Robertson, Keith D.

Background and Aims: Disruption of the epigenome is a hallmark of human disease, including liver cirrhosis and HCC. While genetic heterogeneity is an established effector of pathologic phenotypes, epigenetic heterogeneity is less well understood. Environmental exposures alter the liver-specific DNA methylation landscape and influence the onset of liver cancer. Given that currently available treatments are unable to target frequently mutated genes in HCC, there is an unmet need for novel therapeutics to prevent or reverse liver damage leading to hepatic tumorigenesis, which the epigenome may provide. Approach and Results: We performed genome-wide profiling of DNA methylation, copy number, and gene expression from multiple liver regions from 31 patients with liver disease to examine their crosstalk and define the individual and combinatorial contributions of these processes to liver disease progression. We identified epigenetic heterogeneity hotspots that are conserved across patients. Elevated epigenetic heterogeneity is associated with increased gene expression heterogeneity. Cirrhotic regions comprise 2 distinct cohorts—one exclusively epigenetic, and the other where epigenetic and copy number variations collaborate. Epigenetic heterogeneity hotspots are enriched for genes central to liver function (eg, HNF1A) and known tumor suppressors (eg, RASSF1A). These hotspots encompass genes including ACSL1, ACSL5, MAT1A, and ELFN1, which have phenotypic effects in functional screens, supporting their relevance to hepatocarcinogenesis. Moreover, epigenetic heterogeneity hotspots are linked to clinical measures of outcome. Conclusions: Substantial epigenetic heterogeneity arises early in liver disease development, targeting key pathways in the progression and initiation of both cirrhosis and HCC. Integration of epigenetic and transcriptional heterogeneity unveils putative epigenetic regulators of hepatocarcinogenesis.

Abstract Background DNMT3A and TET2 are epigenetic regulator genes commonly mutated in age-related clonal hematopoiesis (CH). Despite having opposing epigenetic functions, these mutations are associated with increased all-cause mortality and a low risk for progression to hematological neoplasms. While individual impacts on the epigenome have been described using different model systems, the phenotypic complexity in humans remains to be elucidated. Results Here we make use of a natural inflammatory response occurring during coronavirus disease 2019 (COVID-19), to understand the association of these mutations with inflammatory morbidity and mortality. We demonstrate the age-independent, negative impact of DNMT3A mutant CH on COVID-19-related cytokine release severity and mortality. Using single cell proteogenomics we show that DNMT3A mutations involve myeloid and lymphoid cells. Using single cell multiomics sequencing, we identify cell-specific gene expression changes associated with DNMT3A mutations, along with significant epigenomic deregulation affecting enhancer accessibility, resulting in overexpression of IL32, a proinflammatory cytokine that can result in inflammasome activation in monocytes and macrophages. Finally, we show with single cell resolution that the loss of function of DNMT3A is directly associated with increased chromatin accessibility in mutant cells. Conclusions We demonstrate the negative prognostic impact of DNMT3Amt CH on COVID-19 related inflammatory morbidity and mortality. DNMT3Amt CH involves myeloid and lymphoid cells and in the context of COVID-19, was associated with inflammatory transcriptional priming, resulting in overexpression of IL32. This overexpression was secondary to increased chromatic accessibility, specific to DNMT3Amt CH cells. DNMT3Amt CH can serve as a potential biomarker for adverse inflammatory outcomes.

Abstract Clear cell renal carcinoma (ccRCC), which arises from the epithelial cells of the proximal tubule represents 70% of all renal cell carcinoma. Mutations that drive ccRCC in conjunction with near ubiquitous and early VHL loss include, PBRM1 (33 - 45%), BAP1 (15%) and SETD2 (13%). Loss of SETD2 leads to diminished occupancy of H3K36me3 in active gene bodies, which is progressively deregulated in primary RCC and metastasis, linking loss of H3K36me to tumor progression and metastasis. N6-Methyladenosine (m6A) is an important post-transcriptional regulator of RNA transcript stability and half-life. Preliminary findings have shown H3K36me3 (deposited by SETD2), targeting the core m6A methyltransferase machinery (METTL3, METTL14 and WTAP) to the mRNA by virtue of its interaction to METTL14 subunit. The goal of this study is to investigate if H3K36me3 depletion as observed in SETD2 mutant ccRCC, leads to a deregulated m6A methylome. We further investigate how deregulated methylome in key pathways leads to cancer initiation and progression in ccRCC. To accomplish our goals, we generated isogenic ccRCC cell lines with H3K36me3 depletion by knocking out SETD2. RNA-immunoprecipitation and sequencing for m6A (MeRIP) showed global deregulation of m6A methylome in SETD2KO isogenic ccRCC cell lines. We also show a number of pro-tumorigenic genes and pathways including cell invasion, inflammation and stemness being differentially methylated and expressed in SETD2KO ccRCC cell lines. Furthermore, we show SETD2KO cells being more sensitive to METTL3 inhibition suggesting the relevance of m6A hypermethylated transcripts in promoting pro-tumorigenic and growth promoting processes in SETD2KO ccRCC cell lines. Citation Format: Shafiq Shaikh, Ryan Wagner, Ryan Hlady, Keith D. Robertson. Perturbed RNA N6-methyladenosine methylome in SETD2 mutant clear cell renal carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7074.

