Functional Genomic Landscape of Human Breast Cancer Drivers, Vulnerabilities, and Resistance.

Large-scale genomic studies have identified multiple somatic aberrations in breast cancer, including copy number alterations and point mutations. Still, identifying causal variants and emergent vulnerabilities that arise as a consequence of genetic alterations remain major challenges. We performed whole-genome small hairpin RNA (shRNA) "dropout screens" on 77 breast cancer cell lines. Using a hierarchical linear regression algorithm to score our screen results and integrate them with accompanying detailed genetic and proteomic information, we identify vulnerabilities in breast cancer, including candidate "drivers," and reveal general functional genomic properties of cancer cells. Comparisons of gene essentiality with drug sensitivity data suggest potential resistance mechanisms, effects of existing anti-cancer drugs, and opportunities for combination therapy. Finally, we demonstrate the utility of this large dataset by identifying BRD4 as a potential target in luminal breast cancer and PIK3CA mutations as a resistance determinant for BET-inhibitors.

Proteomic Dynamics of Breast Cancer Cell Lines Identifies Potential Therapeutic Protein Targets.

Treatment and relevant targets for breast cancer (BC) remain limited, especially for triple-negative BC (TNBC). We identified 6091 proteins of 76 human BC cell lines using data-independent acquisition (DIA). Integrating our proteomic findings with prior multi-omics datasets, we found that including proteomics data improved drug sensitivity predictions and provided insights into the mechanisms of action. We subsequently profiled the proteomic changes in nine cell lines (five TNBC and four non-TNBC) treated with EGFR/AKT/mTOR inhibitors. In TNBC, metabolism pathways were dysregulated after EGFR/mTOR inhibitor treatment, while RNA modification and cell cycle pathways were affected by AKT inhibitor. This systematic multi-omics and in-depth analysis of the proteome of BC cells can help prioritize potential therapeutic targets and provide insights into adaptive resistance in TNBC.

Computational modeling of ovarian cancer dynamics suggests optimal strategies for therapy and screening.

High-grade serous tubo-ovarian carcinoma (HGSC) is a major cause of cancer-related death. Treatment is not uniform, with some patients undergoing primary debulking surgery followed by chemotherapy (PDS) and others being treated directly with chemotherapy and only having surgery after three to four cycles (NACT). Which strategy is optimal remains controversial. We developed a mathematical framework that simulates hierarchical or stochastic models of tumor initiation and reproduces the clinical course of HGSC. After estimating parameter values, we infer that most patients harbor chemoresistant HGSC cells at diagnosis and that, if the tumor burden is not too large and complete debulking can be achieved, PDS is superior to NACT due to better depletion of resistant cells. We further predict that earlier diagnosis of primary HGSC, followed by complete debulking, could improve survival, but its benefit in relapsed patients is likely to be limited. These predictions are supported by primary clinical data from multiple cohorts. Our results have clear implications for these key issues in HGSC management.

Identification of antigenic epitopes recognized by tumor infiltrating lymphocytes in high grade serous ovarian cancer by multi-omics profiling of the auto-antigen repertoire.

Immunotherapeutic strategies aimed at enhancing tumor cell killing by tumor-specific T cells hold great potential for reducing tumor burden and prolonging survival of cancer patients. Although many potential tumor antigens have been described, identifying relevant targets when designing anti-cancer vaccines or targeted cell therapies remains a challenge. To identify novel, potentially immunogenic candidate tumor antigens, we performed integrated tumor transcriptomic, seromic, and proteomic analyses of high grade serous ovarian cancer (HGSC) patient tumor samples. We identified tumor neo-antigens and over-expressed antigens using whole exome and RNA sequencing and examined these in relation to patient-matched auto-antibody repertoires. Focusing on MHC class I epitopes recognized by CD8+ T cells, HLA-binding epitopes were identified or predicted from the highly expressed, mutated, or auto-antibody target antigen, or MHC-associated peptides (MAPs). Recognition of candidate antigenic peptides was assessed within the tumor-infiltrating T lymphocyte (TIL) population expanded from each patient. Known tumor-associated antigens (TAA) and cancer/testis antigens (CTA) were commonly found in the auto-antibody and MAP repertoires and CD8+ TILs recognizing epitopes from these antigens were detected, although neither expression level nor the presence of auto-antibodies correlated with TIL recognition. Auto-antibodies against tumor-mutated antigens were found in most patients, however, no TIL recognition of the highest predicted affinity neo-epitopes was detected. Using high expression level, auto-antibody recognition, and epitope prediction algorithms, we identified epitopes in 5 novel antigens (MOB1A, SOCS3, TUBB, PRKAR1A, CCDC6) recognized by HGSC patient TILs. Furthermore, selection of epitopes from the MAP repertoire identified 5 additional targets commonly recognized by multiple patient TILs. We find that the repertoire of TIL specificities includes recognition of highly expressed and immunogenic self-antigens that are processed and presented by tumors. These results indicate an ongoing autoimmune response against a range of self-antigens targeted by HGSC TILs.

