NLRC5 Controls Basal MHC Class I Expression in an MHC Enhanceosome-Dependent Manner

This information is current as Andreas Neerincx, Galaxia M. Rodriguez, Viktor Steimle of September 27, 2021. and Thomas A. Kufer J Immunol 2012; 188:4940-4950; Prepublished online 6 April 2012; doi: 10.4049/jimmunol.1103136

http://www.jimmunol.org/content/188/10/4940 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2012/04/06/jimmunol.110313 Material 6.DC1 http://www.jimmunol.org/ References This article cites 53 articles, 21 of which you can access for free at: http://www.jimmunol.org/content/188/10/4940.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 27, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

NLRC5 Controls Basal MHC Class I in an MHC Enhanceosome-Dependent Manner

Andreas Neerincx,* Galaxia M. Rodriguez,†,1 Viktor Steimle,† and Thomas A. Kufer*

Nucleotide-binding domain and leucine-rich repeat (NLR) play important roles in innate immune responses as pattern- recognition receptors. Although most NLR proteins act in cell autonomous immune pathways, some do not function as classical pattern-recognition receptors. One such NLR is the MHC class II transactivator, the master regulator of MHC class II gene transcription. In this article, we report that human NLRC5, which we recently showed to be involved in viral-mediated type I IFN responses, shuttles to the nucleus and activates MHC class I gene expression. Knockdown of NLRC5 in different human cell lines and primary dermal fibroblasts leads to reduced MHC class I expression, whereas introduction of NLRC5 into cell types with very low expression of MHC class I augments MHC class I expression to levels comparable to those found in .

