Abstract Epigenetic regulation is essential for mammary gland development, yet the specific chromatin remodelers that govern mammary epithelial cell fate remain poorly defined. Mutations in SWI/SNF chromatin remodeling complex subunits occur in more than 20% of human cancers, with ARID1A being the most frequently altered. In breast cancer, ARID1A loss of function mutations are enriched in metastatic estrogen receptor-positive (ER+) disease and associated with endocrine therapy resistance. To define the developmental role of Arid1a in vivo, we generated mice with mammary epithelium specific Arid1a deletion. These animals displayed disrupted ductal branching and aberrant terminal end bud formation. Mammary organoids derived from Arid1a deficient tissue further revealed abnormal cystic morphology and impaired differentiation. To dissect the molecular consequences of Arid1a loss, we performed single cell multiomic profiling that combine single nucleus RNA and chromatin accessibility sequencing from the same cells, together with H3K27ac and BRG1 CUT&RUN-seq. Arid1a loss caused a collapse of normal mammary epithelial lineage architecture, with single cell analyses showing failure to maintain basal, luminal progenitor, and alveolar identities. Instead, Arid1a deficient cells were restricted to an undifferentiated luminal hormonal like state characterized by reduced estrogen receptor signaling competence. Chromatin profiling revealed profound remodeling, including decreased accessibility and impaired SWI/SNF targeting at lineage defining transcription factors (TF) such as Foxa1, Gata3, and Sox9. A CRISPR/Cas9 pooled loss of function screen identified Foxa1 and Meis1 as essential regulators whose deletion recapitulated the Arid1a null phenotype, positioning them as downstream effectors required for mammary cell fate specification. Notably, Meis1 emerged as a previously unrecognized regulator of luminal hormonal identity. Our work provides a framework for understanding how Arid1a rewires the chromatin landscape and transcriptional network in normal mammary development. By identifying lineage specific TF motifs and critical regulators like Meis1, we identify new opportunities for targeted therapeutic intervention. Citation Format: Erik Ladewig, Amaia Arruabarrena-Aristorena, Estelle Deby, Srushti Kittane, Fresia Pareja, Ryan Blawski, Yangzhenyu Gao, Laura Baldino, Vito Rebecca, Emiliano Cocco, Hongkai Ji, Pau Castel, Christina Leslie, Wouter Karthaus, Eneda Toska. Arid1a directs lineage specification in mammary epithelial cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3219.

Thetis cells (TCs) are a recently identified lineage of RORγt+ antigen-presenting cells comprising four subsets, TC I to TC IV, including a tolerogenic subset (TC IV) that instructs tolerance to gut microbiota and food antigens1-6. A developmental wave of TCs during early life creates a crucial window of opportunity for establishing intestinal tolerance1,5. The ontogeny of TCs and the cues that shape their abundance and heterogeneity remain unknown, however, limiting efforts to harness their therapeutic potential. Here we identify a population of RORγt+ progenitors, termed Thetis-lymphoid tissue inducer progenitors (TLPs), that give rise to the immediate TC progenitor (TCP) and the lymphoid tissue inducer (LTi) progenitor (LTiP), and identify PU.1 as the transcription factor that governs TC fate. Despite transcriptional similarity to myeloid-derived conventional dendritic cells, we show that TCs descend from the common lymphoid progenitor. Deletion of the plasmacytoid dendritic cell (pDC) lineage-determining transcription factor TCF4 expands TLPs and TCs, indicating a shared developmental branch with pDCs. TLPs are enriched in fetal liver, but, unlike LTi cells, TCs emerge postnatally, indicating that developmentally timed environmental cues promote TCP differentiation. We identify one such cue, RANKL provision by lymphoid tissue organizer cells, that is essential for TC I differentiation. Together, these findings define the ontogeny of TCs and the transcription factors that promote TC differentiation and heterogeneity, facilitating future investigations of these enigmatic cells and their therapeutic potential for tolerance induction in food allergy and autoimmunity.