The spliceosome is a therapeutic vulnerability in MYC-driven cancer.

MYC (also known as c-MYC) overexpression or hyperactivation is one of the most common drivers of human cancer. Despite intensive study, the MYC oncogene remains recalcitrant to therapeutic inhibition. MYC is a transcription factor, and many of its pro-tumorigenic functions have been attributed to its ability to regulate gene expression programs. Notably, oncogenic MYC activation has also been shown to increase total RNA and protein production in many tissue and disease contexts. While such increases in RNA and protein production may endow cancer cells with pro-tumour hallmarks, this increase in synthesis may also generate new or heightened burden on MYC-driven cancer cells to process these macromolecules properly. Here we discover that the spliceosome is a new target of oncogenic stress in MYC-driven cancers. We identify BUD31 as a MYC-synthetic lethal gene in human mammary epithelial cells, and demonstrate that BUD31 is a component of the core spliceosome required for its assembly and catalytic activity. Core spliceosomal factors (such as SF3B1 and U2AF1) associated with BUD31 are also required to tolerate oncogenic MYC. Notably, MYC hyperactivation induces an increase in total precursor messenger RNA synthesis, suggesting an increased burden on the core spliceosome to process pre-mRNA. In contrast to normal cells, partial inhibition of the spliceosome in MYC-hyperactivated cells leads to global intron retention, widespread defects in pre-mRNA maturation, and deregulation of many essential cell processes. Notably, genetic or pharmacological inhibition of the spliceosome in vivo impairs survival, tumorigenicity and metastatic proclivity of MYC-dependent breast cancers. Collectively, these data suggest that oncogenic MYC confers a collateral stress on splicing, and that components of the spliceosome may be therapeutic entry points for aggressive MYC-driven cancers.

A Genomically Characterized Collection of High-Grade Serous Ovarian Cancer Xenografts for Preclinical Testing.

High-grade serous ovarian cancer (HGSC) is the leading cause of morbidity and mortality from gynecologic malignant tumors. Overall survival remains low because of the nearly ubiquitous emergence of platinum resistance and the paucity of effective next-line treatments. Current cell culture-based models show limited similarity to HGSC and are therefore unreliable predictive models for preclinical evaluation of investigational drugs. This deficiency could help explain the low overall rate of successful drug development and the decades of largely unchanged approaches to HGSC treatment. We used gene expression, copy number variation, and exome sequencing analyses to credential HGSC patient-derived xenografts (PDXs) as effective preclinical models that recapitulate the features of human HGSC. Mice bearing PDXs were also treated with standard-of-care carboplatin therapy. PDXs showed similar sensitivity to carboplatin as the patient's tumor at the time of sampling. PDXs also recapitulated the diversity of genomic alterations (copy number variation and mutation profiles) previously described in large data sets that profiled HGSC. Furthermore, mRNA profiling showed that the PDXs represent all HGSC subtypes with the exception of the immunoreactive group. Credentialing of PDX models of HGSC should aid progress in HGSC research by providing improved preclinical models of HGSC that can be used to test novel targets and more accurately evaluate their likelihood of success.

Distinct GAB2 signaling pathways are essential for myeloid and lymphoid transformation and leukemogenesis by BCR-ABL1.

