HNRNPR Regulates the Expression of Classical and Nonclassical MHC Class I

This information is current as Adi Reches, Daphna Nachmani, Orit Berhani, Alexandra of September 29, 2021. Duev-Cohen, Dorin Shreibman, Yael Ophir, Barbara Seliger and Ofer Mandelboim J Immunol 2016; 196:4967-4976; Prepublished online 18 May 2016;

doi: 10.4049/jimmunol.1501550 Downloaded from http://www.jimmunol.org/content/196/12/4967

References This article cites 34 articles, 6 of which you can access for free at: http://www.jimmunol.org/content/196/12/4967.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

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

• No Triage! Every submission reviewed by practicing scientists by guest on September 29, 2021 • 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 © 2016 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

HNRNPR Regulates the Expression of Classical and Nonclassical MHC Class I Proteins

Adi Reches,* Daphna Nachmani,* Orit Berhani,* Alexandra Duev-Cohen,* Dorin Shreibman,* Yael Ophir,* Barbara Seliger,† and Ofer Mandelboim*

MHC class I molecules, in addition to their role in specific activation of the CTL of adaptive immune system, function also as the main ligands for NK cell inhibitory receptors, which prevent NK cells from killing normal, healthy cells. MHC class I proteins are divided into classical and nonclassical proteins. The former group consists of hundreds of HLA-A, B, and C alleles, which are universally expressed, whereas several alleles of the latter group, such as HLA-G, manifest a restricted expression pattern. Despite the important role played by these molecules in innate and adaptive immune responses, their complex expression regulation is not fully known. In our study, we investigated the regulation processes controlling the expression of MHC class I molecules, with a particular focus on their 39 untranslated regions. We identified heterogeneous nuclear ribonucleoprotein R (HNRNPR) as an Downloaded from important positive regulator of classical and nonclassical MHC class I molecules. HNRNPR is a RNA-binding belonging to the heterogeneous nuclear ribonucleoprotein family, which has a known role in processing of precursor mRNA. We demonstrated that HNRNPR binds MHC class I mRNAs in their 39 untranslated regions and enhances their stability and consequently their expression. Furthermore, regulation by HNRNPR modulates the cytotoxic activity of NK cells. In con- clusion, we show that HNRNPR acts as a general positive regulator of MHC class I expression. The Journal of Immunology,

2016, 196: 4967–4976. http://www.jimmunol.org/

ajor histocompatibility complex class I proteins are one family shows a unique affinity to HLA-C subtypes. Practically of the major immune system modulators. They present all of the HLA-C alleles can be divided into two groups, which M peptides to T cells and activate CTLs of the adaptive are recognized by KIR2DL1 or by KIR2DL2 inhibitory NK re- immune system, and, in contrast, serve as inhibitory ligands for NK ceptors. Generally, the KIR2DL1 receptor recognizes HLA-C sub- cell inhibitory receptors (1). NK cells are innate lymphocytes that, types (C2 group), having lysine at position 80, whereas KIR2DL2 besides supporting developmental processes such as fetal growth receptors recognize HLA-C subtypes (C1 group), having asparagine during pregnancy (2), are best known for their ability to dis- atposition80(5). by guest on September 29, 2021 criminate between self and altered self by killing virally infected, MHC class I proteins are divided into classical and nonclassical. transformed, and damaged cells. NK cell activity is governed by In humans, the classical proteins are named HLA-A, -B, and -C, integrating signals derived from a panel of activating and in- each containing hundreds of different alleles. The nonclassical hibitory receptors. The inhibitory receptors are expressed sto- proteins are significantly less polymorphic and include HLA-E, chastically on NK cells, and their main inhibitory ligands are -F, -G, and HFE (6). MHC class I molecules (3, 4). There are several families of HLA-G is one of the most interesting nonclassical MHC class I MHC class I inhibitory receptors; among them, the KIR2DL proteins. It has a very unique expression pattern, in which it is upregulated on cancer cells, whereas on normal tissues its ex- pression is mainly restricted to the extravillous cytotrophoblasts *Lautenberg Center for General and Tumor Immunology, Institute of Medical Re- of the placenta, which is of fetal origin (7). Although HLA-G search Israel-Canada, Faculty of Medicine, Hebrew University Hadassah Medical School, 9112001 Jerusalem, Israel; and †Institute of Medical Immunology, Martin mRNA can be found in several tissues, HLA-G protein is absent, Luther University Halle-Wittenberg, 06112 Halle (Saale), Germany suggesting for regulation of HLA-G expression at the mRNA ORCID: 0000-0002-8489-9986 (D.N.). level. Received for publication July 9, 2015. Accepted for publication April 20, 2016. The transcription processes regulating MHC expression This work was supported by the European Research Council under the European were broadly investigated (8), and a few posttranscription regu- Union’s Seventh Framework Programme (FP/2007-2013)/European Research Council lators were discovered as well (9, 10). However, whether RNA- Grant Agreement 320473-BacNK. Further support came from the Israel Science Foun- dation, the German–Israeli Foundation for Scientific Research and Development (to O.M. binding proteins (RBPs) control the expression of classical and and B.S.), the Lewis Family Foundation, an Israel Cancer Research Fund professorship nonclassical MHC class I proteins is largely unexplored. grant, a Helmholtz Israel grant, and the Rosetrees Trust. This work was also supported RBPs play a part in every aspect of RNA biogenesis, including by Grants GRK1591 and SE-581-22-1 from the Deutscheforschungsgemeinschaft (to B.S.). transcription, pre-mRNA splicing, polyadenylation, RNA modifi- Address correspondence and reprint requests to Prof. Ofer Mandelboim, Lautenberg cation, transport, localization, translation, turnover, and immune ac- Center for General and Tumor Immunology, Hebrew University Hadassah Medical tivities (11–15). One major family of RBPs is the heterogeneous School, P.O. Box 12272, 9112001 Jerusalem, Israel. E-mail address: oferm@ekmd. huji.ac.il nuclear ribonucleoproteins (HNRNPs), which are among the most abundant proteins in the eukaryotic nucleus, taking part in processing Abbreviations used in this article: HNRNP, heterogeneous nuclear ribonucleoprotein; HNRNPR, heterogeneous nuclear ribonucleoprotein R; KD, knockdown; qRT-PCR, of precursor mRNA (16). In this study, we identify a member of this quantitative RT-PCR; RBP, RNA-binding protein; UTR, untranslated region; WB, family, heterogeneous nuclear ribonucleoprotein R (HNRNPR), as Western blot. a general regulator of classical and nonclassical MHC class I Copyright Ó 2016 by The American Association of Immunologists, Inc. 0022-1767/16/$30.00 expression. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1501550 4968 HNRNPR REGULATES THE EXPRESSION OF MHC CLASS I PROTEINS

