Loss of the Immune Checkpoint Cd85j/LILRB1 on Malignant Plasma Cells

Loss of the Immune Checkpoint CD85j/LILRB1 on Malignant Plasma Cells Contributes to Immune Escape in Multiple Myeloma This information is current as of September 28, 2021. Ester Lozano, Tania Díaz, Mari-Pau Mena, Guillermo Suñe, Xavier Calvo, Marcos Calderón, Lorena Pérez-Amill, Vanina Rodríguez, Patricia Pérez-Galán, Gaël Roué, M. Teresa Cibeira, Laura Rosiñol, Ignacio Isola, Luis-Gerardo Rodríguez-Lobato, Beatriz Martin-Antonio, Joan Bladé and Carlos Fernández de Larrea Downloaded from J Immunol published online 12 March 2018 http://www.jimmunol.org/content/early/2018/03/11/jimmun ol.1701622 http://www.jimmunol.org/

Supplementary http://www.jimmunol.org/content/suppl/2018/03/11/jimmunol.170162 Material 2.DCSupplemental

Why The JI? Submit online.

by guest on September 28, 2021 • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

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 © 2018 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published March 12, 2018, doi:10.4049/jimmunol.1701622 The Journal of Immunology

Loss of the Immune Checkpoint CD85j/LILRB1 on Malignant Plasma Cells Contributes to Immune Escape in Multiple Myeloma

Ester Lozano,*,† Tania Dıáz,*,† Mari-Pau Mena,*,† Guillermo Sun˜e,*,‡ Xavier Calvo,* Marcos Calderoń,*,† Lorena Pe´rez-Amill,*,‡ Vanina Rodrı´guez,† Patricia Pe´rez-Galań,† Gae¨l Roue´,†,x M. Teresa Cibeira,* Laura Rosinõl,* Ignacio Isola,* Luis-Gerardo Rodrı´guez- Lobato,* Beatriz Martin-Antonio,*,‡ Joan Blade´,*,‡ and Carlos Fernańdez de Larrea*,†,‡

Mechanisms of immune regulation may control proliferation of aberrant plasma cells (PCs) in patients with monoclonal gammopathy of undetermined significance (MGUS) preventing progression to active multiple myeloma (MM). We hypothesized that Downloaded from CD85j (LILRB1), an inhibitory immune checkpoint for B cell function, may play a role in MM pathogenesis. In this study, we report that patients with active MM had significantly lower levels of CD85j and its ligand S100A9. Decreased CD85j expression could also be detected in the premalignant condition MGUS, suggesting that loss of CD85j may be an early event promoting tumor immune escape. To gain insight into the molecular mechanisms underlying CD85j functions, we next enforced expression of CD85j in human myeloma cell lines by lentiviral transduction. Interestingly, gene expression profiling of CD85j-overexpressing cells revealed a set of downregulated genes with crucial functions in MM pathogenesis. Furthermore, in vitro functional assays http://www.jimmunol.org/ demonstrated that CD85j overexpression increased susceptibility to T cell– and NK-mediated killing. Consistently, ligation of CD85j decreased the number of PCs from individuals with MGUS but not from patients with MM. In conclusion, downregulation of inhibitory immune checkpoints on malignant PCs may provide a novel mechanism of immune escape associated with myeloma pathogenesis. The Journal of Immunology, 2018, 200: 000–000.

ultiple myeloma (MM) is a clonal B cell malignancy marrow (BM) cells and PCs may show aberrant expression of characterized by neoplastic proliferation of a plasma receptors such as CD56 and CD117 (c-Kit). However, the annual M cell (PC) clone. Malignant PCs produce monoclonal rate of malignant transformation from MGUS to MM is 1% (6), Igs, which usually results in organ or tissue impairment (1, 2). MM indicating that mechanisms of control may prevent proliferation of by guest on September 28, 2021 accounts for approximately 13% of hematologic cancers and its aberrant PCs. The molecular mechanisms that maintain the frequency is likely to increase in the near future as the population MGUS state and the mechanisms that trigger progression from ages (1, 3). MM remains incurable although the median survival MGUS to MM are poorly understood. has recently increased due to the introduction of autologous stem- Immune cells must be tightly regulated to mount a speciﬁc cell transplantation and the availability of new agents such as immune response while avoiding autoimmunity. One mechanism of thalidomide, lenalidomide, and bortezomib (4, 5). MM is usually immune regulation is the presence of inhibitory immune check- preceded by the asymptomatic condition monoclonal gammopathy points on the surface of immune cells. Immune checkpoints may of undetermined signiﬁcance (MGUS). In the asymptomatic contain ITIM, which can recruit phosphatases and deliver inhib- MGUS, the frequency of PCs may be up to 10% of the bone itory signals into the cell (7). Upregulation of ligands for inhibitory

*Amyloidosis and Myeloma Unit, Department of Hematology, Hospital Clıńic, Au- cytometry data analysis; V.R., P.P.-G., and G.R. performed gene expression analysis; gust Pi i Sunyer Biomedical Research Institute, University of Barcelona, 08036 B.M.-A., L.P.-A., and G.S. provided reagents and helped with functional NK cyto- Barcelona, Spain; †Division of Hematology and Oncology, August Pi i Sunyer Bio- toxicity assays; M.T.C., L.R., I.I., L.-G.R.-L., J.B., and C.F.d.L. provided patient medical Research Institute, 08036 Barcelona, Spain; ‡Josep Carreras Leukaemia samples and clinical data, designed research, and wrote the manuscript; all authors Research Institute, University of Barcelona, 08036 Barcelona, Spain; and xLabora- reviewed and approved the manuscript. tory of Experimental Hematology, Department of Hematology, Vall d’Hebron Insti- The microarray data presented in this article have been submitted to the Gene Ex- tute of Oncology, Vall d’Hebron University Hospital, 08035 Barcelona, Spain pression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/) under accession ORCIDs: 0000-0002-6307-9807 (E.L.); 0000-0001-7934-9130 (X.C.); 0000- number GSE89793. 0003-0245-2257 (G.R.); 0000-0001-5694-0921 (L.-G.R.-L.); 0000-0003-0612- Address correspondence and reprint requests to Dr. Carlos Fernańdez de Larrea, 2693 (B.M.-A.). Amyloidosis and Myeloma Unit, Department of Hematology, Hospital Clıńic de Received for publication November 27, 2017. Accepted for publication February 13, Barcelona, Institut d’Investigacions Biome`diques August Pi i Sunyer, Villarroel, 2018. 170, 08036 Barcelona, Spain. E-mail address: [email protected] This work was supported in part by Grants RD12/0036/0046 and PI16/00423 from The online version of this article contains supplemental material. Instituto de Salud Carlos III (Ministerio de Economıá y Competitividad, Cofinan- Abbreviations used in this article: BM, bone marrow; CR, complete remission; ciado por Fondo Europeo de Desarrollo Regional, Union Europea, Una Manera de DAVID, Database for Annotation, Visualization, and Integrated Discovery; MGUS, Hacer Europa). This work was partially funded by a Generalitat de Catalunya grant monoclonal gammopathy of undetermined significance; MM, multiple myeloma; PC, (2014SGR-552), the CERCA Programme/Generalitat de Catalunya, and a Josep plasma cell; PD-1, programmed cell death protein 1; VGPR, very good partial Carreras Leukaemia Research Institute grant (CEL029). C.F.d.L. was the recipient response. of an Institut d’Investigacions Biome`diques August Pi i Sunyer starting grant (II040060). Copyright Ó 2018 by The American Association of Immunologists, Inc. 0022-1767/18/$35.00 E.L. performed experiments, analyzed data, and wrote the manuscript; T.D., M.-P.M., and M.C. performed experiments and analyzed results; X.C. performed flow