CAR therapy has transformed the treatment landscape for hematological malignancies but success in solid tumors is at present limited. One important mechanism underlying treatment failure is insufficient functional persistence of the engineered T cells. To overcome the functional decline arising from chronic antigen exposure, we conducted a multi-parameter in vivo screen targeting a curated set of 400 transcription factors. These transcription factors were selected based on their involvement in T cell exhaustion across diverse contexts—including murine acute and chronic stimulation models, CAR designs associated with enhanced persistence, and transcriptional signatures from patients achieving durable clinical responses following CAR therapy. We employed the acute lymphoblastic leukemia (B-ALL) NALM6 mouse model and introduced two sequential tumor rechallenges immediately after primary tumor clearance to simulate persistent CAR activation. The in vivo screen was performed as an enrichment-based approach to identify gRNAs that enhance CAR T cell persistence under chronic in vivo stimulation. Top candidates identified in the in vivo screen were subsequently evaluated in an in vitro chronic activation screen where they were evaluated for multiple functions under persistent activation – expansion, cytotoxicity, differentiation, and inhibitory receptor expression. This screen revealed NFIL3 as a key negative regulator of CAR T cell function. Genetic ablation of NFIL3 in CAR T cells promoted sustained expansion, enhanced IL-2 and Th2-associated cytokine production, and restricted terminal differentiation, overall favoring a transcriptional phenotype that has been associated with superior clinical responses in patients treated with CAR T cells. NFIL3 disruption improved antitumor activity across multiple stress-test models, spanning different CAR designs and tumor types in both, hematologic and solid cancers. These findings underscore the power of comprehensive, multiparametric in vivo genetic screening to uncover novel regulators of CAR T cell function and position NFIL3 as a compelling therapeutic target for enhancing CAR T cell efficacy across diverse cancer indications. Nayan Jain, Yuzhe Shi, Celina May, Sneha Mitra, Philip Bucher, Anton Dobrin, Zeguo Zhao, Sophie Hanina, Vinagolu Rajasekhar, Christina Leslie, Francisco Sánchez-Rivera, Judith Feucht, Michel Sadelain. NFIL3 emerges as a driver of CAR T cell dysfunction from integrated in vivo/in vitro CRISPR screen [abstract]. In: Proceedings of the AACR Immuno-Oncology Conference (AACR IO): Discovery and Innovation in Cancer Immunology: Revolutionizing Treatment through Immunotherapy; 2026 Feb 18-21; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Immunol Res 2026;14(2 Suppl):Abstract nr C008.

Acute cohesin loss causes widespread reorganization of three-dimensional (3D) chromatin architecture but has relatively minor effects on steady-state transcription. It remains unclear whether its role in gene regulation becomes more critical during mitotic exit, when 3D chromatin architecture and transcription are globally re-established. To address this, we acutely depleted RAD21 in mouse embryonic stem cells during mitotic exit under self-renewal or differentiation conditions. Here we show that, although most loops failed to reform without cohesin, the few cohesin-independent loops were linked to active promoters, strong enhancers and H3K27ac mitotic bookmarking. Transcriptional changes were only modest, indicating that gene reactivation largely bypasses cohesin. Sensitive genes showed RAD21 promoter binding, a higher number of structural loops and positioning within well-insulated, gene-poor topologically associating domains. During differentiation, cohesin loss impaired activation of a broader set of developmental genes, partly due to defective de novo regulatory interactions. Together, these findings demonstrate context-specific requirements for cohesin in gene activation.

Metastatic microenvironments vary widely not only in their biochemical composition but also in their mechanical properties. Here, we examined how the mechanical rigidity of the metastatic niche affects metastases seeding and the local efficacy of antitumor immunosurveillance. Cancer cells stiffened in response to increasing environmental rigidity, a biophysical change that mechanically sensitized them to killing by cytotoxic lymphocytes. In immunodeficient mice, rigidity sensing by cancer cells yielded robust bone colonization, accompanied by marked stiffening of the cancer cells themselves. Conversely, in immunocompetent hosts, stiffer cancer cells were selectively eliminated, and bone metastasis was suppressed. In patients, metastatic cell stiffness was associated directly with environmental rigidity and inversely with immune infiltration. Expression of Spp1, encoding the secreted glycoprotein osteopontin, defined a subset of cancer cells that expanded in the bone, and deletion of Spp1 limited environmentally induced cancer cell stiffening, bone colonization, and immune vulnerability. Thus, environmental mechanosensing regulates both metastases seeding and antitumor immunity, providing an immunological basis for metastatic site selection.