Materials and Methods proteins that bound specifically to the RNAs, a SDS gel analysis was Cell culture performed and specific bands were detected with Coomassie brilliant blue G-250 (Sigma-Aldrich). Specific bands were excised and analyzed by mass JEG-3, U87, and Mel1074 cells were maintained in DMEM. The 721.221, spectrometry. Analysis was performed by the Smoler Proteomics Center Jurkat, Bjab, and U937 cells were maintained in RPMI 1640 medium. All (Technion, Haifa, Israel). media were supplemented with 10% FCS. Generation of lentivirus, knockdown, and overexpression RNA affinity purification and mass spectrometry The various RBP-knockdown (KD) vectors and scrambled vector are in The interactions between RNA and RNA-binding proteins were analyzed by the pLKO.1-puro plasmids and were purchased from Sigma-Aldrich. RNA affinity purification, as previously described (17). In short, the 39 Transduced JEG-3, 721.221, Jurkat, Bjab, and U937 cells were grown untranslated regions (UTRs) of HLA-G (sense and antisense orientation, in the presence of 1 mg/ml, 2 mg/ml, 1 mg/ml, 2 mg/ml, and 7 mg/ml National Center for Biotechnology Information Reference Sequence: puromycin, respectively. Cloning of the HNRNPR and HLA-G 39 UTRs NT_167249.2) and a UTR of a control gene with similar length and GC containing 14-bp deletion and insertion was performed into the content (GPR112) were cloned into the pBSII plasmid using the primers pHAGE-DsRED(2)-eGFP(+) lentiviral vector, which also contains (restriction sites are underlined), as follows: HLA-G sense forward (NotI), GFP. Lentiviruses were generated in 293T cells using a transient three- 59-ACGCGGCCGCATTGAAAGGAGGGAGCTA-39 and HLA-G sense plasmid transfection protocol, as previously described (18). Trans- reverse (XbaI), 59-TGTCTAGAAAAGTTCTCATGTCTTCCATTTAT-39; duction efficiency into the cells was assessed by GFP expression, and HLA-G antisense forward (XbaI), 59-ACTCTAGAATTGAAAGGAGG- only cell populations with .90% efficiency were used for experi- GAGCTA-39 and HLA-G antisense reverse (NotI), 59-GTGCGGCCGC- ments. Primer sequences are as follows: HNRNPR forward (NotI), 59- AAAGTTCTCATGTCTTCCATTTAT-39; and GPR112 forward (NotI), 59- ATGCGGCCGCACCATGAAGACCTACAGGCAGAG-39 and HNRNPR TAGCGGCCGCTTTGTGAAGTTGTGCCTAAT-39 and GPR112 reverse reverse (AgeI), 59-TGACCGGTCTAATGGTGATGGTGATGGTGC- (XbaI), 59-AGTCTAGAAGGAGAATTATTCTGACTTTAATATTTATC-39. 9

TTCCACTGTTGCCCA-3 ; HLA-G 14-bp deletion and insertion for- Downloaded from In vitro transcription into RNA was performed using the MEGAscript ward (NotI), 59-GCGGCGGCCGCGCCGCCACCATGGTGGTCATGGC- T7 transcription kit (Life Technologies) after linearization of the plasmids G-39; and HLA-G 14-bp deletion and insertion reverse (AgeI), 59- with PspOMI restriction enzyme (Thermo Scientific [Fermentas]). About GCGACCGGTAAAGTTCTCATGTCTTCCATTTA-39.The14-bpdele- 10% of totally incorporated UTPs were biotin-16-UTPs (GE Healthcare). tion mutations in HLA-G 39 UTR 14-bp insertion were generated by The biotinylated RNAs were coupled to streptavidin-Sepharose beads (GE PCR-based site-directed mutagenesis using the 59 primer 59-AGTGG- Healthcare) and incubated with cytoplasmic extracts prepared from 80% CAAGTCCCTTTGT-39 and the 39 primer 59-ACAAAGGGACTTGC- confluent JEG-3 cells overnight at 4˚C. After purification and elution of CACT-39. http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 1. Identification of RBPs that interact with the 39 UTR of HLA-G. (A) Sequence alignment of the 39 UTRs of HLA-G, HLA-A, HLA-B, HLA-C, HLA-E, and HLA-F. The alignment was performed by using the MultAlin bioinformatics tool (34). Identical nucleotides are indicated in red. Blue nucleotides represent partial nucleotide similarities. (B) Schematic representation of the RNA affinity purification assay scheme. (C) Top three iden- tified candidate RBPs. The Journal of Immunology 4969

