A single-cell atlas characterizes dysregulation of the bone marrow immune microenvironment associated with outcomes in multiple myeloma

A single-cell atlas characterizes dysregulation of the bone marrow immune microenvironment associated with outcomes in multiple myeloma: Cell Annotation Dictionary Author List: William C. Pilcher1#, Lijun Yao2#, Edgar Gonzalez-Kozlova3#, Yered Pita-Juarez4,5,6#, Dimitra Karagkouni4,5,6#, Chaitanya R. Acharya7#, Marina E. Michaud8, Mark Hamilton7, Shivani Nanda4,5,6, Yizhe Song2, Kazuhito Sato2, Julia T. Wang2, Sarthak Satpathy9, Yuling Ma4,5,6, Jessica Schulman7, Darwin D'Souza3, Reyka G. Jayasinghe2, Denis Ohlstrom1, Katherine E. Ferguson10, Giulia Cheloni4,5, Mojtaba Bakhtiari8, Nick Pabustan7, Kai Nie3, Jennifer A. Foltz2, Isabella Saldarriaga4, Rania Alaaeldin8, Eva Lepisto7, Rachel Chen3, Mark A. Fiala11, Beena E. Thomas8, April Cook7, Junia Vieira Dos Santos3, I-ling Chiang2, Igor Figueiredo3, Julie Fortier11, Michael Slade11, Stephen T. Oh12,13,14, Michael P. Rettig15, Emilie Anderson16, Ying Li16, Surendra Dasari16, Michael A. Strausbauch16, Vernadette A. Simon16, Immune Atlas Consortium7, Emir Radkevich3, Adeeb H. Rahman3, Zhihong Chen3, Alessandro Lagana3, John F. DiPersio2, Jacalyn Rosenblatt4,5,17, Seunghee Kim-Schulze3, Sagar Lonial8,20, Shaji Kumar16, Swati S. Bhasin8, Taxiarchis Kourelis16, Madhav V. Dhodapkar18,19 , Ravi Vij11,21, David Avigan4,5,17, Hearn J. Cho3,7, George Mulligan7$, Li Ding2,21$, Sacha Gnjatic3$, Ioannis S. Vlachos4,5,6,22,23$, Manoj Bhasin1,8,9,20$ Affiliations: 1Coultier Department of Biomedical Engineering, Georgia Institute of Technology, Atlanta, GA, USA 2Department of Medicine, Washington University in St. Louis, St. Louis, MO, USA 3Tisch Cancer Institute, Department of Immunology and Immunotherapy, Icahn School of Medicine at Mount Sinai, New York, NY, USA 4Beth Israel Deaconess Medical Center, Boston, MA, USA 5Harvard Medical School, Boston, MA, USA 6Broad Institute of MIT and Harvard, Cambridge, MA, USA 7MMRF, Norwalk, CT, USA 8Department of Pediatrics, Emory School of Medicine, Atlanta, GA, USA 9Department of Biomedical Informatics, Emory School of Medicine, Atlanta, GA, USA 10School of Biological Sciences, Georgia Institute of Technology, Atlanta, GA, USA 11Bone Marrow Transplantation & Leukemia Section, Division of Oncology, Washington University School of Medicine, St. Louis, MO, USA 12Division of Hematology, Department of Medicine, Washington University School of Medicine, St. Louis, MO, USA Last Updated: September 26th, 2025 2 13Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO, USA 14Immunomonitoring Laboratory, Center for Human Immunology and Immunotherapy Programs, Washington University School of Medicine, St. Louis, MO, USA 15Division of Oncology, Washington University School of Medicine, St. Louis, MO, USA 16Mayo Clinic, Rochester, MN, USA 17Cancer Center & Cancer Research Institute, Beth Israel Deaconess Medical Center, Boston, MA, USA 18Department of Hematology Oncology, Emory School of Medicine, Atlanta, GA, USA 19Winship Cancer Institute, Emory School of Medicine, Atlanta, GA, USA 20Aflac Cancer and Blood Disorders Center, Children's Healthcare of Atlanta, Atlanta, GA, USA 21Siteman Cancer Center, Washington University in St. Louis, St. Louis, MO, USA 22Spatial Technologies Unit, Harvard Medical School Initiative for RNA Medicine, Boston, MA, USA 23Cancer Center & Cancer Research Institute, Beth Israel Deaconess Medical