Tyrosine kinase inhibitors (TKIs) directed against BCR-ABL1, the product of the Philadelphia (Ph) chromosome, have revolutionized treatment of patients with chronic myeloid leukemia (CML). However, acquired resistance to TKIs is a significant clinical problem in CML, and TKI therapy is much less effective against Ph(+)B-cell acute lymphoblastic leukemia (B-ALL). BCR-ABL1, via phosphorylated Tyr177, recruits the adapter GRB2-associated binding protein 2 (GAB2) as part of a GRB2/GAB2 complex. We showed previously that GAB2 is essential for BCR-ABL1-evoked myeloid transformation in vitro. Using a genetic strategy and mouse models of CML and B-ALL, we show here that GAB2 is essential for myeloid and lymphoid leukemogenesis by BCR-ABL1. In the mouse model, recipients of BCR-ABL1-transducedGab2(-/-)bone marrow failed to develop CML-like myeloproliferative neoplasia. Leukemogenesis was restored by expression of GAB2 but not by GAB2 mutants lacking binding sites for its effectors phosphatidylinositol 3-kinase (PI3K) or SRC homology 2-containing phosphotyrosine phosphatase 2 (SHP2). GAB2 deficiency also attenuated BCR-ABL1-induced B-ALL, but only the SHP2 binding site was required. The SHP2 and PI3K binding sites were differentially required for signaling downstream of GAB2. Hence, GAB2 transmits critical transforming signals from Tyr177 to PI3K and SHP2 for CML pathogenesis, whereas only the GAB2-SHP2 pathway is essential for lymphoid leukemogenesis. Given that GAB2 is dispensable for normal hematopoiesis, GAB2 and its effectors PI3K and SHP2 represent promising targets for therapy in Ph(+)hematologic neoplasms.

The mammalian-membrane two-hybrid assay (MaMTH) for probing membrane-protein interactions in human cells.

Cell signaling, one of key processes in both normal cellular function and disease, is coordinated by numerous interactions between membrane proteins that change in response to stimuli. We present a split ubiquitin-based method for detection of integral membrane protein-protein interactions (PPIs) in human cells, termed mammalian-membrane two-hybrid assay (MaMTH). We show that this technology detects stimulus (hormone or agonist)-dependent and phosphorylation-dependent PPIs. MaMTH can detect changes in PPIs conferred by mutations such as those in oncogenic ErbB receptor variants or by treatment with drugs such as the tyrosine kinase inhibitor erlotinib. Using MaMTH as a screening assay, we identified CRKII as an interactor of oncogenic EGFR(L858R) and showed that CRKII promotes persistent activation of aberrant signaling in non-small cell lung cancer cells. MaMTH is a powerful tool for investigating the dynamic interactomes of human integral membrane proteins.

A negative genetic interaction map in isogenic cancer cell lines reveals cancer cell vulnerabilities.

Improved efforts are necessary to define the functional product of cancer mutations currently being revealed through large-scale sequencing efforts. Using genome-scale pooled shRNA screening technology, we mapped negative genetic interactions across a set of isogenic cancer cell lines and confirmed hundreds of these interactions in orthogonal co-culture competition assays to generate a high-confidence genetic interaction network of differentially essential or differential essentiality (DiE) genes. The network uncovered examples of conserved genetic interactions, densely connected functional modules derived from comparative genomics with model systems data, functions for uncharacterized genes in the human genome and targetable vulnerabilities. Finally, we demonstrate a general applicability of DiE gene signatures in determining genetic dependencies of other non-isogenic cancer cell lines. For example, the PTEN(-/-) DiE genes reveal a signature that can preferentially classify PTEN-dependent genotypes across a series of non-isogenic cell lines derived from the breast, pancreas and ovarian cancers. Our reference network suggests that many cancer vulnerabilities remain to be discovered through systematic derivation of a network of differentially essential genes in an isogenic cancer cell model.

COLT-Cancer: functional genetic screening resource for essential genes in human cancer cell lines.

Genome-wide pooled shRNA screens enable global identification of the genes essential for cancer cell survival and proliferation and provide a 'functional genetic' map of human cancer to complement genomic studies. Using a lentiviral shRNA library targeting approximately 16,000 human genes and a newly developed scoring approach, we identified essential gene profiles in more than 70 breast, pancreatic and ovarian cancer cell lines. We developed a web-accessible database system for capturing information from each step in our standardized screening pipeline and a gene-centric search tool for exploring shRNA activities within a given cell line or across multiple cell lines. The database consists of a laboratory information and management system for tracking each step of a pooled shRNA screen as well as a web interface for querying and visualization of shRNA and gene-level performance across multiple cancer cell lines. COLT-Cancer Version 1.0 is currently accessible at http://colt.ccbr.utoronto.ca/cancer.

Characterization of innate lymphoid cell subsets infiltrating melanoma and epithelial ovarian tumors.