RNA immunoprecipitation cycler (Life Technologies) with primers targeting GAPDH, MICB, TAP1, TAP2, Tapasin, HLA-A, HLA-B, HLA-C, and HLA-G, as follows: GAPDH The assay was performed as previously described (19). Briefly, Mel1074 forward, 59-GAGTCAACGGATTTGGTCGT-39; GAPDH reverse, 59-GA- cells expressing HIS-tagged HNRNPR cells were used for the validation TCTCGCTCCTGGAAGATG-39; MICB forward, 59-CTGCTGTTTCTG- of the HNRNPR–HLA–mRNA interactions. Total cell lysate was pre- GCCGTC-39; MICB reverse, 59-ACAGATCCATCCTGGGACAG-39;TAP1 pared using an ice-cold lysis buffer. The lysate was precleared using forward, 59-TCAGGGCTTTCGTACAGGAG-39;TAP1reverse,59-TCCG- protein A/G PLUS-agarose beads sc-2003 (Santa Cruz Biotechnology). GAAACCGTGTGTACTT-39;TAP2forward,59-ACTGCATCCTGGATCT- Then the cleared lysate was incubated at 4˚C overnight with HIS-tag Ab CCC-39;TAP2reverse,59-TCGACTCACCCTCCTTTCTC-39;Tapasinfor- (R&D Systems; catalogue MAB050) or Ig control (eBioscience; catalogue ward, 59-AAGCTCAAGTCCAGCAGAGC-39; Tapasin reverse, 59-CAGCA- 14-4714-85). Protein A/G PLUS-agarose beads were then added for an GGAGCCTGTTCTCAT-39; HLA-A forward, 59-GGGTCATATGTGTCTT- additional 4 h. After several washing steps, RNA was isolated using TRI GGGG-39; HLA-A reverse, 59-GCAGTTGAGAGCCTACCTGG-39;HLA-B reagent (T9424; Sigma-Aldrich), and the presence of the HLA transcripts forward, 59-CTCATGGTCAGAGATGGGGT-39; HLA-B reverse, 59-TCC- was detected via quantitative RT-PCR (qRT-PCR). Enrichment was cal- GCAGATACCTGGAGAAC-39; HLA-C forward, 59-GTGGCCTCATGGT- culated relative to the levels of the specific transcript in the Ig control CAGAGAG-39; HLA-C reverse, 59-TCCGCAGATACCTGGAGAAC-39;HLA- samples. Gforward,59-AGTCAAAGACAGGGTGGTGG-39;andHLA-Greverse,59- GGAGTGGCTCCACAGATACC-39. RNA extraction and cDNA preparation Fusion proteins Total RNA was extracted with TRI reagent (T9424; Sigma-Aldrich) and was treated with Turbo-DNase (Ambion), and a poly(A) tail was added KIR2DL1-Ig and KIR2DL2-Ig fusion proteins were generated in 293T cells using the poly(A) kit (Ambion). For generation of cDNA libraries, the and were purified on a protein G column, as previously described (20). The Moloney murine leukemia virus reverse transcriptase (Invitrogen) was fusion proteins used in this work were regularly assayed by SDS protein gels used for reverse transcription (according to the manufacturer’s in- to ensure the proteins were not degraded. Protein purity of all Ig fusion structions), in the presence of an adapter primer. Detection of the var- proteins used in this study was ∼100%. Downloaded from ious transcripts was performed with quantitative real-time PCR (see below). Western blot analysis Quantitative real-time PCR Lysates of the cells were prepared, and SDS gel electrophoresis was executed. Proteins were transferred onto a nitrocellulose membrane For quantitative real-time PCR, freshly prepared cDNAs were used for with the tank blot procedure, and specific protein bands were detected SYBR Green-based detection in a QuantStudio 12k Flex real-time PCR using Abs detecting HNRNPR (sc-16541, Santa Cruz Biotechnology http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 2. HNRNPR controls classical MHC class I expression. (A) KD of the RBPs PABPC4 (left), DDX3X (middle), and HNRNPR (right) in JEG-3 cells was performed using specific short hairpin RNA (indicated as shRNA). WBs were performed with the appropriate specific mAbs, and expression was compared with JEG-3 cells expressing either scrambled shRNA (indicated as Scramble) or empty vector (indicated as Empty). hGAPDH was used as loading control. Figure shows one representative experiment of three performed. Contrasts in the WB figures were altered for better clarity. (B)FACS analysis of the expression of MHC class I using the anti-class I mAb W6/32 on JEG-3 cells transduced with the indicated shRNAs (black empty histograms) in comparison with scrambled shRNA (gray empty histograms). The filled gray histogram represents staining of the scrambled shRNA with secondary mAb only. The background of the shRNA was similar to the scrambled one and is not shown in the figure. Figures show one representative experiment of three performed. (C) WBs were performed on the indicated cell lines using anti-HNRNPR mAb. hGAPDH was used as control. Figure shows one representative experiment of three performed. Contrasts in the WB figures were altered for better clarity. (D) FACS analysis of the expression of MHC class I on U87 cells overexpressing HNRNPR (black empty histograms). The gray empty histogram represents U87 transduced with empty vector. The filled gray histogram represents staining of the empty vector with secondary mAb only. The background of the overexpression HNRNPR was similar to the empty vector one and is not shown in the figure. 4970 HNRNPR REGULATES THE EXPRESSION OF MHC CLASS I PROTEINS

