posvitz-Fejfar, F. Rebeaud, M. Guzzardi, innate and adaptive immunity to cancer. Annu. Rev. Immunol. 29: 235–271. E. M. Penas, J. Dierlamm, W. C. Chan, L. M. Staudt, and M. Thome. 2009. 37. Gobin, S. J., A. Peijnenburg, V. Keijsers, and P. J. van den Elsen. 1997. Site alpha Essential role of MALT1 protease activity in activated B cell-like diffuse large is crucial for two routes of IFN gamma-induced MHC class I transactivation: the B-cell lymphoma. Proc. Natl. Acad. Sci. USA 106: 19946–19951. ISRE-mediated route and a novel pathway involving CIITA. Immunity 6: 601– 27. Guarda, G., C. Dostert, F. Staehli, K. Cabalzar, R. Castillo, A. Tardivel, 611. P. Schneider, and J. Tschopp. 2009. T cells dampen innate immune responses 38. Gobin, S. J., M. van Zutphen, A. M. Woltman, and P. J. van den Elsen. 1999. through inhibition of NLRP1 and NLRP3 inflammasomes. Nature 460: 269– Transactivation of classical and nonclassical HLA class I genes through the IFN- 273. stimulated response element. J. Immunol. 163: 1428–1434. 28. Guarda, G., M. Braun, F. Staehli, A. Tardivel, C. Mattmann, I. Fo¨rster, M. Farlik, 39. Donaldson, J. G., and D. B. Williams. 2009. Intracellular assembly and traf- T. Decker, R. A. Du Pasquier, P. Romero, and J. Tschopp. 2011. Type I interferon ficking of MHC class I molecules. Traffic 10: 1745–1752. inhibits interleukin-1 production and inflammasome activation. Immunity 34: 40. Petersen, J. L., C. R. Morris, and J. C. Solheim. 2003. Virus evasion of MHC 213–223. class I molecule presentation. J. Immunol. 171: 4473–4478. 29. Janssens, S., A. Tinel, S. Lippens, and J. Tschopp. 2005. PIDD mediates 41. Yewdell, J. W., and A. B. Hill. 2002. Viral interference with antigen presentation. NF-kappaB activation in response to DNA damage. Cell 123: 1079–1092. Nat. Immunol. 3: 1019–1025. by guest on October 1, 2021. Copyright 2012 Pageant Media Ltd. http://classic.jimmunol.org Downloaded from