Center, Boston, MA MEM, MH, SN, YZ, KS, JTW, SS contributed equally as co-second authors. Author List Footnotes: #: These authors contributed equally $: These authors jointly supervised the work $ Senior and Co-corresponding authors: Manoj Bhasin, PhD Health Sciences Research Building, Room N320 1760 Haygood drive Atlanta, GA 30322 phone: 404-712-9849 email: [email protected] Ioannis Vlachos, PhD 330 Brookline Ave, 519A, Dana Building, BIDMC, Boston, MA 02115 Phone: (617)-667-4143 Email: [email protected]d.edu Sacha Gnja4c, PhD 1470 Madison Avenue, Hess s5-105, Box 1044A, New York NY 10029 Phone: 212-824-8438 Li Ding, PhD 4444 Forest Park Avenue St. Louis, MO 63108 Phone: 314-286-1848 Last Updated: September 26th, 2025 3 E-mail: [email protected] Email: [email protected] George Mulligan, PhD Mulfple Myeloma Research Foundafon 383 Main Avenue, 5th Floor. Norwalk, CT 06851 Phone: 203-652-0458 E-mail: [email protected] Last Updated: September 26th, 2025 4 Table of Contents Cell Population Annotation Dictionary 1 NK and T cells 1 CD4 T Cells 1 CD8 T Cells 4 Natural Killer (NK) Cells 5 B cells, Erythroblasts, and Progenitors 7 B Cells 7 Erythroblasts: 8 Myeloid cells 9 Fibroblasts 11 Erythrocytes 11 Plasma Cells 12 Last Updated: September 26th, 2025 5 Note: The descriptions below correspond to the “Discovery” cohort of the Immune Atlas object. This refers to the 361 aliquot, 263 patient iteration of the object used for initial clustering and annotation. Cells from samples in the additional 122 aliquot ‘Validation’ cohort were labeled via label transfer. Cell counts and percentages below will only be representative of the original “Discovery” cohort. On the publicly available Zenodo objects, the internal cluster name will correspond to ‘minor_cell_types’ in the metadata, while the annotated name, as seen in the paper, will correspond to the ‘cellID_short’ column. On the MMRF VLAB version of the original metadata, the internal subcluster name will be listed under subcluster_V03072023, while the annotated name will be under the same cellID_short header. For visualizations of the populations listed below, please refer to Main Figure 2 and Extended Data Figure 3 in the Nature Cancer publication. Cell Population Annotation Dictionary The immune cell atlas is composed of 1,149,344 cells, partitioned into 5 major compartments: NK and T cells (k=629,877, 54.80% of all cells), B cells and erythroblasts (k=232,056, 20.19% of all cells), Erythrocytes (k=20,300, 1.77% of all cells), Myeloid cells (k=168,874, 14.69% of all cells), Fibroblasts (k=946, 0.08% of all cells), and Plasma cells (k=97,291, 8.46% of all cells). The following list will provide information for on the 106 clusters, including plasma cell and doublet populations. Percentages refer to the fraction of all cells collectively, including baseline and followup timepoints, without weighing by patient. For each cluster, the internal cluster name will be listed in the format of [Compartment].[Cluster].