Introduction: Recently, we demonstrated transcriptional down-regulation of hypertrophy pathways in myectomy tissue derived from patients with obstructive hypertrophic cardiomyopathy (HCM) despite translational activation of associated hypertrophy pathways. Previous studies of other forms of cardiac hypertrophy demonstrate hypertrophy pathways are activated via histone acetylation and histone deacetylase (HDAC) inhibitors prevented development of hypertrophy. Whether histone modifications regulate hypertrophy pathways in HCM remains unexplored. Methods: Chromatin immunoprecipitation sequencing (ChIP-seq) was performed on myectomy tissues from 40 patients with HCM and 23 controls to characterize differential histone marks for H3K4me3 (active promoter), H3K9ac (active promoter), H3K9me3 (heterochromatin), H3K27ac (active enhancer), H3K27me3 (repression) across the genome. The differential histone marks were integrated with RNA-sequencing (RNA-seq) data derived from the same patients. Results: Overall, we identified 2,912 (7%) differential H3K4me3 peaks impacting 5 pathways, 23,339 (21%) differential H3K9ac peaks impacting 87 pathways, 33 (0.05%) differential H3K9me3 peaks, and 58,837 (42%) differential H3K27ac peaks impacting 50 pathways. There was increased H3K27ac (q-value = 0.002) and H3K9ac (q-value = 0.009) at RAS-MAPK loci. Univariate analysis of concordance between H3K9ac with RNA-seq data showed activation of cardiac hypertrophy signaling (z-score=4.5; -log [BH p-value] = 3.1), while H3K27me showed down-regulation of cardiac hypertrophy signaling (z-score=-2.4; -log [BH p-value] = 2.4). Multivariate integration of the ChIP-seq data with RNA-seq data identified 3,189 transcripts concordant with all five histone marks impacting 254 pathways, 18 (7%) of which are involved in cardiac hypertrophy. Conclusions: Myectomy tissue from patients with obstructive HCM shows enrichment of histone modifications at cardiac hypertrophy loci with activation via H3K27 and H3K9 acetylation and repression via H3K27 methylation. Further studies are needed to understand the balance and temporal relationship between these competing regulators in HCM pathobiology and the potential for epigenetic therapies.

Abstract Genes encoding SWI/SNF chromatin remodeling complex subunits are recurrently mutated across human cancers. In cholangiocarcinoma (CCA), AT-rich interaction domain-1A (ARID1A) and polybromo-1 (PBRM1) loss of function mutations occur in approximately 18 and 10% of cases respectively. Yet our mechanistic understanding and ability to therapeutically exploit ARID1A and PBRM1 deficiency are immature. Genetically accurate preclinical models are needed to define the role of ARID1A and PBRM1 and test therapeutic approaches in CCA. Analysis of human CCA tumors showed that loss of function mutations in ARID1A and PBRM1 frequently co-occur with each other and with PI3K/AKT pathway activation. Therefore, we developed novel murine models of cholangiocarcinoma initiated by Arid1a knockout or dual Arid1a/Pbrm1 knockout in the context of AKT activation, employing a surgical biliary transfection method for somatic gene editing with CRISPR/Cas9. Mice with cholangiocyte Arid1a knockout and AKT activation (AAC) or dual Arid1a/Pbrm1 knockout and AKT activation (APAC) formed numerous liver tumors within 8 months. Histologically, tumors were predominantly CCA and mixed HCC/CCA, with a similar spectrum of tumors across AAC and APAC mice. Gene editing with CRISPR/Cas9 was confirmed at the DNA level, and loss of protein expression was confirmed in tumor cells by immunohistochemistry. Similar to human CCA, many tumors were densely infiltrated with cancer-associated fibroblasts and a variable degree of CD45+ immune cells. To generate scalable models for preclinical testing, tumors from AAC and APAC mice were dissociated to create cell lines for in vitro studies and propagated by orthotopic transplantation for in vivo studies. Cell lines and orthotopic tumors were validated and retained the genetic features and histologic characteristics of the parental tumors. Collectively these studies have produced a suite of novel, well-characterized, genetically relevant mouse models of epigenetic dysregulation in CCA due to ARID1A and PBRM1 deficiency, which can be used broadly for discovery research and preclinical testing of novel treatment approaches. Citation Format: Caitlin B. Conboy, Jennifer L. Tomlinson, Ryan D. Watkins, Jayla T. Millender, Danielle M. Carlson, Rory L. Smoot, Keith D. Robertson, Greg J. Gores. Novel murine models of ARID1A and PBRM1-deficient cholangiocarcinoma for preclinical discovery and development [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr A048.