Mutations in the pioneer transcription factor FOXA1 occur in 10%-40% of prostate cancers and broadly alter chromatin accessibility. In a cohort of 874 primary and metastatic tumors, we confirm frequent Wing2 missense mutations and indels, as well as C-terminal truncating frameshifts. To define their functional impact, we performed single-nucleus multiome profiling in mouse prostate organoids expressing representative alleles, including overexpressed wild-type FOXA1. Each subgroup produces distinct chromatin and transcriptional changes, but all perturb epithelial lineage specification. Indel mutants promote basal-like states, whereas C-terminal truncations, Wing2 missense mutations, and elevated wild-type FOXA1 drive secretory L1-like luminal fates. Integrated RNA-seq, ATAC-seq, and ChIP-seq reveal that L1-like specification involves a hybrid androgen receptor/FOXA1 motif and cooperation with POU2F1. In vivo, these same alleles, combined with Trp53/Pten loss, shift tumor histology from basal-like to secretory luminal phenotypes.

Abstract By leveraging a sequence-based deep learning framework, we seek to uncover mechanisms by which mutations arise in noncoding regulatory regions, potentially leading to the discovery of novel targets in metastatic breast cancer. Research around metastatic breast cancer has largely focused on analyzing coding mutations to characterize progression. Though noncoding mutations are known to affect transcription factor binding and regulation of gene expression, few noncoding mutation drivers have been identified. Previously published work has shown that metastatic mutation rate correlates with open chromatin in the cells-of-origin. However, this work has mostly been done at a coarse, region-level scale, identifying mutational hotspot regions. In this work, we aim to uncover genomic positions in regulatory regions of metastatic breast cancer with elevated mutation rates, and identify their potential mutation mechanism. We propose a novel deep learning model that uncovers the sequence context-based covariates of per-base mutation rate in regulatory regions of metastatic breast cancer. With access to over 500,000 mutations from the Hartwig Medical Foundation cohort, the neural network is trained on sequences from regulatory regions in normal breast epithelium, and predicts per-base mutation rate profiles for the region. As a result, the model learns how sequence features change mutation likelihood at particular genomic positions. Analysis of the saliency map of the model allows for identification of specific sites with higher-than-expected mutation rates, which is potentially indicative of increased transcription factor binding that extends beyond selection by pro-metastatic regulatory programs. These findings provide a method to connect noncoding mutation patterns to mutation mechanism and regulatory effects in metastatic genomes. Future work aims at expanding this model to a pan-cancer level, revealing shared and cancer-specific noncoding mutations that have potential to reveal patterns of metastasis. This scalable sequence model framework provides an advantage over existing methods particularly due to its base-pair level resolution modeling of the metastatic cancer genome. Thus, its implications for modeling and uncovering regulatory mechanisms of cancer is key. Citation Format: Ariaki Dandawate, Christina Leslie, Ekta Khurana. Modeling base-pair level mutation rate in metastatic breast cancer using a sequence-based deep learning model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1500.

Hi-C is a chromosome conformation capture assay used to study three-dimensional (3D) genome organization. Single-cell Hi-C technologies now enable the examination of 3D chromatin organization in individual cells, although these approaches often suffer from low-coverage libraries and data sparsity. Here, we introduce HiC2Self, a self-supervised framework for denoising Hi-C contact maps that requires only low-coverage data as input. HiC2Self reconstructs key structures such as topologically associating domains (TADs) and significant loops from bulk libraries, including cell-type-specific Hi-C structures, without the generalization challenges faced by supervised models. HiC2Self can also accurately reconstruct significant loops from Micro-C data at 1-kilobase resolution. When applied to single-nucleus methyl-3C data, HiC2Self successfully reconstructs local TAD structures around specific genes at 10-kilobase resolution with as few as 50 cells. Last, HiC2Self enables the examination of single-cell structures at 50-kilobase resolution in individual cells of the same cell type. HiC2Self thus provides a general tool for denoising bulk, pseudobulk, and single-cell 3D contact maps to enable downstream analyses.