The innate lymphoid cell (ILC) family is composed of heterogeneous innate effector and helper immune cells that preferentially reside in tissues where they promote tissue homeostasis. In cancer, they have been implicated in driving both pro- and anti-tumor responses. This apparent dichotomy highlights the need to better understand differences in the ILC composition and phenotype within different tumor types that could drive seemingly opposite anti-tumor responses. Here, we characterized the frequency and phenotype of various ILC subsets in melanoma metastases and primary epithelial ovarian tumors. We observed high PD-1 expression on ILC subsets isolated from epithelial ovarian tumor samples, while ILC populations in melanoma samples express higher levels of LAG-3. In addition, we found that the frequency of cytotoxic ILCs and NKp46+ILC3 in tumors positively correlates with monocytic cells and conventional type 2 dendritic cells, revealing potentially new interconnected immune cell subsets in the tumor microenvironment. Consequently, these observations may have direct relevance to tumor microenvironment composition and how ILC subset may influence anti-tumor immunity.

PD-1 regulates ILC3-driven intestinal immunity and homeostasis.

Interleukin-(IL) 22 production by intestinal group 3 innate lymphoid cells (ILC3) is critical to maintain gut homeostasis. However, IL-22 needs to be tightly controlled; reduced IL-22 expression is associated with intestinal epithelial barrier defect while its overexpression promotes tumor development. Here, using a single-cell ribonucleic acid sequencing approach, we identified a core set of genes associated with increased IL-22 production by ILC3. Among these genes, programmed cell death 1 (PD-1), extensively studied in the context of cancer and chronic infection, was constitutively expressed on a subset of ILC3. These cells, found in the crypt of the small intestine and colon, displayed superior capacity to produce IL-22. PD-1 expression on ILC3 was dependent on the microbiota and was induced during inflammation in response to IL-23 but, conversely, was reduced in the presence of Notch ligand. PD-1+ ILC3 exhibited distinct metabolic activity with increased glycolytic, lipid, and polyamine synthesis associated with augmented proliferation compared with their PD-1- counterparts. Further, PD-1+ ILC3 showed increased expression of mitochondrial antioxidant proteins which enable the cells to maintain their levels of reactive oxygen species. Loss of PD-1 signaling in ILC3 led to reduced IL-22 production in a cell-intrinsic manner. During inflammation, PD-1 expression was increased on natural cytotoxicity receptor (NCR)- ILC3 while deficiency in PD-1 expression resulted in increased susceptibility to experimental colitis and failure to maintain gut barrier integrity. Collectively, our findings uncover a new function of the PD-1 and highlight the role of PD-1 signaling in the maintenance of gut homeostasis mediated by ILC3 in mice.

Ex vivo activation of the GCN2 pathway metabolically reprograms T cells, leading to enhanced adoptive cell therapy.

The manipulation of T cell metabolism to enhance anti-tumor activity is an area of active investigation. Here, we report that activating the amino acid starvation response in effector CD8+ T cells ex vivo using the general control non-depressible 2 (GCN2) agonist halofuginone (halo) enhances oxidative metabolism and effector function. Mechanistically, we identified autophagy coupled with the CD98-mTOR axis as key downstream mediators of the phenotype induced by halo treatment. The adoptive transfer of halo-treated CD8+ T cells into tumor-bearing mice led to robust tumor control and curative responses. Halo-treated T cells synergized in vivo with a 4-1BB agonistic antibody to control tumor growth in a mouse model resistant to immunotherapy. Importantly, treatment of human CD8+ T cells with halo resulted in similar metabolic and functional reprogramming. These findings demonstrate that activating the amino acid starvation response with the GCN2 agonist halo can enhance T cell metabolism and anti-tumor activity.

Identification of acquired Notch3 dependency in metastatic Head and Neck Cancer.

During cancer development, tumor cells acquire changes that enable them to invade surrounding tissues and seed metastasis at distant sites. These changes contribute to the aggressiveness of metastatic cancer and interfere with success of therapy. Our comprehensive analysis of "matched" pairs of HNSCC lines derived from primary tumors and corresponding metastatic sites identified several components of Notch3 signaling that are differentially expressed and/or altered in metastatic lines and confer a dependency on this pathway. These components were also shown to be differentially expressed between early and late stages of tumors in a TMA constructed from over 200 HNSCC patients. Finally, we show that suppression of Notch3 improves survival in mice in both subcutaneous and orthotopic models of metastatic HNSCC. Novel treatments targeting components of this pathway may prove effective in targeting metastatic HNSCC cells alone or in combination with conventional therapies.

B Cells Promote T Cell Immunosenescence and Mammalian Aging Parameters.

