Merkel cell carcinoma (MCC) is a poorly differentiated neuroendocrine carcinoma with limited treatment options, primarily immunotherapy, to which only ~50% of patients respond. Lineage plasticity drives its poorly differentiated phenotype, in turn promoting tumor aggressiveness and treatment resistance. Targeting the mechanisms underlying lineage plasticity could help induce differentiation, reduce proliferation, and potentially sensitize the tumor to existing therapies, yet such strategies are underdeveloped in MCC. Here, we integrate scRNA-seq and bulk ATAC-seq data to generate Boolean networks, simulate their dynamics, and predict a key regulator of differentiation, which we validated in vitro. Using CytoTRACE2 across two independent datasets, we revealed the existence of tumor subpopulations with distinct developmental potency states. We then constructed and refined transcription factor regulatory networks using BooleaBayes and expanded them with ATAC-seq inferred regulatory interactions. Across multiple network constructions, simulations of single-gene perturbations consistently identified the Notch effector RBPJ as the key regulator predicted to shift MCC cells toward a more differentiated state. Experimental knockdown of RBPJ in an MCC cell line altered expression of differentiation-associated genes, reduced the expression of MCC markers, and drastically reduced cell growth. These findings identify RBPJ as a regulator of MCC lineage plasticity and candidate for targeted treatment, while highlighting the utility of probabilistic network modeling for prioritizing therapeutic targets in translational cancer research.

e15073 Background: Colorectal cancer (CRC) remains a significant clinical challenge due to its intrinsic heterogeneity, the paucity of personalized treatment options, and the commonality of chemoresistance. Methods: To better understand drug response dynamics at the single-cell level, we have developed a high-throughput, image-based phenotypic profiling pipeline using CRC patient-derived organoid (PDO) monolayer cultures. Diverse CRC PDOs representing multiple Consensus Molecular Subtypes were treated with a combination of 5-Fluorouracil and Oxaliplatin (FOX), imaged utilizing a biomarker panel identified using transcriptomic data, and analyzed using our newly developed phenotypic profiling pipeline. To further elucidate individual tumor response in vivo, we have also developed protocols to implant our PDOs into mice to create Patient Derived Xenograft (PDX) models. Results: Using our new analysis pipeline, we have identified a subpopulation of CRC cells that, after acute exposure to FOX treatment, exhibit elevated AKT signaling and increased expression of cancer stem cell markers. We observe this subpopulation across multiple PDOs, including across distinct CRC molecular subtypes, suggesting that chemotherapy itself may contribute to the enrichment of drug-resistant phenotypes driven by PI3K/AKT survival signaling. Targeting this survival phenotype, we also demonstrate that transient pre-treatment with the clinically utilized PI3K/mTOR dual inhibitor Dactolisib effectively sensitizes CRC PDOs to FOX, synergizing with this standard-of-care chemotherapy regimen and reducing the fraction of chemoresistant cells. Synergy score profiling revealed that this combination was broadly effective across diverse PDOs, with the strongest response noted in the organoid exhibiting the highest baseline AKT signaling. Preliminary data in our PDX models suggest this combination is synergistic in vivo as well. Conclusions: Our findings highlight the power of personalized organoid-based phenotypic profiling for dissecting molecular mechanisms of therapeutic resistance and support the rationale for transient PI3K/mTOR inhibition as a strategy to improve CRC treatments and outcomes.

Motivation:Gene regulatory networks undergo dynamic restructuring during development and disease. Identifying when and how these networks change is crucial for understanding developmental and disease transitions, yet existing change-point detection methods often ignore network structure or lack interpretable community assignments. Results:We present PARROT (Phase-Altering Regulatory Rewiring Over Time), a framework for detecting change-points in dynamic networks using Stochastic Block Models. PARROT jointly estimates change-point locations and community structure across four network classes: unipartite and bipartite with either Gaussian or Bernoulli edge models. Simulations demonstrate improved performance and community recovery compared to other methods. Applications to human cardiac differentiation and mouse lung development data successfully recovered known phase boundaries. PARROT identifies both which genes are reassigned across modules and how the connections change between states. Availability:PARROT is available as an R package at https://github.com/cchen22/PARROT. Contact:chenchen9945@gmail.com. Supplementary information:Supplementary data are available at Bioinformatics online.