Regulatory T (Treg) cells are a specialized CD4+ T cell lineage with essential anti-inflammatory functions. Analysis of Treg cell adaptations to non-lymphoid tissues that enable their specialized immunosuppressive and tissue-supportive functions raises questions about the underlying mechanisms of these adaptations and whether they represent stable differentiation or reversible activation states. Here, we characterize distinct colonic effector Treg cell transcriptional programs. Attenuated T cell receptor (TCR) signaling and acquisition of substantial TCR-independent functionality seems to facilitate the terminal differentiation of a population of colonic effector Treg cells that are distinguished by stable expression of the immunomodulatory cytokine IL-10. Functional studies show that this subset of effector Treg cells, but not their expression of IL-10, is indispensable for colonic health. These findings identify core features of the terminal differentiation of effector Treg cells in non-lymphoid tissues and their function. The authors show that terminally differentiated colonic Treg cells are required for maintaining colonic health and, although these cells are major producers of this cytokine, IL-10 is dispensable for their suppressive function.

Malignant T-cell transformation after chimeric antigen receptor (CAR) T-cell therapy has been described, but the contribution of CAR integration to oncogenesis is not clear. Here we report a case of a T-cell lymphoma harboring a lentiviral integration in a known tumor suppressor, TP53, which developed in a patient with multiple myeloma after B-cell maturation antigen (BCMA) CAR T-cell therapy.

Single-cell RNA sequencing studies have revealed the heterogeneity of cell states present in the rheumatoid arthritis (RA) synovium. However, it remains unclear how these cell types interact with one another in situ and how synovial microenvironments shape observed cell states. Here, we use spatial transcriptomics (ST) to define stable microenvironments across eight synovial tissue samples from six RA patients and characterize the cellular composition of ectopic lymphoid structures (ELS). To identify disease-relevant cellular communities, we developed DeepTopics, a scalable reference-free deconvolution method based on a Dirichlet variational autoencoder architecture. DeepTopics identified 22 topics across tissue samples that were defined by specific cell types, activation states, and/or biological processes. Some topics were defined by multiple colocalizing cell types, such as CD34+ fibroblasts and LYVE1+ macrophages, suggesting functional interactions. Within ELS, we discovered two divergent cellular patterns that were stable across ELS in each patient and typified by the presence or absence of a "germinal-center-like" topic. DeepTopics is a versatile and computationally efficient method for identifying disease-relevant microenvironments from ST data, and our results highlight divergent cellular architectures in histologically similar RA synovial samples that have implications for disease pathogenesis.

Human embryonic stem cells (hESCs) are notable for their ability to self-renew and to differentiate into all tissue types in the body. NANOG is a core regulator of hESC identity, and dynamic control of its expression is crucial to maintain the balance between self-renewal and differentiation. Transcriptional regulation depends on enhancers, but NANOG enhancers in hESCs are not well characterized. Here, we report two NANOG enhancers discovered from a CRISPR interference screen in hESCs. Deletion of a single copy of either enhancer significantly reduced NANOG expression, compromising self-renewal and increasing differentiation propensity. Interestingly, these two NANOG enhancers are involved in a tandem duplication event found in certain primates including humans but not in mice. However, the duplicated counterparts do not regulate NANOG expression. This work expands our knowledge of functional enhancers in hESCs and highlights the sensitivity of the hESC state to the dosage of core regulators and their enhancers.