[Subcluster], along with the shorthand cell cell-type ID referred to in the rest of the main text. For non-doublet populations, a short description of the population is provided, along with markers. Refer to Figure 2 and Supplemental Figure 10 in the text for UMAPs and DotPlots for these populations. NK and T cells NK and T cells formed the largest compartment (k=629.877, 54.80% of all cells) and were divided into 32 cell types, including CD4+ and CD8+, NK T cells and NK cells. The T cell compartment was mainly characterized based on the expression of canonical T cell markers indicating distinct T cell states from naïve/central memory to early effector memory T cells that give rise to highly activated and cytotoxic effector memory T cells, NK T cells and pro-inflammatory phenotypes. The NK compartment, identified through lack of CD3 expression combined with canonical NK markers, was mainly characterized based on the expression of CD56 (NCAM1) canonical marker, ranging from immature CD56-high cells to mature CD56-dim cells. CD4 T cells: The CD4 cell compartment (k=306,883, 55.55% of T cells) comprised 11 cell states. Last Updated: September 26th, 2025 6 NkT.0 (CD4_Tn): A CD4 naïve T cell population (k=144,092, 26.08% of T cells) with intermediate to low expression of activation markers (CD44, CD69) and strong expression of naïve/central memory-related markers (SELL, CD7, LEF1, IL7R, TCF7, CCR7). NkT.1.0 (CD4_Tcm_KLRB1): A CD4 central memory T cell population (k=49,246, 8.95% of T cells) with intermediate expression of CD69 and CD44 activation markers. Strong expression of central memory-related markers IL7R, TCF7 and of killer cell lectin-like receptor B1 (KLRB1). NkT.1.1 (CD4_Teff): A CD4 effector population (k=21,296, 3.86% of T cells), in an intermediate state highly expressing activation markers (CD44, CD69, DUSP), as well as naïve/central memory markers including IL7R. Strong expression of KLRB1 and GZMK was also observed. NkT.1.2 (CD4_Tcm_NFKBIA): A CD4 central memory T cell population (k=15,462, 2.80% of T cells) with intermediate expression of CD69 and CD44 activation markers. Strong expression of central memory-related markers CXCR4 and FOXO1. A co-stimulatory signal was detected with the high expression of ICOS and an NF-κB signaling through the high expression of NFKBIA. NkT.1.4 (CD4_Tem_IFN): A CD4 effector memory T cell population (k=4,412, 0.80% of T cells), with high expression of memory markers including IL7R, FAS, TCF7 and intermediate levels of SELL. Intermediate CD44 expression indicating semi-activation. The cluster was highly characterized by an interferon-induced phenotype with the high expression of interferon genes, specifically ones related to ISG15 antiviral pathways, including IFI6, IFI44, IFI44L, IFIT1, IFITM1, and IFITM2. NkT.1.5 (CD4_Th): A CD4 effector helper population (k=2,596, 0.47% of T cells), highly expressing activation markers (CD44, CD69, DUSP, DUSP4), as well as classical T helper markers including GATA3 and STAT1. Intermediate to high expression of interferon genes, including IFNG and IFITM2. NkT.2.2 (CD4_CTL): A CD4 cytotoxic T lymphocyte population (k= 13,731, 2.49% of T cells) following a lineage differentiation from CD4 Th1 cells. Characterized by high expression of both CD4-related and cytotoxic markers, including granzymes (GZMA, GZMB, GZMK, GZMH) and GNLY. NkT.7 (CD4_Th_LEF1): A CD4 T cell population (k= 19,603, 3.55% of T cells) presenting an intermediate phenotype between naïve/memory and effector helper T cells. High expression of the early T cell activation marker CD44 and of naïve and central memory-related markers (LEF1, TCF7, CCR7, FOXP1, FOXO1). Upregulation of markers related to a T helper phenotype, including RORA, STAT3, STAT4, and STAT6, as well as a co-stimulatory activity with high expression of CD28. NkT.8 (Treg): A T regulatory cell population (k=18,469, 3.34% of T cells) strongly expressing canonical T reg markers, including