[1:200]; 15018-1-AP, Proteintech [1:1000]), DDX3X (09-860, Millipore catalogue 709-136-098). All staining was analyzed by FACS using the [1:1000]), PABPC4 (14960-1-AP, Proteintech [1:1000]), or GAPDH CellQuest software. (SC-32233, Santa Cruz Biotechnology [1:1000]) as loading control. All was diluted in 5% BSA in PBS. Chemiluminescence caused by de- NK cell cytotoxicity assays tection Ab-linked HRP (Jackson ImmunoResearch Laboratories) was detected. NK cells were isolated from healthy donors using a MACS separation kit (Miltenyi Biotec) and grown in the presence of IL-2 (PeproTech). Target Flow cytometry cells were incubated overnight in the presence of 35S-methionine added to a methionine-free media (Sigma-Aldrich). NK cells were incubated or not Flow cytometry for the following Abs was used: HLA-G (purchased from with 1 mg anti-KIR2DL2 Ab for blocking for 1 h on ice and then added to AbD Serotec; catalogue MCA2044), HLA-A*02 (mAb clone BB7.2), the target cells. The level of 35S release was measured after 5 h of incu- HLA-B*07 (mAb clone BB7.1), MHC class I (mAb clone W6/32), ULBP1 bation with effectors using a beta counter TopCount (Packard). (mAb clone 170818), ULBP2 (mAb clone 165903), ULBP3 (mAb clone 166510), ULBP4 (mAb clone 6E6), ICAM-1 (mAb clone HCD54), CD48 RNA fluorescence in situ hybridization (mAb clone 4H9), MICA (mAb clone 159227), MICB (mAb clone 236511), KIR2DL1, KIR2DL2, and CD56 (purchased from BioLegend; The locked nucleic acid oligonucleotide 300500 (/5TYE665/ sequence 59- catalogue 339504, 312604, and 318310, respectively). Staining was performed TCGCTCTGGTTGTAGTAGC-39; Exiqon) probe that recognizes all MHC with 0.2 mg mAbs per 100,000 cells. KIR2DL1-Ig and KIR2DL2-Ig class I proteins was used. The assay was performed according to the fusion protein staining was performed with concentrations ranging be- manufacturer’s instructions for in situ detection of mRNA in fixed cells. tween 1 mg and 10 mg/well. Binding was detected by the appropriate Results were analyzed using Nikon 90i confocal microscope and photo- secondary Ab (Alexa Flour 647 catalogue 115-606-062 or allophycocyanin graphed with Nikon D-eclipse C1 camera. Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 3. HNRNPR KD reduces HLA-G expression and is specific. (A) FACS analysis of HLA-G expression on JEG-3 cells using the HLA-G–specific mAb MEM-G/9 transduced with the indicated short hairpin RNAs against the various RBPs (black empty histograms, indicated above the histograms), in comparison with scrambled short hairpin RNA (gray empty histograms). The filled gray histogram represents staining with the scrambled short hairpin RNA with secondary mAb only. The background of the short hairpin RNA was similar to the scrambled one and is not shown in the figure. Figure shows one representative experiment of three performed. (B) FACS analysis of HLA-G expression using the HLA-G–specific mAb MEM-G/9 on 721.221 cells transduced with HLA-G containing the 39 UTR 14-bp insertion (left) or 14-bp deletion (right). The cells are also transduced with short hairpin RNAs against HNRNPR (black empty histograms). The gray empty histograms show scrambled short hairpin RNA. The filled gray histogram represents staining of the scrambled short hairpin RNA with secondary mAb only. The background of the short hairpin RNA was similar to the scrambled one and is not shown in the figure. Figure shows one representative experiment of three performed. (C) FACS analysis of various NK ligand expression using specific mAb as indicated on Bjab cells transduced with short hairpin RNAs against HNRNPR (black empty histograms), in comparison with scrambled short hairpin RNA (gray empty histograms). The filled gray histogram represents staining of the scrambled short hairpin RNA with secondary mAb only. The background of the short hairpin RNA was similar to the scrambled one and is not shown in the figure. Figure shows one representative experiment of three performed. The Journal of Immunology 4971

Results cells using short hairpin RNA. KD efficiency was verified using Identifying RBPs that regulate MHC class I expression Western blot (WB) assays (Fig. 2A). MHC class I expression on the surface of various JEG-3 KD cells was assayed by FACS using Little is known about whether RBPs regulate class I MHC protein pan anti-MHC class I mAb W6/32. Reduced MHC class I ex- expression. Comparison of the 39 UTRs of various classical and pression was observed only in cells with HNRNPR KD, whereas nonclassical MHC class I proteins revealed that the 39 UTRs of little or no alteration of MHC class I expression was observed in the MHC class I proteins are highly similar except for the 39 UTR of HLA-F (Fig. 1A). This suggests that MHC class I pro- the other KDs (Fig. 2B). These findings led us to conclude that teins might be regulated by shared, posttranscriptional regulation HNRNPR, but not DDX3 and PABPC4, is a positive regulator of mechanisms. MHC class I expression. To further confirm this, we screened To identify RBPs that regulate MHC class I expression, we various tumor cell lines for HNRNPR expression using WB transcribed in vitro the 39 UTR of HLA-G, using biotinylated (Fig. 2C). Then, we overexpressed HNRNPR in the U87 cell line oligonucleotides. We then incubated the biotinylated 39 UTR of that is negative for HNRNPR expression (Fig. 2C) and observed HLA-G with protein lysates that were derived from the HLA- elevation of MHC class I (Fig. 2D). G–expressing cell line JEG-3 (21), followed by incubation with In parallel, we also stained the various JEG-3 KD cells for HLA- streptavidin beads. The unbound proteins were washed, and the G expression using an anti–HLA-G–specific mAb and observed bound proteins were eluted and run using SDS-PAGE. Specific reduced expression of HLA-G following HNRNPR KD (Fig. 3A). 9 protein bands, which did not appear in the control experiments, HLA-G contains in its 3 UTR 14-bp insertion polymorphism that was shown to play a role in regulation of HLA-G expression

were excised and sent to mass spectrometry analysis for identifi- Downloaded from cation (Fig. 1B). Three candidates were chosen from the mass (23, 24). To test whether a 14-bp insertion/deletion plays a role in 9 spectrometry results (Fig. 1C), based on their statistical signi- HNRNPR regulation, we expressed HLA-G with its 3 UTR that ficance score and known function: DDX3X, a RNA helicase; includes either the 14-bp insertion or deletion in 721.221 cells. We PABPC4, a poly(A)-binding protein; and HNRNPR, which asso- then knocked down HNRNPR and tested HLA-G expression. As ciates with pre-mRNAs (22). can be seen in Fig. 3B, KD of HNRNPR led to reduced HLA-G expression, irrespective of whether the 39 UTR included the 14-bp HNRNPR positively regulates MHC class I protein expression insertion/deletion. http://www.jimmunol.org/ To test whether the identified RBPs indeed regulate MHC class I To confirm that the KD HNRNPR is specific, we stained Bjab expression, the three candidate RBPs were knocked down in JEG-3 cells for an array of NK ligands before and after KD of HNRNPR. by guest on September 29, 2021

