Transcriptome sequencing of Hodgkin lymphoma Hodgkin and Reed-Sternberg cells reveals escape from NK cell recognition and an unfolded protein response

doi:10.21203/rs.3.rs-7957952/v1

Download PDF

Article

Transcriptome sequencing of Hodgkin lymphoma Hodgkin and Reed-Sternberg cells reveals escape from NK cell recognition and an unfolded protein response

https://doi.org/10.21203/rs.3.rs-7957952/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Classic Hodgkin lymphoma (cHL) shares mutations with primary mediastinal B cell lymphoma (PMBL) but differs in histology, clinical behavior, and phenotype. To define transcriptional programs underlying these differences, we performed flow cytometric cell sorting and low-input RNA sequencing of Hodgkin and Reed-Sternberg (HRS) cells from eighteen primary tumors, paired intra-tumoral B cells, and four cHL cell lines, and compared them with RNA-sequencing data from 40 PMBL cases. Transcriptomic profiling revealed that HRS cells undergo abortive plasma cell differentiation with robust activation of the unfolded protein response (UPR), a feature shared with multiple myeloma but absent in diffuse large B cell lymphoma and PMBL. HRS cells also demonstrated profound immune evasion, including suppression of B cell identity genes and loss of natural killer cell recognition through downregulation of SLAM family ligands such as CD48. Comparative analysis with PMBL highlighted shared oncogenic programs and key distinctions: HRS cells exhibited greater loss of B cell identity, absence of GCB- and plasma cell markers, and unique upregulation of cytoskeletal and mitotic pathways consistent with their multinucleated morphology. These findings establish HRS cells as aberrantly differentiated GCB cells with partial plasmacytic features, UPR activation and distinct immune evasion strategies.

Health sciences/Medical research

Health sciences/Pathogenesis/Oncogenesis

Hodgkin and Reed-Sternberg (HRS) cells, the malignant cells of classic Hodgkin lymphoma (cHL) are rare and embedded within a dense infiltrate of benign immune and stromal cells, with > 99% of bulk lymph node derived from non-neoplastic cells (1). This complicates isolation of tumor-specific nucleic acids, which requires enrichment of HRS cells. Limited availability of pure HRS cells has historically hindered comprehensive molecular characterization. Cultured cHL cell lines have served as surrogates, revealing distinct gene expression profiles from normal B cells and other B cell lymphomas, including recurrent CIITA gene fusions in a subset of cases (2, 3). Studies of microdissected HRS cells identified chromosomal imbalances and transcriptional differences, notably downregulation of genes regulating cytokinesis and genomic stability, which may contribute to their multinucleated, genomically unstable phenotype (4–7).

Our group previously performed whole exome and genome sequencing of purified HRS cells, identifying somatic mutations, indels and copy number alterations, and mapping the evolutionary timing of these events in pediatric and adult cHL (8, 9). These analyses showed that some alterations arise before germinal center entry, while others occur early within the germinal center, but terminal HRS cells do not complete the full germinal center program, as evidenced by the absence of SBS9 mutational signature seen in other germinal center-derived lymphomas. Characterization of HRS DNA from cell lines, primary tissue (including from subsequent studies using flow-sorted HRS cells) and circulating tumor DNA revealed a spectrum of mutations overlapping with diffuse large B cell lymphomas (DLBCL) and primary mediastinal B cell lymphomas (PMBL) (10–21).

Based on these findings, we hypothesized that the HRS transcriptome could explain at least some of the major differences between these related malignancies. To address this, we performed whole-transcriptome sequencing of HRS cells from primary cases, matched intra-tumoral B cells, and four HL cell lines, compared with 40 PMBL cases. Our analysis reveals that HRS cells exhibit abortive plasma cell differentiation with elevated unfolded protein response, alongside novel alterations in oncogenic, mitotic and immune evasion pathways, providing molecular insight into the unique biology of cHL.

Tissue specimens

Eighteen cHL cases underwent RNA sequencing, including cases previously reported for exome sequencing (8) and by Maura et al. (9) (Supplementary Table S1). Specimens originated from Weill Cornell Medical College, Mount Sinai Medical Center, Children’s Hospital of Los Angeles, Children’s Healthcare of Atlanta, Children’s Hospital of Philadelphia, Roswell Park Cancer Center, and Memorial Sloan Kettering Cancer Center. Case 1 from the prior study(8) lacked residual material, and was excluded, though numbering was preserved for consistency. Seventeen cases (2–18) had paired intra-tumoral B cell cells, and one (case 19) did not. Cohort size was limited by availability of fresh-frozen viable cell suspensions and HRS cell recovery. Specimens were mechanically dissociated and cryopreserved after lymph node biopsy. Formalin-fixed paraffin-embedded (FFPE) biopsies were used for immunohistochemical validation. For comparison, we performed RNA sequencing on 40 PMBL cases, using tissue from FFPE blocks. All PBCL cases were obtained at Weill Cornell. All cHL and PMBL were deidentified and obtained with IRB approval.

Immunohistochemistry

IHC was performed on tissue sections (cases 2–10), a cHL microarray containing 16 additional cases and a PMBL TMA containing 50 cases. Immunohistochemical staining was done using the procedures and antibodies described in the Supplementary Methods.