FOXP3, IL2RA, IL2RB, and the dysfunctional markers CTLA4, TOX, and TIGIT. Last Updated: September 26th, 2025 7 NkT.10.1 (CD4_Teff_TNF): A CD4 effector T cell population (k=5,108, 0.92% of T cells), with high expression of memory markers including CD69 and CD44. The cell population exhibited an interferon-stimulated/induced pro-inflammatory phenotype by highly expressing TNF, IFNG, IFI44, IFIT1, IFIT2, IFIT3, and NFKBIA. NkT.10.0 (CD8_Teff_TNF) is a CD8+ equivalent population which displays many similar characteristics to this cluster. NkT.12 (CD4_Tcm_IFN): A CD4 central memory T cell population (k=12,688, 2.30% of T cells) with naive/central memory markers, including SELL, CCR7, FOXO1, CD7, and TCF7. Intermediate levels of activation markers (CD69, CD44). An interferon-stimulated phenotype, primarily related to ISG15 antiviral pathways, was observed with the high expression of IFI6, IFI44, IFI44L, IFIT1, and IFITM1. CD8 T cells: The CD8 cell compartment (k= 245,519, 44.45% of T cells) comprised 17 cell states. NkT.6 (CD8_Tn): A CD8 naïve T cell population (k=31,682, 5.74% of T cells) with low expression of the activation markers CD44 and CD69 markers. Strong expression of naïve/central memoryrelated markers (SELL, CD27, LEF1, IL7R, and TCF7) NkT.1.3 (CD8_Tcm): A CD8 central memory T-cell population (k=5,896, 1.07% of T cells) following a lineage differentiation from CD8 naïve cells (CD8_Tn). Low expression of naïve T cell markers (TCF7, SELL, CD27, CD28, CCR7) along with memory markers (TRADD, IL7R, TIMP1). Lack of activation (CD69, CD44), cytotoxicity or chemokine production related markers. NkT.2.0 (CD8_Teff): A CD8 cytotoxic effector T cell population (k= 50,293, 9.10% of T cells). Strong expression of cytotoxic markers including (GZMH, GNLY, PRF1, FGFBP2, NKG7) as well as IFNG. Lack of CD27 and CD28, indicating an endpoint in CD8 T cell development1. No footprints of exhaustion were observed. NkT.2.1 (CD8_Teff_HLA): A CD8 transitioning-effector T cell population (k=20,614, 3.73% of T cells). following a lineage differentiation from GZMK+ Central Memory cells (CD8_Tcm_ GZMK) to GZMB+ Effector cells (CD8_Teff). High expression of specific granzymes and activation markers including, GZMA, GZMM, GZMH, GZMK, CTSW and KLRK1. Certain cytotoxicity markers such as FGFBP2, GNLY, and PRF1, are present, but lower relative to other CD8+ Cytotoxic populations. High expression of IFNG along with MHC-I and MHC-II class markers (CD74, HLA-DRA, HLA-DRB1, HLA-DPB1), previously discussed as late activation markers in CD8 T cells2. This population does not appear to have a direct mouse equivalent3. Exhaustion markers, including LAG3 and TIGIT are also expressed. NkT.2.3 (CD8_Teff_b): A CD8 effector T cell population (k=1,245, 0.23% of T cells). Strong expression of cytotoxicity markers (GZMB, GZMH, GNLY, PRF1, FGFBP2, NKG7). Intermediate expression of erythroid marker genes in the background (HBA, HBB). Population is not patient nor site specific. Last Updated: September 26th, 2025 8 NkT.2.4 (CD8_Teff_c): A CD8 effector T cell population (k=985, 0.18% of T cells) with intermediate to high levels of cytotoxic genes (GZMB, GZMH, GNLY, PRF1, FGFBP2, NKG7). Patient specific cytotoxic cluster, abundant in plasma cell immunoglobulins (IGLC1). NkT.3.0 (CD8_Tem): A CD8 effector memory T cell population (k=36,489, 6.61% of T cells) with a pro-inflammatory phenotype, highly expressing early activation markers (CD69, CD44), the cytotoxic marker GZMK