Expression of NLRC5 positively correlates with MHC class I expression in human tissues. Functionally, we show that both the N- Downloaded from terminal effector domain of NLRC5 and its C-terminal leucine-rich repeat domain are needed for activation of MHC class I expression. Moreover, nuclear shuttling and function depend on a functional Walker A motif. Finally, we identified a promoter sequence in the MHC class I promoter, the X1 box, to be involved in NLRC5-mediated MHC class I gene activation. Taken together, this suggested that NLRC5 acts in a manner similar to class II transactivator to drive MHC expression and revealed NLRC5 as an important regulator of basal MHC class I expression. The Journal of Immunology, 2012, 188: 4940–4950. http://www.jimmunol.org/ ost cells have evolved a variety of pathogen-recognition most prominent members are Nod1 and Nod2, trigger the ca- receptors, such as the membrane-bound TLRs (1). A nonical NF-kB pathway in a receptor interacting kinase 2- and H second important group of pathogen-recognition recep- inhibitor of IkB kinase-dependent manner upon exposure to bac- tors, the nucleotide-binding domain leucine-rich repeat (NLR) terial peptidoglycan (4). proteins mediate recognition of pathogen-associated molecular The recognition of viral infection is mediated by TLRs, NLRs, patterns (PAMPs) in the cytosol. NLR proteins share a typical other cytoplasmic receptors (e.g., Rig I-like helicases), and DNA- tripartite structure, consisting of an N-terminal death fold effector binding proteins (e.g., AIM-2) (5). A common feature shared by domain, a central NACHT ATPase domain, and a C-terminal most of these viral receptors is the activation of type I IFN pathways leucine-rich repeat (LRR) domain (2). NLRs can be divided into to induce antiviral immunity (6). We recently showed that Sendai by guest on September 27, 2021 two main functional groups: the inflammasome and nodosome - and polyinosinic:polycytidylic acid [poly(I:C)]-mediated families. The pyrin-like domain (PYD)-containing inflammasome type I IFN responses are partially dependent on NLRC5 in hu- NLR proteins, of which the two most prominent members are man cells (7). NLRC5 (alternatively named NOD27 or CLR16.1) NLRP1 and NLRP3, react to PAMPs and danger-associated mo- shows the typical structural organization of an NLR protein, but it lecular patterns by inducing caspase-1 activation and subsequent contains an N-terminal non-CARD non-PYD effector domain of IL-1b processing (3). In contrast, the caspase activation and re- unknown function. Independently, NLRC5 was reported to be in- cruitment domain (CARD)-containing NLRs, of which the two volved in human CMV-mediated type I IFN responses (8). Addi- tionally, it was shown that NLRC5 might negatively regulate NF- *Institute for Medical Microbiology, Immunology, and Hygiene, University of Co- kB–dependent responses and Rig I-like helicase-mediated type I logne, 50931 Cologne, Germany; and †De´partement de Biologie, Universite´ de Sher- IFN responses (9) and may act in synergy with NLRP3 inflam- ´ brooke, Sherbrooke, Quebec J1K 2R1, Canada masome signaling (10). Although most of these findings await in- 1 Current address: Section d’Immunologie, De´partement de Pe´diatrie, Universite´ de dependent verification, they mainly agree in that NLRC5 expres- Sherbrooke, Sherbrooke, Que´bec, Canada. sion is induced by IFN-g and PAMPs, and suggest a role for NLRC5 Received for publication November 3, 2011. Accepted for publication February 28, 2012. in innate immune responses to (11). This work was supported by the German Research Foundation (Grant SFB670 to In addition to innate immune responses, containment of viral T.A.K.). V.S. is supported by a research grant (262845) from the Natural Sciences infection relies on destruction of infected cells via Ag presentation and Engineering Research Council, Canada. by the MHC class I molecules to cytotoxic T lymphocytes of the Address correspondence and reprint requests to Dr. Thomas A. Kufer or Dr. Viktor adaptive . Two recent reports analyzed a potential Steimle, Institute for Medical Microbiology, Immunology, and Hygiene, University of Cologne, Joseph-Stelzmann Strasse 9, Geb. 37, 50931 Cologne, Germany (T.A.K.) involvement of NLRC5 in MHC class I gene expression, but they or De´partement de Biologie, Universite´ de Sherbrooke, 2500 Boulevard Universite´, came to opposite conclusions. Although Meissner et al. (12) re- Sherbrooke, QC J1K 2R1, Canada (V.S.). E-mail addresses: thomas.kufer@uk-koeln. ported that NLRC5 is able to activate MHC class I gene expres- de (T.A.K.) and [email protected] (V.S.) sion in Jurkat, HEK293T, and HeLa cells, Benko et al. (13) found The online version of this article contains supplemental material. that knockdown of NLRC5 in the murine macrophage-like cell Abbreviations used in this article: CARD, caspase activation and recruitment domain; CIITA, class II transactivator; LepB, leptomycin B; LRR, leucine-rich repeat; NLR, line RAW264.7 increased IFN-g–induced MHC class I expres- nucleotide-binding domain and leucine-rich repeat; NLS, nuclear localization signal; sion. NLRC5-deficient mice have been generated, but MHC class PAMP, pathogen-associated molecular pattern; poly(I:C), polyinosinic:polycytidylic I expression was not analyzed in these animals (14). acid; PYD, pyrin-like domain; siRNA, small interfering RNA; wt, wild-type. Although a role for NLRC5 in transcriptional regulation is Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 unusual for NLR proteins, it is not unheard of; there is at least one www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103136 The Journal of Immunology 4941 very well-characterized example of another NLR protein with a Materials and Methods nuclear function (2). Class II transactivator (CIITA), the first de- Cells and cell culture scribed mammalian NLR protein, is the main transcriptional ac- Primary human dermal fibroblasts were obtained and maintained as de- tivator of MHC class II gene expression (15). Mutations in CIITA scribed (7). cause bare syndrome (15), a hereditary, generally fatal HEK293T, HeLa, and B16-F10 cells were cultured in DMEM, and THP1 primary immunodeficiency caused by the complete absence of cells were cultured in VLE RPMI 1640 (Biochrom) containing 10% heat- MHC (in humans: HLA) class II gene expression (16). CIITA is an inactivated FCS, 2 mM glutamine, and 100 U/ml each penicillin and atypical transcriptional activator that does not bind directly to streptomycin in 5% CO2 at 37˚C. DNA but is rather recruited to the highly conserved MHC class II Plasmids promoters via protein–protein interactions with a multisubunit Plasmids encoding myc-tagged NLRC5 full-length and deletion constructs DNA-binding protein complex termed the MHC enhanceosome were cloned into pcDNA3.1-3xmyc-B using restriction digestion cloning (16, 17). The MHC enhanceosome contains the X1 box-binding (EcoRV; XhoI) from the previous described FLAG-tagged NLRC5 (7). RFX complex, the X2 box-binding CREB, and the Y box-binding Plasmids encoding truncated versions of NLRC5 were generated by PCR NF-Y complex (16, 17). Binding of these factors to the S/X/Y cloning in pcDNA3.1-3xmyc-B. These truncated versions contain the following amino acids (referred to NP_115582): myc-NLRC5 isoform 3 aa element in proximal promoters of MHC class II is highly 1–720, myc-NLRC5 DDD-nuclear localization signal (NLS) aa 110–1866, cooperative, with an absolute dependence on RFX and NF-Y (18, myc-NLRC5-DD aa 1–133, and myc-NLRC5 LRR aa 589–1866. N-termi- 19). MHC class II genes are expressed in a complex pattern with nally 2xSV40-NLS–tagged NLRC5 was cloned by PCR into pCDNA3.1- professional APCs showing constitutive expression, whereas most 3xmyc-B using the following primer: fwd: 59-GGGGATATCAAGGATC- somatic cell types are negative for MHC class II expression. CAAAAAAGAAGAGAAAGGTAGATCCAAAAAAGAAGAGAAAGG- Downloaded from TAATGGACCCCGTTGGCCT-39 and rev: 59-GGGCCCCTCGAGGTC- However, expression in the latter can be induced by various GACTCATCAAGTACCCCAAGGGG-39. stimuli, most notably IFN-g. Both constitutive and inducible MHC CIITA constructs in EBV-based expression vector EBS-NPL were de- class II expression is strictly dependent, qualitatively and quan- scribed (15, 31, 32). EGFP-NLRC5 and EGFP-NLRC5 K234A were titatively, on CIITA expression (15, 17, 18, 20–22). cloned into EBSB-PL, a derivative of the EBS-PL expression vector (33), in which the hygromycin-resistance cassette was replaced by a blastici- In contrast to MHC class II genes, MHC class I genes are con- dine-resistance cassette. stitutively expressed on the majority of nucleated cells (23, 24). Luciferase reporter constructs for MHC class I promoter activation were http://www.jimmunol.org/ However, MHC class I gene expression regulation is not simple, described (34). The HLA class II DRA promoter reporter plasmid contains because expression levels vary widely between different tissues, a fragment from position 2150 to +31 bp with respect to the transcription and expression is modulated by different stimuli, such as type I start site amplified from Raji genomic DNA and cloned into the polylinker of pGL3 (Promega). Generation of an HLA-B250”TCGCA” luciferase and II IFNs and other signals (23–25). MHC class I promoters are reporter plasmid was conducted, as described in Gobin et al. (34). Plasmids composed of a core promoter and upstream regulatory elements encoding b-galactosidase or empty vector plasmid were described (35). containing binding sites for NF-kB and IRF1 (24). Site-a, another All plasmids were verified by sequencing. cis upstream regulatory element, contains a -regulatory element Luciferase assays similar to the S/X/Y region, and both constitutive and IFN-g in- 4 ducible MHC class I expression was reported to depend, at least A total of 3 3 10 HeLa or HEK293T cells/well was seeded in 96-well by guest on September 27, 2021 in part, on the MHC enhanceosome and CIITA (26–28). In vivo plates. Cells were transfected with 10 ng (HEK293T) or 20 ng (HeLa) b-galactosidase reporter plasmid, 20 ng HLA-A230, HLA-B250, HLA- binding of endogenous CIITA to the HLA class I C gene promoter B250”TCGCA,” or HLA-DRA reporter luciferase, and various amounts in B cells and dendritic cells was also shown (29). In contrast, many of expression vectors for NLRC5 or CIITA, as indicated in the figures. cell types, including murine and human resting T cells, express The total amount of transfected DNA was adjusted to 51 ng/well with high levels of MHC class I in the absence of MHC class II and pCDNA3.1. The transfection was performed using X-tremeGene 9 (Roche), as indicated in the manufacturer’s protocol. Luciferase activity was CIITA, and CIITA-deficient mice showed normal MHC class I measured 16–24 h after transfection and normalized to b-galactosidase expression levels in all tissues analyzed (30). Thus, both consti- activity (36). tutive and inducible MHC class I expression mechanisms remain to be fully understood. Indirect immunofluorescence microscopy In this study, we identified a function for NLRC5 in activating HeLa cells were seeded on coverslips in 24-well plates. Cells were MHC class I expression in a site-a/S/X/Yand MHC enhanceosome- transfected with 1 mg myc-tagged NLRC5 using Lipofectamine 2000 dependent manner. Endogenous NLRC5 expression levels correlate (Invitrogen). After overnight incubation, cells were treated with 50 nM Leptomycin B (LepB; Sigma-Aldrich) or left untreated. After 4 h of LepB with constitutive MHC class I expression levels in various human treatment, cells were fixed using 4% paraformaldehyde (Roth) in PBS tissues. Exogenous expression of NLRC5 in the murine melanoma for 10 min at room temperature. Cells were permeabilized using 0.5% cell line B16-F10, in which MHC class I expression is low to un- Triton X-100 (Roth) in ice-cold PBS for 5 min. After permeabilization, detectable, induces endogenous MHC class I expression to levels cells were blocked with 3% BSA for 20 min at room temperature. Myc- tagged protein was detected using mouse anti-myc Ab 9E10 (Sigma) di- found in primary murine lymphocytes. Overexpression of NLRC5 luted 1:1000 in 3% BSA, followed by incubation with rabbit anti-mouse increases MHC class I reporter gene expression in human cell Alexa Fluor 488 (1:500) in 3% PBS. DNA was stained using DAPI. Cells lines, whereas knockdown of NLRC5 reduces expression. NLRC5- were mounted in ProLong Gold Antifade reagent (Invitrogen). Images R activated MHC class I expression is dependent on an intact X1 were recorded with a Cell microscope (Olympus) and processed using element in the HLA class I B gene promoter, and dominant- ImageJ software. negative mutants of CIITA competitively inhibit NLRC5- Transfection of B16-F10 cells dependent MHC class I gene activation, strongly suggesting that B16-F10 cells were transfected using PEI-max (Polysciences). Transient NLRC5 and CIITA work, at least in part, through the same inter- transfections were analyzed 2–4 d after transfection. For the selection of action partners in the MHC enhanceosome. Functionally, we show stable transfectants, cells were selected with 200 mg/ml hygromycin B in this article that the N-terminal effector domain, the central (EBS-NPL–based expression vectors) or 7.5 mg/ml blasticidine (EBSB- NACHT domain with an intact ATPase motif, and the C-terminal PL–based vectors) for at least 2 wk. CIITA-transfected cells were sorted for MHC class II expression with anti-IA/IE mAb 2G9 (BD Biosciences) LRR domains are all required for activation of the MHC class I and anti-rat IgG-coupled Dynabeads (Invitrogen), while NLRC5-transfected promoter. Furthermore, our data suggest that constant cytoplasmic/ cells were sorted with anti H-2k(b) Ab AF6-88.5 (BD Biosciences) and nuclear exchange is needed for full activity of NLRC5. anti-mouse IgG-coupled Dynabeads. 4942 NLRC5 INDUCES MHC I