Cell Sorting

HRS cells and intra-tumoral B cells were isolated from cHL tissues by FACS, as described (8, 22–24). Briefly, thawed cell suspensions were washed in RPMI 1640/20% FBS containing DNase A, stained for 15 minutes on ice with an antibody panel CD64-FITC(22, Beckman Coulter (BC), Miami FL), CD30-PE (BerH83, Beckton-Dickinson (BD), San Jose, CA, RRID:AB_400238), CD5-ECD (BL1a, BC, RRID:AB_3678603), CD40-PE-Cy5.5 (custom conjugate, gift of Jonathan Fromm) or CD40-PerCP-eFluor 710 (1C10, Ebiosciences, San Diego, CA), CD20-PC7 (B9E9, BC), CD15-APC (HI98, BD, RRID:AB_10893192), CD45 APC-H7 (2D1,BD, RRID:AB_1645480) or CD45-Krome Orange (J.33, BC, RRID:AB_2888654), and CD95-Pacific Blue (DX2, Life Technologies, Grand Island, NY), resuspended in FACS buffer. Sorting was performed on a FACSAria (130µm nozzle, 12 psi) to separate HRS, B, and T cells. Sorted cells were collected in N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid buffer solution containing 50% FBS.

Flow cytometry

NK cells were quantified in 43 cHL tumors and 10 reactive nodes using a panel including CD45, CD7, CD3 and CD56 (8). NK cells were defined as CD7⁺CD56⁺CD3^- mononuclear cells. Differences were assessed with two-tailed Student t-test (Excel, Microsoft, Bellevue, WA, RRID:SCR_016137). CD48 expression (clone TU145, BD, RRID:AB_396099) was analyzed on primary HRS cells was examined in 5 cases using a panel containing CD30, CD15, CD20, CD40, CD3, CD95, CD64, and CD45, as previously described (8).

RNA Extraction, Library Construction and Next-Generation Sequencing

cHL: Flow-sorted cHL cells were pelleted, washed and RNA extracted using Arcturus PicoPure RNA Isolation kit (KIT0204, Life Technologies, Carlsbad, CA) with RNAse-Free DNase (79254, Qiagen, Venlo, Netherlands) treatment. RNA quality and concentration were assessed with Bioanalyzer RNA Pico (5067 − 1513, Agilent, Santa Clara, CA). Libraries were prepared from 1-4ng total RNA using the SMARTer Ultra Low Input RNA Kit (Clontech, Mountain View, CA), followed by Illumina-compatible library construction with the KAPA LTP kit (KK8221, Woburn, MA). Libraries were sequenced on Illumina HiSeq in paired-end 101bp mode. Data are deposited in NCBI GEO (accession: GSE301492).

PMBL: Total RNA was extracted from FFPE tissue sections using an in-house bead-based protocol. A 10% MCT solution (Nature’s Way coconut oil, NaOH) was prepared to generate a proteinase K mix for deparaffinization. Samples were incubated at 56C for 1 hour with intermittent shaking (30s at 1000rpm every 5 minutes). After incubation, the supernatant (RNA-containing fraction) and treated with DNase. RNA was subsequently purified using 0.8x SPRI beads (Agencourt AMPure XP Beads (Beckman Coulter, A63882) and eluted in nuclease-free water. Pooled RNA libraries were prepared using the KAPA Stranded RNA-Seq Kit library preparation kit (Roche 07962169001) and xGen Hybridization and Capture kit (IDT 1080577) in accordance with the manufacturer’s instructions.

Differential Gene expression analysis

Raw data were mapped to the human genome GRCh38 using HISAT2. FeatureCounts from the Rsubread package (version1.24.7) was used for read counting after which genes without a counts per million reads (CPM) in at least 3 samples were excluded from downstream analysis (25). Count data were normalized using the trimmed mean of M-values (TMM) method and differential gene expression analysis was performed using the limma-voom pipeline (limma version 3.40.6) (26–28). GSEA-4.3.2 was used for Gene set enrichment analysis (GSEA) (29, 30). pheatmap and ggplot2 (version 3.2.1, RRID:SCR_014601) were used to plot the heatmap and cluster-specific trends (31, 32). Comparative analyses of PMBL, ABC-DLBCL and GCB-DLBCL expression profiles used the dataset of Rosenwald et al. (12).

Data Input and Preprocessing and Methodology for Virus Identification

The computational methods used for this analysis are presented in the Supplementary Methods section.

Distinct gene expression profiles of HRS cells compared to intra-tumoral B cells

To examine the transcriptome of HRS cells, we conducted bulk-RNA sequencing on primary HRS cells from 18 cHL cases, including paired intra-tumoral non-neoplastic B cells from 17 cases and four cHL cell lines (KMH2, HDLM2, L1236, and L428). Cases were numbered 2–19, where 2 through 10 match those reported for exome sequencing (8) and 11–16 and 19 those reported for exome or full genome sequencing (9) (Supplementary Table S1). Principal component analysis (PCA) and unsupervised clustering analysis revealed two distinct clusters, separating intra-tumoral B cells from both primary HRS cells and cHL cell lines (Fig. 1A-B).

