NLR family member NLRC5 is a transcriptional regulator of MHC class I genes

Abstract

MHC class I plays a critical role in the immune defense against viruses and tumors by presenting antigens to CD8 T cells. An NLR protein, class II transactivator (CIITA), is a key regulator of MHC class II gene expression that associates and cooperates with transcription factors in the MHC class II promoter. Although CIITA also transactivates MHC class I gene promoters, loss of CIITA in humans and mice results in the severe reduction of only MHC class II expression, suggesting that additional mechanisms regulate the expression of MHC class I. Here, we identify another member of the NLR protein family, NLRC5, as a transcriptional regulator of MHC class I genes. Similar to CIITA, NLRC5 is an IFN-γ–inducible nuclear protein, and the expression of NLRC5 resulted in enhanced MHC class I expression in lymphoid as well as epithelial cell lines. Using chromatin immunoprecipitation and reporter gene assays, we show that NLRC5 associates with and activates the promoters of MHC class I genes. Furthermore, we show that the IFN-γ–induced up-regulation of MHC class I requires NLRC5, because knockdown of NLRC5 specifically impaired the expression of MHC class I. In addition to MHC class I genes, NLRC5 also induced the expression of β2-microglobulin, transporter associated with antigen processing, and large multifunctional protease, which are essential for MHC class I antigen presentation. Our results suggest that NLRC5 is a transcriptional regulator, orchestrating the concerted expression of critical components in the MHC class I pathway.

MHC class I and class II play essential roles in the activation of adaptive immune responses by presenting antigens to T lymphocytes. MHC class I molecules are composed of MHC-encoded heavy chains and the invariant subunit β2-microglobulin (β2M) (1). Humans have three classical MHC class Ia molecules (HLA-A, HLA-B, and HLA-C), which are vital to the detection and elimination of viruses, cancerous cells, and transplanted cells. In addition, there are three nonclassical MHC class Ib molecules (HLA-E, HLA-F, and HLA-G), which have immune regulatory functions (2, 3). Antigen-derived peptides are presented by MHC class I–β2M complexes at the cell surface to CD8 T cells carrying an antigen-specific T cell receptor. Peptides are mostly produced from the degradation of cytoplasmic proteins by a specialized proteasome or immunoproteasome, which is optimized to generate MHC class I peptides and contains several IFN-γ–inducible subunits, such as large multifunctional protease 2 (LMP2) and LMP7 (4). Peptide loading onto MHC class I is carried out by the peptide-loading complex (PLC), which includes the MHC class I heavy chain, β2M, tapasin, endoplasmic reticulum (ER)p57, calreticulin, and transporter associated with antigen processing 1 (TAP1)/TAP2, a transporter that translocates peptides from the cytoplasm into the ER (4, 5).

Unlike MHC class II, which is found mainly in antigen-presenting cells, MHC class Ia is ubiquitously expressed in almost all nucleated cells (1, 6). Both MHC class I and class II genes are highly inducible by IFN-γ stimulation and share similar cis-regulatory elements in their promoters, termed W/S, X1, X2, and Y-box motifs, which also associate with similar transcription factor complexes (7, 8). These transcription factors include the X-box binding trimeric RFX protein complex (composed of RFX5, RFXAP, and RFXANK), the ×2-box binding cAMP response element binding protein/activating transcription factor (CREB/ATF), and the Y-box binding nuclear factor-Y protein (composed of NF-YA, NF-YB, and NF-YC) (9). Together, they form a macromolecular nucleoprotein complex called the MHC enhanceosome (10).

Class II transactivator (CIITA), a member of the NLR or nucleotide binding domain (NBD) leucine-rich repeat (LRR) family of proteins (11, 12), regulates the transcription of MHC class II by associating with the MHC enhanceosome (10, 13). The expression of CIITA is induced in B cells and dendritic cells as a function of developmental stage and is inducible by IFN-γ in most cell types (14 –16). There are 22 NLR proteins in humans, which share three characteristic functional domains: an N-terminal protein–protein interaction domain such as a caspase recruitment domain (CARD) or a PYRIN domain, a centrally located NBD (or NACHT, NAIP, CIITA, HET-E and TP1 domain), and C-terminal LRRs (11, 12). Aside from CIITA, NLR proteins are localized in the cytoplasm and contribute to the innate immune response by recognizing microbial products and exogenous danger signals, leading to inflammation and/or cell death (11, 12).