A dysregulated adaptive immune system is a key feature of aging, and is associated with age-related chronic diseases and mortality. Most notably, aging is linked to a loss in the diversity of the T cell repertoire and expansion of activated inflammatory age-related T cell subsets, though the main drivers of these processes are largely unknown. Here, we find that T cell aging is directly influenced by B cells. Using multiple models of B cell manipulation and single-cell omics, we find B cells to be a major cell type that is largely responsible for the age-related reduction of naive T cells, their associated differentiation towards pathogenic immunosenescent T cell subsets, and for the clonal restriction of their T cell receptor (TCR). Accordingly, we find that these pathogenic shifts can be therapeutically targeted via CD20 monoclonal antibody treatment. Mechanistically, we uncover a new role for insulin receptor signaling in influencing age-related B cell pathogenicity that in turn induces T cell dysfunction and a decline in healthspan parameters. These results establish B cells as a pivotal force contributing to age-associated adaptive immune dysfunction and healthspan outcomes, and suggest new modalities to manage aging and related multi-morbidity.

Overproduction of IFNγ by Cbl-b-Deficient CD8+ T Cells Provides Resistance against Regulatory T Cells and Induces Potent Antitumor Immunity.

Regulatory T cells (Treg) are an integral component of the adaptive immune system that negatively affect antitumor immunity. Here, we investigated the role of the E3 ubiquitin ligase casitas B-lineage lymphoma-b (Cbl-b) in establishing CD8+ T-cell resistance to Treg-mediated suppression to enhance antitumor immunity. Transcriptomic analyses suggested that Cbl-b regulates pathways associated with cytokine signaling and cellular proliferation. We showed that the hypersecretion of IFNγ by Cbl-b-deficient CD8+ T cells selectively attenuated CD8+ T-cell suppression by Tregs. Although IFNγ production by Cbl-b-deficient T cells contributed to phenotypic alterations in Tregs, the cytokine did not attenuate the suppressive function of Tregs. Instead, IFNγ had a profound effect on CD8+ T cells by directly upregulating interferon-stimulated genes and modulating T-cell activation. In murine models of adoptive T-cell therapy, Cbl-b-deficient T cells elicited superior antitumor immune response. Furthermore, Cbl-b-deficient CD8+ T cells were less susceptible to suppression by Tregs in the tumor through the effects of IFNγ. Collectively, this study demonstrates that the hypersecretion of IFNγ serves as a key mechanism by which Cbl-b-deficient CD8+ T cells are rendered resistant to Tregs. See related Spotlight by Wolf and Baier, p. 370.

Ontogeny and Vulnerabilities of Drug-Tolerant Persisters in HER2+ Breast Cancer.

Resistance to targeted therapies is an important clinical problem in HER2-positive (HER2+) breast cancer. "Drug-tolerant persisters" (DTP), a subpopulation of cancer cells that survive via reversible, nongenetic mechanisms, are implicated in resistance to tyrosine kinase inhibitors (TKI) in other malignancies, but DTPs following HER2 TKI exposure have not been well characterized. We found that HER2 TKIs evoke DTPs with a luminal-like or a mesenchymal-like transcriptome. Lentiviral barcoding/single-cell RNA sequencing reveals that HER2+ breast cancer cells cycle stochastically through a "pre-DTP" state, characterized by a G0-like expression signature and enriched for diapause and/or senescence genes. Trajectory analysis/cell sorting shows that pre-DTPs preferentially yield DTPs upon HER2 TKI exposure. Cells with similar transcriptomes are present in HER2+ breast tumors and are associated with poor TKI response. Finally, biochemical experiments indicate that luminal-like DTPs survive via estrogen receptor-dependent induction of SGK3, leading to rewiring of the PI3K/AKT/mTORC1 pathway to enable AKT-independent mTORC1 activation. SIGNIFICANCE: DTPs are implicated in resistance to anticancer therapies, but their ontogeny and vulnerabilities remain unclear. We find that HER2 TKI-DTPs emerge from stochastically arising primed cells ("pre-DTPs") that engage either of two distinct transcriptional programs upon TKI exposure. Our results provide new insights into DTP ontogeny and potential therapeutic vulnerabilities. This article is highlighted in the In This Issue feature, p. 873.

Coenzyme A fuels T cell anti-tumor immunity.