Comparative analysis between HRS cells and intra-tumoral B cells identified 2144 upregulated genes (logFC ≥ 2; adjp ≤ 0.01) and 2451 downregulated genes (logFC≥-2; adjp ≤ 0.01) (Fig. 1C, Supplementary Datatable S1 and S2). Based on these results, we defined the top 200 upregulated and top 200 downregulated genes (Supplementary Datatable S3). Similarly, we compared differentially expressed genes (DEGs) between HL cell lines and intra-tumoral B cells, as well as between primary HRS cells with HL cell lines (Figure S1A-B). Although cHL cell lines and primary HRS cells clustered together in the PCA plot (Fig. 1A), a substantial number of DEGs were observed between the two groups (Figure S1D), indicating that in vitro cell lines only partially recapitulate the transcriptional landscape of primary HRS cells, potentially due to adaptation to culture conditions and loss of microenvironmental cues.

Our results were compared with published Affymetrix array profiles of microdissected HRS cells (29 cases, 5 cell lines)(6). Despite platform differences, gene-level fold changes were strongly concordant (Pearsons's coefficient 0.86; Spearman’s correlation 0.73), with only 20 genes showing discordant direction (Figure S1C-D, Supplementary Table S2). Compared to arrays, RNA-sequencing in this study provided greater sensitivity and dynamic range.

HRS cells exhibit an unfolded protein response

We first examined genes significantly upregulated in HRS cells compared to intra-tumoral B cells (logFC ≥ 2; adj-pvalue ≤ 0.01). Gene ontology analysis revealed enrichment of mitotic cell cycle, tube morphogenesis and cell development pathways (Fig. 2A), consistent with previous reports linking abortive mitosis and incomplete cytokinesis to the multinucleated phenotype of HRS cells (33, 34). GSEA further demonstrated enrichment of hallmark pathways dysregulated in cHL, including G2M checkpoint, IL2-STAT5 signaling, MYC targets, TNFa signaling via NFKB and inflammatory response (Fig. 2B-C). Transcription factor target analysis highlighted over-representation of cHL-associated regulators, particularly E2F and NFKB family members (Fig. 2D) (35, 36). Together, these results validate our analytical approach and confirm that the transcriptional profile of HRS cells is consistent with established features of cHL.

Beyond these expected findings, our analysis uncovered novel pathway enrichment in HRS cells. Specifically, GSEA revealed significant upregulation of the unfolded protein response (UPR), a pathway not previously associated with cHL (Fig. 2B, 2E-F). To assess whether UPR activation is unique to HRS cells, we evaluated UPR signature expression in DLBCL using established gene sets (11, 12, 17, 18). Multiple myeloma (MM), a plasma cell malignancy characterized by strong UPR activity, was included as positive control. As expected, MM samples exhibited elevated UPR signature scores and increased expression of UPR-related genes (Supplementary Figure S2A-C). In contrast, neither subtype of DLBCL, including activated B-cell (ABC) or germinal center B-cell (GCB), exhibited consistent evidence of UPR activation (Supplementary Figure S2B-D).

Among the top upregulated UPR genes, we identified PDIA6, an endoplasmic reticulum-localized disulfide isomerase essential for protein folding and prevention of aggregation through catalysis of disulfide bond formation and breakage (Fig. 2F-G). Immunohistochemistry (IHC) for PDIA6 in cHL cases 2–10, as well as a tissue microarray of 16 additional cases, demonstrated strong and specific staining in HRS cells in all cases with minimal background in surrounding lymphoid cells (Fig. 2H). Staining intensity was comparable to, or greater than, that observed in residual plasma cells, which are characterized by high UPR activity. Apart from plasma cells, which are easily recognized morphologically, PDIA6 expression was highly specific for HRS cells, indicating its potential as a sensitive and specific diagnostic marker for cHL.

Downregulation of NK cell recognition pathways in HRS cells

We next examined genes downregulated in HRS cells relative to intra-tumoral B cells (logFC≤-2; adj-pval ≤ 0.01). As expected, GSEA and GO analysis revealed suppression of immune regulatory pathways, including B cell receptor signaling, antigen processing and presentation, and B cell activation, processes known to be impaired in HRS cells (Fig. 3A, Figure S3A-C). Transcription factor target analysis further supported these findings, revealing reduced activity of key B cell regulators such as IRF8, PAX5 and POU2F2, consistent with the dedifferentiated phenotype of HRS cells (Figure S3D).

Beyond these established findings, GSEA identified novel downregulated pathways, particularly those related to leukocyte activation and degranulation (Fig. 3A). Cell type-specific signature analysis highlighted marked suppression of natural killer (NK) cell-mediated cytotoxicity, with consistent downregulation of genes involved in cytotoxicity, leukocyte-mediated killing and immune cell effector function (Fig. 3B-D).

Given the central role of the signaling lymphocytic activation molecule family (SLAMF) receptors in NK cell function, we examined their expression in HRS cells. Six of nine activating SLAM family receptors - SLAMF2 (CD48), SLAMF3 (LY9), SLAMF4 (CD244), SLAMF5 (CD84), SLAMF6 and SLAMF7 – were significantly downregulated on HRS cells (Fig. 3E-F), suggesting that reduced expression of these ligands may contribute to NK cell evasion. Flow cytometry confirmed loss of CD48 (SLAMF2) in both cell lines and five primary cases (Fig. 3G). In parallel, NK cell frequencies were significantly reduced in cHL tumors compared to reactive lymph nodes (median 0.6% vs. 1.4%, p < 0.01) (Fig. 3H). Finally, IHC validated consistent loss of CD48 across all sequenced cases (cases 2–10) and 22 additional specimens (Fig. 3I).