Previous studies have shown that, although to a lesser extent, CIITA also has a role in the transactivation of MHC class I genes (6 –9, 17). However, the expression of CIITA is generally restricted to lymphocytes and professional antigen-presenting cells and thus, is unlikely to account for the ubiquitous expression of MHC class I (6, 18). Furthermore, although mutations of the CIITA gene can cause bare lymphocyte syndrome (BLS), an immunodeficiency characterized by the lack of MHC class II expression, a subgroup of BLS patients that lack CIITA retains the expression of MHC class I but not MHC class II (19, 20). Similarly, in mice deficient for CIITA, both constitutive and IFN-γ–induced expression of MHC class I molecules is intact (21 –23). These findings indicate that, in addition to CIITA, other molecules or mechanisms are involved in the regulation of MHC class I expression.

Here, we show that another NLR protein, NLRC5 (NLR family, CARD domain containing 5/NOD27/CLR16.1), regulates the expression of MHC class I. Similar to CIITA, NLRC5 is highly inducible by IFN-γ and can translocate into the nucleus. We show that NLRC5 activates the promoters of MHC class I genes and induces the transcription of MHC class I as well as related genes involved in MHC class I antigen presentation, suggesting that NLRC5 is a critical regulator of MHC class I-mediated immune responses.

Results

NLRC5 Contains a Nuclear Localization Signal and Shuttles Between the Cytosol and Nucleus.

To study the function of NLRC5, we investigated its cellular distribution using a GFP fusion protein. To our surprise, NLRC5 was found not only in the cytosol but also in the nucleus (Fig. 1A Upper). We checked the stability of the fusion protein by Western blot analysis, confirming that its nuclear localization was not caused by a smaller, GFP-containing cleavage product (Fig. S1A). It has been shown that CIITA, which also displays a heterogeneous steady-state distribution, shuttles between the nucleus and cytosol as a result of nuclear localization signal (NLS)-mediated nuclear import and exportin-1 (CRM1)-dependent nuclear export (24 –26). Similar to CIITA, which is a closely related member of the NLR protein family (Fig. 1B), NLRC5 could be trapped in the nucleus on treatment with the CRM1 inhibitor leptomycin B (LMB) (Fig. 1A Lower and Fig. S1B). Quantification of the cellular distribution in a blind manner revealed that under steady-state conditions, NLRC5 localized exclusively in the cytosol in ∼15% of the cells. The majority of the cells showed an intermediate distribution (80%), and about 5% of the cells displayed an exclusively nuclear localization (Fig. S1B). LMB treatment resulted in nuclear localization of NLRC5 in more than 75% of the cells. Of note, we observed that in cells highly expressing the protein, NLRC5 was predominantly localized to the cytosol, whereas NLRC5 was found more frequently in the nucleus in cells with lower expression levels, indicating that the nuclear localization of NLRC5 is not a result of overexpression (Fig. 1A Upper). In addition to human NLRC5, similar results were obtained using the murine Nlrc5, which can also be trapped in the nucleus on LMB treatment (Fig. S2).

Fig. 1.

NLRC5 contains an N-terminal bipartite NLS and can translocate into the nucleus. (A, D, and F) HEK293T cells were transfected with expression plasmids coding for GFP or the indicated GFP fusion proteins; 48 h posttransfection, cells were treated with 10 nM leptomycin B (LMB) for 90 min or left untreated. Fixed cells were stained with Hoechst 33342 to indicate the nuclei. (Scale bar: 10 μm.) (A) Cellular localization of NLRC5 and CIITA on LMB treatment. (B) Phylogenetic tree of CARD-containing NLRs. (C) Schematic representation of the NLRC5 deletion mutant constructs used to map the nuclear localization signal. The position of the NLS is indicated by an asterisk. (D) Cellular localization of NLRC5 deletion mutants on LMB treatment. (E) Sequence of the bipartite NLS found in the N terminus of NLRC5. Alanine substitution of the right or left arm of the NLS was used to construct the NLSI and NLSII import mutant-expression plasmids. (F) Cellular localization of the NLSI and NLSII mutant forms of NLRC5 on LMB treatment.