Introduction Despite advances in Hodgkin lymphoma (HL) treatment, effective strategies for relapsed/refractory (R/R) HL, especially after PD-1 blockade failure, remain a critical need. Monocyte-driven inflammation and dysfunctional CD4⁺ T cells have been implicated in resistance to PD-1 inhibitors (Paczkowska et al., Nat Commun, 2024). Based on this, we hypothesized that HDAC inhibitors (HDACi) may restore PD-1 sensitivity by enhancing antigen presentation and promoting a tumor microenvironment conducive to anti-tumor immunity. In the present study, we leveraged single-cell combined indexing of transcriptome and epitopes (CITE)-sequencing and high-resolution spatial transcriptomics to characterize the circulating and microenvironmental features associated with response to combined HDACi and PD-1 blockade in R/R HL. Methods Our group recently completed a phase II trial evaluating entinostat plus pembrolizumab in R/R HL (Stuver et al., ASH 2024). To investigate mechanisms of response and resistance, we analyzed 20 longitudinally collected peripheral blood samples from 9 patients and 6 paired (pre- and on-treatment) tumor biopsies using single-cell RNA, TCR, and surface epitope profiling (Single Cell Immune Profiling, 10x Genomics) and spatial transcriptomics (Visium HD, 10x Genomics). Cell types in spatial data were annotated using an optimal transport-based approach leveraging a published HL reference atlas (Stewart et al., Blood, 2023). Cell-cell communication analysis was conducted to identify ligand-receptor interactions influencing treatment sensitivity and resistance. Results 39 patients were enrolled, including 22 (56%) who had previously failed anti-PD1 therapy. Among 38 evaluable patients, complete and objective response rates were 47% and 63%, respectively. Peripheral blood analysis was performed on 9 patients (median age, 36 [21–77]) with samples collected at Cycle 1 Day 1, Cycle 2 Day 1, and, if applicable, at progression. Of these, 5 (55%) achieved complete remission, 1 (11%) partial remission, and 3 (34%) experienced progression. After filtering, 68,988 cells were analyzed: 37,710 monocytes and 21,013 T cells. We identified 11 T cell subsets, including CD8⁺ and CD4⁺ T cells exhibiting a stress-associated transcriptional program (RUNX1, TOX, tyrosine phosphatases). Myeloid profiling revealed enrichment of inflammatory classical monocytes (IL-1β, NLRP3, NF-κB targets) in patients who had failed PD-1 monotherapy (p < 0.0001 for IL1B in monocytes). Longitudinal profiling showed that entinostat plus pembrolizumab upregulated antigen presentation genes (B2M, HLA-DRA, HLA-A, HLA-B, CIITA, CD74) in both responders and non-responders. Resistance was associated with persistent inflammatory monocytes, reduced clonal expansion of effector CD8⁺ T cells (6.97% vs. 24% in responders), and a skewed CD4⁺ T cell compartment marked by elevated frequencies of stressed and Treg CD4⁺ cells (LFC > 5, p < 0.05). These findings suggest that inflammatory monocytes may suppress anti-tumor immunity by impairing antigen-dependent CD8⁺ T cell priming and expansion. Spatial profiling confirmed that treatment resistance was associated with tumor infiltration by IL1B⁺ monocytes, T follicular helper (Tfh) and Treg cells, highlighting immunosuppressive monocyte-CD4+ T cell niches as key drivers of resistance. Furthermore, we identified the CCL17–CCR4 axis as a central mediator of immune evasion and resistance to PD1 blockade in HL. Specifically, sustained CCL17 expression by local networks of HRS cells and inflammatory monocytes was found to recruit Tfh and Treg cells, thereby preventing CD8⁺ T cell surveillance. Co-expression of STAT6 and CCL17 was observed in all pre-treatment samples but persisted only in non-responder patients on-treatment. These findings suggest that HDAC inhibition may overcome resistance by suppressing STAT6 activity in HRS cells, reducing CCL17 expression, disrupting the monocyte-HRS feedback loop, and restoring effective CD8⁺ T cell-mediated tumor immunity. Conclusion: Our integrative single-cell and spatial analysis reveals that adaptive immune networks and localized immunosuppressive niches critically shape the response to checkpoint blockade in Hodgkin lymphoma. From these findings, the spatial organization of the tumor microenvironment emerges as a critical determinant of therapeutic sensitivity and a promising target for intervention.