Metabolic programming is intricately linked to the anti-tumor properties of T cells. To study the metabolic pathways associated with increased anti-tumor T cell function, we utilized a metabolomics approach to characterize three different CD8+ T cell subsets with varying degrees of anti-tumor activity in murine models, of which IL-22-producing Tc22 cells displayed the most robust anti-tumor activity. Tc22s demonstrated upregulation of the pantothenate/coenzyme A (CoA) pathway and a requirement for oxidative phosphorylation (OXPHOS) for differentiation. Exogenous administration of CoA reprogrammed T cells to increase OXPHOS and adopt the CD8+ Tc22 phenotype independent of polarizing conditions via the transcription factors HIF-1α and the aryl hydrocarbon receptor (AhR). In murine tumor models, treatment of mice with the CoA precursor pantothenate enhanced the efficacy of anti-PDL1 antibody therapy. In patients with melanoma, pre-treatment plasma pantothenic acid levels were positively correlated with the response to anti-PD1 therapy. Collectively, our data demonstrate that pantothenate and its metabolite CoA drive T cell polarization, bioenergetics, and anti-tumor immunity.

IL6 Induces an IL22+ CD8+ T-cell Subset with Potent Antitumor Function.

CD8+ T cells can be polarized into several different subsets as defined by the cytokines they produce and the transcription factors that govern their differentiation. Here, we identified the polarizing conditions to induce an IL22-producing CD8+ Tc22 subset, which is dependent on IL6 and the aryl hydrocarbon receptor transcription factor. Further characterization showed that this subset was highly cytolytic and expressed a distinct cytokine profile and transcriptome relative to other subsets. In addition, polarized Tc22 were able to control tumor growth as well as, if not better than, the traditional IFNγ-producing Tc1 subset. Tc22s were also found to infiltrate the tumors of human patients with ovarian cancer, comprising up to approximately 30% of expanded CD8+ tumor-infiltrating lymphocytes (TIL). Importantly, IL22 production in these CD8+ TILs correlated with improved recurrence-free survival. Given the antitumor properties of Tc22 cells, it may be prudent to polarize T cells to the Tc22 lineage when using chimeric antigen receptor (CAR)-T or T-cell receptor (TCR) transduction-based immunotherapies.

High expression of B7-H3 on stromal cells defines tumor and stromal compartments in epithelial ovarian cancer and is associated with limited immune activation.

BACKGROUND: B7-H3 and B7-H4 are highly expressed by many human malignancies making them attractive immunotherapeutic targets. However, their expression patterns and immune contexts in epithelial ovarian cancer have not been well characterized. METHODS: We used flow cytometry, immunohistochemistry, and genomic analyses to determine the patterns of B7-H3, B7-H4, and PD-L1 expression by tumor, stromal, and immune cells in the ovarian tumor microenvironment (TME). We analyzed immune cell frequency and expression of PD-1, TIM3, LAG3, ICOS, TIA-1, granzyme B, 2B4, CD107a, and GITR on T cells; CD20, CD22, IgD, BTLA, and CD27 on B cells; CD16 on monocytes; and B7-H3, B7-H4, PD-L1, PD-L2, ICOSL, CD40, CD86, and CLEC9a on antigen-presenting cells by flow cytometry. We determined intratumoral cellular location of immune cells using immunohistochemistry. We compared differences in immune infiltration in tumors with low or high tumor-to-stroma ratio and in tumors from the same or unrelated patients. RESULTS: On non-immune cells, B7-H4 expression was restricted to tumor cells whereas B7-H3 was expressed by both tumor and stromal cells. Stromal cells of the ovarian TME expressed high levels of B7-H3 compared to tumor cells. We used this differential expression to assess the tumor-to-stroma ratio of ovarian tumors and found that high tumor-to-stroma ratio was associated with increased expression of CD16 by monocytes, increased frequencies of PD-1high CD8+ T cells, increased PD-L1 expression by APCs, and decreased CLEC9a expression by APCs. We found that expression of PD-L1 or CD86 on APCs and the proportion of PD-1high CD4+ T cells were strongly correlated on immune cells from tumors within the same patient, whereas expression of CD40 and ICOSL on APCs and the proportion of PD-1high CD8+ T cells were not. CONCLUSIONS: This study provides insight into the expression patterns of B7-H3 and B7-H4 in the ovarian TME. Further, we demonstrate an association between the tumor-to-stroma ratio and the phenotype of tumor-infiltrating immune cells. We also find that some but not all immune parameters show consistency between peritoneal metastatic sites. These data have implications for the design of immunotherapies targeting these B7 molecules in epithelial ovarian cancer.