Comparative transcriptomic analysis of cHL reveals key similarities to CD30⁺ B cells, and both normal and malignant plasma cells.

We found that HRS cells exhibit upregulation of the UPR signaling, a hallmark of plasma cells, raising the question of which B cell subset they most closely resemble transcriptomically. Previous studies suggested that HRS cells may originate from CD30⁺ cells within the germinal center (7). To test this, we compared the published CD30⁺ B cell gene signature (7) to our HRS samples. Indeed, genes upregulated in CD30⁺ cells were significantly enriched (NES:2.21; p-val:0.000) (Figure S4A). To further examine relationships with other non-malignant B cell subsets, we analyzed the top 200 genes upregulated in HRS cells relative to intra-tumoral B cells across diverse B cell populations. Consistent with UPR signaling, the HRS signature was significantly enriched in bone marrow plasma cells (BMPCs), but not in tonsillar plasma cells, despite their shared plasma cell identity (Fig. 4A). Cibersortx deconvolution further demonstrated that HRS cells exhibit a transcriptomic profile more like BMPCs and GCBs than other B cell subsets (Fig. 4B-C, S4B).

Given the overlap between HRS cells and plasma cell programs, we next examined HRS signature expression in MM. Genes upregulated in HRS relative to intra-tumoral B cells were significantly enriched in MM compared with GCBs (NES: 1.64; p = 0.002), while genes downregulated in HRS were negatively enriched in MM (NES: -2.45, p = 0.000) (Fig. 4D).

Virus discovery

Having defined the transcriptional landscape of HRS cells and their relationship to plasma cells, we next asked whether known and unknown infectious agents contribute to cHL pathogenesis. To address this, we screened HRS cells (cases 2–9) and four cell lines using the Pandora(37), and Virdetect pipeline (38). Epstein-Barr virus (EBV) was detected only in case 8, yielding 19 contigs (> 500bp, max 1829bp) with expression of LMP1 and LMP2, consistent with latency II pattern (Figure S5). The only other virus identified was bovine viral diarrheal virus (BVDV), detected in HRS cells of (cases 2–4) and B cells (cases 3 and 7), likely reflecting fetal calf serum contamination (Supplementary Datatable S5). Low-stringency alignment of non-human contigs revealed no credible novel viral sequences, suggesting undiscovered viruses in cHL are either rare or highly divergent.

Comparative transcriptomic analysis of cHL reveals key differences between cHL and PMBL

Having established the transcriptional landscape of HRS cells and confirmed a limited contribution from viral infection, we next compared HRS gene expression profiles with PMBL to define shared and distinct molecular features between these related B cell malignancies. cHL and PMBL share clinical and pathological features, including GCB-cell origin and overlapping mutational landscapes (8, 9, 16, 18, 39, 40). However, they exhibit distinct clinical behavior and histology, making a comparative analysis critical to understanding the transcriptional programs underlying both their commonalities and differences.

To delineate the transcriptional programs distinguishing these entities, we compared the transcriptomes of HRS cells, intra-tumoral B cells and 40 PMBL samples. Unsupervised clustering positioned PMBL samples between intra-tumoral B cells and HRS cells, (Fig. 5A). As the sequencing of HRS and PMBL cases was performed independently, we compared expression using log fold-change (logFC) of HRS vs intra-tumoral B cells relative to mean PMBL expression. This identified four major categories: 1523 genes upregulated in HRS and highly expressed in PMBL samples; 1585 genes downregulated in HRS and lowly expressed in PMBL; 2131 genes upregulated in HRS but lowly expressed in PMBL; and 2193 genes downregulated in HRS but highly expressed in PMBL (Fig. 5B). Among the top 20 genes downregulated in HRS but highly expressed in PMBL were LMO2, a GCB-associated gene, and TNFRFS17 (encoding BCMA), a plasma cell marker. IHC confirmed robust expression of BCMA and LMO2 in PMBL cases (BCMA 37/45, LMO2 41/41), but absence in HRS cells from cHL biopsies (BCMA 0/17; LMO2 0/15) (Fig. 5C).

GO analysis of genes downregulated in HRS but highly expressed in PMBL highlighted BCR signaling, immune regulation, leukocyte activation and cytokine signaling pathways (Fig. 5D). ssGSEA analysis corroborated immune activation in both malignancies, but at significantly lower levels in HRS. Consistent with our earlier finding that HRS cells suppress NK-mediated cytotoxicity via downregulation of SLAMF receptors, we observed broad downregulation of NK cytotoxicity genes in HRS compared with PMBL (Fig. 5F). Except for SLAMF4 (CD244), which is expressed at low levels in PMBL, all other SLAMF receptors (CD48, CD84, LY9, SLAMF6 and SLAMF7) are highly expressed in PMBL (Fig. 5G). IHC for CD48 validated these results: in PMBL, 36 cases showed strong positivity (in > 50% of the tumor cells), 6 cases showed partial positivity (5–50% of tumor cells) and 2 cases were negative. In contrast, in 22 cHL cases were negative for CD48 expression in HRS cells (Fig. 5H).