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1701622 2 LOSS OF CD85j IN MULTIPLE MYELOMA checkpoints on the surface of tumor cells is one escape mecha- BM mononuclear cells from patients with MGUS or MM were obtained + nism that cancer cells may develop to inhibit the host immune by density gradient centrifugation (Ficoll; Sigma-Aldrich). Fresh CD138 system (8, 9). Blockade of inhibitory immune checkpoints has PCs were isolated with anti-CD138 mAb-coated immunomagnetic beads (Miltenyi Biotec, San Diego, CA) using an AutoMacs cell sorter (Miltenyi been approved for the treatment of solid tumors and hematologic Biotec). RNA from PCs was isolated using Trizol reagent and total RNA malignancies (10–12). However, the roles of inhibitory immune was retrotranscribed using High Capacity cDNA Reverse Transcription kit checkpoints when immune cells become tumor cells remain un- (Thermo Fisher Scientific). TaqMan Universal PCR Master Mix and known. We hypothesized that inhibitory immune checkpoints on probes were from Thermo Fisher Scientific. Reactions were run on a 7900 Real-Time PCR System (Thermo Fisher Scientific). Values are represented the surface of PCs may play a role in maintaining immune control as the difference in cycle threshold values normalized to endogenous in the premalignant condition MGUS. Thus, loss of these mech- control b-glucuronidase for each sample as per the following formula: anisms may confer a selective advantage to the aberrant clone relative RNA expression = 22d cycle threshold. promoting progression to MM. Flow cytometry analysis In this study, we focus on CD85j (also known as LILRB1 and ILT2), which is expressed predominantly on B cells, monocytes, Flow cytometry analysis was performed following procedures standardized by dendritic cells, NK, and T cells (13–15). CD85j is a receptor for the EuroFlow Consortium (23). To identify PCs with aberrant phenotypes, an eight-color panel of Abs was used: CD38-FITC (clone HB-7), CD56-PE class I MHC Ags including HLA-A, HLA-B, HLA-C, and HLA-G (clone MY31), CD19-PerCPCy5.5 (clone SJ25C1), CD81-APCH7 (clone (16). Its cytoplasmic tail contains four ITIM motifs (17) and li- JS-81), CD45-V450 (clone 2D1), CD138-BV510 (clone MI15), which were gation of CD85j inhibits IFN production by NK and T cells (13, from BD Biosciences, CD117-PC7 (clone 104D2D1) from Beckman Coul- 14). It has also been reported that binding of HLA-G to CD85j ter; and human ILT2/CD85j APC-conjugated Ab (clone 292305) was purchased from R&D Systems A minimum of 500,000 events was acquired for suppresses B cell responses (15). CD85j is also expressed on BM each sample using a BD FACSCanto II flow cytometer with FACSDiva Downloaded from PCs from healthy donors (18). MM cell line growth may be re- software (BD Biosciences). Data were analyzed with FlowJo Software. duced by HLA-G/CD85j interaction (19). Interestingly, S100A9 ELISA (also known as MRP14) has been recently identified as a novel ligand for CD85j (20) and its role in MM remains to be elucidated. Human S100A9 levels were determined in BM plasma and peripheral The aim of this study is to investigate the contribution of the plasma using ELISA kits DY5578 and DY008 according to the manufacturer’s instructions (R&D Systems). CD85j-S100A9 axis to immune regulation in MGUS and pro- http://www.jimmunol.org/ gression from MGUS to MM. To this end, we first analyzed CD85j overexpression by lentiviral transduction primary PCs from patients with MGUS, active MM, and with MM Lentiviral particles were generated using HEK293T cells and the Lenti- in complete remission (CR). This work has established that vpak packaging kit (packaging plasmids and MegaTran1.0 transfection malignant PCs downregulate the expression of both the inhibitory reagent) with Lenti open reading frame clone of human LILRB1, transcript receptor CD85j and its ligand S100A9 in patients with MM. To variant 1, pLenti-C-mGFP vector purchased from OriGene Technologies. gain insight into the mechanisms underlying the loss of immune MM cell lines were lentivirally transduced to express CD85j-GFP or GFP- control in the presence of 8 mg/ml polybrene (Sigma-Aldrich). After 3 d, regulation in MM, we enforced expression of CD85j in myeloma cells were FACS-sorted on a BDAria sorter (BD Biosciences). cells and explored changes in gene expression profiling and functional activity in immune assays. Transcriptional analysis Gene expression profiling and gene set enrichment analysis by guest on September 28, 2021 showed that CD85j overexpression was associated to downreg- RNA samples had an RNA integrity number above eight determined using a ulation of genes with essential roles in the pathogenesis of MM, Bioanalyzer 2100 instrument (Agilent Technologies). cRNA was hybridized on suggesting that loss of CD85j in MM may confer selective the HT HG-U219 GeneChip (Affymetrix) following standardized protocols. advantage for the pathogenic clone. Furthermore, we demon- Scanning was processed in a Gene Titan instrument and analyzed with Gen- eChip Command Console Software (Affymetrix). Raw data were normalized strated that specific killing by NK cells is increased against using the Robust Multichip Analysis algorithm implemented in the Expression CD85j-overexpressing myeloma cells. Accordingly, ligation of Console Software v1.1 (Affymetrix). For identification of differentially CD85j with a specific Ab decreased PC survival in MGUS but not expressed genes, MultiExperiment Viewer platform (v4.9) and a Rank Products , in MM. In summary, loss of this inhibitory axis CD85j-S100A9 test were used, applying a paired analysis with a p value 0.001. Functional annotation of enrichment analysis was performed using Database for Anno- may promote tumor cell functions and immune escape in mye- tation, Visualization and Integrated Discovery (DAVID) v6.7 (National Insti- loma. Thus, downregulation of inhibitory checkpoints on the tute of Allergy and Infectious Diseases, National Institutes of Health) (24). surface of immune-derived cancer cells may provide a novel Further pathway analysis was done using gene set enrichment analysis soft- mechanism to escape immune control in human malignancies. ware provided by the Broad Institute. The microarray data have been deposited in the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/) under accession number GEO:GSE89793. Materials and Methods Patients Cell proliferation assays BM aspiration samples and peripheral blood were collected from 69 patients MM cell lines RPMI-8226, U266, and ARP-1 were cultured in RPMI 1640 with MGUS, 75 patients with MM, and 17 patients with MM in CR di- medium supplemented with 10–15% FBS, 2 mM glutamine, 100 U/ml agnosed at the Amyloidosis and Myeloma Unit in the Department of penicillin, and 100 mg/ml streptomycin, and maintained at 37˚C and 5% Hematology (Hospital Clıńic of Barcelona). Clinical characteristics of the CO2. MM cell lines were obtained from the Division of Hematology and recruited patients are summarized in Table I. Sample collection and clin- Oncology, Institut d’Investigacions Biome`diques August Pi i Sunyer (Dr. ical record review were performed after informed written consent in ac- D. Colomer). CD85j-overexpressing cells were cultured at 50,000 cells per cordance with the Declaration of Helsinki. Study protocol was approved by well in 100 ml of RPMI 1640 medium supplemented with 10% FBS in the Institutional Review Board at Hospital Clıńic of Barcelona. Patients the presence of indicated concentrations of dexamethasone (Merck, S.L.), were diagnosed according to standard International Myeloma Working lenalidomide (Selleck Chemicals), and bortezomib (Janssen-Cilag Inter- Group criteria (21). national). After 48 h viable cells were analyzed using the MTT colori- metric assay (Sigma-Aldrich). MTT reagent was added for one additional Gene expression analysis hour before cell lysis and spectrophotometric measurement. The Gene Expression Omnibus database was used to find previously Functional cytotoxicity assays published human PC data sets such as GSE47552, including BM PCs from 5 healthy donors, 20 patients with MGUS, and 41 newly diagnosed untreated Given that cord blood–derived NK cells are being investigated as a cellular patients with MM, and analyzed by hybridizing RNA to Human Gene 1.0 therapy against MM (25), we isolated cord blood–derived NK cells and ST Array (Affymetrix) (22). cocultured them with MM target cells in 96-well U-bottom plates for 4 h. The Journal of Immunology 3