Supplementary Table (Table S1) and Figure Legends

Diffuse large B cell lymphomas (DLBCL) are the most common lymphoid malignancies in adults. Despite advances in molecular classification, the pathogenesis of DLBCL, particularly of the BN2 subtype, remains poorly understood, which limits the advancement of tailored and more effective therapeutic strategies. BN2-DLBCL are characterized by alterations in BCL6 and NOTCH2, lack an AICDA mutational signature, and are presumed to arise outside germinal centers (GC). Among its defining alterations, truncating mutations in SPEN (SPENTRUNC) are significantly enriched, but the effects and clinical relevance of these alterations remain unexplored. Here, we found that SPENTRUNC likely represent loss-of-function (LOF) events, as they led to reduced SPEN protein levels (p=0.04). Clinically, SPENTRUNC mutations correlated with significantly worse overall survival (OS), especially in non-GCB DLBCL patients (HR: 1.82; p<0.0001). Co-occurring truncating mutations in NOTCH2 (NOTCH2TRUNC), which confer gain-of-function (GOF) effects, further worsened prognosis when in combination with SPENTRUNC (HR: 3.33; p<0.0001). Patients with dual SPENTRUNC/NOTCH2TRUNC (SN2) mutations were also older (p=0.03) and had poorer ECOG performance status (p=0.02), defining a high-risk subgroup urgently needing targeted therapies. To understand how SN2 mutations shape disease biology, we introduced B cell–specific SpenLOF and Notch2GOF mutations in mice. The SN2 genotype led to a cumulative expansion of autoimmune/aged B cells (AiBCs), a hyper-reactive inflammatory B cell subset implicated in autoimmunity and lymphomagenesis. Given the hypothesized extra-follicular origin of BN2-DLBCL, we tested whether AiBCs could arise in SN2 mice lacking Bcl6, which is essential for GC formation. Indeed, AiBC expansion occurred independently of GC formation, as SN2;Bcl6–/– and SN2 mice showed comparable AiBC levels, supporting their extra-follicular derivation. To further evaluate their malignant potential, we assessed clonality and proliferation in SN2 AiBCs versus their wild-type (WT) counterparts. SN2 AiBCs exhibited significantly higher clonality and proliferation (p<0.05). Notably, female SN2 AiBCs showed even greater proliferation (KI67+) than male SN2 AiBCs (p<0.0001), a difference not observed in WT mice. This disproportionate fitness of female SN2 cells translated to a competitive advantage, as shown by bone marrow chimera assays (p=0.04), regardless of the hormonal sex status of the recipient animal. Accordingly, female SN2 mice had significantly reduced survival due to lymphoma development than their male counterparts (OS HR: 13.4; p=0.001), a trend mirrored in human SN2-DLBCL patients (OS HR: 4.14; p=0.07). This sex-bias is of particular interest, as SPEN is known to be essential in X-chromosomal inactivation (XCI), a process happening in females to equilibrate X chromosomal gene dosage to male cells. Although XIST RNA-FISH did not reveal changes in XCI (p=0.2), SN2 lymphomas showed significant hypomethylation of the X chromosome compared to WT B cells and N2 lymphomas (p=5.9e-4). Autosomal regions, in contrast, were hypermethylated (p<2.2e-18), suggesting X-chromosome–specific dysregulation due to SPEN loss. Among X-linked genes, TLR7, a known AiBC driver, emerged as a candidate mediator. Indeed, female SN2 lymphomas expressed higher TLR7 levels than males (p=0.05). To test whether the TLR7 pathway confers an exploitable therapeutic vulnerability, we treated SN2 lymphoma cells with AZ1495, an IRAK1/4 inhibitor acting downstream of TLR7, and found that only female cells showed in vitro and in vivo sensitivity (p<0.05). Efficacy was further confirmed using a female human SN2-DLBCL PDX model (p=0.01), supporting the translational potential of targeting this axis. In summary, SPENTRUNC defines a poor-prognosis marker in DLBCL, and cooperates with NOTCH2TRUNC to drive aggressive, extra-follicular lymphomas via expansion of pathogenic AiBCs. We identify a novel, sex-biased pathogenic mechanism involving X-chromosomal dysregulation and TLR7 overexpression, offering a rationale for precision therapy in BN2-DLBCL.