Conversely, genes upregulated in HRS but lowly expressed in PMBL were enriched for developmental and microtubule cytoskeletal pathways (Fig. 5I). ssGSEA analysis confirmed a significant enrichment of microtubule cytoskeleton organization in HRS relative to PMBL (Fig. 5J), consistent with their aberrant mitotic progression and characteristic multinucleated morphology, which is largely absent in PMBL.

We present the first RNA sequencing dataset of HRS cells isolated from primary cases of cHL, providing an integrated view of their transcriptional programs relative to intra-tumoral B cells, other B cell malignancies and plasma cell populations. Our transcriptomic analysis reveals that HRS shares greater similarity with plasma cells than with mature B cells and exhibits features reminiscent of plasma cell malignancies such as MM. While prior studies have established a GCB cell origin for cHL, supported by the presence of ongoing somatic hypermutation (41–44), our findings provide molecular evidence for partial plasmacytic differentiation, extending prior immunohistochemical observations of abortive plasma cell features in HRS cells (45, 46).

HRS cells show widespread downregulation of B cell identity genes, including BCR components and transcription factors such as PAX5 and IRF8. At the same time, they display features of incomplete plasma cell differentiation, including expression of IRF4/MUM1, but relatively low levels of PRDM1/BLIMP1 and absence of CD138 in most cases, although exceptions have been reported (47, 48). This imbalance likely explains the paradox of robust UPR activation despite the absence of immunoglobulin production. Comparison with MM further underscores the aberrant plasmacytic features of HRS cells. Like MM, HRS cells show strong UPR activation, including overexpression of XBP1, ATF6 and other UPR components (49). However, unlike MM, HRS cells fail to undergo full plasma cell maturation, lacking consistent CD138 expression and immunoglobulin secretion. Thus, cHL appears locked in a non-productive, partially plasmacytic state. While plasma cell differentiation in MM is tightly coupled to BLIMP1 activity, in cHL UPR activation appears to arise through alternative mechanisms, possibly linked to FOXO1 downregulation (23, 50). Among UPR genes, PDIA6 was consistently overexpressed and highly specific to HRS cells and plasma cells within cHL biopsies, suggesting a potential role as a diagnostic marker.

These findings support a model in which HRS cells aberrantly exit the germinal center through an abortive plasma cell differentiation program further compromised by failure to express immunoglobulin. Consistent with this, HRS cells show greater transcriptional similarity to BMPCs than to tonsil plasma cells, aligning with the notion that BMPCs arise from cells that have transited through the germinal center (51). Interestingly, the dissociation we previously reported between the canonical AID signature and SBS9 signature in cHL, where AID is present but SBS9 is absent, was again observed here (9). Although SBS9 was initially attributed to non-canonical AID activity, recent studies link it to replicative stress in GCB cells (52). Such dissociation, though rarely seen in B cell lymphomas, occurs in a subset of MM cases with MAF rearrangements (53), suggesting that cHL and certain MM subsets may share aberrant GC exit mechanisms.

Our comparative transcriptomic analysis highlights both shared and divergent features between cHL and PMBL. Both malignancies activate proliferative and oncogenic pathways, but HRS cells are distinguished by a more profound loss of B cell identity, absence of GCB-associated genes such as LMO2 and enrichment for cytoskeletal and mitotic programs consistent with their multinucleated morphology. Unlike PMBL, HRS cells do not express LMO2 nor BCMA, reinforcing the notion that the two tumors diverge at the level of differentiation despite overlapping mutational landscapes. Furthermore, HRS cells display unique enrichment of neuronal and cytoskeletal gene programs, reflecting possible lineage infidelity or aberrant differentiation processes not observed in other B cell malignancies.

Our dataset also refines the immune evasion strategies of HRS cells. In addition to impaired antigen presentation through B2M mutations (8), overexpression of checkpoint ligands (PDL1, PDL2 and TIM3), and secretion of immunosuppressive cytokines, we identify a novel mechanism - loss of NK cell-mediated cytotoxicity. Specifically, downregulation of multiple SLAM family ligands, including CD48, was observed in both primary HRS cells and cell lines and validated at the protein level. This suppression coincided with significantly reduced NK cell infiltration in cHL tumors compared with reactive lymph nodes, suggesting spatial exclusion of NK cells as an additional immune evasion strategy. Taken together with frequent loss of MHC class I, these findings indicate that HRS cells evade both T cell and NK cell mediated surveillance, offering a more comprehensive view of their immune escape repertoire. This finding has therapeutic implications, as CAR-NK strategies are now entering clinical development.

Finally, we investigated the possibility of viral involvement in cHL pathogenesis beyond EBV. Our unbiased viral search revealed no novel viral sequences, and EBV transcripts were detected only in known EBV-positive cases (54–56). This is consistent with the prevailing model in which EBV contributes to cHL pathogenesis through NFKB activation in a subset of cases, while EBV-negative disease relies on somatic alterations, such as mutations in TNFAIP3 (A20), to activate similar pathways (8, 9, 57). Thus, EBV positive tumors rely on viral oncogenic programs, whereas EBV-negative tumors acquire genetic lesions to converge on similar signaling outcomes.