FIGURE 1. Loss of inhibitory immune checkpoint CD85j and its ligand S100A9 in active MM. Microarray data from the Gene Ex- pression Omnibus database, accession number GSE47552 was used to analyze (A) CD85j (LILRB1) mRNA expression and its ligand S100A9 mRNA expression in normal PCs (nPCs) (n = 5), PCs from individuals with MGUS (n = 20), and patients with MM (n = 41). Data are pre- Downloaded from sented as mean 6 SEM. (B) CD85j (LILRB1) and S100A9 gene expression was quantified in isolated CD138+ PCs and CD1382 cells of BM from MM patients (n = 20) and individuals with MGUS (n = 20) by http://www.jimmunol.org/ real-time PCR. (C) Confirmation study at protein level with a second cohort of MM patients (n = 21) and individuals with MGUS (n = 26). Representative dot plots of BM samples and their corresponding histograms with the mean fluorescence intensity values of CD85j expression determined by eight- color flow cytometry. Complete by guest on September 28, 2021 gating strategy is not shown. (D) Statistical analysis of CD85j expression on PCs. (E) S100A9 con- centration measured by ELISA in BM plasma and in peripheral plasma from patients with MGUS (n = 20) compared with active MM (n = 20). *p , 0.05, **p , 0.01, ***p , 0.001.

Cytotoxicity was analyzed in europium-release assays following the Ex vivo BM functional assays manufacturer’s instructions (DELFIA EuTDA Cytotoxicity kit; Perki- nElmer). Specific cytotoxicity was calculated as follows: % cytotoxicity = BM mononuclear cells were isolated by Ficoll density gradient centrifu- (experimental release – spontaneous release) 3 100/(maximal release gation and cultured in the presence of 10 mg/ml of functional human ILT2/ – spontaneous release). T cell cytotoxicity assays were performed with CD85j Ab (clone 292319) or IgG2B Isotype Control (clone 20116) from CD3+ T cells isolated from PBMC of healthy donors with CD3 R&D Systems. After 18 h, the PC number was measured in triplicate Microbeads (Miltenyi Biotec). Isolated T cells were then activated with acquiring 50 ml per run on an Attune Acoustic Focusing Cytometer anti-CD3 (1 mg/ml) and anti-CD28 (1 mg/ml) in the presence of IL2 (10 U/ml) (Thermo Fisher Scientific). for 72 h. After polyclonal T cell stimulation, we performed the cytotoxicity Statistical analysis assay at a 10:1 ratio. To assess S100A9 blockade, 20 mg/ml of neutralizing Ultra-LEAF purified anti-human S100A8/A9 heterodimer Ab (clone Kruskal–Wallis test followed by Dunn multiple comparison tests were used A15105B) from BioLegend or isotype control was added 2 h prior to coculture to analyze differences in gene expression between independent groups of with NK cells. The number of target cells per well was analyzed by flow patients. One-way ANOVA followed by Tukey multiple comparison test was cytometry with the addition of 50 ml of CountBright Absolute Counting Beads used to analyze differences in protein expression between independent (Thermo Fisher Scientific) per sample. groups of patients. Differences in CD85j expression on normal and aberrant 4 LOSS OF CD85j IN MULTIPLE MYELOMA