FIGURE 4. HNRNPR endogenously binds MHC class I mRNAs. (A) Correlation between the indicated HLA expression and HNRNPR expression in choriocarcinoma cell lines, based on data derived from Omnibus accession GDS2241. (B) GFP levels of Mel1074 cells transduced with a lentivirus that encodes for GFP and for HIS-tagged HNRNPR (indicated in the figure as HNRNP OE) as comparison with Mel1074 cells transduced with lentivirus encoding for the empty vector (indicated in the figure as Empty vector). (C) WBs were performed with anti-HNRNPR mAbs, and expression was compared between Mel1074 cells transduced with lentivirus encoding for the empty vector (indicated in the figure as Empty vector) and Mel1074 cells transduced with lentivirus encoding for His-tagged HNRNPR (indicated in the figure as HNRNPR-OE). hGAPDH was used as control. Contrasts in the figure were altered for clarity. Figure shows one representative experiment of three performed. (D and E) RIP was performed on whole-cell lysate of Mel1074 cells overexpressing the HIS-tagged HNRNPR using anti HIS-tag mAb. The relative levels of HLA-A (D), or HLA-G (E) mRNA were evaluated using qRT-PCR. The figure shows fold enrichment of the mRNAs of the indicated HLAs bound by HNRNPR as compared with hGAPDH (used as en- dogenous control). The abundance of each HLA control pulldown was set as 1. Shown are mean 6 SEM of triplicates. Figure shows one representative experiment of three performed. **p , 0.005 (D), *p , 0.05 (E). 4972 HNRNPR REGULATES THE EXPRESSION OF MHC CLASS I PROTEINS

We used Bjab cells and not Jeg3 cells, because they express many was detected (8- and 5.5-fold, respectively; Fig. 4D, 4E). NK cell ligands, whereas Jeg3 cells are negative for expression of These combined results indicate that HNRNPR endogenously most NK ligands. No change in NK ligand expression was observed binds the mRNA of classical and nonclassical (HLA-G) MHC (Fig. 3C). In contrast, HNRNPR KD led to a decrease in MHC class I. class I expression (Fig. 3C). HNRNPR positively regulates the expression of specific MHC HNRNPR endogenously bind MHC class I mRNA class I proteins We next analyzed a Gene Expression Omnibus dataset (GDS2241) To test whether HNRNPR affects the expression of each of the of various choriocarcinoma cell lines (25) and observed a direct classical MHC class I proteins, HLA-A, HLA-B, and HLA-C, we correlation between HNRNPR expression and the expression of knocked down HNRNPR in BJAB and Jurkat cells that express HLA-A, HLA-B, HLA-C, and HLA-G (Fig. 4A). To further in- HLA-A*02 and HLA-B*07, respectively (26, 27). Staining of the vestigate whether HNRPNR endogenously binds the 39 UTRs of appropriate cells with specific Abs against HLA-A*02 and HLA- MHC class I proteins, we performed RNA immunoprecipitation B*07 revealed that both proteins were downregulated in HNRNPR assays. Because none of the commercially available anti-HNRNPR KD cells (Fig. 5A, 5B). mAbs were able to precipitate this protein, we cloned HNRNPR To the best of our knowledge, no reliable commercial anti–HLA- fused to a HIS-tag. Cloning was done into lentivirus vector that C–specific mAb is available. Thus, to test whether HNRNPR af- also expresses GFP. We then expressed the his-tag HNRNPR in fects HLA-C expression, we used a fusion protein composed of Mel1074 cells that express little or no endogenous HNRNPR the extracellular portion of KIR2DL2 fused to human IgG1,

(Fig. 2C). Expression was verified by FACS, using GFP as in- named KIR2DL2-Ig. As mentioned in the introduction, HLA-C Downloaded from dicator for the transduction efficiency (Fig. 4B) and by WB proteins can be divided into two groups in terms of NK receptor (Fig. 4C). We next precipitated HNRNPR using an anti–HIS-tag recognition. The inhibitory receptor KIR2DL1 recognizes mem- Ab and performed qRT-PCR on the HNRNPR-bound RNA. A bers of the C2 group, such as, HLA-C*02,*04,*05, and *06, significant enrichment for both HLA-A and HLA-G mRNAs whereas KIR2DL2 recognizes members of the C1 group, such as, http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 5. HNRNPR regulates the expression of HLA-A, B, and C. (A and B) FACS analysis of the expression of HLA-A*02 using the HLA-A*02–specific mAb BB7.2 on Bjab cells (A), and of HLA-B*07 on Jurkat cells (B) using the anti–HLA-B*07–specific mAb BB7.1. Analysis was performed on cells expressing short hairpin RNA against HNRNPR (black empty histograms) and on cells expressing scrambled short hairpin RNA (gray empty histograms). The filled gray histogram represents staining of the scrambled short hairpin RNA with secondary mAb only. The background of the short hairpin RNA was similar to the scrambled one and is not shown in the figure. Figure shows one representative experiment of three performed. (C) Schematic representation of HLA-C ligands for the KIR2DL1 and KIR2DL2 inhibitory receptors. (D) HLA typing of the U937 cell line. (E) FACS analysis of the expression of MHC class I on U937 cells transduced with HNRNPR-short hairpin RNA (black empty histograms) in comparison with scrambled short hairpin RNA (gray empty histograms). The filled gray histogram represents staining of the scrambled short hairpin RNA with secondary mAb only. The background of the short hairpin RNA was similar to the scrambled one and is not shown in the figure. Figure shows one representative experiment of three performed. (F) FACS staining with the KIR2DL2-Ig of U937 cells. Cells transduced with HNRNPR-short hairpin RNA (black empty histograms) in comparison with cells transduced with scrambled short hairpin RNA(gray empty histograms). The filled gray histogram represents staining of the scrambled short hairpin RNA with secondary mAb only. The background of the short hairpin RNA was similar to the scrambled one and is not shown in the figure. Figure shows one representative experiment of three performed. The Journal of Immunology 4973