Given the predicted size of the NLRC5 fusion protein (∼230 kDa), passive diffusion through the nuclear pore is not possible. Active transport, however, requires the presence of an NLS that is recognized by nuclear import receptors (27). To identify the NLS of NLRC5, we performed deletion mutant analysis. As depicted in Fig. 1C, we expressed the deletion mutants of NLRC5 as GFP fusion proteins. Although all fusion constructs containing an intact N-terminal CARD (WT, CARD, and ΔLRR) were found in the nucleus, deletion of the CARD (ΔCARD and LRR) resulted in a strictly cytosolic localization (Fig. 1D). Similar to free GFP, the NACHT domain fusion protein was found in both the nucleus and cytosol, presumably because of passive diffusion as a result of its smaller size. These results suggested that an NLS may be located in the N-terminal CARD. Indeed, sequence analysis of NLRC5 revealed a putative bipartite NLS at the transition between the CARD and the NBD (Fig. 1E) (25, 26). As predicted, mutation of the NLS abolished nuclear localization under steady-state conditions, and treatment of the cells with LMB failed to trap the NLS mutants of NLRC5 in the nucleus (Fig. 1F). Taken together, our results show that, similar to the transcriptional coregulator CIITA, NLRC5 shuttles between the cytosol and nucleus and thus, is likely to have a nuclear function.

NLRC5 Transcriptionally Induces the Expression of MHC Class I and Functionally Related Genes.

The observation that NLRC5 is also found in the nucleus and shares significant sequence similarity to the transcriptional coregulator CIITA (Fig. 1 A and B) prompted us to perform a gene array to identify putative target genes of NLRC5. For this purpose, we generated Jurkat T cell lines that stably express either the wild-type or mutant forms of NLRC5 harboring mutations in the NBD: Walker A (deficient in nucleotide binding), Walker B (deficient in nucleotide hydrolysis), and the combined Walker AB, carrying both mutations (Fig. S3A) (28). Gene-chip analysis using these mutant Jurkat T cells showed that a surprisingly limited number of genes were differentially regulated (Fig. S3). As predicted, clustering analysis grouped the active forms of NLRC5 (WT and Walker B) together, and they show a strikingly different pattern of gene expression compared with cells expressing either GFP alone or the catalytically inactive forms of NLRC5 (Walker A and Walker AB). Among the genes most up-regulated by the active forms of NLRC5 were the various members of the MHC class I (HLA-A, -B, -C, and -E) family as well as other genes involved in class I antigen presentation and processing, such as β2M, LMP2, and TAP1 (Fig. S3 B and C). Quantitative real-time PCR and Western blot analysis confirmed elevated levels of the corresponding transcripts and proteins, respectively, in cells expressing the WT and Walker B mutant NLRC5, but not GFP alone, or the inactive forms of NLRC5 (Walker A and Walker AB) (Fig. 2 A and B). Furthermore, flow-cytometry analysis using a pan HLA-A, -B, or -C antibody and an antibody specific for HLA-E confirmed an increase in MHC class I surface expression in cells expressing NLRC5 WT or the Walker B mutant (Fig. 2C Left). As previously shown, MHC class I and related genes are inducible by IFN-γ (Fig. 2C bottom row ) (5, 29). However, we did not observe elevated levels of IFN-γ expression in our gene-array analysis, and the expression level of signal transducer and activator of transcription (STAT1), an IFN-γ–inducible gene, did not vary between the different cell lines (Fig. 2A). These findings, along with the observation that overexpression of NLRC5 does not activate NF-κB-, activator protein 1 (AP-1), interferon-sensitive response element (ISRE), or interferon regulatory factor 3 (IRF3)-dependent reporter genes or the promoters of IFN-α and IFN-β (Fig. S4), rule out the role of these other pathways in NLRC5-mediated MHC class I induction. Instead, NLRC5 might directly regulate the expression of MHC class I genes.

Fig. 2.