Small interfering RNA knockdown 38), and CIITA is the only other NLR protein with a clearly Small interfering RNA (siRNA) duplexes specific for NLRC5 and control defined function in transcriptional activation. This suggested a siRNA were described (7). For siRNA-based knockdown in HeLa cells, functional analogy between these two NLR proteins; therefore, we 50,000 cells were seeded into 24-well plates, 6 h prior to siRNA trans- compared the ability of NLRC5 and CIITA to activate HLA class I fection. After incubation, cells were transfected with 10 nM siRNA using and II reporter gene constructs in HEK293T cells. NLRC5 ex- HiPerfect (QIAGEN), as described in the manufacturer’s protocol. Seventy- pression resulted in a significant and dose-dependent HLA class I two hours posttransfection, triplicates were pooled, and RNA was extracted using the RNeasy extraction kit (QIAGEN). induction, whereas its effect on an HLA class II DRA promoter Knockdown in THP1 cells and primary human dermal fibroblasts was reporter construct was negligible (Fig. 1B). In contrast, CIITA performed as described (7). RNA was extracted using the RNeasy extrac- activated the HLA class II DRA promoter strongly and showed tion kit (QIAGEN). For quantitative RT-PCR analysis, 1 mg RNA was a low, but reproducible, activation of the HLA class I B promoter reverse transcribed using the First-Strand cDNA Synthesis Kit (Fermentas). (Fig. 1B). Quantitative PCR analysis Knockdown of NLRC5 in human myeloid macrophage-like For quantitative PCR analyses, 50 ng cDNA was analyzed in a total volume differentiated THP1 cells via siRNA robustly reduced mRNA of 25 ml using the iQ SYBR Green Supermix (Bio-Rad), as described in levels of endogenous NLRC5 and led to a decreased expression of the manufacturer’s protocol. For amplification of fragments, the following HLA-B mRNA (Fig. 1C) and HLA class I cell surface expression primers were used: NLRC5_fwd: 59-CTGCAGCCAAGTTCTTAGGG, (Supplemental Fig. 1A). Of note, two independent NLRC5- NLRC5_rev: 59-TCAGCTGAGGGAGTTGAGGT-39;HLA-B_fwd:59- specific siRNA duplexes yielded similar results (Fig. 1C). We CTACCCTGCGGAGATCA-39, HLA-B_rev: 59-ACAGCCAGGCCAGC- AAC-39; and GAPDH_fwd: 59-CAACGACCACTTTGTCAAGC-39, GA- obtained similar results using HeLa cells (Fig. 1D). In all cases, PDH_rev: 59-TCTTCCTCTTGTGCTCTTGC-39. transfection of control siRNA had no effect on NLRC5 or MHC Downloaded from For gene-expression profiling, multiple tissue cDNA panels (Clontech) class I gene expression compared with untreated cells (Fig. 1C, were used. All quantitative PCR reactions were run on a Bio-Rad iQ5 1D). We also analyzed nontransformed primary human dermal cycler, and data were evaluated by the iQ5 system software (version 2.0) using the DDCT method. fibroblasts in which we induced high levels of NLRC5 expression by treatment with poly(I:C) (7). For both NLRC5-specific siRNA Flow cytometry duplexes, knockdown of NLRC5 correlated with a reduction in

For staining of surface HLA class I molecules, THP-1 cells were incubated endogenous HLA class I B mRNA expression levels (Fig. 1E). http://www.jimmunol.org/ with anti–HLA-ABC Ab (BD Pharmingen) or control Ab allophycocyanin- Moreover, a similar result was obtained with cells derived from mouse IgG1 (BD Pharmingen). FACS analysis of THP1 cells was per- a second donor and when NLRC5 expression was induced by formed using a BD FACSCanto and FACSDiva analysis software. B16-F10 cells were stained for MHC class II expression with Alexa Fluor 647- Sendai virus (Supplemental Fig. 1B). coupled anti-IA/IE mAb 2G9 (BD Biosciences), for MHC class I ex- Taken together, this strongly suggests that NLRC5 is involved in pression with biotinylated anti H-2k(b) Ab AF6-88.5 (BD Biosciences), the control of constitutive MHC class I gene expression. followed by secondary staining with Streptavidin-allophycocyanin (BD Biosciences), or appropriate controls. Cells were analyzed on a FACSCa- MHC class I gene expression correlates with NLRC5 libur instrument using CellQuest software (BD Biosciences). expression levels

Protein-stability assay MHC class I expression is near ubiquitous on nucleated cells, but it by guest on September 27, 2021 For inhibition of protein neosynthesis, HEK293T cells were treated with shows strong differences in expression levels between different cell 30 mg/ml cycloheximide for 1, 3, or 6 h and subsequently lysed in lines and tissues. Based on the previous findings, we asked whether Laemmli buffer, followed by Western blot analysis. NLRC5 levels correlate with basal HLA class I gene expression in Immunoprecipitation of endogenous NLRC5 human tissues. In agreement with our previous report (7), we found NLRC5 expression levels to vary widely between different tissues ∼ HeLa cells were seeded to a confluence of 80% and stimulated overnight (Fig. 2A). Comparison of HLA class I B mRNA expression levels using 100 mg/ml poly(I:C) to induce NLRC5 expression (7). After stim- ulation with poly(I:C), cells were treated with 50 nM LepB for 4 h. Cells with those of NLRC5 revealed a close correlation between the were fractionated using the QIAGEN QProteome Kit, as described in the expression levels of both genes in tissues with very low (skeletal manufacturer’s protocol. The immunoprecipitation was performed using muscle and heart), intermediate (liver, kidney, prostate, thymus, anti-NLRC5 3H8 Ab (7) bound to Protein G Sepharose Beads (GE and colon), or high expression (CD8+ T cells). However, in some Healthcare) for 4 h at room temperature. Cytoplasmic and nuclear immu- noprecipitation samples were analyzed by SDS-PAGE and subsequent tissues, such as brain and lung, the expression levels were more Western blot analysis. Detection Abs used were anti-NLRC5 3H8 (7), anti- disparate (Fig. 2A). These results are compatible with a positive tubulin (T7186; Sigma-Aldrich), and anti-lamin A/C (2032; Cell Signaling). role for NLRC5 in MHC class I gene expression but also suggest that, at least in certain tissues, other factors contribute to the Statistical analysis overall MHC class I expression levels. Data are presented as mean + SD. Significance was assessed with a two- Several tissues and cell lines show very low to undetectable MHC sided Student t test. class I expression (24, 28). One example is the murine melanoma cell line B16 and its various sublines (28, 39). B16-F10 cells were Results transfected with an empty vector or with vectors expressing EGFP- NLRC5 contributes to MHC class I gene expression NLRC5 or EGFP-CIITA. Analysis of transiently transfected cells Several recent reports analyzed NLRC5 with conflicting results for showed that NLRC5 robustly induced MHC class I expression, many of the reported functions, including its ability to activate whereas CIITA induced MHC class II expression. Expression of a MHC class I transcription. To address this unresolved issue, we first Walker A mutant of NLRC5 (GFP-NLRC5 K234A) had no effect cotransfected various amounts of an NLRC5 expression vector in on MHC class I expression (data not shown). HEK293T cells together with HLA class I A and B gene promoter After 2 wk of selection with appropriate antibiotics, NLRC5- reporter constructs (27, 36). Expression of NLRC5 significantly transfected B16-F10 cells were sorted once with Ab and mag- activated HLA-A and HLA-B luciferase reporter expression in this netic beads for MHC class I+ cells, whereas CIITA-transfected system (Fig. 1A). cells were enriched for MHC class II+ cells (Fig. 2B, 2C). In NLRC5 and CIITA are most closely related to each other within parallel experiments, we were unable to enrich for MHC class I+ the NLR gene family in their NACHT and LRR domains (12, 37, or class II+ cells in vector- or NLRC5-K234A–transfected cells, The Journal of Immunology 4943

FIGURE 1. NLRC5 has a role in MHC class I gene expression. HEK293T cells were transiently trans- fected with increasing amounts of NLRC5 (A)or CIITA and NLRC5 (B) expression constructs in the presence of the indicated luciferase reporter and a b-galactosidase expression vector. Luciferase activity normalized to b-galactosidase activity is shown as fold activation to mock-transfected cells (set to 1). Data Downloaded from represent mean + SD (n = 3). (C) PMA-differentiated THP1 cells were transfected with 50 nM of the indi- cated siRNA for 72 h. Quantitative RT-PCR with NLRC5 (open bars) and HLA-B–specific primers (fil- led bars) is shown as fold relative expression normal- ized to GAPDH (mean + SD; n = 3). mRNA levels in D C untreated cells were set to 1. ( ) Same as in ( ), using http://www.jimmunol.org/ HeLa cells transfected with 10 nM of NLRC5_1 siRNA or scrambled control siRNA (siCTL) for 72 h. (E) Primary human dermal fibroblasts were transfected with 10 nM of the indicated siRNA duplexes for 72 h and subsequently treated with 10 mg/ml poly(I:C). Quantitative RT-PCR of NLRC5 and HLA-B expres- sion is shown as the mean of the relative expression to GAPDH + SD (n = 3). mRNA levels in poly(I:C) treated controls were set to 1. *p , 0.05, **p , 0.005. by guest on September 27, 2021