PCs were statistically evaluated using the Mann–Whitney U test. Two-tailed eight-color panel (23) containing CD38, CD138, CD45, CD19, CD81, paired t test was used to compare data in active MM patients before and after CD56, CD117, and CD85j. Aberrant PCs may show multiple com- , treatment. Differences were considered statistically significant at p values binations of abnormal phenotypes such as decreased CD38, CD452, 0.05. All statistical analyses were performed using GraphPad Prism, v6.02 2 + + 2 2 (GraphPad Software). CD19 ,CD56,CD117,CD27 ,andCD81 . In BM samples from patients with MGUS may coexist normal and aberrant PCs. How- Results ever, given that only a small fraction of individuals with MGUS Downregulation of the inhibitory receptor CD85j and its will develop MM, we used the terms aberrant or phenotypically ligand S100A9 in malignant PCs from patients with MM abnormal PCs but not malignant for PCs in patients with MGUS. Microarray expression analysis performed by Davies et al. (26) As shown in Fig. 2, flow cytometry analysis identified three dis- identified 74 genes differentially expressed in PCs from patients tinct patterns of expression in MGUS: 1) normal PCs expressing high with MGUS (n = 7) compared with newly diagnosed patients with only CD85j ; 2) PCs coexpressing both CD85j and CD56; and MM (n = 24). Among them, CD85j (LILRB1) was one of the most 3) PCs expressing CD56, which lost CD85j expression like most downregulated genes in PCs from patients with MM compared with of the pathologic PCs in MM (Fig. 2A). Similar results were MGUS (26). To confirm these data, we first analyzed another obtained with aberrant expression of CD117 (Fig. 2B). Given that publicly available human PC data set GSE47552, comparing PCs some aberrant PCs may express CD85j, no significant differ- from healthy donors (n = 5), individuals with MGUS (n = 20) and ences were found comparing expression of CD85j in normal PCs patients with MM (n = 41) (22). According to this data set, CD85j versus aberrant PCs (Fig. 2C, 2D). However, when the fre- expression on normal PCs was significantly higher compared with quency of aberrant PCs was superior to 40% of total PCs, we both MGUS and MM (Fig. 1A) but no significant differences were observed a significant decrease in CD85j expression on aberrant Downloaded from found in MM compared with MGUS. Interestingly, its recently PCs (Fig. 2E), suggesting that the loss of this inhibitory discovered ligand S100A9 was significantly downregulated in pa- immune checkpoint may be an early event associated with an tients with MM compared with MGUS (Fig. 1A). To better un- increased number of aberrant PCs in asymptomatic individuals derstand these discrepant results, we isolated PCs from BM with MGUS. aspirates from individuals with MGUS and patients with active + 2 CD85j expression recovered in patients with MM after MM, and both fractions CD138 and CD138 were analyzed by http://www.jimmunol.org/ effective treatment real-time PCR. Thus, we found that CD85j expression was significantly downregulated in BM PCs from patients with MM com- We next wanted to investigate the axis CD85j-S100A9 in pa- pared with MGUS (Fig. 1B, Table I), consistent with the results tients in CR defined by absence of monoclonal protein in the reported by Davies et al. (26). Accordingly, S100A9 was also sig- serum and urine by immunofixation, along with ,5% BM PCs nificantly decreased in PCs from MM compared with MGUS (27). As shown in Fig. 3, patients with MM in CR showed (Fig. 1B). Moreover, CD1382 BM cells also showed significantly significantly higher levels of CD85j expression than patients lower expression of S100A9 in patients with MM compared with with active MM. Additionally, patients who needed a second BM MGUS. aspirate allowed us to analyze CD85j expression before and after To validate our findings at a protein level, we next analyzed therapeutic intervention. CD85j expression significantly increased by guest on September 28, 2021 CD85j on PC cell surface expression by flow cytometry in a after treatment compared with the first sample in patients who second cohort of patients with MGUS and active MM. Consis- achieved CR or very good partial response (VGPR) after tently, CD85j expression was significantly decreased on malig- bortezomib-based induction therapy (Fig. 3B). Conversely, CD85j nant PCs from patients with MM compared with MGUS (Fig. 1C, 1D). expression did not recover but continued decreasing in patients that In addition, CD85j expression was also analyzed on the other subsets were in progression despite the induction treatment (Fig. 3B). Thus, of the B cell linage in these BM aspirates. Thus, pre–B cells showed CD85j expression on malignant PCs may increase in patients with low expression of CD85j whereas mature B cells showed significantly VGPR, and CD85j expression on normal PCs from patients in CR is higher levels of CD85j compared with pre–B cells in MGUS and MM, similar to the values found in individuals with MGUS. Taken to- indicating that high CD85j expression is associated to late differenti- gether, our data indicate that CD85j could be a useful marker to ation stages in the B cell linage (Supplemental Fig. 1A–C). We also assess malignant PCs versus normal PCs in patients treated with studied other BM cell subsets that may express CD85j, such as NK therapeutic Abs. cells, monocytes, and dendritic cells but no significant differences were observed in MM compared with MGUS (Supplemental Fig. Downregulation of inhibitory receptors CD85d and CD85a 1D–G). expression in MM Interestingly, levels of S100A9 were significantly reduced in Genes encoding members of the LILR family (LILRB1-5)are both BM plasma and peripheral plasma from patients with active located on chromosome 19, in the leukocyte receptor complex, MM compared with individuals with MGUS (Fig. 1E). Therefore, which comprises a large cluster of cell surface receptors such as downregulation of the inhibitory axis CD85j-S100A9 could rep- KIRs,OSCAR,LAIR,FCAR,andGPV1(28)(Fig.4A).We resent a novel mechanism to avoid intrinsic negative signals into next wanted to investigate whether these ITIM-bearing LILRB the myeloma cell to escape immune surveillance. receptors are also downregulated in MM. Indeed, LILRB2 (CD85d) and LILRB3 (CD85a) were significantly decreased in CD85j downregulation in individuals with MGUS was PCs from patients with MM (Fig. 4B). No significant differ- associated with increased frequency of aberrant PCs ences were found in LILRB4 and LILRB5 expression (data not Based on these results, we hypothesized that loss of the inhib- shown). Accordingly, flow cytometry analysis confirmed that itory receptor CD85j may be one of the molecular mechanisms expression of both inhibitory receptors CD85d and CD85a was involved in immune escape in the premalignant state MGUS, significantly lower on PCs from patients with MM (Fig. 4C, which precedes MM. To investigate whether expression of CD85j 4D), suggesting that downregulation of this inhibitory receptor was decreased in the aberrant PCs in MGUS, we analyzed family may contribute to impairment of immune regulation phenotypically normal and aberrant PCs in MGUS based on an in MM. The Journal of Immunology 5

Table I. Clinical characteristics of patients with MGUS, active MM, and patients with MM in CR

MGUS (n = 69) MM (n = 75) CR (n =17) Median age, y (range) 69 (41–88) 67 (38–88) 60 (41–83) Gender (female/male) 34/35 37/38 9/8 Immunological subtype, n (%) IgG 50 (73%) 38 (51%) 8 (47%) IgA 9 (13%) 22 (29%) 8 (47%) Only light chains 3 (4%) 7 (9%) 1 (6%) Others 7 (10%) 8 (11%) — Light chain subtype, n (%) k 34 (49%) 52 (69%) 12 (71%) l 35 (51%) 23 (31%) 5 (29%) Median serum M-spike, g/l (range) 11 (4.8–27) 31 (3.8–76.6) — International stage system, n (%) I — 16 (21%) 6 (35%) II — 27 (36%) 6 (35%) III — 32 (43%) 5 (29%) Median BM PCs % (range) 5 (2–9) 36 (4–94) 3 (1–4) —, not applicable. Downloaded from

CD85j overexpression negatively regulated genes playing that loss of CD85j found in MM may be associated with tran- essential roles in the pathogenesis of MM scriptional changes promoting MM pathogenesis. To gain insight into the molecular mechanisms associated with CD85j expression on MM cells, we enforced CD85j expression by CD85j expression rendered myeloma cells more susceptible to lentiviral transduction in MM cell lines and analyzed changes in NK and T cell–mediated antitumor activity http://www.jimmunol.org/ gene expression proﬁle compared with GFP-transduced cells. The We next analyzed the functional relevance of CD85j overexpression microarray data have been deposited in the Gene Expression in myeloma cells. As depicted in Fig. 6, CD85j-overexpressing cells Omnibus database under the accession number GEO:GSE89793. did not show reduced viability or increased sensitivity to current Similar to primary myeloma cells, MM cell lines showed very low treatments. Given that NK cells play a prominent role in con- basal levels of CD85j expression before transduction (Fig. 5A). trolling myeloma cell growth through cytotoxic activity and the Robust Multiarray Average analysis and a Rank Products test release of inﬂammatory cytokines, we wanted to investigate indicated that 121 transcripts (90 genes mapped through UniGene) whether loss of CD85j would provide a selective advantage to