In summary, our findings refine the molecular identity of HRS cells as aberrantly differentiated GCB-derived cells with plasmacytic features, robust UPR activation, and unique immune evasion strategies. The distinction from PMBL underscores the divergent transcriptional trajectories of related B cell lymphomas, while transcriptomic similarities to MM highlight a shared yet incomplete plasmacytic program. The identification of UPR activation and NK cell evasion as defining features of HRS cells points to potential diagnostic markers and novel therapeutic vulnerabilities in cHL.

ACKNOWLEDGEMENTS

This project was supported by the Department of Pathology and Laboratory Medicine of Weill Cornell Medicine and its Center for Translational Pathology. It was funded in part by NIH grant R01-CA068939 to EC. JR was partially funded by the Tri-I Training Program in Computational Biology and Medicine (5T32GM083937). MR and JR were supported by MSK Cancer Center Support Grant/Core Grant (P30 CA008748).

AUTHORSHIP CONTRIBUTIONS

EC, MR, and LGR conceived of the experiments, advised on every aspect, conducted validation experiments and wrote the manuscript; MR and JR sorted primary cases, optimized and constructed libraries. JR, IYK, WD, FW, SZ, BB, and AC analyzed data with LM, OE, and RR. JB, SIP, AEK, MJO, NG, MSL, MJB contributed patient samples. All authors reviewed the manuscript.

DISCLOSURE OF CONFLICTS OF INTEREST

We have no conflicts of interest to report.