showing that B16-F10 cells are negative or nearly negative for expression is very low in these stable transfectants, because EGFP MHC class I and II (Fig. 2B1, 2C1) and that NLRC5-K234A does expression could not be detected (Fig. 2B5, 2C5). However, we not activate MHC class I expression (data not shown). did detect EGFP expression readily in the same cells 2 d after Stable expression of EGFP-NLRC5 induced endogenous MHC transfection, demonstrating the integrity of the construct (data not class I gene and protein expression levels very substantially in two shown). EGFP-NLRC5 expression had no effect on MHC class II independently transfected and selected cell populations (Fig. 2B3, expression in B16-F10 cells (Fig. 2C3, 2C4; compare EGFP+ and 2B4). Remarkably, MHC class I levels in the EGFP+ cells were EGFP2 cells). comparable to those found on freshly isolated murine splenic These results show that endogenous NLRC5 and MHC class I lymphocytes (Fig. 2B2). The distinctive kink in the dot plots mRNA expression levels are closely correlated in various cell lines showing EGFP-NLRC5 expression on the x-axis and MHC class I and tissues and that NLRC5, on its own, is able to induce high levels [H2K(b)] on the y-axis (Fig. 2B3, 2B4) suggests a threshold of endogenous MHC class I gene expression in B16-F10 cells. mechanism, where only a limited amount of NLRC5 expression is needed to fully activate endogenous MHC class I gene expression. NLRC5-mediated MHC class I expression is dependent on EGFP-CIITA also induced MHC class I expression in B16-F10 nuclear shuttling and the integral protein fold of NLRC5 cells, albeit less efficiently than did NLRC5 (Fig. 2B5 versus Fig. NLRC5 shares the canonical tripartite NLR structure consisting of 2B1; see Supplemental Fig. 1C for a histogram overlay). As ex- an N-terminal effector domain, a central nucleotide-binding do- pected, MHC class II expression was strongly induced by EGFP- main, and a C-terminal LRR region (7). However, our knowledge CIITA in the same cells (Fig. 2C5). The level of EGFP-CIITA of the contribution of these domains to the subcellular localiza- 4944 NLRC5 INDUCES MHC I

FIGURE 2. NLRC5 levels correlate with basal MHC class I expression. (A) Relative mRNA expression of MHC class I and NLRC5 in selected human tissues. HLA-B and NLRC5 expres- sion, normalized to GAPDH expres- sion is shown. The level in kidney was set to 1. Bars represent means + SD (n = 3). Tissues are ranked according Downloaded from to detected HLA-B expression. (B) Surface expression of MHC class I in B16-F10 cells, murine splenocytes, and B16-F10 cells lines stably expressing EGFP-NLRC5 or EGFP-CIITA. Flow cytometry analysis showing EGFP and C

MHC class II signals are shown. ( ) http://www.jimmunol.org/ The same as in (B), but showing MHC class II expression. by guest on September 27, 2021

tion, nuclear trafficking, and MHC class I activation of NLRC5 which lacks the entire LRR domain and is expressed in several remains fragmentary. Therefore, we generated various NLRC5 human tissues (7). NLRC5 isoform 3 localized predominantly in deletions and point mutants and analyzed their ability to shuttle the nucleus, even in the absence of LepB (Fig. 3A), yet it com- through the nucleus by immunofluorescence analysis, as well as pletely failed to activate MHC class I transcription (Fig. 3B). This their capacity to activate MHC class I transcription by HLA class I indicates that the C-terminal LRR domain is involved in nuclear B promoter reporter assays in HeLa cells (Fig. 3). Wild-type (wt) export of NLRC5, as well as in transcriptional activation. NLRC5 was localized very predominantly in the cytoplasm, but CIITA contains, at its N-terminal end, a “classical” acidic- inhibition of Crm1-dependent nuclear export with LepB led to activation domain that activates transcription in yeast when teth- a strong nuclear accumulation of NLRC5 (Fig. 3A). These ex- ered to a promoter (40, 41). CIITA is expressed from three in- periments show that NLRC5 transits through the , dependent promoters, generating three isoforms (FI, FIII, and which is in agreement with recently reported findings (12, 13). FIV) (42). Although CIITA FIII and FIV start at their N-terminal The requirement of a functional ATPase domain in NLRC5 for end with the acidic activation domain, CIITA FI contains an ad- subcellular localization and transport and the ability to activate ditional domain N-terminal of the acidic domain, which has ho- MHC class I transcription were analyzed by mutating the conserved mology to a CARD domain and was reported to contribute to lysine residue (K234) in the predicted Walker A motif of NLRC5 MHC class II gene activation (43). The N-terminal domain of to alanine (K234A) (7, 38). In contrast to wt NLRC5, NLRC5 NLRC5 adopts a death domain fold, but it lacks obvious homol- K234A was localized exclusively in the cytoplasm, even after ogy to the CARD and PYD found in other NLRs, including the LepB treatment (Fig. 3A). In agreement with this exclusively CARD of CIITA FI, or known transcriptional activation domains cytoplasmic localization, we found that NLRC5 K234A was un- (7). In contrast, the results shown above suggested that NLRC5 able to activate the HLA-B250 promoter in HeLa cells (Fig. 3B), functions as a transcriptional activator. Accordingly, we investi- as we also observed in HEK293T cells (Supplemental Fig. 2A) gated whether the N-terminal NLRC5 effector domain is neces- and in B16-F10 cells (data not shown). This suggests that nuclear sary for MHC class I transcriptional activation. A truncated form shuttling of NLRC5 is essential for its activity on MHC class I of NLRC5, lacking the N-terminal part of the effector domain but gene expression. still containing a recently identified nuclear localization signal at We showed previously that NLRC5 is expressed in various amino acid positions 122–134 (NLRC5DDD NLS) (12), was isoforms. In this study, we focused our attention on isoform 3, found in the nucleus in ∼50% of HeLa cells in the absence of The Journal of Immunology 4945

FIGURE 3. Requirements for NLRC5 nuclear Downloaded from shuttling and MHC class I activation. (A) Photo- micrographs of indirect immunofluorescence analysis of HeLa cells transfected with the indicated plasmid for 24 h. Cells were fixed untreated (left panels)or after treatment with 50 nM LepB for 4 h to block nuclear export (right panels). Signals obtained with the myc tag-specific 9E10 Ab are shown in green. DNA http://www.jimmunol.org/ staining with DAPI is shown in blue. Scale bar, 10 mm. (B) HLA-B250 reporter luciferase assays in HeLa cells in which increasing amounts of the corresponding constructs were transiently transfected. Luciferase ac- tivity, normalized to b-galactosidase activity, is shown as fold activation to cells transfected with empty vector (“0” set to 1). Data represent mean + SD (n = 3). All constructs were tested in a single transfection series. (C) Schematic representation of the constructs. The N- terminal NLS is highlighted in red. Numbers indicate by guest on September 27, 2021 amino acid residues. ED, Effector domain; LRR, LRR domain.

LepB. Nuclear localization of this construct was increased by domain alone showed an exclusively nuclear localization in the inhibiting nuclear export (Fig. 3A); however, it was unable to absence of LepB, whereas the LRR region alone was completely activate the MHC class I promoter (Fig. 3B, Supplemental Fig. cytoplasmic, irrespective of the presence or absence of LepB (Fig. 2A). 3A). Neither construct was able to activate MHC class I tran- These experiments show that the N-terminal death-fold domain, scription in HeLa or HEK293T cells (Fig. 3B, Supplemental Fig. a functional ATPase domain, and the C-terminal LRR domain of 2A). All constructs were expressed at comparable levels in NLRC5 are all required to induce MHC class I gene expression, HEK293T and HeLa cells, showing that the differences in reporter whereas the requirements for nuclear shuttling are different. As gene activation reflect protein activity and not protein levels (data expected from the previous experiments, the N-terminal death not shown, Supplemental Fig. 2B). 4946 NLRC5 INDUCES MHC I