and multiple chemokine and cytokine related genes (CCL3, CCL4, CMC1, XCL1, XCL2). This cluster appears on a separate trajectory from the other cytotoxic populations, branching off from the CD8 Central Memory populations. Expresses some exhaustion markers, such as TIGIT. NkT.3.1 (CD8_Tcm_GZMK): A CD8 central Memory T cell population (k=26,878, 4.87% of T cells) with high expression of GZMK cytotoxic marker. Low expression of naïve markers (TCF7, CD27, CD28), and upregulation of memory markers (TRADD, IL7R, TIMP1). Low expression of activation (CD69 and CD44 low), and lack of expression on cytokine/chemokine, and cytotoxicityrelated markers. The cell population presented as a branch point between the CD8 naïve T cells and the cytotoxic/GZMB+, the activated GZMK+, and the MAIT lineage. NkT.3.2 (CD8_Tem_NFKB): A CD8 Effector Memory T cell population (k=10,944, 1.98% of T cells) highly expressing NFKB related genes. High expression of early activation markers (CD69, CD44) and the cytotoxic marker GZMK. High expression of markers related to NFKB signaling, including NFKB1, REL, NR4A2, and TNFAIP3. NkT.5.0 (CD8_T_adp): A CMV Adaptive, NK-Like, CD8 T Cell Population (k=17,630, 3.19% of T cells). This is a cytotoxic CD8+ population, with high expression of cytotoxicity markers (GZMB, GZMH, GNLY, PRF1, FGFBP2, NKG7). High expression of FCGR3A (CD16), low expression of NK receptor NCR3, killer-like receptor KLRC2, and the regulatory T cell marker IKZF2 (Helios). High expression of the alpha-beta TCRs, TRAC and TRBC1, and the gamma-delta TCRs, TRDC and TRGC2. Appears to be related to previously described ‘CMV-Adaptive’ CD8+ populations4. Similar profiles to the CMV Adaptive NK cells (NK_adp). NkT.5.2 (CD8_T_adp_b): A patient specific CMV Adaptive, NK-Like, CD8 T Cell Population (k=366, 0.07% of T cells). Similar expression profiles to CD8_T_adp, including high expression of cytotoxicity molecules, and expression of the inhibitory NK receptor KLRC2. High expression of plasma cell immunoglobulins IGLC2 and IGLC3. NkT.10.0 (CD8_Teff_TNF): A TNF+ CD8 T cell population, with markers related to NFKB pathways (k=10,394, 1.98% of T cells). Intermediate expression of cytotoxicity markers (GZMB, GZMH, GNLY), with high expression of TNF and IFNG. High expression of activation markers (CD69, CD44) and markers related to NFKB activation (NFKB1, REL, NR4A2, TNFAIP3). Enriched in multiple interferon-inducible markers, including IFIT2 and IFIT3. NkT.10.1 (CD4_Teff_TNF) is the CD4+ equivalent population and shows many similar characteristics to this cluster. NkT.11 (CD8_T_Apoptotic): A stressed, apoptotic T cell population, with enriched expression of Mitochondrial markers relative to the rest of the T cell compartment (k=14,477, 2.62% of T cells). NkT.13 (MAIT): Mucosal Associated Invariant T cells (k=11,610, 2.10% of T cells), highly expressing canonical markers including NCR3, SLC4A10, ZBTB16, and KLRB1. Cluster also Last Updated: September 26th, 2025 9 appears to contain a subpopulation of double-negative gamma-delta T cells, defined by expression of TRGC2 and TRDC. NkT.14 (CD8_Tem_IFN): A CD8 T effector memory cell population (k=4,729, 0.86% of T cells) highly expressing cytotoxic markers (GZMA, GZMH, GZMK, GZMM, NKG7) and interferon related genes (IFI16, IF35, IFI44, IFI44L, IFI6, IFIT1, IFIT5, IFITM1). High expression of some exhaustion markers (LAG3, TIGIT, BATF). NkT.15 (Dbl11): Doublet Population (k=806, 0.15% of T cells). NkT.16 (Dbl12): Doublet Population (k=481, 0.09% of T cells). Natural Killer (NK) Cells: The NK cell compartment (k=77,475, 6.74% of all cells) is comprised of four different populations. NK cells were originally found in the NK/T compartment, and were distinguished by negative expression of T cell lineage markers (CD3-, CD8A-, CD4-), along with positive expression of canonical NK markers, including NCAM1 (CD56), CD247, and NCR3. The cell populations were divided into CD56-bright and CD56-dim. NkT.4 (NK_CD56dim): CD56 Dim NK cell population (k=45,913, 59.26% of NK cells). This is a standard cytotoxic CD56 dim NK population. The cluster presented low NCAM1 expression and high levels of FCGR3A (CD16) and GZMB. A highly cytotoxic NK cluster with high expression of cytotoxic markers such as GNLY, PRF1, and FGFBP2. The population upregulated the activating receptors KLRB1 and KLRF1. Distinguished from other CD56 Dim clusters by the high expression of FCER1G and SH2DB15,6. NkT.5.1 (NK_adp): A CMV Adaptive, CD56-Dim NK cell population (k=13,752, 17.75% of NK cells). High expression of cytotoxicity markers such as GNLY, PRF1, NKG7, FGFBP2. This cell population highly expressed markers associated with an ‘Adaptive NK’ phenotype, downregulating the KLRB1 and KLRF1 receptors, lacking the FCER1G expression, and upregulating the killer-like receptors KLRC2 and KLRC37,8. NkT.9.0 (NK_CD56bright): CD56 Bright NK cell population (k=10,566, 13.64% of NK cells). This is an immature, CD56 bright NK population, distinguished by high expression of NCAM1, negative expression of FCGR3A (CD16), and expression of GZMK instead of GZMB. The population has lower expression of cytotoxicity molecules (PRF1, NKG7), and higher expression of chemokines (CMC1, XCL1, XCL2). Cluster is distinguished from other CD56 bright populations by the expression of TCF7 and SELL5,6. NkT.9.1 (NK_resident): CD56 Bright, Bone Marrow Resident NK cell population (k=7,244, 9.35% of NK cells). This is a variant of a CD56 bright NK population, distinguished by the expression of CD69, CD160, TIGIT, and IKZF3, with no expression of naïve markers TCF7 and SELL. This cluster is negative for FCGR3A (CD16) and GZMB and has a high expression of GZMK. Low expression of markers related to cytotoxicity (PRF1, NKG7) with a high pro-inflammatory activity by highly expressing chemokine genes such as CMC1, XCL1, XCL2, CCL3, CCL45,9. Last Updated: September 26th, 2025 16 Plasma.12 (Pc_IFI27): Plasma cell IFI27 expressing cell population (k=880, 0.90% of plasma cells). Top DE markers: ISG15, IFI27, IFI614. The presence of these markers may represent an interferon responsive population. Plasma.13 (Pc_CD74): Plasma cell CD74 expressing population (k=860, 0.88% of plasma cells). Top DE markers: CD74, HLA-DRA, RPL30. MHC-II Expression on malignant plasma cells has been observed in other studies.15 Plasma.14 (Pc_PPM1K): Patient specific plasma cell PPM1K expressing population (k=732, 0.75% of plasma cells). Top DE markers: PPM1K, RNGTT, FCRLA. Plasma.15 (Pc_BCMA): Patient specific plasma cell BCMA expressing population (k=451, 0.46% of plasma cells). Top DE markers: IGHG1, IGLC2, TNFRSF17. Plasma.16 (Pc_NOL4): Patient specific plasma cell NOL4 expressing population (k=399, 0.41% of plasma cells). Top DE markers: RPS5, NOL4, RPS3A. Plasma.17 (Pc_PUS3): Patient specific plasma cell PUS3, TMEM60 expressing population (k=244, 0.25% of plasma cells). Top DE markers: PUS3, TMEM60. Plasma.18 (Pc_IGHV3): Patient