Induction of MHC class I and functionally related genes by NLRC5. (A) RNA isolated from Jurkat T cells stably expressing the indicated GFP fusion proteins was analyzed by quantitative real-time PCR for the expression of the indicated genes. GFP, empty vector; WT, wild-type NLRC5; A, Walker A mutant; B, Walker B mutant; AB, Walker AB mutant. (B) The same Jurkat T cell lines were examined for the expression of MHC class I heavy chain (HC), β2M, TAP1, and LMP2 by Western blot analysis. Actin levels are shown as a loading control. (C) Surface expression of MHC class I in Jurkat T cell lines expressing GFP (gray line) or the indicated GFP-NLRC5 fusion proteins (black line) was examined by flow cytometry using anti–pan-MHC class I (HLA-A, -B, and -C) and HLA-E antibodies. IFN-γ (100 U/mL) treatment was used as a positive control. Data obtained with an isotype control antibody are indicated by the shaded area. (D) HEK293T cells were transiently transfected with the expression plasmids for GFP fused to NLRC5, CIITA (black line), or GFP only (gray line). The expression of MHC class I (HLA-A, -B, and -C) or class II (HLA-DR) was analyzed by flow cytometry 48 h posttransfection. Data obtained with an isotype control antibody are indicated by the shaded area.

Because MHC class I is ubiquitously expressed in all nucleated cells, we sought to determine whether the observed up-regulation of MHC class I genes was limited to lymphoid cells or could be extended to other cell types. As shown in Fig. 2D, transient expression of NLRC5 in an epithelial cell line (HEK293T cells) also increased MHC class I expression nearly 4-fold. In comparison, expression of CIITA only moderately increased MHC class I expression but in agreement with previous reports (6, 9), strongly induced the expression of MHC class II. Expression of the Walker A and B mutants in HEK293T cells (Fig. S5B) again showed that nucleotide binding, but not nucleotide hydrolysis, was required for the activity of NLRC5 and the induction of MHC class I. Importantly, this transcriptional effect seems to be specific for the nuclear NLRs, because none of the cytosolic CARD-containing NLRs tested (NOD1, NOD2, and NLRC3) increased the expression of MHC class I as shown by flow cytometry and Western blot analysis (Fig. S5 A and B). In summary, these data indicate that NLRC5 induces the expression of MHC class I and related genes involved in MHC class I antigen presentation and thus, can substitute for IFN-γ stimulation of cells.

NLRC5 Binds to MHC Class I Promoters and Induces Their Expression.

To investigate whether NLRC5 directly acts on the promoters of the MHC class I genes, we performed luciferase-reporter gene assays with the promoters of the corresponding genes. Transient expression of NLRC5 in HEK293T cells is sufficient to induce luciferase expression from the promoters of HLA-A, -B, -C, -F, -G, and β2M (Fig. 3A). Similar levels of induction on the same promoters were observed when CIITA was overexpressed as has been reported previously (8, 29 –31). Only a minor induction was observed on the promoter of TAP1, and NLRC5 failed to induce luciferase expression on the TAP2 promoter and any of the MHC class II reporter constructs analyzed (HLA-DRA, -DQA, and -DPA). In contrast, CIITA transfection strongly activated the promoters of MHC class II genes. Next, we examined if NLRC5 also physically associates with the MHC class I promoters using the stable Jurkat T cell lines described earlier in a ChIP assay. We immunoprecipitated the corresponding wild-type and mutant NLRC5 proteins, and the associated DNA fragments were quantified by qPCR using gene-specific primers covering the immediate upstream region of the HLA genes (Fig. 3B). As seen in Fig. 3C, a 6- to 8-fold enrichment in promoter occupancy was observed for NLRC5 WT and the Walker B mutant on the promoters of HLA-A and HLA-B compared with the cell line expressing GFP only. In agreement with the data obtained from the gene-expression analyses, no promoter binding was detected for the inactive forms of NLRC5 (Walker A and Walker AB) as well as on the promoter of the MHC class II (HLA-DRA) and an unrelated gene (GAPDH). Furthermore, ChIP analysis in nonhematopoietic cells using transiently transfected HEK293T cells revealed that NLRC5 can associate with HLA-A and -B promoters to a similar extent as CIITA (Fig. S6), which has previously been reported to bind to MHC class I promoters (10). Taken together, the luciferase assay and the ChIP experiment show that NLRC5 not only associates with the promoters of the MHC class I genes with remarkable specificity but also has the capacity to transactivate their expression.