HLA-C*01,*03, *07, and*08 (Fig. 5C). To test whether HNRNPR were then used in killing assays with two targets: wild-type U937 can affect HLA-C expression, we haplotyped the U937 cell line cells and HNRNPR KD U937 cells. As can be seen in Fig. 6B, we and determined that both HLA-C proteins expressed by this cell observed an increase in killing of HNRNPR KD U937 cells. To line are ligands for NK inhibitory receptor KIR2DL2 (Fig. 5D). demonstrate that the increased killing is due to HLA-C down- When we knocked down HNRNPR in U937 cells, we observed a regulation, we blocked the KIR2DL2 receptor of NK clone D reduction in MHC class I protein expression (from 1232.51 to using a specific Ab and observed that the killing of the wild-type 804.83 median fluorescent intensity; Fig. 5E). cells was restored to the level observed in the HNRNPR KD cells. Next, we used the KIR2DL2-Ig protein (which recognizes HLA- In contrast, blocking of KIR2DL2 receptor on the HNRNPR KD C*01 and *07; Fig. 5C) (28) to stain the U937 cells. Although the cells had no effect (Fig. 6C). control U937 cells were stained with KIR2DL2-Ig, this staining HNRNPR stabilizes MHC class I mRNA was abolished in the HNRNPR KD cells (Fig. 5F), indicating that HNRNPR also affects HLA-C expression. Finally, to determine the mechanism by which HNRNPR positively regulates MHC class I expression, we performed a mRNA stability The HNRNPR-mediated HLA-C downregulation abolishes NK assay. In this procedure, cells were treated with the transcription cell inhibition inhibitor, actinomycin D, and the level of mRNA degradation was To test whether reduced HLA-C expression observed in HNRNPR determined using qRT-PCR at two time points. We found that the KD cells will affect NK cell function, we isolated NK clones mRNAs of all HLAs were less stable in HNRNPR KD cells expressing the KIR2DL2 receptor. These clones were identified by compared with the controls (Fig. 7A), indicating that HNRNPR double staining for CD56 together with anti-KIR2DL1 or anti- stabilizes the mRNAs of MHC class I molecules. HLA-A seemed Downloaded from KIR2DL2 mAbs (Fig. 6A). Three of the abovementioned clones to be more stable than the other MHC class I proteins, but it was http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 6. HNRNPR KD abolishes the KIR2DL2-mediated inhibition. (A) FACS analysis of the expression of CD56 and KIR2DL1 (left dot blots)and KIR2DL2 (right dot blots)onfourNKclones(namedA-D).(B) 35S-methionine–labeled U937 cells untransduced (WT) or transduced with HNRNPR-short hairpin RNA were incubated for 5 h with three KIR2DL2 NK clones, and killing levels were assessed. The killing levels are normalized to those of untransduced cells, which were set to 1. Figure shows one representative experiment of three performed. Shown is the relative average killing 6 SD, **p , 0.005, ***p , 0.0005. (C) Killing of untransduced (WT) or HNRNPR-short hairpin RNA–transduced labeled U937 cells by KIR2DL2-positive NK clone D, which was pre- incubated with either an isotype-matched control Ab (IgG ctrl) or an Ab to KIR2DL2 (anti-KIR2DL2, DX27). The killing levels are normalized to those of untransduced cells, which were set to 1. Shown is the relative average killing 6 SD. Figure shows one representative experiment of three performed. **p , 0.005. 4974 HNRNPR REGULATES THE EXPRESSION OF MHC CLASS I PROTEINS Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 7. HNRNPR stabilizes MHC class I mRNAs. (A) Relative mRNA abundance of the indicated HLAs or MICB in short hairpin RNA transduced in comparison with scrambled short hairpin RNA-transduced JEG-3 cells treated with 5 ml/ml ActD. mRNA levels were evaluated by qRT-PCR at time points, as indicated. The abundance of each mRNA at time point 0 was set as 1, and the abundance after treatment is shown relative to it, for each cell type. hGAPDH was used as an endogenous control. Shown are mean 6 SEM of triplicates. Data are representative of three independent experiments. *p , 0.05, **p , 0.005. (B) Relative mRNA abundance of the indicated MHC class I processing/loading machinery components in short hairpin RNA transduced in comparison with scrambled short hairpin RNA-transduced JEG-3 cells. mRNA levels were evaluated by qRT-PCR. The abundance of each mRNA at scrambled short hairpin RNA-transduced cells was set as 1, and the abundance after treatment is shown relative to it, for each cell type. hGAPDH was used as an endogenous control. Shown are mean 6 SEM of triplicates. Data are representative of three independent experiments. (C) Representative images of RNA fluorescence in situ hybridization with an antisense probe compatible to a common part of all MHC class I mRNA performed in JEG-3 cells transduced with short hairpin RNAs against HNRNPR (lower) and with scrambled short hairpin RNA (upper). Nuclear DAPI staining appears in blue, and MHC class I mRNA probe appears in red. Original magnification 31000.

still less stable under HNRNPR KD conditions. The RNA stability Finally, to determine whether some of the HNRNPR stabilization of MICB, which was set as control, was not affected by KD of mechanism involves mislocalization of the MHC class I mRNA, HNRNPR (Fig. 7A). Further conformation of the specificity of the we performed RNA fluorescence in situ hybridization assays. As HNRNPR was the observation that the RNA expression of compo- can be seen, the majority of the MHC class I mRNA is perinuclear nents of the MHC class I processing/loading pathway such as TAP1, and the KD of HNRNPR, as expected, decreased the mRNA ex- TAP2, and Tapasin was not affected by the HNRNPR KD (Fig. 7B). pression but did not alter its localization (Fig. 7C). The Journal of Immunology 4975