were upregulated in CD85j-overexpressing MM cell lines com- escape immune attack. Our functional cytotoxicity assays with by guest on September 28, 2021 pared with control-GFP MM cell lines (p , 0.001) (Supplemental CD85j-transduced cells showed that overexpression of CD85j on Table I). DAVID’s algorithm for pathway and functional analysis MM cells significantly enhanced NK-mediated killing (Fig. 6B). revealed that the list of upregulated genes was enriched in anno- In accordance with our findings in NK-mediated cytotoxicity as- tations related to GO:0060333 ∼ IFN-g–mediated signaling says, T cell–mediated cytotoxicity assays also demonstrated that pathway (HLA-DRA, HLA-C, HLA-DPA1, NMI, CD44, MT2A), CD85j-transduced cells were more susceptible to specific killing GO:0031295 ∼ T cell costimulation (CD80, CD86, TRAC), than GFP-transduced control cells (Fig. 6C), confirming that and GO:0006955 ∼ immune response (CD70, CD74, GPR183, loss of CD85j in MM may promote resistance to cytotoxic cell– CCL3, IL7, CCL22, TNFRSF10D) (Fig. 5B, Supplemental mediated tumor killing. Table II). To provide direct evidence of the role of S100A9, we also Furthermore, 158 transcripts (116 genes mapped through performed cytotoxicity assays in the presence of blocking anti- UniGene) were downregulated in CD85j-overexpressing MM cells S100A9. We first evaluated intracellular S100A9 expression in (p , 0.001) (Supplemental Table I). Functional annotation analysis NK cells by flow cytometry. Before the cytotoxicity assay, of downregulated genes showed a significant enrichment for GO: 20–50% of NK cells were positive for S100A9 (Fig. 6D). NK 0019731 ∼ antibacterial humoral response (Bonferroni test p = cells were preincubated with neutralizing anti-human S100A8/ 6.7 3 10212) (Supplemental Table II). Of note, the list of down- A9 heterodimer Ab or isotype control. As shown in Fig. 6E, regulated genes contained 13 out of 116 genes whose products are CD85j expression increased the specific lysis compared with involved in the pathogenesis and progression of MM (UCHL1, control target cells. However, this increase was partially FGFR3, ADM, CAV1, MAF, FOXO3, PRDM1, ITGB1, DEPTOR, overcome by blocking S100A9, suggesting that S100A9 is in- SDC1, ST3GAL6, HSPB1, FAM46C) (Fig. 5C). Indeed, gene set volved in the increase in cytotoxicity against CD85j-expressing enrichment analysis showed that CD85j overexpression negatively myeloma cells (Fig. 6D, 6E). regulated genes found elevated in MM gene sets (Fig. 5D). Finally, to further investigate whether expression of CD85j For instance, caveolin-1 (CAV1) is required for vascular endo- on primary normal PCs may be associated with a better immune thelial growth factor–triggered MM cell migration (29) and was control by cytotoxic cells, we next cultured BM cells from found downregulated in CD85j-transduced MM cell lines by real- patients with MGUS and MM in the presence of the anti-CD85j time PCR (Fig. 5E). Moreover, this gene set also contained genes Ab or isotype control. After 18 h, no differences in viability frequently overexpressed in MM cells and required for their sur- were observed by Trypan blue staining in the presence of anti- vival such as PRDM1 (Blimp1) (30), DEPTOR (mTOR inhibitor) CD85j mAb (data not shown). In line with our previous results, (31), syndecan-1 (SDC1), and HspB1 (Hsp27) (Fig. 5E). Taken flow cytometry analysis showed that CD85j ligation signifi- together, CD85j overexpression negatively regulated genes cantly decreased the number of PCs from MGUS patients playing essential roles in the pathogenesis of MM, suggesting (Fig. 6F). All BM samples from MGUS (eight samples with 6 LOSS OF CD85j IN MULTIPLE MYELOMA Downloaded from FIGURE 2. Flow cytometry analysis of CD85j expression in aberrant PCs from individuals with asymptomatic MGUS. (A) Flow analysis of coexpression of CD85j and aberrant expression of CD56 on PCs from representative patients with B

MGUS and active MM. ( ) Flow data of http://www.jimmunol.org/ coexpression of CD85j and aberrant expression of CD117 on PCs in MGUS and MM. Representative dot plots of BM samples from representative patients with MGUS and active MM. (C) Representa- tive analysis of BM samples from individuals with MGUS comparing normal phenotype PCs with aberrant PCs. Com- plete gating strategy is not shown. (D)

Statistical analysis of CD85j mean ﬂuo- by guest on September 28, 2021 rescence intensity values in normal versus aberrant PCs from individuals with asymptomatic MGUS (n =27).(E) Expression of CD85j on aberrant PCs comparing samples with ,40% and $40% of abnormal PCs (n = 27). *p , 0.05.

normal PCs and two samples with aberrant PCs) were positive at the same conditions as functional assays with BM cells, and for CD85j. Additionally, to evaluate whether the Ab binding to no signiﬁcant differences were found in viability or prolifer- PCs could mediate direct effects on the PCs, we incubated ation, suggesting that this Ab did not have an agonistic func- CD85j-overexpressing cells with anti-CD85j or isotype control tion (data not shown). Taken together, our results suggest that The Journal of Immunology 7

FIGURE 3. CD85j expression increased in patients with MM in CR but not in progression. (A) CD85j Downloaded from expression on PCs from patients with active MM (n = 25), VGPR (n = 13), and CR (n = 17). (B) Longi- tudinal flow analysis of patients with MM before and after therapeutic intervention. Percentage of PCs and CD85j mean fluorescence intensity, http://www.jimmunol.org/ first sample (filled histogram), and second sample (open histogram) are depicted. Data summary of CD85j expression in patients with MM in CR (n = 4), VGPR (n = 3), and MM in progression (n = 3). *p , 0.05, **p , 0.01. by guest on September 28, 2021

the decrease in PC number could be mediated by induction of (12), such as blockade of the inhibitory programmed cell death Ab-dependent cellular cytotoxicity. On the contrary, malignant protein 1 (PD-1) in Hodgkin’s lymphoma (34). In MM, although PCs from patients with MM were resistant to treatment with PCs may express PD-1 ligand 1, BM cytotoxic T cells express low anti-CD85j Ab (Fig. 6F). Thus, ex vivo functional assays in- levels of PD-1, suggesting that PD-1 blockade may not be sufﬁ- dicated that CD85j expression in normal PCs is associated cient to activate T cells as a single agent (35–37). Accordingly, with immune control in asymptomatic MGUS. In contrast, loss several clinical trials evaluating PD-1 blockade in patients with of CD85j expression could contribute to survival and immune MM have achieved better responses in combination with escape of malignant PCs in patients with MM. approved therapies (lenalidomide/dexamethasone, pomalidomide) than as a monotherapy (38, 39). Combinatorial approaches Discussion based on blockade of negative immune checkpoints offer ex- Costimulatory and coinhibitory immune checkpoints tightly reg- traordinary opportunities to improve the treatment of hematologic ulate the immune response upon activation (32). It has recently malignancies. become clear that tumor cells can upregulate ligands for inhibitory To explore new targets for checkpoint blockade in MM, in this immune receptors to avoid immune attack (33). Therapeutic ma- study we focus on investigating the role of the ITIM-bearing receptor nipulation of inhibitory immune checkpoints has proven to be CD85j in controlling PC functions. CD85j delivers a cell-intrinsic effective against solid tumors (10) and hematologic malignancies inhibitory signal for NK cells (13), T cells (14), and B cells (15). 8 LOSS OF CD85j IN MULTIPLE MYELOMA

FIGURE 4. Downregulation of inhibitory receptors CD85d and CD85a in MM. (A)

Schematic gene cluster of LILRB family at Downloaded from chromosomal region 19q13.4. (B) Gene expression of LILRB2 (CD85d) and LILRB3 (CD85a) was quantiﬁed in isolated CD138+ PCs and CD1382 cells of BM from patients with MGUS and MM (n = 20) by real-time PCR. (C) Con- ﬁrmation study at the protein level with a second

cohort of patients with MGUS and MM. Ex- http://www.jimmunol.org/ pression of LILRB2 (CD85d) was determined by ﬂow cytometry by measuring its mean ﬂuores- cence intensity (MFI) on CD38+ cells. (D) Rep- resentative dot plots gating on CD38+ cells and showing their corresponding histograms with the MFI values of LILRB3 (CD85a) expression. **p , 0.01, ***p , 0.001. by guest on September 28, 2021