Fig. 3.

NLRC5 binds and transactivates MHC class I gene promoters. (A) NLRC5-mediated transactivation of MHC class I and functionally related genes. HEK293T cells were transiently transfected with either expression vectors for GFP, GFP-NLRC5, or GFP-CIITA, along with luciferase reporter constructs of the indicated gene promoters. Cell lysates were analyzed 48 h posttransfection by dual-luciferase assay. Data are a representative of three independent experiments performed in duplicates, and error bars represent ± SD. (B) Schematic representation of the W/SXY module found in the promoters of MHC class I and class II genes. The position of the primers used in the ChIP assay is indicated with arrows (P1 and P2). (C) NLRC5 occupancy, in terms of fold enrichment, at the HLA-A, -B, or -DRA promoters as determined by ChIP. Jurkat T cells stably expressing the indicated GFP fusion proteins were analyzed by ChIP assay using an anti-GFP antibody for immunoprecipitation and the indicated qPCR primers (B). GFP, empty vector; WT, wild-type NLRC5; A, Walker A mutant; B, Walker B mutant; AB, Walker AB mutant. Error bars indicate SEM (± SEM) from three independent experiments.

NLRC5 Is Rapidly Induced by IFN-γ and Is Required for IFN-γ–Induced Expression of MHC Class I.

It has been shown that rapid induction of CIITA mediates the up-regulation of MHC class II on IFN-γ stimulation (15, 32). Because NLRC5 is also an IFN-γ–inducible gene (33), we explored the possibility that NLRC5 may mediate the IFN-γ–induced transcription of MHC class I genes. First, we compared the expression kinetics of NLRC5 and a MHC class I gene on IFN-γ treatment. HLA-A transcript levels reach a maximum only 12–24 h after IFN-γ stimulation in HeLa cells, but similar to the IFN-γ–response gene STAT1, NLRC5 is induced early after IFN-γ treatment (Fig. 4A), which is also a characteristic of CIITA induction by IFN-γ (15, 32). Similar kinetics of NLRC5 and HLA-A expression were observed in Jurkat T cells (Fig. S7 A and B).

Fig. 4.

Knockdown of NLRC5 results in decreased up-regulation of MHC class I on IFN-γ treatment. (A) HeLa cells were stimulated with IFN-γ (100 U/mL) for the indicated time points, and the kinetics of NLRC5, HLA-A, and STAT1 expression were analyzed by quantitative real-time PCR. (B and C) HeLa cells were transfected with NLRC5-specific or control siRNAs; 16 h posttransfection, cells were stimulated with IFN-γ for 24 h. (B) Knockdown efficiency of NLRC5 was determined by quantitative real-time PCR using gene-specific primers, and data were normalized to the expression of the GAPDH gene. Scr, control scrambled siRNA. Error bars represent the ± SD from one representative of three independent experiments performed in duplicates. *P < 0.05. (C) Surface expression of MHC class I and β1-integrin was analyzed by flow cytometry. (D) Model depicting the role of NLRC5 in the IFN-γ–induced up-regulation of MHC class I genes. Discussion has further details.

Next, we analyzed the effect of NLRC5 depletion by RNA interference on the expression of MHC class I after IFN-γ stimulation. We had observed earlier that surface expression of MHC class I is readily induced on IFN-γ stimulation (Fig. S7C), and in agreement with our hypothesis, transfection of HeLa cells with two different NLRC5-specific siRNAs, but not a scrambled control siRNA, significantly reduced the IFN-γ–induced up-regulation of MHC class I (Fig. 4C Left and Fig. S8). In contrast, an unrelated but IFN-γ–inducible surface receptor, β1-integrin, was not affected by the depletion of NLRC5 (Fig. 4C Right). Similarly, the IFN-γ–induced expression of CIITA and HLA-DR was not affected by the depletion of NLRC5 (Fig. S8), strongly suggesting that NLRC5 is required for the efficient induction of MHC class I observed on IFN-γ stimulation.