Swerdlow SH, Campo E, Harris NL, Jaffe ES, Pileri SA, Stein H, et al. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. Lyon, France2008.
Schwering I, Brauninger A, Klein U, Jungnickel B, Tinguely M, Diehl V, et al. Loss of the B-lineage-specific gene expression program in Hodgkin and Reed-Sternberg cells of Hodgkin lymphoma. Blood. 2003;101(4):1505-12.
Steidl C, Shah SP, Woolcock BW, Rui L, Kawahara M, Farinha P, et al. MHC class II transactivator CIITA is a recurrent gene fusion partner in lymphoid cancers. Nature. 2011;471(7338):377-81.
Hartmann S, Martin-Subero JI, Gesk S, Husken J, Giefing M, Nagel I, et al. Detection of genomic imbalances in microdissected Hodgkin and Reed-Sternberg cells of classical Hodgkin's lymphoma by array-based comparative genomic hybridization. Haematologica. 2008;93(9):1318-26.
Tiacci E, Doring C, Brune V, van Noesel CJ, Klapper W, Mechtersheimer G, et al. Analyzing primary Hodgkin and Reed-Sternberg cells to capture the molecular and cellular pathogenesis of classical Hodgkin lymphoma. Blood. 2012;120(23):4609-20.
Steidl C, Diepstra A, Lee T, Chan FC, Farinha P, Tan K, et al. Gene expression profiling of microdissected Hodgkin Reed-Sternberg cells correlates with treatment outcome in classical Hodgkin lymphoma. Blood. 2012;120(17):3530-40.
Weniger MA, Tiacci E, Schneider S, Arnolds J, Ruschenbaum S, Duppach J, et al. Human CD30+ B cells represent a unique subset related to Hodgkin lymphoma cells. J Clin Invest. 2018;128(7):2996-3007.
Reichel J, Chadburn A, Rubinstein PG, Giulino-Roth L, Tam W, Liu Y, et al. Flow sorting and exome sequencing reveal the oncogenome of primary Hodgkin and Reed-Sternberg cells. Blood. 2015;125(7):1061-72.
Maura F, Ziccheddu B, Xiang JZ, Bhinder B, Rosiene J, Abascal F, et al. Molecular Evolution of Classic Hodgkin Lymphoma Revealed Through Whole-Genome Sequencing of Hodgkin and Reed Sternberg Cells. Blood Cancer Discov. 2023;4(3):208-27.
Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, Rosenwald A, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature. 2000;403(6769):503-11.
Savage KJ, Monti S, Kutok JL, Cattoretti G, Neuberg D, De Leval L, et al. The molecular signature of mediastinal large B-cell lymphoma differs from that of other diffuse large B-cell lymphomas and shares features with classical Hodgkin lymphoma. Blood. 2003;102(12):3871-9.
Rosenwald A, Wright G, Leroy K, Yu X, Gaulard P, Gascoyne RD, et al. Molecular diagnosis of primary mediastinal B cell lymphoma identifies a clinically favorable subgroup of diffuse large B cell lymphoma related to Hodgkin lymphoma. J Exp Med. 2003;198(6):851-62.
Monti S, Savage KJ, Kutok JL, Feuerhake F, Kurtin P, Mihm M, et al. Molecular profiling of diffuse large B-cell lymphoma identifies robust subtypes including one characterized by host inflammatory response. Blood. 2005;105(5):1851-61.
Juskevicius D, Jucker D, Dietsche T, Perrina V, Rufle A, Ruiz C, et al. Novel cell enrichment technique for robust genetic analysis of archival classical Hodgkin lymphoma tissues. Lab Invest. 2018;98(11):1487-99.
Spina V, Bruscaggin A, Cuccaro A, Martini M, Di Trani M, Forestieri G, et al. Circulating tumor DNA reveals genetics, clonal evolution, and residual disease in classical Hodgkin lymphoma. Blood. 2018;131(22):2413-25.
Wienand K, Chapuy B, Stewart C, Dunford AJ, Wu D, Kim J, et al. Genomic analyses of flow-sorted Hodgkin Reed-Sternberg cells reveal complementary mechanisms of immune evasion. Blood Adv. 2019;3(23):4065-80.
Agirre X, Meydan C, Jiang Y, Garate L, Doane AS, Li Z, et al. Long non-coding RNAs discriminate the stages and gene regulatory states of human humoral immune response. Nat Commun. 2019;10(1):821.
Mottok A, Hung SS, Chavez EA, Woolcock B, Telenius A, Chong LC, et al. Integrative genomic analysis identifies key pathogenic mechanisms in primary mediastinal large B-cell lymphoma. Blood. 2019;134(10):802-13.
Noerenberg D, Briest F, Hennch C, Yoshida K, Hablesreiter R, Takeuchi Y, et al. Genetic Characterization of Primary Mediastinal B-Cell Lymphoma: Pathogenesis and Patient Outcomes. J Clin Oncol. 2024;42(4):452-66.
Heger JM, Mammadova L, Mattlener J, Sobesky S, Cirillo M, Altmuller J, et al. Circulating Tumor DNA Sequencing for Biologic Classification and Individualized Risk Stratification in Patients With Hodgkin Lymphoma. J Clin Oncol. 2024;42(35):4218-30.
Alig SK, Shahrokh Esfahani M, Garofalo A, Li MY, Rossi C, Flerlage T, et al. Distinct Hodgkin lymphoma subtypes defined by noninvasive genomic profiling. Nature. 2024;625(7996):778-87.
Fromm JR, Kussick SJ, Wood BL. Identification and purification of classical Hodgkin cells from lymph nodes by flow cytometry and flow cytometric cell sorting. Am J Clin Pathol. 2006;126(5):764-80.
Vogel MJ, Xie L, Guan H, Tooze RM, Maier T, Kostezka U, et al. FOXO1 repression contributes to block of plasma cell differentiation in classical Hodgkin lymphoma. Blood. 2014;124(20):3118-29.
Reichel JB, McCormick J, Fromm JR, Elemento O, Cesarman E, Roshal M. Flow-sorting and Exome Sequencing of the Reed-Sternberg Cells of Classical Hodgkin Lymphoma. J Vis Exp. 2017(124).
Anders S, Pyl PT, Huber W. HTSeq--a Python framework to work with high-throughput sequencing data. Bioinformatics. 2015;31(2):166-9.
Robinson MD, Oshlack A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010;11(3):R25.
Law CW, Chen Y, Shi W, Smyth GK. voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014;15(2):R29.
Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47.
Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545-50.
Liberzon A, Birger C, Thorvaldsdottir H, Ghandi M, Mesirov JP, Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015;1(6):417-25.
Kolde R. pheatmap: Pretty Heatmaps. R package version 1.0.13. 2025.
Wickham H. ggplot2: Elegant Graphics for Data Analysis.: Springer-Verlag, New York; 2016.
Rengstl B, Newrzela S, Heinrich T, Weiser C, Thalheimer FB, Schmid F, et al. Incomplete cytokinesis and re-fusion of small mononucleated Hodgkin cells lead to giant multinucleated Reed-Sternberg cells. Proc Natl Acad Sci U S A. 2013;110(51):20729-34.
Xavier de Carvalho A, Maiato H, Maia AF, Ribeiro SA, Pontes P, Bickmore W, et al. Reed-Sternberg cells form by abscission failure in the presence of functional Aurora B kinase. PLoS One. 2015;10(5):e0124629.
Georgiadi EC, Dimtsas GS, Vassilakopoulos TP, Pangalis GA, Kittas C, Doussis-Anagnostopoulou IA. Functional p53 can modulate the relationship between E2F-1 expression and tumor kinetics in Hodgkin lymphoma. Leuk Lymphoma. 2015;56(3):748-54.
Fernandez S, Diaz E, Rita CG, Estevez M, Montalban C, Garcia JF. BET inhibitors induce NF-kappaB and E2F downregulation in Hodgkin and Reed-Sternberg cells. Exp Cell Res. 2023;430(2):113718.
Rabadan R.
Selitsky SR, Marron D, Hollern D, Mose LE, Hoadley KA, Jones C, et al. Virus expression detection reveals RNA-sequencing contamination in TCGA. BMC Genomics. 2020;21(1):79.
Mottok A, Woolcock B, Chan FC, Tong KM, Chong L, Farinha P, et al. Genomic Alterations in CIITA Are Frequent in Primary Mediastinal Large B Cell Lymphoma and Are Associated with Diminished MHC Class II Expression. Cell Rep. 2015;13(7):1418-31.
Lacy SE, Barrans SL, Beer PA, Painter D, Smith AG, Roman E, et al. Targeted sequencing in DLBCL, molecular subtypes, and outcomes: a Haematological Malignancy Research Network report. Blood. 2020;135(20):1759-71.
Kanzler H, Kuppers R, Hansmann ML, Rajewsky K. Hodgkin and Reed-Sternberg cells in Hodgkin's disease represent the outgrowth of a dominant tumor clone derived from (crippled) germinal center B cells. J Exp Med. 1996;184(4):1495-505.
Kuppers R. Identifying the precursors of Hodgkin and Reed-Sternberg cells in Hodgkin's disease: role of the germinal center in B-cell lymphomagenesis. J Acquir Immune Defic Syndr. 1999;21 Suppl 1:S74-9.
Marafioti T, Hummel M, Foss HD, Laumen H, Korbjuhn P, Anagnostopoulos I, et al. Hodgkin and reed-sternberg cells represent an expansion of a single clone originating from a germinal center B-cell with functional immunoglobulin gene rearrangements but defective immunoglobulin transcription. Blood. 2000;95(4):1443-50.
Liso A, Capello D, Marafioti T, Tiacci E, Cerri M, Distler V, et al. Aberrant somatic hypermutation in tumor cells of nodular-lymphocyte-predominant and classic Hodgkin lymphoma. Blood. 2006;108(3):1013-20.
Buettner M, Greiner A, Avramidou A, Jack HM, Niedobitek G. Evidence of abortive plasma cell differentiation in Hodgkin and Reed-Sternberg cells of classical Hodgkin lymphoma. Hematol Oncol. 2005;23(3-4):127-32.
Garcia JF, Roncador G, Garcia JF, Sanz AI, Maestre L, Lucas E, et al. PRDM1/BLIMP-1 expression in multiple B and T-cell lymphoma. Haematologica. 2006;91(4):467-74.
Tzankov A, Zimpfer A, Pehrs AC, Lugli A, Went P, Maurer R, et al. Expression of B-cell markers in classical hodgkin lymphoma: a tissue microarray analysis of 330 cases. Mod Pathol. 2003;16(11):1141-7.
Carbone A, Gloghini A, Gaidano G, Franceschi S, Capello D, Drexler HG, et al. Expression status of BCL-6 and syndecan-1 identifies distinct histogenetic subtypes of Hodgkin's disease. Blood. 1998;92(7):2220-8.
Chang KC, Chen PC, Chen YP, Chang Y, Su IJ. Dominant expression of survival signals of endoplasmic reticulum stress response in Hodgkin lymphoma. Cancer Sci. 2011;102(1):275-81.
Tellier J, Shi W, Minnich M, Liao Y, Crawford S, Smyth GK, et al. Blimp-1 controls plasma cell function through the regulation of immunoglobulin secretion and the unfolded protein response. Nat Immunol. 2016;17(3):323-30.
Dilosa RM, Maeda K, Masuda A, Szakal AK, Tew JG. Germinal center B cells and antibody production in the bone marrow. J Immunol. 1991;146(12):4071-7.
Machado HE, Mitchell E, Obro NF, Kubler K, Davies M, Leongamornlert D, et al. Diverse mutational landscapes in human lymphocytes. Nature. 2022;608(7924):724-32.
Rustad EH, Yellapantula V, Leongamornlert D, Bolli N, Ledergor G, Nadeu F, et al. Timing the initiation of multiple myeloma. Nat Commun. 2020;11(1):1917.
Gallagher A, Perry J, Shield L, Freeland J, MacKenzie J, Jarrett RF. Viruses and Hodgkin disease: no evidence of novel herpesviruses in non-EBV-associated lesions. Int J Cancer. 2002;101(3):259-64.
Wilson KS, Gallagher A, Freeland JM, Shield LA, Jarrett RF. Viruses and Hodgkin lymphoma: no evidence of polyomavirus genomes in tumor biopsies. Leuk Lymphoma. 2006;47(7):1315-21.
Wilson KS, Freeland JM, Gallagher A, Cosby SL, Earle JA, Alexander FE, et al. Measles virus and classical Hodgkin lymphoma: no evidence for a direct association. Int J Cancer. 2007;121(2):442-7.
Schmitz R, Hansmann ML, Bohle V, Martin-Subero JI, Hartmann S, Mechtersheimer G, et al. TNFAIP3 (A20) is a tumor suppressor gene in Hodgkin lymphoma and primary mediastinal B cell lymphoma. J Exp Med. 2009;206(5):981-9.