Chromosomal instability (CIN) is a hallmark of prostate cancer that strongly correlates with metastatic burden and appears prominently in both primary cancer and metastatic disease. Low Gleason score primary prostate tumors display pervasive centrosome loss, a known mechanistic driver of CIN, that disrupts normal spindle assembly and increases mitotic errors. Previously, we found that transient depletion of centrosomes in immortalized, non-tumorigenic prostate epithelial cells (PrEC) induced a burst of CIN, generating cell lines capable of forming xenograft tumors. We used a multi-omics approach that integrates genomic and transcriptomic data to identify the oncogenic alteration signatures caused by transient centrosome loss. We identified a consensus set of focal copy-number variations (CNVs) induced by centrosome loss in cultured cells that are also detectable within a subset of samples from a prostate cancer patient cohort. Using this CNV signature, we were able to derive a unique transcriptomic signature (CIN9) from prostate cancer patient samples that showed strong predictive value for adverse clinical outcomes. Our experimental system uses centrosome loss to promote a punctuated burst of genomic crisis that is characteristic of genome evolution during prostate cancer progression. Consequently, this prostate cancer model produced recurrent structural variations that are detectable in patient samples and associate with worse outcomes.

BackgroundLung adenocarcinoma shows distinct differences between males and females in incidence, prognosis, and treatment response, suggesting unique molecular mechanisms that remain underexplored. This study aims to identify sex-specific molecular signatures and therapeutic targets in lung adenocarcinoma using multi-omics approaches to inform personalized treatment strategies.MethodsWe conducted an integrative analysis of transcriptomic and proteomic data from the Clinical Proteomic Tumor Analysis Consortium (CPTAC) and The Cancer Genome Atlas (TCGA) datasets, comparing male and female lung adenocarcinoma profiles. Transcription factor activity was assessed using TIGER on gene expression data, while kinase activity was evaluated with PTM-SEA on proteomic data. These results were combined to build a kinase-transcription factor signaling network. Potential sex-specific drugs were identified using the PRISM drug screening database.ResultsThe analysis revealed significant sex-based differences in transcription factor and kinase activity. Notably, NR3C1, AR, and AURKA exhibited sex-biased expression and activity. The constructed signaling network highlighted druggable pathways linked to cancer-related processes, with distinct profiles in males and females. PRISM screening identified glucocorticoid receptor agonists and aurora kinase inhibitors as promising sex-specific therapeutic candidates.ConclusionsOur findings underscore the importance of considering sex differences in lung adenocarcinoma molecular profiles. The integration of transcriptomic and proteomic data reveals sex-specific pathways and potential therapies, paving the way for personalized treatment approaches tailored to male and female patients.

The human cytomegalovirus (HCMV) UL135 and UL138 genes play opposing roles regulating latency and reactivation in CD34+ human progenitor cells. We designed an RNA sequencing study to compare the transcriptional profile of HCMV infection in the presence and absence of these genes using the Tohoku Hospital Pediatrics-1 (THP-1) monocytic cell line model for latency. Relative to primary cell models, THP-1 cells offer the strength of a homogenous population that uniformly silences gene expression and will synchronously reexpress viral genes following stimulation to differentiate, which models early phases of viral reactivation. The loss of UL138 resulted in elevated levels of viral gene expression and in spontaneous adhesion of distinct cell populations that support HCMV gene expression and genome synthesis. The loss of UL135 resulted in diminished viral gene expression during an initial burst that occurs as latency is established and in no expression of eleven viral genes from the ULb' region even following differentiation and reexpression of viral genes. Transcriptional network analysis revealed host transcription factors (TFs) with potential to regulate the ULb' genes in coordination with pUL135. We show that the cellular TF peroxisome proliferator-activated receptor gamma binds to the viral genome and influences the expression of UL133-UL138 locus genes. Our results define roles for UL135 and UL138 in regulation of patterns of viral gene expression for the establishment of latency and reexpression of viral genes for reactivation and reveal insights into differentiation-linked mechanisms of transcriptional control of the HCMV genome.