Discussion

Since the complementation cloning of CIITA from MHC class II-deficient patients in 1993, CIITA has been often referred to as a master regulator of MHC class II expression, because CIITA is required for both constitutive and IFN-γ–inducible transcription of MHC class II genes (15, 20, 32). However, the contribution of CIITA to MHC class I expression is less clear. In this study, we identify NLRC5 as a regulator of MHC class I genes in addition to CIITA. NLRC5 and CIITA share important characteristics in their structure and function. First, as related members of the NLR family (Fig. 1B), both have the same tripartite architecture, although expression of the CARD-containing isoform of CIITA is limited to dendritic cells (34). Interestingly, both proteins require an active NBD for their function. It has been shown that the NTP binding motif in CIITA is essential for transactivation of MHC class II genes (28, 35, 36). Similarly, the Walker A mutation, which prevents NTP binding, but not the Walker B mutation, which abolishes NTP hydrolysis, resulted in a loss of NLRC5 function (Fig. 2). Second, both proteins can localize to the nucleus. CIITA carries three NLSs, including an N-terminal NLS, which is found at a similar position to that required for NLRC5 nuclear translocation (Fig. 1 E and F) (24 –26). In addition, multiple nuclear export signals (NES) are predicted in the C-terminal LRRs of CIITA, and our deletion mutant analysis suggests that the C-terminal LRRs of NLRC5 are also involved in the regulation of nuclear export, although we have not further mapped the exact position of the NES (26). Recently, it was shown that cytosolic NLRC5 negatively regulates the NF-κB and type I IFN signaling pathway by direct binding to IκB kinase (IKK) and retinoic acid-inducible gene I (RIG-I) (37). Our findings do not rule out a function of NLRC5 in the cytoplasm but rather show its role in the nucleus as a transcriptional regulator of MHC class I genes. Third, despite the lack of a DNA binding domain, both NLRC5 and CIITA can associate with and transactivate MHC class I promoters (Fig. 3 A and C and Fig. S6) (10, 17, 29). CIITA is known to associate with a set of transcription-factor complexes, or MHC enhanceosome, on the WXY motif of the MHC class I and class II gene promoters. The results of our ChIP and reporter gene assays indicate that NLRC5 may use a similar platform to activate MHC class I gene promoters. Finally, both NLRC5 and CIITA are highly inducible on IFN-γ stimulation (Fig. 4A) (15, 32, 33), and binding sites for STAT1, which are activated on IFN-γ stimulation, have been mapped in the promoters of both genes (33, 38 –40). This suggests that both proteins are involved in mediating IFN-γ–induced changes in gene expression. In particular, CIITA and NLRC5 seem to orchestrate the concerted expression of sets of functionally related genes critical for antigen presentation. CIITA, in addition to the classical MHC class II genes, induces the invariant chain Ii and the nonclassical MHC class II genes HLA-DM and HLA-DO, which play accessory roles in MHC class II antigen presentation (16). We show here that NLRC5, beyond the induction of MHC class I genes, up-regulates β2M, TAP1, and LMP2, which are essential for antigen presentation by the MHC class I pathway (Fig. 2A).

However, in spite of these similarities and overlapping functions, there are also noticeable differences between NLRC5 and CIITA. A unique feature of NLRC5 is its striking specificity for the induction of genes involved in the MHC class I pathway as opposed to CIITA, which can induce both MHC class I and class II genes. In this study, we found that expression of NLRC5 in epithelial and lymphoid cells was sufficient to induce MHC class I but not MHC class II genes, despite their similar promoter architecture (Fig. 2 C and D). Furthermore, our findings also suggest that NLRC5 is exclusively associated with the promoters of MHC class I (Fig. 3C) and NLRC5 transactivated promoters of MHC class I and related genes but not those of MHC class II genes (Fig. 3A). A possible explanation for this specificity could lie in the structural differences between the two proteins. NLRC5, unlike CIITA, lacks N-terminal acidic and proline/serine/threonine-rich domains, which are required for MHC class II promoter activation (41). NLRC5 will, thus, require additional cofactors to interact with and activate the enhanceosome found on the MHC class I promoters.