Indeed, HLA-G binds to CD85j with high affinity, promoting im- In this study, we found that loss of inhibitory CD85j may take mune tolerance (40). In patients with MGUS and MM, serum HLA- place in early stages of MM pathogenesis in patients with G concentrations were higher compared with healthy donors, but no asymptomatic MGUS associated with a higher frequency of prognostic value was observed (41). In contrast, high expression of aberrant PCs. In the early steps of MM evolution (43, 44), HLA-G on the PC surface was associated with poor prognosis in avoiding immune regulation may be one of the first challenges MM (42). Based on studies with MM cell lines, the interaction of that the pathologic plasma clone has to overcome to proliferate. CD85j and HLA-G has been recently proposed as a suitable target Considering that interaction with the BM microenvironment is in B cell malignancies (19). However, we characterized expression crucial for MM clone proliferation (45, 46), loss of inhibitory of CD85j from a large cohort of patients from asymptomatic stages immune checkpoints may provide a selective advantage for the to active MM and demonstrated that malignant PCs significantly malignant cell clone promoting immune escape in early stages downregulated not only CD85j (LILRB1) but also other inhibitory of MM. members of the LILR family such as LILRB2 and LILRB3. Our Previous functional assays showed that neutralizing CD85j on findings exclude these negative checkpoints as suitable targets and NK cells did not increase cytotoxicity against MM cell lines (47). uncover a novel mechanism of immune escape by downregulating In this study, we focus on the role of CD85j on PC function. Due inhibitory checkpoints on the surface of malignant cells. to the paucity of normal BM PCs and their low proliferation rate The Journal of Immunology 9 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 5. Gene expression profiling of CD85j-overexpressing myeloma cells. (A) Enforced expression of CD85j in three myeloma cell lines (U266, RPMI-8226, ARP-1) by lentiviral transduction and specific selection by FACS. (B) Heat map of 121 transcripts found upregulated in CD85j-overexpressing myeloma cells with p , 0.001. Annotations related to GO:0060333 ∼ IFN-g–mediated signaling pathway (green lines) using DAVID. (C) Heat map of 158 transcripts found downregulated in CD85j-overexpressing myeloma cells with p , 0.001. Significant enrichment for GO:0019731 ∼ antibacterial humoral response (blue lines) and genes involved in MM (red lines). (D) Gene set enrichment analysis of CD85j-overexpressing cells differentially expressed genes significantly correlated with gene sets upregulated in MM such as SHAFFER_IRF4_TARGETS_IN_MYELOMA_VS_MATURE_B_LYM-PHOCYTE and ZHAN_MULTIPLE_MYELOMA_MS_UP. C2 curated gene sets from online pathway databases, publications in PubMed, and knowledge of domain experts. The enrichment score (ES), nominal enrichment score (NES), nominal p value (NOM p-val), and false discovery rate (FDR) for each gene set are shown. (E) Real-time PCR validation of selected differentially expressed genes between CD85j-overexpressing cells and GFP-control cells (mean 6 SEM, n = 3). *p , 0.05. in vitro, knockdown experiments on normal PCs were not feasible. S100A9 blockade partially abrogates the increase in cytotoxicity On the contrary, we demonstrated that enforced expression of against CD85j-expressing myeloma cells, suggesting that S100A9 CD85j on MM cell lines was associated with downregulation of may contribute to the immune control of aberrant PCs from the genes involved in MM and a higher susceptibility to cell MGUS state to active MM. cytotoxic-mediated killing. Our data indicate that loss of CD85j To the best of our knowledge, this is the first report showing on malignant PCs may eliminate an inhibitory cell-intrinsic signal downregulation of inhibitory immune checkpoints on tumor cells. increasing PC resistance to NK attack and promoting immune This novel mechanism of immune escape may be particularly escape in myeloma patients. Recently, S100A9 has been identified relevant in hematological malignancies as cancer cells derive from as a novel ligand for CD85j and its interaction was implicated in immune cells tightly regulated by stimulatory and inhibitory im- the control of HIV type 1 replication by NK cells (20). Our results mune checkpoints. Therefore, although therapeutic manipulation show that S100A9 was significantly decreased at the mRNA and of inhibitory immune checkpoints on cancer cells with agonistic protein level in both BM and peripheral plasma from patients with Abs could represent an attractive approach, it is first necessary to active MM, revealing a new immune-regulatory role for S100A9 carefully analyze their expression on primary cancer cells from in myeloma. Dysregulated expression of S100 proteins is a com- patients at different stages of disease. Our data support the concept mon feature of human cancers promoting proliferation, metastasis, that PCs may negatively regulate their inhibitory receptors when and immune evasion (48). Expression of S100A9 is upregulated transforming into malignant cells, which raises the question of in breast cancer, melanoma, thymus, and prostate cancers (48). whether the loss of inhibitory checkpoints on cancer cells is a In contrast, S100A9 was one of the most downregulated genes in common mechanism contributing to tumor progression in other he- MLL-associated leukemia (49). In this study, we demonstrate that matological malignances such as chronic lymphocytic leukemia. 10 LOSS OF CD85j IN MULTIPLE MYELOMA

FIGURE 6. CD85j expression on myeloma cells increased specific killing by NK and T cells. (A) Cell viability was assessed by MTT assay. Cells were treated with dexamethasone (0.05 and 0.1 mM), lenalidomide/dexamethasone (L/D) (5 and 0.05 mM), and bortezomib (5 and 10 nM) for 48 h. Data represent mean 6 SEM from three independent experiments. (B) Cytolytic activity of Downloaded from cord-blood NK cells against CD85j- overexpressing myeloma cell lines (U266, RPMI-8226, ARP-1) target cells at indicated E:T ratios analyzed by a standard 4 h europium release assay. Data from three independent experiments with CB-NK cells prepared from different http://www.jimmunol.org/ donors were combined. (C) Cytotoxicity assays of CD3+ T cells from healthy donors against CD85j-overexpressing myeloma cell lines (U266, RPMI-8226, ARP-1) target cells at an E:T ratio of 10:1 were analyzed by flow cytometry. (D) Representative histogram showing % of intracellular S100A9 expression in NK cells (open histogram) and isotype control (filled histogram). (E) NK cells were by guest on September 28, 2021 preincubated with 20 mg/ml of blocking anti-human S100A8/A9 heterodimer (clone A15105B) or isotype control. Af- ter 2 h, target cells ARP1-GFP or CD85j- expressing ARP1 were added at a ratio of 10:1 for 4 h (n = 3). (F) Ex vivo BM cells from patients with MGUS (n = 10) and MM (n = 6) cultured in the presence of 10 mg/ml of anti-CD85j Ab (clone 292319) or isotype control. After 18 h, CD38+ PC number was quantified by flow cytometry. Ligation of CD85j selectively decreased PC number in MGUS but not in MM. Differences in percentage of change in PC cell number are depicted. *p , 0.05, ***p , 0.001. The Journal of Immunology 11