The previous experiments were carried out by ectopic expression of epitope-tagged protein constructs. Overexpression can easily lead to saturation of transport pathways and, thus, may yield ar- tifactual results. Therefore, we also analyzed the subcellular lo- calization of endogenous NLRC5 in HeLa cells by cytoplasmic and nuclear protein fractionation and Western blotting (Fig. 4, upper panel). To obtain detectable amounts of NLRC5, cells were pretreated with poly(I:C) to induce NLRC5 expression. This revealed a strong nuclear accumulation of endogenous NLRC5 after blocking of nuclear export by LepB, whereas NLRC5 was mainly cytoplasmic in untreated cells (Fig. 4). Blotting for a- tubulin and a-Lamin A/C demonstrated the absence of cross- contamination of cytoplasmic and nuclear fractions (Fig. 4, middle and lower panels). Of note, induction of NLRC5 expression by poly(I:C) treatment did not correlate with accumulation of NLRC5 in the nucleus. This is reminiscent of our observation that over- expressed NLRC5 strongly activated MHC class I gene expres- sion, beside being mainly localized in the cytoplasm also in HeLa cells (Fig. 3). Downloaded from Taken together, these results show that NLRC5 shuttles through the nucleus and that MHC class I promoter activation is dependent on a functional ATPase domain and on the presence of both the death domain and LRR region of NLRC5, suggesting that the integral NLRC5 fold is needed for subcellular transport and

transactivation function. http://www.jimmunol.org/ Nuclear–cytoplasmic shuttling of NLRC5 enhances MHC class I promoter activation To analyze the role of nuclear shuttling of NLRC5 in MHC class I signaling, we asked whether forcing nuclear localization of NLRC5 would result in increased MHC class I promoter activation. Therefore, we generated NLRC5 constructs containing either a myc epitope tag or a myc epitope followed by a double SV40 NLS sequence at the N terminus (myc-2xNLS-NLRC5). As shown in by guest on September 27, 2021 Fig. 5A, this construct showed a predominantly nuclear localiza- tion, even in cells that had not been treated with LepB. Unex- pectedly, myc-2xNLS-NLRC5 showed a reduced capacity to activate the HLA class I A and B promoters and did not show a dose-dependent induction of the MHC class I promoter, as ob-

FIGURE 5. Nuclear/cytoplasmic shuttling is important for NLRC5 function. (A) Photomicrographs of indirect immunofluorescence analysis of HeLa cells transfected with the indicated plasmid for 24 h. Signals obtained with the myc tag-specific 9E10 Ab are shown in green. DNA staining with DAPI is shown in blue. Scale bar, 10 mm. (B and C) HEK293T cells were transfected with increasing amounts (0, 1, 5, 10 ng) of wt NLRC5, a Walker A mutant (K234A), and NLRC5 containing ad- ditional NLS (2x-NLS-NLRC5 wt). The expression constructs were transfected in the presence of the indicated luciferase reporter construct and a b-galactosidase expression vector. Luciferase activity, normalized to b-galactosidase activity, is shown as nRLU. Data represent mean + SD (n = 3). (D) HEK293T cells were transfected with either wt myc-NLRC5 FIGURE 4. Endogenous NLRC5 shuttles to the nucleus. Western blot- or myc-2xNLS-NLRC5. After 24 h, protein stability was assayed by ting analysis of immunoprecipitations of endogenous NLRC5 from the Western analysis in cells in which protein neosynthesis was blocked with nuclear and cytoplasmic fraction of HeLa cells. HeLa cells were treated 30 mg/ml cycloheximide (CHX) for the indicated time. with 100 mg/ml poly(I:C) for 20 h to increase NLRC5 expression and subsequently with 50 nM LepB for 4 h where indicated. About 1 3 107 cells were lysed and fractionated. Endogenous NLRC5 was precipitated served for wt myc-NLRC5 (Fig. 5B, 5C). This was not due to a from the cytosolic and nuclear fractions (each concentrated to ∼2 mg/ml) change in protein stability or turnover, as assessed by Western with the NLRC5-specific 3H8 Ab and detected in the immunoprecipitate (upper panel). Probing for a-tubulin (middle panel) and lamin A/C (lower blotting after blocking of protein synthesis using cycloheximide panel) served as quality control for the purity of the fractionation used for (Fig. 5D). Although, we cannot formally exclude that the intro- the precipitation. About 100 mg total protein of these fractions was loaded duction of the NLS sequences affected the protein function, block- on the gel. Immunoprecipitates correlate to ∼10-fold enrichment over in- ing nuclear export by LepB also led to decreased MHC class I gene put. activation in HEK293T cells by untagged NLRC5 (data not shown). The Journal of Immunology 4947

This indicates that nuclear localization, as well as constant box element (“TCGCA”) exhibits a strongly reduced activity for cytoplasmic/nuclear shuttling, is essential for full NLRC5 activity. both basal (CIITA-independent) and CIITA-dependent MHC class I expression. Therefore, we tested whether NLRC5 could activate NLRC5-mediated MHC class I promoter activation is mediated an HLA class I B promoter reporter construct containing a mu- by the MHC enhanceosome tated X1 box element (“TCGCA”) (Fig. 6A). Luciferase reporter Activation of MHC class I genes is dependent on a core promoter assays in HEK293T cells revealed an inability for overexpressed and upstream regulatory elements containing NF-kB– and ISRE- NLRC5 to activate this defective X1-Box HLA class I B250 binding elements and a regulatory region originally called site-a construct (“TCGCA”), whereas NLRC5 robustly turned on the wt (26). Site-a was shown to contain a region homologous to the S/X/ B250-luciferase construct in the same experiment (Fig. 6B, left Y composite cis-acting crucial for MHC class II gene activation panel). The parallel experiment with CIITA confirmed the results (26, 28, 34). The MHC enhanceosome complex, and in particular of Gobin and colleagues (26) (Fig. 6B, right panel). Interestingly, the RFX complex, was shown to be required both for basal MHC we found in this and similar experiments that CIITA isoform I class I expression (in absence of CIITA) and for CIITA-dependent activated MHC class I expression more efficiently compared with activated MHC class I expression (26). Given the close homology isoform III (compare Fig. 6B, right panel with Fig. 1B, left panel, of NLRC5 and CIITA, we anticipated that the site-a/S/X/Y ele- and Fig. 6D). Taken together, these findings suggest that MHC ment is also involved in NLRC5-mediated MHC class I promoter class I activation by both NLRC5 and CIITA is dependent on activation. Gobin and colleagues (26) showed that a mutated X1 a functional X1 box. Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 6. MHC class I promoter activation by NLRC5 relies on a functional X-box motif. (A) Schematic representation of the B250 promoter, highlighting the wt (B250) and mutated X1 box (B250”TCGCA”) mutation used. (B) HEK293T cells were transfected with increasing amounts of wt NLRC5 or CIITA (isoform 1) in the presence of either the HLA-B250-luciferase construct or the X1 box mutant (HLA-B250”TCGCA”) luciferase construct. Luciferase activity, normalized to b-galactosidase activity, is shown as fold activation to mock-transfected cells (set to 1). Data represent mean + SD (n = 3). (C) HEK293T cells were transfected with 1 ng NLRC5 or 1 ng CIITA alone or together with 1, 5, or 10 ng of CIITA L335. B250 (left panel) and DRA (right panel) luciferase activity, normalized to b-galactosidase activity, is shown as fold activation to mock-transfected cells (set to 1). Data represent mean + SD (n = 3). (D) HEK293T cells were transfected with 1 ng NLRC5 and increasing amounts of CIITA. HLA-B250 reporter-mediated luciferase activity, normalized to b-galactosidase activity, is shown as fold activation to mock-transfected cells (set to 1). Data represent mean + SD (n = 3). *p , 0.05, **p , 0.005. 4948 NLRC5 INDUCES MHC I

CIITA is recruited to the MHC enhanceosome through multiple Furthermore, and quite remarkably, stable expression of NLRC5 protein–protein interactions with RFX5, RFXAP, CREB, NF-YB, in the murine melanoma cell line B16-F10, which shows low to and NF-YC (32, 44–46). Most of the interaction surfaces in CIITA undetectable basal MHC class I expression, is sufficient to induce were mapped to the NACHT and LRR regions (reviewed in Ref. endogenous MHC class I expression to the same level as found in 18), which are the regions that show the highest sequence simi- freshly isolated lymphocytes. It appears that low levels of NLRC5 larity between NLRC5 and CIITA. We reasoned that NLRC5 and are sufficient for full activation of MHC class I expression in CIITA might share, at least in part, similar interaction surfaces a threshold-dependent manner. This is in contrast to the mode of and, therefore, be able to compete with each other for binding to action of CIITA, which activates endogenous MHC class II tran- the MHC enhanceosome. Therefore, we coexpressed NLRC5 with scription in a dose-dependent manner over a wide range of con- a strongly dominant-negative N-terminally truncated CIITA mu- centrations (22, 31). tant. CIITA-NLS-L335 is expressed at 30–50-fold higher levels The integrity of the RFX-binding X1 box within the site-a/S/X/Y compared with wt CIITA because of its greatly increased stability, element of the HLA class I B promoter is necessary for NLRC5- and it displaces CIITA efficiently from the MHC class II promoter dependent transcription. So far, only the requirement for RFX (33, 47, 48). As we showed before, CIITA-NLS-L335 very effi- binding to the site-a/S/X/Y has been rigorously demonstrated (26). ciently suppressed CIITA-dependent MHC class II gene expres- However, because the sequence, spacing, and orientation of the S/ sion in a dose-dependent manner (Fig. 6C, right panel). Very W/Z, X1, X2, and Y elements are highly conserved between MHC interestingly, we also observed a strong reduction in NLRC5- class I and class II promoters, we can extrapolate that the complete dependent MHC class I promoter activation when CIITA-NLS- RFX/CREB/NF-Y enhanceosome binds to the site-a–/S/X/Y ele-