There is a paucity of human in vitro models available to study the effects of alcohol in hepatocellular carcinoma (HCC). HCC cell lines are often deficient in either alcohol dehydrogenase (ADH) or cytochrome P450 2E1 (CYP2E1). VL‐17a is currently the most widely studied model available. VL‐17a is derived from a HepG2 cell line engineered to overexpress mouse ADH and human CYP2E1. LH86 is a novel HCC cell line previously developed in our laboratory. Here we compare the LH86 cell line with the current model to identify a novel tool to study the effects of alcohol in HCC. The aim of this study is to evaluate the LH86 cell line as a novel human in vitro model to study the effects of alcohol in HCC. The cell lines were cultured with alcohol and gene expression and epigenomic changes were identified. The LH86 cell line was observed to have unique functional properties indicating that it is a potential novel tool to study the effects of alcohol in HCC. Support or Funding Information R01 AA19976 TO KR AND CL

Histone‐lysine N‐methyltransferase SETD2 (SETD2), the sole histone methyltransferase that catalyzes trimethylation of lysine 36 on histone H3 (H3K36me3), is often mutated in clear cell renal cell carcinoma (ccRCC). SETD2 mutation and/or loss of H3K36me3 is linked to metastasis and poor outcome in ccRCC patients. Epithelial‐to‐mesenchymal transition (EMT) is a major pathway that drives invasion and metastasis in various cancer types. Here, using novel kidney epithelial cell lines isogenic for SETD2, we discovered that SETD2 inactivation drives EMT and promotes migration, invasion, and stemness in a transforming growth factor‐beta‐independent manner. This newly identified EMT program is triggered in part through secreted factors, including cytokines and growth factors, and through transcriptional reprogramming. RNA‐seq and assay for transposase‐accessible chromatin sequencing uncovered key transcription factors upregulated upon SETD2 loss, including SOX2, POU2F2 (OCT2), and PRRX1, that could individually drive EMT and stemness phenotypes in SETD2 wild‐type (WT) cells. Public expression data from SETD2 WT/mutant ccRCC support the EMT transcriptional signatures derived from cell line models. In summary, our studies reveal that SETD2 is a key regulator of EMT phenotypes through cell‐intrinsic and cell‐extrinsic mechanisms that help explain the association between SETD2 loss and ccRCC metastasis.

Abstract Clear cell renal cell carcinoma (ccRCC) is the most common subtype of renal cell carcinoma (RCC) and accounts for majority of the deaths from renal cancer. SETD2, the sole histone methyltransferase that catalyzes trimethylation of lysine 36 on histone H3 (H3K36me3), is frequently mutated in ccRCC. Studies have shown that SETD2 mutations are often subclonal and the mutation rate dramatically increases from 25% in primary ccRCC to more than 60% in metastatic ccRCC, suggesting SETD2 loss plays an important role in renal cancer metastasis. However, mechanisms underpinning how SETD2 mutation drives invasion and metastasis in ccRCC is largely unknown. Here, by performing RNA-seq and ATAC-seq in a non-tumorigenic renal epithelial cell line isogenic for SETD2 loss-of-function, we found that SETD2 inactivation induces dramatic changes in the transcriptome and drastically enhances chromatin accessibility, which allows binding of a variety of transcription factors. We also discovered that SETD2 loss induces a TGFβ-independent EMT program, which is a major pathway linked to metastasis. Correlation of RNA-seq with ATAC-seq identifies SOX2, OCT2 and PRRX1 as key transcription factors promoting the SETD2 loss-driven EMT program. Furthermore, we found that this distinct EMT program is in part triggered by secreted factors, which suggests that SETD2 mutation promotes a metastasis-permissive tumor microenvironment through paracrine signaling. In summary, our findings reveal that SETD2 is a key regulator of EMT phenotypes that may explain the role of SETD2 mutations in ccRCC metastasis and further uncover potential therapeutic targets for SETD2-mutant ccRCC patients. Citation Format: Tianchu Wang, Ryan Wagner, Ryan Hlady, Sungho Kim, Xiaoyu Pan, Thai Ho, Keith Robertson. SETD2 loss in renal epithelial cells induces epithelial to mesenchymal transition in a TGFβ-independent manner [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1226.