Discussion to NK cell–mediated elimination, so HNRNPR can theoretically The MHC class I proteins can be divided into classical and non- be manipulated in the opposite direction, to facilitate evasion from classical proteins. Although the classical HLA-A, -B, and -C proteins, NK cells. and some of the nonclassical ones, such as HLA-E, are expressed on all nucleated cells, other nonclassical MHC class I proteins, such as Disclosures HLA-G, manifest a restricted expression pattern (6). Despite years of The authors have no financial conflicts of interest. investigation, the factors controlling the ubiquitous expression of classical MHC class I proteins and those controlling the expression References of HLA-G are far from being completely understood. Discovering 1. Niederkorn, J. Y. 2009. Immune escape mechanisms of intraocular tumors. Prog. these factors is essential, as it may lead to a better understanding of Retin. Eye Res. 28: 329–347. MHC class I deficiencies (29) and MHC manipulation by cancer or 2. Hanna, J., D. Goldman-Wohl, Y. Hamani, I. Avraham, C. Greenfield, S. Natanson-Yaron, D. Prus, L. Cohen-Daniel, T. I. Arnon, I. Manaster, et al. viruses (1, 30). 2006. Decidual NK cells regulate key developmental processes at the human To identify RBP candidates that might regulate MHC class I ex- fetal-maternal interface. Nat. Med. 12: 1065–1074. pression, we used a RNA-AP method and identified HNRNPR. We 3. Seidel, E., A. Glasner, and O. Mandelboim. 2012. Virus-mediated inhibition of natural cytotoxicity receptor recognition. Cell. Mol. Life Sci. 69: 3911–3920. discovered that HNRNPR is a general positive regulator of classi- 4. Koch, J., A. Steinle, C. Watzl, and O. Mandelboim. 2013. Activating natural cal MHC class I proteins and that it controls HLA-G expression cytotoxicity receptors of natural killer cells in cancer and infection. Trends irrespective of whether the HLA-G 39 UTR contains 14-bp deletion/ Immunol. 34: 182–191. 5. Mandelboim, O., H. T. Reyburn, M. Vale´s-Go´mez, L. Pazmany, M. Colonna, insertion. We verified that HNRNPR endogenously binds to the 39 G. Borsellino, and J. L. Strominger. 1996. Protection from lysis by natural killer UTRs of MHC class I and also observed a strong positive correlation cells of group 1 and 2 specificity is mediated by residue 80 in human histo- Downloaded from compatibility leukocyte antigen C alleles and also occurs with empty major between HNRNPR expression and all of the HLA subtype expres- histocompatibility complex molecules. J. Exp. Med. 184: 913–922. sion using a Gene Expression Omnibus dataset of choriocarcinoma 6. Pratheek, B. M., T. K. Nayak, S. S. Sahoo, P. K. Mohanty, S. Chattopadhyay, cell lines, which may hint at a functional effect of HNRNPR in N. G. Chakraborty, and S. Chattopadhyay. 2014. Mammalian non-classical major histocompatibility complex I and its receptors: important contexts of gene, cancer. evolution, and immunity. Indian J. Hum. Genet. 20: 129–141. In the Gene Expression Omnibus database, HLA-C seems to be 7. Rebmann, V., F. da Silva Nardi, B. Wagner, and P. A. Horn. 2014. HLA-G as a