specific plasma cell IGHV3 expressing population (k=242, 0.25% of plasma cells). Top DE markers: IGHV3-11, IGLV1-44, BIRC3. Plasma.19 (Pc_IGKV4): Patient specific, low quality plasma cell IGKV4 expressing population (k=185, 0.19% of plasma cells). Top DE markers: IGKV4-1, MED8, RNF34. Plasma.20 (Pc_MPO): Patient specific, low quality plasma cell MPO expressing population (k=170, 0.17% of plasma cells). Top DE markers: MPO, LRRC75A, SPON2. The presence of MPO may represent neutrophilic contamination. Plasma.21 (Pc_IGHG3): Patient specific plasma cell IGHG3 expressing population (k=140, 0.14% of plasma cells). Top DE markers: IGHG3, RPLP1, KRTCAP2. Plasma.22 (Pc_CCDC88A): Patient specific plasma cell CCDC88A expressing population (k=104, 0.11% of plasma cells). Top DE markers: CCDC88A, JSRP1, SPAG4. Plasma.23 (Pc_IGHA1): Patient specific plasma cell IGHA1 expressing population (k=91, 0.09% of plasma cells). Top DE markers: IGHA1, IGKV1-6, PLPP5. Last Updated: September 26th, 2025 17 Cell Population Dictionary References 1. Effros, R. B. Loss of CD28 expression on T lymphocytes: A marker of replicative senescence. Dev. Comp. Immunol. 21, 471–478 (1997). 2. Saraiva, D. P. et al. Expression of HLA-DR in cytotoxic T lymphocytes: A validated predictive biomarker and a potential therapeutic strategy in breast cancer. Cancers (Basel) 13, 3841 (2021). 3. Holling, T. M., Schooten, E. & van Den Elsen, P. J. Function and regulation of MHC class II molecules in T-lymphocytes: of mice and men. Hum. Immunol. 65, 282–290 (2004). 4. Sottile, R. et al. Human cytomegalovirus expands a CD8 + T cell population with loss of BCL11B expression and gain of NK cell identity. Sci. Immunol. 6, (2021). 5. Melsen, J. E. et al. Human bone marrow-resident natural killer cells have a unique transcriptional profile and resemble resident memory CD8+ T cells. Front. Immunol. 9, (2018). 6. de Andrade, L. F. et al. Discovery of specialized NK cell populations infiltrating human melanoma metastases. JCI Insight 4, (2019). 7. Schlums, H. et al. Cytomegalovirus infection drives adaptive epigenetic diversification of NK cells with altered signaling and effector function. Immunity 42, 443–456 (2015). 8. Yang, C. et al. Heterogeneity of human bone marrow and blood natural killer cells defined by single-cell transcriptome. Nat. Commun. 10, 3931 (2019). 9. Lugthart, G. et al. Human lymphoid tissues harbor a distinct CD69+CXCR6+ NK cell population. J. Immunol. 197, 78–84 (2016). 10. Onieva, J. L. et al. High IGKC-expressing intratumoral Plasma cells predict response to Immune Checkpoint Blockade. Int. J. Mol. Sci. 23, 9124 (2022). 11. Fan, F. & Podar, K. The role of AP-1 transcription factors in plasma cell biology and multiple myeloma pathophysiology. Cancers (Basel) 13, 2326 (2021). Last Updated: September 26th, 2025 18 12. Cho, S.-F. et al. MALAT1 long non-coding RNA is overexpressed in multiple myeloma and may serve as a marker to predict disease progression. BMC Cancer 14, (2014). 13. Baez, A. et al. Myelomatous plasma cells display an intermediate gene expression pattern between a normal plasma cell and a memory B cell. Blood 122, 1892–1892 (2013). 14. Boiarsky, R. et al. Single cell characterization of myeloma and its precursor conditions reveals transcriptional signatures of early tumorigenesis. Nat. Commun. 13, (2022). 15. Burton, J. D. et al. CD74 is expressed by multiple myeloma and is a promising target for therapy. Clin. Cancer Res. 10, 6606–6611 (2004).