Further investigation is needed to discover whether other inhibitory 24. Huang da, W., B. T. Sherman, and R. A. Lempicki. 2009. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4: 44–57. molecules are also involved in this new mechanism of immune 25. Martin-Antonio, B., A. Najjar, S. N. Robinson, C. Chew, S. Li, E. Yvon, evasion in human malignancies. M. W. Thomas, I. Mc Niece, R. Orlowski, C. Munõz-Pinedo, et al. 2015. Transmissible cytotoxicity of multiple myeloma cells by cord blood-derived NK cells is mediated by vesicle trafficking. Cell Death Differ. 22: 96–107. Acknowledgments 26. Davies, F. E., A. M. Dring, C. Li, A. C. Rawstron, M. A. Shammas, We thank all the members of the department of hematology (Hospital Clıńic S. M. O’Connor, J. A. Fenton, T. Hideshima, D. Chauhan, I. T. Tai, et al. 2003. of Barcelona). Insights into the multistep transformation of MGUS to myeloma using microarray expression analysis. Blood 102: 4504–4511. 27. Blade´, J., D. Samson, D. Reece, J. Apperley, B. Bjo¨rkstrand, G. Gahrton, M. Gertz, Disclosures S. Giralt, S. Jagannath, and D. Vesole, Myeloma Subcommittee of the EBMT. Euro- The authors have no financial conflicts of interest. pean Group for Blood and Marrow Transplant. 1998. Criteria for evaluating disease response and progression in patients with multiple myeloma treated by high-dose therapy and haemopoietic stem cell transplantation. Br. J. Haematol. 102: 1115–1123. 28. Barrow, A. D., and J. Trowsdale. 2008. The extended human leukocyte receptor References complex: diverse ways of modulating immune responses. Immunol. Rev. 224: 1. Palumbo, A., and K. Anderson. 2011. Multiple myeloma. N. Engl. J. Med. 364: 98–123. 1046–1060. 29. Podar, K., R. Shringarpure, Y. T. Tai, M. Simoncini, M. Sattler, K. Ishitsuka, 2. Blade´,J.,C.Fernańdez de Larrea, L. Rosinõl,M.T.Cibeira,R.Jimeńez, and R. Powles. P. G. Richardson, T. Hideshima, D. Chauhan, and K. C. Anderson. 2004. Caveolin- 2011. Soft-tissue plasmacytomas in multiple myeloma: incidence, mechanisms of 1 is required for vascular endothelial growth factor-triggered multiple myeloma cell extramedullary spread, and treatment approach. J. Clin. Oncol. 29: 3805–3812. migration and is targeted by bortezomib. Cancer Res. 64: 7500–7506. 3. Palumbo, A., S. Bringhen, H. Ludwig, M. A. Dimopoulos, J. Blade´, M. V. Mateos, 30. Jourdan, M., M. Cren, N. Robert, K. Bollore´, T. Fest, C. Duperray, F. Guilloton, L. Rosinõl, M. Boccadoro, M. Cavo, H. Lokhorst, et al. 2011. Personalized therapy D. Hose, K. Tarte, and B. Klein. 2014. IL-6 supports the generation of human in multiple myeloma according to patient age and vulnerability: a report of the long-lived plasma cells in combination with either APRIL or stromal cell-soluble Downloaded from European Myeloma Network (EMN). Blood 118: 4519–4529. factors. Leukemia 28: 1647–1656. 4. Kumar, S. K., A. Dispenzieri, M. Q. Lacy, M. A. Gertz, F. K. Buadi, S. Pandey, 31. Peterson, T. R., M. Laplante, C. C. Thoreen, Y. Sancak, S. A. Kang, P. Kapoor, D. Dingli, S. R. Hayman, N. Leung, et al. 2014. Continued im- W. M. Kuehl, N. S. Gray, and D. M. Sabatini. 2009. DEPTOR is an mTOR provement in survival in multiple myeloma: changes in early mortality and inhibitor frequently overexpressed in multiple myeloma cells and required for outcomes in older patients. Leukemia 28: 1122–1128. their survival. Cell 137: 873–886. 5. Blade´, J., C. F. de Larrea, and L. Rosinõl. 2012. Incorporating monoclonal an- 32. Sharpe, A. H., and A. K. Abbas. 2006. T-cell costimulation–biology, therapeutic tibodies into the therapy of multiple myeloma. J. Clin. Oncol. 30: 1904–1906. potential, and challenges. N. Engl. J. Med. 355: 973–975. 6. Kyle,R.A.,T.M.Therneau,S.V.Rajkumar,J.R.Offord,D.R.Larson,M.F.Plevak, 33. Fox, B. A., D. J. Schendel, L. H. Butterfield, S. Aamdal, J. P. Allison, P. A. Ascierto,

and L. J. Melton, III. 2002. A long-term study of prognosis in monoclonal M. B. Atkins, J. Bartunkova, L. Bergmann, N. Berinstein, et al. 2011. Defining the http://www.jimmunol.org/ gammopathy of undetermined significance. N.Engl.J.Med.346: 564–569. critical hurdles in cancer immunotherapy. J. Transl. Med. 9: 214. 7. Dae¨ron, M., S. Jaeger, L. Du Pasquier, and E. Vivier. 2008. Immunoreceptor tyrosine- 34. Ansell, S. M., A. M. Lesokhin, I. Borrello, A. Halwani, E. C. Scott, based inhibition motifs: a quest in the past and future. Immunol. Rev. 224: 11–43. M. Gutierrez, S. J. Schuster, M. M. Millenson, D. Cattry, G. J. Freeman, et al. 8. Pizzi, M., M. Boi, F. Bertoni, and G. Inghirami. 2016. Emerging therapies 2015. PD-1 blockade with nivolumab in relapsed or refractory Hodgkin’s lym- provide new opportunities to reshape the multifaceted interactions between the phoma. N. Engl. J. Med. 372: 311–319. immune system and lymphoma cells. Leukemia 30: 1805–1815. 35. Dhodapkar, M. V., R. Sexton, R. Das, K. M. Dhodapkar, L. Zhang, R. Sundaram, 9. Lonberg, N., and A. J. Korman. 2017. Masterful antibodies: checkpoint block- S. Soni, J. J. Crowley, R. Z. Orlowski, and B. Barlogie. 2015. Prospective ade. Cancer Immunol. Res. 5: 275–281. analysis of antigen-specific immunity, stem-cell antigens, and immune check- 10. Pardoll, D. M. 2012. The blockade of immune checkpoints in cancer immuno- points in monoclonal gammopathy. Blood 126: 2475–2478. therapy. Nat. Rev. Cancer 12: 252–264. 36. Suen, H., R. Brown, S. Yang, P. J. Ho, J. Gibson, and D. Joshua. 2015. The 11. Sharma, P., and J. P. Allison. 2015. Immune checkpoint targeting in cancer therapy: failure of immune checkpoint blockade in multiple myeloma with PD-1 inhib-