Given its specificity for MHC class I induction, it is also possible that NLRC5 plays a dominant role in the regulation of MHC class I gene expression. This view is supported by the results of our knockdown analyses, which clearly show that the IFN-γ–induced up-regulation of CIITA cannot compensate for the reduction in MHC class I expression observed on NLRC5 depletion (Fig. 4C and Fig. S8). Furthermore, no reduction in MHC class I expression has been observed in CIITA-deficient mice (21 –23). Taken together, our findings show that NLRC5 is necessary and sufficient for the induction of MHC class I expression. NLRC5 may, thus, act as a counterpart to CIITA in its function as an MHC class I transactivator (CITA). Future analyses of the in vivo function of NLRC5 are required to reveal if these two molecules play redundant or more exclusive roles in MHC class I-dependent immune responses.

We, thus, propose the following model of NLRC5 function in the expression of MHC class I genes. On IFN-γ stimulation, activated STAT1 acts on the promoters of NLRC5 and CIITA and rapidly induces these genes (Fig. 4D). Subsequently, CIITA may activate the promoters of both MHC class I and class II genes by associating with the MHC enhanceosome, which includes the RFX, CREB/ATF, and NF-Y protein complexes on the conserved WXY module in the MHC promoters (Fig. 4D). NLRC5 may also associate with a similar enhanceosome on the MHC class I promoter, consisting of the same or similar components as those described for the CIITA enhanceosome. However, unlike the CIITA enhanceosome, the NLRC5 enhanceosome is specific to promoters of MHC class I and related genes (Fig. 4D).

In conclusion, we have found that an NLR protein, NLRC5, acts as a transactivator for MHC class I and related genes. Because MHC class I is crucial for the activation of CD8 T cells, NLRC5 may play an important role in the immune response against viral infection or cancers. Future studies are required to unveil the function of NLRC5 in vivo under these pathological conditions.

Materials and Methods

Cell Lines and Reagents.

Human embryonic kidney 293T (HEK293T) cells (CRL-11268; ATCC) and HeLa cells (CCL-2; ATCC) were cultured in DMEM supplemented with 10% FBS and penicillin (100 U/mL)/streptomycin (100 μg/mL; Gibco). Jurkat T cells (TIB152; ATCC) were maintained in RPMI-1640 (Thermo Scientific) supplemented with 10% FBS and penicillin/streptomycin. HEK293T was transiently transfected using FuGENE 6 Transfection Reagent (Roche) in serum-free media according to the manufacturer's protocol. Recombinant human IFN-γ is from BioLegend. LMB was obtained from LC Laboratories.

Quantitative Real-Time PCR Analysis.

Quantitative real-time PCR analysis was performed as described and is detailed in SI Materials and Methods (42).

Flow Cytometry.

Antibodies against human HLA-A, -B, -C (W6/32), HLA-E (3D12), HLA-DR (L243; Biolegend), and β1-integrin (TS2/16; Martin Hemler, DFCI, Boston, MA) were used in this study. Cells were stained, washed, resuspended in PBS/1% FBS/0.05% NaN₃, and analyzed by FACSCalibur (Becton Dickinson) followed by analysis using FlowJo software.

Knockdown of NLRC5 by RNA Interference.

HeLa cells (0.5 × 10⁶/well) were transfected with 20 nM siRNA using Hyperfect (Qiagen) according to the manufacturer's instructions. Cells were stimulated 16 h posttransfection with 100 U/mL IFN-γ (BioLegend). After 24 h of stimulation, cells were harvested and analyzed by flow cytometry and quantitative real-time PCR. The control siRNA (scrambled) as well as siRNAs targeting NLRC5 were obtained from Ambion.

Luciferase Assay.

HEK293T cells were split into 24 wells and cotransfected with 300 ng of GFP-, GFP-NLRC5–, or GFP-CIITA–expression plasmids and 100 ng of the indicated luciferase reporter constructs; 50 ng/well of promoterless Renilla luciferase vector (pRL-null; Promega) were included for normalization of transfection efficiency. Cells were harvested 48 h posttransfection, and cell lysates were analyzed using the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions. The reporter gene constructs were previously described (31).

ChIP Assay.

Chromatin immunoprecipitation of NLRC5 was performed as previously described and is detailed in SI Materials and Methods (43).

Statistical Analysis.

Data were subjected to Student t test for analysis of statistical significance, and a P value of <0.05 was considered to be significant.