Abstract Background: COVID-19 causes significant morbidity and mortality, albeit with considerable heterogeneity among affected individuals. It remains unclear which host factors determine disease severity and survival. Given the propensity of clonal hematopoiesis (CH) to promote inflammation in healthy individuals, we investigated its effect on COVID-19 outcomes. Methods: We performed a multi-omics interrogation of the genome, epigenome, transcriptome, and proteome of peripheral blood mononuclear cells from COVID-19 patients (n=227). We obtained clinical data, laboratory studies, and survival outcomes. We determined CH status and TET2-related DNA methylation. We performed single-cell proteogenomics to understand clonal composition in relation to cell phenotype. We interrogated single-cell gene expression in isolation and in conjunction with DNA accessibility. We integrated these multi-omics data to understand the effect of CH on clonal composition, gene expression, methylation of cis-regulatory elements, and lineage commitment in COVID-19 patients. We performed shRNA knockdowns to validate the effect of one candidate transcription factor in myeloid cell lines. Results: The presence of CH was strongly associated with COVID-19 severity and all-cause mortality, independent of age (HR 3.48, 95% CI 1.45-8.36, p=0.005). Differential methylation of promoters and enhancers was prevalent in TET2-mutant, but not DNMT3A-mutant CH. TET2-mutant CH was associated with enhanced classical/intermediate monocytosis and single-cell proteogenomics confirmed an enrichment of TET2 mutations in these cell types. We identified cell-type specific gene expression changes associated with TET2 mutations in 102,072 single cells (n=34). Single-cell RNA-seq confirmed the skewing of hematopoiesis towards classical and intermediate monocytes and demonstrated the downregulation of EGR1 (a transcription factor important for monocyte differentiation) along with up-regulation of the lncRNA MALAT1 in monocytes. Combined scRNA-/scATAC-seq in 43,160 single cells (n=18) confirmed the skewing of hematopoiesis and up-regulation of MALAT1 in monocytes along with decreased accessibility of EGR1 motifs in known cis-regulatory elements. Using myeloid cell lines for functional validation, shRNA knockdowns of EGR1 confirmed the up-regulation of MALAT1 (in comparison to wildtype controls). Conclusions: CH is an independent prognostic factor in COVID-19 and skews hematopoiesis towards monocytosis. TET2-mutant CH is characterized by differential methylation and accessibility of enhancers binding myeloid transcriptions factors including EGR1. The ensuing loss of EGR1 expression in monocytes causes MALAT1 overexpression, a factor known to promote monocyte differentiation and inflammation. These data provide a mechanistic insight to the adverse prognostic impact of CH in COVID-19. Citation Format: Moritz Binder, Terra L. Lasho, Wazim Mohammed Ismail, Nana A. Ben-Crentsil, Jenna A. Fernandez, Minsuk Kim, Susan M. Geyer, Amelia Mazzone, Christy M. Finke, Abhishek A. Mangaonkar, Jeong-Heon Lee, Kwan Hyun Kim, Vernadette A. Simon, Fariborz Rakhshan Rohakthar, Amik Munankarmy, Susan M. Schwager, Jonathan J. Harrington, Melissa R. Snyder, Nathalie M. Droin, Eric Solary, Keith D. Robertson, Eric D. Wieben, Eric Padron, Nicholas Chia, Alexandre Gaspar-Maia, Mrinal M. Patnaik. Enhancer deregulation inTET2-mutant clonal hematopoiesis is associated with increased COVID-19 severity and mortality [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 806.

Disrupted liver regeneration following hepatectomy represents an “undruggable” clinical challenge associated with poor patient outcomes. Yes-associated protein (YAP), a transcriptional coactivator that is repressed by the Hippo pathway, is instrumental in liver regeneration. We have previously described an alternative, Hippo-independent mechanism of YAP activation mediated by downregulation of protein tyrosine phosphatase nonreceptor type 11 (PTPN11, also known as SHP2) inhibition. Herein, we examined the effects of YAP activation with a selective SHP1/SHP2 inhibitor, NSC-87877, on liver regeneration in murine partial hepatectomy models. In our studies, NSC-87877 led to accelerated hepatocyte proliferation, improved liver regeneration, and decreased markers of injury following partial hepatectomy. The effects of NSC-87877 were lost in mice with hepatocyte-specific Yap/Taz deletion, and this demonstrated dependence on these molecules for the enhanced regenerative response. Furthermore, administration of NSC-87877 to murine models of nonalcoholic steatohepatitis was associated with improved survival and decreased markers of injury after hepatectomy. Evaluation of transcriptomic changes in the context of NSC-87877 administration revealed reduction in fibrotic signaling and augmentation of cell cycle signaling. Cytoprotective changes included downregulation of Nr4a1 , an apoptosis inducer. Collectively, the data suggest that SHP2 inhibition induces a pro-proliferative and cytoprotective enhancement of liver regeneration dependent on YAP.