tolerogenic molecule in transplantation and pregnancy. J. Immunol. Res. 2014: http://www.jimmunol.org/ less affected; however, in our stabilization assays or in the KD it 297073. was very efficiently downregulated. Therefore, at the present we 8. Agrawal, S., and M. C. Kishore. 2000. MHC class I gene expression and reg- cannot determine whether some HLA alleles are more susceptible ulation. J. Hematother. Stem Cell Res. 9: 795–812. 9. Castelli, E. C., L. C. Veiga-Castelli, L. Yaghi, P. Moreau, and E. A. Donadi. to HNRNPR regulation. 2014. Transcriptional and posttranscriptional regulations of the HLA-G gene. J. The binding site of HNRNPR is unknown, but the binding site of Immunol. Res. 2014: 734068. its closest relatives, HNRNPQ and HNRNPA/B, was suggested to 10. Manaster, I., D. Goldman-Wohl, C. Greenfield, D. Nachmani, P. Tsukerman, Y. Hamani, S. Yagel, and O. Mandelboim. 2012. MiRNA-mediated control of be a certain triple RNA-recognition motif (RRMx3) (31, 32). These HLA-G expression and function. PLoS One 7: e33395. motifs are not found in the 39 UTR of the MHC class I proteins. 11. Glisovic, T., J. L. Bachorik, J. Yong, and G. Dreyfuss. 2008. RNA-binding proteins and post-transcriptional gene regulation. FEBS Lett. 582: 1977–1986. Therefore, it is still possible that HNRNPR indirectly affects MHC 12. Moore, M. J., and N. J. Proudfoot. 2009. Pre-mRNA processing reaches back to by guest on September 29, 2021 class I expression. We think, however, that this is unlikely because transcription and ahead to translation. Cell 136: 688–700. other components of the MHC class I processing pathway did not 13. Sonenberg, N., and A. G. Hinnebusch. 2009. Regulation of translation initiation in eukaryotes: mechanisms and biological targets. Cell 136: 731–745. appear in the mass spectrometry data of the RNA-AP results and 14. Martin, K. C., and A. Ephrussi. 2009. mRNA localization: gene expression in the because no change was observed in the RNA levels of some MHC spatial dimension. Cell 136: 719–730. class I processing components in the HNRNPR KD. 15. Nachmani, D., T. Gutschner, A. Reches, S. Diederichs, and O. Mandelboim. 2014. RNA-binding proteins regulate the expression of the immune activating Thus, HNRNPR is a general positive regulator of both classical ligand MICB. Nat. Commun. 5: 4186. and nonclassical MHC class I. Indeed, the 39 UTR of classical 16. Martinez-Contreras, R., P. Cloutier, L. Shkreta, J.-F. Fisette, T. Revil, and B. Chabot. 2007. hnRNP proteins and splicing control. Adv. Exp. Med. Biol. 623: MHC class I molecules and HLA-G share a striking resemblance. It 123–147. therefore seems as if an evolutionary conservation maintained a high 17. Ha¨mmerle, M., T. Gutschner, H. Uckelmann, S. Ozgur, E. Fiskin, M. Gross, degree of identity of the 39 UTR regulatory sequence of the classical B. Skawran, R. Geffers, T. Longerich, K. Breuhahn, et al. 2013. Posttranscrip- tional destabilization of the liver-specific long noncoding RNA HULC by the MHC class I proteins and HLA-G. Because HLA-E carries the same IGF2 mRNA-binding protein 1 (IGF2BP1). Hepatology 58: 1703–1712. resemblance, we predict that HNRNPR would influence its ex- 18. Stern-Ginossar, N., N. Elefant, A. Zimmermann, D. G. Wolf, N. Saleh, M. Biton, 9 E. Horwitz, Z. Prokocimer, M. Prichard, G. Hahn, et al. 2007. Host immune pression as well. In contrast, the 3 UTR of HLA-F is considerably system gene targeting by a viral miRNA. Science 317: 376–381. different that all other UTRs, and therefore it does not carry that 19. Peritz, T., F. Zeng, T. J. Kannanayakal, K. Kilk, E. Eirı´ksdo´ttir, U. Langel, and resemblance and most probably would not be affected by HNRNPR. J. Eberwine. 2006. Immunoprecipitation of mRNA-protein complexes. Nat. Protoc. 1: 577–580. It is expected that HNRNP interacts with additional mRNAs, 20. Mandelboim, O., P. Malik, D. M. Davis, C. H. Jo, J. E. Boyson, and similar to probably all other RBPs that bind several mRNA (33). J. L. Strominger. 1999. Human CD16 as a lysis receptor mediating direct natural Still, we show in this work that other NK ligands are not affected killer cell cytotoxicity. Proc. Natl. Acad. Sci. USA 96: 5640–5644. 21. Kovats, S., E. K. Main, C. Librach, M. Stubblebine, S. J. Fisher, and R. DeMars. by HNRNPR and the expression of several components of the class 1990. A class I antigen, HLA-G, expressed in human trophoblasts. Science 248: I processing pathway are also not affected. Furthermore, HNRNPR 220–223. 22. UniProt Consortium. 2015. UniProt: a hub for protein information. Nucleic Acids KD, which led to reduced expression of specific HLA-C subtypes, Res. 43: D204–D212. led to an increased killing by NK cells. 23. Ge, Y. Z., Q. Ge, M. H. Li, G. M. Shi, X. Xu, L. W. Xu, Z. Xu, T. Z. Lu, R. Wu, The endogenous mechanisms controlling HNRNPR expression L. H. Zhou, et al. 2014. Association between human leukocyte antigen-G 14-bp insertion/deletion polymorphism and cancer risk: a meta-analysis and systematic are currently unknown. In many types of transformed and virally review. Hum. Immunol. 75: 827–832. infected cells, downregulation of MHC class I expression occurs to 24. Iversen, A. C., O. T. D. Nguyen, L. F. Tømmerdal, I. P. Eide, V. M. Landsem, evade the adaptive immune system (1, 30). It will be interesting and N. Acar, R. Myhre, H. Klungland, and R. Austgulen. 2008. The HLA-G 14bp gene polymorphism and decidual HLA-G 14bp gene expression in pre-eclamptic important to investigate in the future whether virus infection leads and normal pregnancies. J. Reprod. Immunol. 78: 158–165. to reduced HNRNPR expression, and consequently to a reduction of 25. Burleigh, D. W., C. M. Kendziorski, Y. J. Choi, K. M. Grindle, R. L. Grendell, R. R. Magness, and T. G. Golos. 2007. Microarray analysis of BeWo and JEG3 MHC class I expression, and subsequent escape from CTL attack. trophoblast cell lines: identification of differentially expressed transcripts. Pla- Conversely, downregulation of MHC class I also exposes these cells centa 28: 383–389. 4976 HNRNPR REGULATES THE EXPRESSION OF MHC CLASS I PROTEINS

26. Litwin, V., J. Gumperz, P. Parham, J. H. Phillips, and L. L. Lanier. 1993. 30. Hewitt, E. W. 2003. The MHC class I antigen presentation pathway: strategies Specificity of HLA class I antigen recognition by human NK clones: evidence for viral immune evasion. Immunology 110: 163–169. for clonal heterogeneity, protection by self and non-self alleles, and influence of 31. Yang, X. C., M. P. Torres, W. F. Marzluff, and Z. Dominski. 2009. Three proteins of the target cell type. J. Exp. Med. 178: 1321–1336. the U7-specific Sm ring function as the molecular ruler to determine the site of 39- 27. Adams, S., F.-M. Robbins, D. Chen, D. Wagage, S. L. Holbeck, H. C. Morse, III, end processing in mammalian histone pre-mRNA. Mol. Cell. Biol. 29: 4045–4056. D. Stroncek, and F. M. Marincola. 2005. HLA class I and II genotype of the NCI- 32. Afroz, T., Z. Cienikova, A. Cle´ry, and F. H. T. Allain. 2015. One, two, three, 60 cell lines. J. Transl. Med. 3: 11. four! How multiple RRMs read the genome sequence. Methods Enzymol. 558: 28. Roberts, C. H., S. Molina, P. Makalo, H. Joof, E. M. Harding-Esch, S. E. Burr, 235–278. D. C. W. Mabey, R. L. Bailey, M. J. Burton, and M. J. Holland. 2014. Con- 33. Hafner, M., M. Landthaler, L. Burger, M. Khorshid, J. Hausser, P. Berninger, junctival scarring in trachoma is associated with the HLA-C ligand of KIR and is A. Rothballer, M. Ascano, Jr., A. C. Jungkamp, M. Munschauer, et al. 2010. exacerbated by heterozygosity at KIR2DL2/KIR2DL3. PLoS Negl. Trop. Dis. 8: Transcriptome-wide identification of RNA-binding protein and microRNA target e2744. sites by PAR-CLIP. Cell 141: 129–141. 29. Hanna, S., and A. Etzioni. 2014. MHC class I and II deficiencies. J. Allergy Clin. 34. Corpet, F. 1988. Multiple sequence alignment with hierarchical clustering. Immunol. 134: 269–275. Nucleic Acids Res. 16: 10881–10890. Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021