L335 was coexpressed with NLRC5, suggesting a competition ment in a manner similar to the MHC class II promoters. This Downloaded from between the two factors for the same binding sites (Fig. 6C, left notion is reinforced by the observation that a strongly dominant- panel). Because the previous experiments indicated that both negative CIITA mutant (CIITA-NLS-L335) inhibits NLRC5- NLRC5 and CIITA act via site-a/S/X/Y and the MHC enhan- dependent MHC class I activation in a dose-dependent manner. ceosome in the MHC class I promoters, we wanted to test for This also suggests that NLRC5 and CIITA, which are the two most a possible cross-talk between these two factors. Therefore, we closely related NLR proteins, interact with the MHC enhanceo-

cotransfected a constant amount of NLRC5 with increasing some via overlapping protein interfaces. This is further supported http://www.jimmunol.org/ amounts of CIITA and measured the activation of the HLA class I by the observation of an additive effect on MHC class I gene B reporter construct. As shown in Fig. 6D, we observed an ad- expression when both CIITA and NLRC5 are coexpressed. Thus, ditive effect of CIITA on NLRC5-dependent MHC class I ex- the results presented in this article strongly suggest that NLRC5 is pression. This further supports the hypothesis that NLRC5 and involved in the constitutive activation of MHC class I genes. CIITA occupy the same binding interface. NLRC5 and CIITA show important similarities with respect to Taken together, these results show that NLRC5 activates MHC MHC enhanceosome-dependent MHC gene activation, as well as class I gene expression via MHC enhanceosome binding to the site- with regard to their intracellular transport and localization. Both a/S/X/Y element and also suggest that NLRC5 interacts, at least in show a highly complex nucleo-cytoplasmic shuttling behavior that part, with the same protein partners as does CIITA via overlapping is necessary for transactivation function and that involves all by guest on September 27, 2021 interaction sites. domains of both proteins. We confirmed the initial observation that ectopically expressed NLRC5 is localized predominantly in the Discussion cytoplasm, but transits through the nucleus in a Crm-1–dependent We show in this study that human NLRC5 plays a role in basal or manner (12, 13). Very importantly, the same is true for endoge- constitutive MHC class I expression in a manner quite similar to nous NLRC5, as shown in this study by subcellular protein frac- the mode of action of CIITA. Of note, in an independent study, tionation after LepB treatment. Both the integrity of the Walker A Meissner et al. (49) obtained very similar results that strongly motif within the ATPase domain (NACHT) and an NLS in the N- support this conclusion. Despite the very different expression pat- terminal effector domain are required for nuclear localization, terns of MHC class I and II genes, there is a certain overlap in the whereas the C-terminal LRR region appears to be involved in molecular mechanisms controlling the expression of these genes. nuclear export, as shown by the predominantly nuclear localiza- The MHC enhanceosome binds to the site-a/S/X/Y element in the tion of NLRC5 isoform 3 lacking the LRR region and the ex- MHC class I promoter, and Gobin et al. (26) showed a strongly clusively cytoplasmic localization of the NLRC5 LRR domain reduced constitutive MHC class I expression in RFX-deficient fi- when expressed on its own. broblast cell lines, indicating that the MHC enhanceosome plays Endogenous CIITA shows a relatively even distribution between an important role in the constitutive expression of MHC class I the cytoplasm and nucleus, and it is also exported from the nucleus genes in the absence of CIITA. Although CIITA is indispensable in a Crm-1–dependent manner (31). At least three potential NLS for MHC class II gene expression, it has been well established that motifs have been described in CIITA (reviewed in Ref. 50); this factor is not required for constitutive MHC class I expression, however, none of these motifs behaves like a conventional NLS even though it is clearly capable of increasing it. This begged the motif, and many other mutations, especially within the NACHT question whether a CIITA-equivalent factor binding to the MHC domain and LRR regions of CIITA, lead to an inhibition of nu- enhanceosome might be involved in MHC class I gene activation, clear import and loss of transactivation potential (31). Our pre- at least in tissues and cells that are negative for CIITA expression. vious studies suggested that the efficiency of nuclear import of The results presented in this article provide compelling evidence CIITA, rather than its steady-state concentration in the nucleus, that NLRC5 is such a factor. NLRC5 mRNA expression levels are is rate limiting for MHC class II gene activation (31, 32). This correlated to constitutive MHC class I expression levels in several is largely reminiscent of the results obtained in this study for cell lines and tissues; ectopic NLRC5 expression activates HLA NLRC5, which show that increasing NLRC5 protein levels do class I promoter reporter constructs in a dose-dependent manner, correlate with enhanced MHC class I expression but not with as well as increasing HLA class I cell surface expression in THP1 nuclear accumulation of NLRC5. Moreover, we found in this cells; and knockdown of NLRC5 reduces MHC class I transcription study that an N-terminal fusion of an exogenous SV40-derived both in immortalized cell lines and primary cells. NLS to myc-tagged NLRC5 was very efficient in increasing nu- The Journal of Immunology 4949 clear localization but paradoxically reduced the MHC class I 2. Kufer, T. A., and P. J. Sansonetti. 2011. NLR functions beyond pathogen rec- ognition. Nat. Immunol. 12: 121–128. gene-transactivation potential of myc-2xNLS-NLRC5. Although 3. Mariathasan, S., and D. M. Monack. 2007. Inflammasome adaptors and sensors: we cannot formally exclude that the added 2xNLS element dis- intracellular regulators of infection and inflammation. Nat. Rev. Immunol. 7: 31– turbed the integrity of the N-terminal domain of NLRC5, we do 40. 4. Elinav, E., T. Strowig, J. Henao-Mejia, and R. A. Flavell. 2011. Regulation of the not think that this is very likely, because N-terminal addition of antimicrobial response by NLR proteins. Immunity 34: 665–679. the myc epitope tag or even EGFP did not appear to have an effect 5. Wilkins, C., and M. Gale, Jr. 2010. Recognition of viruses by cytoplasmic on the activity and transport behavior of NLRC5 (Figs. 2, 3, data sensors. Curr. Opin. Immunol. 22: 41–47. 6. Schro¨der, M., and A. G. Bowie. 2007. An arms race: innate antiviral responses not shown). Furthermore, blocking of nuclear export of NLRC5 and counteracting viral strategies. Biochem. Soc. Trans. 35: 1512–1514. with LepB also reduced MHC class I transactivation (data not 7. Neerincx, A., K. Lautz, M. Menning, E. Kremmer, P. Zigrino, M. Ho¨sel, shown). There are several possible explanations for this complex H. Bu¨ning, R. Schwarzenbacher, and T. A. Kufer. 2010. A role for the human nucleotide-binding domain, leucine-rich repeat-containing family member transport behavior, which are not mutually exclusive. Trans- NLRC5 in antiviral responses. J. Biol. Chem. 285: 26223–26232. activation potential might be dependent on cytoplasmic mod- 8. 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 ifications of the proteins that are not carried out correctly if oligomerization domain-like receptor NLRC5 is involved in IFN-dependent nuclear import is mediated via a “classical” importin-a–dependent antiviral immune responses. J. Immunol. 