toward combination strategies with curative potential. Cell 161: 205–214. itors in a phase 1 study. Leukemia 29: 1621–1622. by guest on September 28, 2021 12. Armand, P. 2015. Immune checkpoint blockade in hematologic malignancies. 37. Paiva, B., A. Azpilikueta, N. Puig, E. M. Ocio, R. Sharma, B. O. Oyajobi, Blood 125: 3393–3400. S. Labiano, L. San-Segundo, A. Rodriguez, I. Aires-Mejia, et al. 2015. PD-L1/ 13.Morel,E.,andT.Belloń. 2008. HLA class I molecules regulate IFN-gamma pro- PD-1 presence in the tumor microenvironment and activity of PD-1 blockade in duction induced in NK cells by target cells, viral products, or immature dendritic multiple myeloma. Leukemia 29: 2110–2113. cells through the inhibitory receptor ILT2/CD85j. J. Immunol. 181: 2368–2381. 38. Lesokhin, A. M., S. M. Ansell, P. Armand, E. C. Scott, A. Halwani, 14. Nikolova, M., P. Musette, M. Bagot, L. Boumsell, and A. Bensussan. 2002. M. Gutierrez, M. M. Millenson, A. D. Cohen, S. J. Schuster, D. Lebovic, et al. Engagement of ILT2/CD85j in Se´zary syndrome cells inhibits their CD3/TCR 2014. Preliminary results of a phase I study of Nivolumab (BMS-936558) in signaling. Blood 100: 1019–1025. patients with relapsed or refractory lymphoid malignancies. Blood 124: 291. 15. Naji, A., C. Menier, F. Morandi, S. Agaugue´, G. Maki, E. Ferretti, S. Bruel, V. Pistoia, 39. Jelinek, T., and R. Hajek. 2016. PD-1/PD-L1 inhibitors in multiple myeloma: the E. D. Carosella, and N. Rouas-Freiss. 2014. Binding of HLA-G to ITIM-bearing Ig- present and the future. OncoImmunology 5: e1254856. like transcript 2 receptor suppresses B cell responses. J. Immunol. 192: 1536–1546. 40. Carosella, E. D., N. Rouas-Freiss, D. Tronik-Le Roux, P. Moreau, and J. LeMaoult. 16. Kochan, G., D. Escors, K. Breckpot, and D. Guerrero-Setas. 2013. Role of non-classical 2015. HLA-G: an immune checkpoint molecule. Adv. Immunol. 127: 33–144. MHC class I molecules in cancer immunosuppression. OncoImmunology 2: e26491. 41. Leleu, X., G. Le Friec, T. Facon, L. Amiot, R. Fauchet, B. Hennache, V. Coiteux, 17. Colonna, M., F. Navarro, T. Belloń, M. Llano, P. Garcıá, J. Samaridis, I. Yakoub-Agha, S. Dubucquoi, H. Avet-Loiseau, et al; Intergroupe Francophone L. Angman, M. Cella, and M. Lo´pez-Botet. 1997. A common inhibitory receptor du Mye´lome. 2005. Total soluble HLA class I and soluble HLA-G in multiple for major histocompatibility complex class I molecules on human lymphoid and myeloma and monoclonal gammopathy of undetermined significance. Clin. myelomonocytic cells. J. Exp. Med. 186: 1809–1818. Cancer Res. 11: 7297–7303. 18.Banham,A.H.,M.Colonna,M.Cella,K.J.Micklem,K.Pulford, 42. Brown, R., K. Kabani, J. Favaloro, S. Yang, P. J. Ho, J. Gibson, P. Fromm, A. C. Willis, and D. Y. Mason. 1999. Identification of the CD85 antigen as H. Suen, N. Woodland, N. Nassif, et al. 2012. CD86+ or HLA-G+ can be ILT2, an inhibitory MHC class I receptor of the immunoglobulin superfamily. transferred via trogocytosis from myeloma cells to T cells and are associated J. Leukoc. Biol. 65: 841–845. with poor prognosis. Blood 120: 2055–2063. 19. Naji, A., C. Menier, G. Maki, E. D. Carosella, and N. Rouas-Freiss. 2012. Neoplastic 43. Dhodapkar, M. V. 2016. MGUS to myeloma: a mysterious gammopathy of B-cell growth is impaired by HLA-G/ILT2 interaction. Leukemia 26: 1889–1892. underexplored significance. Blood 128: 2599–2606. 20. Arnold, V., J. S. Cummings, U. Y. Moreno-Nieves, C. Didier, A. Gilbert, 44. Das, R., T. Strowig, R. Verma, S. Koduru, A. Hafemann, S. Hopf, M. H. Kocoglu, F. Barre´-Sinoussi, and D. Scott-Algara. 2013. S100A9 protein is a novel ligand C. Borsotti, L. Zhang, A. Branagan, et al. 2016. Microenvironment-dependent growth for the CD85j receptor and its interaction is implicated in the control of HIV-1 of preneoplastic and malignant plasma cells in humanized mice. Nat. Med. 22: 1351– replication by NK cells. Retrovirology 10: 122. 1357. 21. Dimopoulos, M., R. Kyle, J. P. Fermand, S. V. Rajkumar, J. San Miguel, 45. Bianchi, G., and N. C. Munshi. 2015. Pathogenesis beyond the cancer clone(s) in A. Chanan-Khan, H. Ludwig, D. Joshua, J. Mehta, M. Gertz, et al; International multiple myeloma. Blood 125: 3049–3058. Myeloma Workshop Consensus Panel 3. 2011. Consensus recommendations for 46. Kawano, Y., M. Moschetta, S. Manier, S. Glavey, G. T. Go¨rgu¨n, A. M. Roccaro, standard investigative workup: report of the international myeloma workshop K. C. Anderson, and I. M. Ghobrial. 2015. Targeting the bone marrow micro- consensus panel 3. Blood 117: 4701–4705. environment in multiple myeloma. Immunol. Rev. 263: 160–172. 22. Lo´pez-Corral, L., L. A. Corchete, M. E. Sarasquete, M. V. Mateos, R. Garcıá- 47. Heidenreich, S., C. Zu Eulenburg, Y. Hildebrandt, T. Stu¨big, H. Sierich, Sanz, E. Ferminãń, J. J. Lahuerta, J. Blade´, A. Oriol, A. I. Teruel, et al. 2014. A. Badbaran, T. H. Eiermann, T. M. Binder, and N. Kro¨ger. 2012. Impact of the Transcriptome analysis reveals molecular profiles associated with evolving steps NK cell receptor LIR-1 (ILT-2/CD85j/LILRB1) on cytotoxicity against multiple of monoclonal gammopathies. Haematologica 99: 1365–1372. myeloma. Clin. Dev. Immunol. 2012: 652130. 23. van Dongen, J. J., L. Lhermitte, S. Bo¨ttcher,J.Almeida,V.H.vanderVelden, 48. Bresnick, A. R., D. J. Weber, and D. B. Zimmer. 2015. S100 proteins in cancer. J. Flores-Montero, A. Rawstron, V. Asnafi, Q. Lećrevisse,P.Lucio,etal;EuroFlow Nat. Rev. Cancer 15: 96–109. Consortium (EU-FP6, LSHB-CT-2006-018708). 2012. EuroFlow antibody panels 49.Li,Z.,R.T.Luo,S.Mi,M.Sun,P.Chen,J.Bao,M.B.Neilly,N.Jayathilaka, for standardized n-dimensional flow cytometric immunophenotyping of normal, D. S. Johnson, L. Wang, et al. 2009. Consistent deregulation of gene expression between reactive and malignant leukocytes. Leukemia 26: 1908–1975. human and murine MLL rearrangement leukemias. Cancer Res. 69: 1109–1116.