There is NO conflict of interest to disclose.

SupplementaryDatatablesBCG.xlsx
Supplemental datatables
RNAseqsupplementalmethodstablesandfiguresforBCGFinal.docx
Supplemental methods and figures

Download PDF

Review #2 received at journal
20 Dec, 2025
Reviewer #2 agreed at journal
12 Dec, 2025
Reviewer #1 agreed at journal
24 Nov, 2025
Reviewers invited by journal
03 Nov, 2025
Editor assigned by journal
27 Oct, 2025
Submission checks completed at journal
27 Oct, 2025
First submitted to journal
26 Oct, 2025

You are reading this latest preprint version

Transcriptome sequencing of Hodgkin lymphoma Hodgkin and Reed-Sternberg cells reveals escape from NK cell recognition and an unfolded protein response

Status:

Version 1

Abstract

Figures

INTRODUCTION

MATERIALS/SUBJECTS AND METHODS

Tissue specimens

Immunohistochemistry

Cell Sorting

Flow cytometry

RNA Extraction, Library Construction and Next-Generation Sequencing

Differential Gene expression analysis

Data Input and Preprocessing and Methodology for Virus Identification

RESULTS

Distinct gene expression profiles of HRS cells compared to intra-tumoral B cells

HRS cells exhibit an unfolded protein response

Downregulation of NK cell recognition pathways in HRS cells

Virus discovery

Comparative transcriptomic analysis of cHL reveals key differences between cHL and PMBL

DISCUSSION

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1