184: 1990–2000. pathway. It is possible that nuclear import is indirect and depen- 9. 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 dent on the interaction with protein partners in the cytoplasm. The and type I signaling pathways. Cell 141: 483–496. involvement of the LRR domain in this process makes this plau- 10. Davis, B. K., R. A. Roberts, M. T. Huang, S. B. Willingham, B. J. Conti, sible, because LRR domains are best known as protein-interaction W. J. Brickey, B. R. Barker, M. Kwan, D. J. Taxman, M. A. Accavitti-Loper, et al. 2011. Cutting edge: NLRC5-dependent activation of the inflammasome. domains. Lastly, it is possible that nuclear export is somehow J. Immunol. 186: 1333–1337. Downloaded from linked to the transactivation function of NLRC5 and CIITA. 11. Lamkanfi, M., and T. D. Kanneganti. 2012. Regulation of immune pathways by Interestingly, it was recently shown for several plant NLRs that the NOD-like receptor NLRC5. Immunobiology 217: 13–16. 12. Meissner, T. B., A. Li, A. Biswas, K. H. Lee, Y. J. Liu, E. Bayir, D. Iliopoulos, nuclear/cytoplasmic transition is needed for full activity, analogous P. J. van den Elsen, and K. S. Kobayashi. 2010. NLR family member NLRC5 is to our findings for NLRC5 (51). In plants, nucleotide-binding and a transcriptional regulator of MHC class I genes. Proc. Natl. Acad. Sci. USA 107: LRR domain-containing proteins (also called R proteins) also play 13794–13799. 13. Benko, S., J. G. Magalhaes, D. J. Philpott, and S. E. Girardin. 2010. NLRC5 pivotal roles in the detection of invading pathogens. For example, limits the activation of inflammatory pathways. J. Immunol. 185: 1681–1691. http://www.jimmunol.org/ the Arabidopsis thaliana R protein RPS4 has to translocate to 14. Kumar, H., S. Pandey, J. Zou, Y. Kumagai, K. Takahashi, S. Akira, and T. Kawai. 2011. NLRC5 deficiency does not influence cytokine induction by virus and the nucleus to develop complete functionality (52, 53). Moreover, bacteria infections. J. Immunol. 186: 994–1000. Cheng and colleagues (53, 54) reported a cytoplasmic and nuclear 15. Steimle, V., L. A. Otten, M. Zufferey, and B. Mach. 1993. Complementation localization of the Arabidopsis thaliana RproteinSNC1 and cloning of an MHC class II transactivator mutated in hereditary MHC class II deficiency (or bare lymphocyte syndrome). Cell 75: 135–146. components of the nuclear pore control nuclear shuttling of these 16. Reith, W., and B. Mach. 2001. The bare lymphocyte syndrome and the regulation proteins (53). This raises the exciting possibility that nuclear of MHC expression. Annu. Rev. Immunol. 19: 331–373. shuttling of the human NLRs CIITA and NLRC5 is also controlled 17. Choi, N. M., P. Majumder, and J. M. Boss. 2011. Regulation of major histo- compatibility complex class II genes. Curr. Opin. Immunol. 23: 81–87. by unidentified components of the nuclear pore. 18. Reith, W., S. LeibundGut-Landmann, and J. M. Waldburger. 2005. Regulation of Based on our previous study, which showed an involvement of MHC class II gene expression by the class II transactivator. Nat. Rev. Immunol. by guest on September 27, 2021 NLRC5 in type I IFN responses and the data discussed in this 5: 793–806. 19. Moreno, C. S., G. W. Beresford, P. Louis-Plence, A. C. Morris, and J. M. Boss. article, we propose that NLRC5 has a dual function. We assume 1999. CREB regulates MHC class II expression in a CIITA-dependent manner. that NLRC5 contributes to cytosolic control of type I IFN re- Immunity 10: 143–151. 20. Steimle, V., C. A. Siegrist, A. Mottet, B. Lisowska-Grospierre, and B. Mach. sponses, as well as to the basal induction of MHC class I gene 1994. Regulation of MHC class II expression by interferon-gamma mediated by expression. Together, both responses likely enhance innate and the transactivator gene CIITA. Science 265: 106–109. adaptive immune responses to viral and bacterial infection. Many 21. Chang, C. H., J. D. Fontes, M. Peterlin, and R. A. Flavell. 1994. Class II transactivator (CIITA) is sufficient for the inducible expression of major histo- pathogens interfere with MHC class I-dependent Ag presentation to compatibility complex class II genes. J. Exp. Med. 180: 1367–1374. escape CD8 responses. Moreover, MHC class I downreg- 22. Otten, L. A., V. Steimle, S. Bontron, and B. Mach. 1998. Quantitative control of ulation is often observed in malignant tumors, which rely on similar MHC class II expression by the transactivator CIITA. Eur. J. Immunol. 28: 473– 478. immune-evasive strategies as do infectious pathogens. 23. Burke, P. A., and K. Ozato. 1989. Regulation of major histocompatibility The findings presented in this article indicate that NLRC5 is an complex class I genes. Year Immunol. 4: 23–40. important player in the control of MHC class I-dependent immune 24. van den Elsen, P. J., T. M. Holling, H. F. Kuipers, and N. van der Stoep. 2004. Transcriptional regulation of antigen presentation. Curr. Opin. Immunol. 16: 67– responses and, thus, represents a valuable new target for the de- 75. velopment of novel strategies to regulate MHC class I expression in 25. Howcroft, T. K., A. Raval, J. D. Weissman, A. Gegonne, and D. S. Singer. 2003. Distinct transcriptional pathways regulate basal and activated major histocom- viral infection and cancer immunotherapy. patibility complex class I expression. Mol. Cell. Biol. 23: 3377–3391. 26. Gobin, S. J., A. Peijnenburg, M. van Eggermond, M. van Zutphen, R. van den Acknowledgments Berg, and P. J. van den Elsen. 1998. The RFX complex is crucial for the con- stitutive and CIITA-mediated transactivation of MHC class I and beta2- We thank Martin Kro¨nke for continuous support of this research, Paola microglobulin genes. Immunity 9: 531–541. Zigrino and Cornelia Mauch (Department of Dermatology and Venerology, 27. Gobin, S. J., M. van Zutphen, S. D. Westerheide, J. M. Boss, and P. J. van den University of Cologne) for providing primary human dermal fibroblasts, Elsen. 2001. The MHC-specific enhanceosome and its role in MHC class I and Harald Bielig and Alain Lavigueur for critical reading of the manuscript, beta(2)-microglobulin gene transactivation. J. Immunol. 167: 5175–5184. 28. Martin, B. K., K. C. Chin, J. C. Olsen, C. A. Skinner, A. Dey, K. Ozato, and Jens Seeger for help with flow cytometry, and Maureen Menning for J. P. Ting. 1997. Induction of MHC class I expression by the MHC class II excellent technical assistance. transactivator CIITA. Immunity 6: 591–600. 29. Krawczyk, M., Q. Seguı´n-Este´vez, E. Leimgruber, P. Sperisen, C. Schmid, P. Bucher, and W. Reith. 2008. Identification of CIITA regulated genetic module Disclosures dedicated for antigen presentation. PLoS Genet. 4: e1000058. The authors have no financial conflicts of interest. 30. Otten, L. A., S. Leibundgut-Landmann, J. Huarte, I. C. Kos-Braun, C. Lavanchy, E. Barras, B. Borisch, V. Steimle, H. Acha-Orbea, and W. Reith. 2006. Revisiting the specificity of the MHC class II transactivator CIITA in vivo. Eur. J. Immunol. 36: 1548–1558. References 31. Camacho-Carvajal, M. M., S. Klingler, F. Schnappauf, S. B. Hake, and V. Steimle. 1. Kawai, T., and S. Akira. 2011. Toll-like receptors and their crosstalk with other 2004. Importance of class II transactivator leucine-rich repeats for dominant- innate receptors in infection and immunity. Immunity 34: 637–650. negative function and nucleo-cytoplasmic transport. Int. Immunol. 16: 65–75. 4950 NLRC5 INDUCES MHC I

32. Hake, S. B., K. Masternak, C. Kammerbauer, C. Janzen, W. Reith, and class II transactivator contains a caspase recruitment domain that confers potent V. Steimle. 2000. CIITA leucine-rich repeats control nuclear localization, in vivo transactivation activity. J. Biol. Chem. 276: 19089–19093. recruitment to the major histocompatibility complex (MHC) class II enhan- 44. Scholl, T., S. K. Mahanta, and J. L. Strominger. 1997. Specific complex for- ceosome, and MHC class II gene transactivation. Mol. Cell. Biol. 20: 7716– mation between the type II bare lymphocyte syndrome-associated transactivators 7725. CIITA and RFX5. Proc. Natl. Acad. Sci. USA 94: 6330–6334. 33. Bontron, S., C. Ucla, B. Mach, and V. Steimle. 1997. Efficient repression of 45. DeSandro, A. M., U. M. Nagarajan, and J. M. Boss. 2000. Associations and endogenous major histocompatibility complex class II expression through interactions between bare lymphocyte syndrome factors. Mol. Cell. Biol. 20: dominant negative CIITA mutants isolated by a functional selection strategy. 6587–6599. Mol. Cell. Biol. 17: 4249–4258. 46. Zhu, X. S., M. W. Linhoff, G. Li, K. C. Chin, S. N. Maity, and J. P. Ting. 2000. 34. Gobin, S. J., A. Peijnenburg, V. Keijsers, and P. J. van den Elsen. 1997. Site alpha Transcriptional scaffold: CIITA interacts with NF-Y, RFX, and CREB to cause is crucial for two routes of IFN gamma-induced MHC class I transactivation: the stereospecific regulation of the class II major histocompatibility complex pro- ISRE-mediated route and a novel pathway involving CIITA. Immunity 6: 601– moter. Mol. Cell. Biol. 20: 6051–6061. 611. 47. Schnappauf, F., S. B. Hake, M. M. Camacho Carvajal, S. Bontron, B. Lisowska- 35. Kufer, T. A., E. Kremmer, D. J. Banks, and D. J. Philpott. 2006. Role for erbin in Grospierre, and V. Steimle. 2003. N-terminal destruction signals lead to rapid bacterial activation of Nod2. Infect. Immun. 74: 3115–3124. degradation of the major histocompatibility complex class II transactivator 36. Zurek, B., H. Bielig, and T. A. Kufer. 2011. Cell-based reporter assay to analyze CIITA. Eur. J. Immunol. 33: 2337–2347. activation of Nod1 and Nod2. Methods Mol. Biol. 748: 107–119. 48. Masternak, K., A. Muhlethaler-Mottet, J. Villard, M. Peretti, and W. Reith. 2000. 37. Istomin, A. Y., and A. Godzik. 2009. Understanding diversity of human innate Molecular genetics of the Bare lymphocyte syndrome. Rev. Immunogenet. 2: immunity receptors: analysis of surface features of leucine-rich repeat domains 267–282. in NLRs and TLRs. BMC Immunol. 10: 48. 49. Meissner, T. B., Y.-J. Liu, K.-H. Lee, A. Li, A. Biswas, M. C. J. A. van 38. Proell, M., S. J. Riedl, J. H. Fritz, A. M. Rojas, and R. Schwarzenbacher. 2008. Eggermond, P. J. van den Elsen, and K. S. Kobayashi. 2012. NLRC5 cooperates The Nod-like receptor (NLR) family: a tale of similarities and differences. PLoS with the RFX transcription factor complex to induce MHC class I gene ex- ONE 3: e2119. pression. J. Immunol. 188: 4951–4958. 39. Gorelik, E., S. Peppoloni, R. Overton, and R. B. Herberman. 1985. Increase in 50. van den Elsen, P. J., S. J. Gobin, M. C. van Eggermond, and A. Peijnenburg. H-2 antigen expression and immunogenicity of BL6 melanoma cells treated with 1998. Regulation of MHC class I and II gene transcription: differences and N-methyl-N9-nitronitrosoguanidine. Cancer Res. 45: 5341–5347. similarities. Immunogenetics 48: 208–221. Downloaded from 40. Riley, J. L., S. D. Westerheide, J. A. Price, J. A. Brown, and J. M. Boss. 1995. 51. Garcı´a, A. V., and J. E. Parker. 2009. Heaven’s Gate: nuclear accessibility and Activation of class II MHC genes requires both the X box region and the class II activities of plant immune regulators. Trends Plant Sci. 14: 479–487. transactivator (CIITA). Immunity 2: 533–543. 52. Burch-Smith, T. M., M. Schiff, J. L. Caplan, J. Tsao, K. Czymmek, and 41. Zhou, H., and L. H. Glimcher. 1995. Human MHC class II gene transcription S. P. Dinesh-Kumar. 2007. A novel role for the TIR domain in association with directed by the carboxyl terminus of CIITA, one of the defective genes in type II pathogen-derived elicitors. PLoS Biol. 5: e68. MHC combined immune deficiency. Immunity 2: 545–553. 53. Wiermer, M., H. Germain, Y. T. Cheng, A. V. Garcı´a, J. E. Parker, and X. Li. 42. Muhlethaler-Mottet, A., L. A. Otten, V. Steimle, and B. Mach. 1997. Expression 2010. Nucleoporin MOS7/Nup88 contributes to plant immunity and nuclear

of MHC class II molecules in different cellular and functional compartments is accumulation of defense regulators. Nucleus 1: 332–336. http://www.jimmunol.org/ controlled by differential usage of multiple promoters of the transactivator 54. Cheng,Y.T.,H.Germain,M.Wiermer,D.Bi,F.Xu,A.V.Garcı´a, L. Wirthmueller, CIITA. EMBO J. 16: 2851–2860. C. Despre´s, J. E. Parker, Y. Zhang, and X. Li. 2009. Nuclear pore complex com- 43. Nickerson, K., T. J. Sisk, N. Inohara, C. S. Yee, J. Kennell, M. C. Cho, ponent MOS7/Nup88 is required for innate immunity and nuclear accumulation P. J. Yannie, II, G. Nunez, and C. H. Chang. 2001. Dendritic cell-specific MHC of defense regulators in Arabidopsis. Plant Cell 21: 2503–2516. by guest on September 27, 2021 Figure S1

A 25000 THP1

20000

15000

10000

5000 HLA-ABC [surface expression]

0 siCTL siNLRC5_1 siNLRC5_4

B 1.2 HLA-B hFibr (donor 2) NLRC5 1.0

0.8

0.6

0.4 [relative expression] 0.2

0 siCTL siNLRC5_4 siCTL siNLRC5_4 siCTL siNLRC5_4

SeV poly[I:C]

C

B16F10 70 B16F10-CIITA 60

50 40

Counts 30 20 10

0 4 110102 103 10 MHC class I (H-2k b )

Figure S1 (relates to Figures 1 and 2):

A. PMA differentiated THP1 cells were transfected with 50 nM of siRNA for 72 h. Flow cytometry analysis of MHC class I surface expression in these cells is shown. Data is shown as median fluorescence intensity + S.D. (see also Figure 1C).

B. Primary human dermal fibroblasts from donor 2 were transfected with 10 nM of NLRC5 specific siRNA or scrambled control siRNA (siCTL) for 72 h and subsequently treated with 10 g/ml poly(I:C) or 80 HAU/ml Sendai virus for 16 h. Quantitative RT-PCR of NLRC5 and HLA-B expression is shown as the mean of the relative expression to GAPDH + S.D. (n=3). (relates to Figure 1E).

C. Histogramm visualizing the effect of CIITA and MHC class I expression, representing the data shown in Figure 2B (panels 1 and 5) as a histogram overlay. (see also Figure 2B). Figure S2

A

15.0 15.0 15.0 12.5 12.5 12.5 10.0 10.0 10.0 7.5 7.5 7.5 5.0 5.0 5.0 2.5 2.5 HLA-B250 [fold] 2.5 0 0 0 01510 01510 01510 NLRC5 wt [ng] NLRC5 K234A [ng] NLRC5 Iso3 [ng]

15.0 15.0 15.0 12.5 12.5 12.5 10.0 10.0 10.0 7.5 7.5 7.5 5.0 5.0 5.0

HLA-B250 [fold] 2.5 2.5 2.5 0 0 0 01510 0150 1 5 10 01510 NLRC5 DD NLS [ng] NLRC5 DD [ng] NLRC5 LRR [ng]

B

DD NLS wt KA  Iso3 DD LRR NLRC5 NLRC5-LRR * Iso3

DD

Tubulin

Figure S2 (relates to Figure 3):

A. HLA-B250 reporter luciferase assays in HEK293T cells where increasing amounts of the corresponding constructs where transiently transfected. Luciferase activity, normalized to -galactosidase activity is shown as fold activation to cells transfected with empty vector ("0" set to 1). Data represents mean + S.D. (n=3). All constructs were tested in a single transfection series. (relates to Figure 3B).

B. Western blot analysis showing comparable expression of the NLRC5-constructs in HEK293T. Probing for a-tubulin served as control for equal loading. * indicates unspecific background.