Merkel Cell Carcinoma (MCC) is an aggressive neuroendocrine cutaneous malignancy arising from either ultraviolet-induced mutagenesis or Merkel cell polyomavirus (MCPyV) integration. Despite extensive research, our understanding of the molecular mechanisms driving the transition from normal cells to MCC remains limited. To address this knowledge gap, we assessed the impact of inducible MCPyV T antigens on normal human fibroblasts by performing RNA-seq. Our data uncovered changes in expression and regulation of Wnt signaling pathway members. Building on this observation, we bioinformatically evaluated various Wnt pathway perturbagens for their ability to reverse the MCC gene expression signature and identified pyrvinium pamoate, an FDA-approved anthelminthic drug known for its antitumor activity in other cancers. Leveraging transcriptomic, network, and molecular analyses, we found that pyrvinium targets multiple MCC vulnerabilities. Pyrvinium not only reverses the neuroendocrine features of MCC by modulating canonical and noncanonical Wnt signaling but also inhibits cancer cell growth by activating p53-mediated apoptosis, disrupting mitochondrial function, and inducing endoplasmic reticulum stress. Finally, we demonstrated that pyrvinium reduces tumor growth in an MCC mouse xenograft model. These findings offer a deeper understanding of the role of Wnt signaling in MCC and highlight the utility of pyrvinium as a potential treatment for MCC.

The E7 oncoprotein of mouse papillomavirus (MmuPV1) plays a pivotal role in both viral infection and cancer development. While earlier studies have identified key cellular targets of MmuPV1 E7, such as pRB and PTPN14, the broader impact of MmuPV1 E7 on keratinocyte homeostasis and shared activities with human papillomavirus (HPV) E7 remains unclear. In this study, we employed proteomic and transcriptomic analyses using our established mouse keratinocyte model-previously instrumental in uncovering a novel function of MmuPV1 E6-to investigate the biological consequences of MmuPV1 E7 expression in mouse keratinocytes. Our findings reveal that MmuPV1 E7 induces cellular changes reminiscent of those driven by "high-risk" HPV infection implicated in cervical cancer. Notably, MmuPV1 E7 did not activate canonical E2F-responsive gene expression or promote proliferation, reinforcing the idea that MmuPV1 E6 is the primary driver of cell cycle activation. However, MmuPV1 E7 expression led to a significant accumulation of stress keratin 17, a marker associated with immune evasion and elevated in both HPV16 transgenic models and MmuPV1 infections. Additionally, we observed enhanced PI3K-AKT-mTOR signaling, with increased levels of phosphorylated S6 kinase and heightened sensitivity to epidermal growth factor stimulation. Collectively, these results underscore the role of MmuPV1 E7 in promoting oncogenic phenotypes and highlight its relevance as a model for studying the molecular underpinnings of "high-risk" HPV-driven disease. IMPORTANCE:In this study, we determined the ability of the MmuPV1 E7 oncoprotein in promoting disruption of keratinocyte homeostasis in mouse keratinocytes. Using a multiomics approach, we observed that MmuPV1 E7 promoted several phenotypes associated with "high-risk" human papillomavirus (HPV) infection. Specifically, we confirmed that MmuPV1 E7 does not increase E2F-responsive gene expression and proliferation of mouse keratinocytes. We did find that MmuPV1 E7 was able to increase the expression of stress keratin 17, which promotes immune evasion in papillomavirus infections. Finally, MmuPV1 E7 showed increased expression of genes associated with PI3K-AKT-mTOR signaling. Consistent with this observation, MmuPV1 E7-expressing mouse keratinocytes had elevated phosphorylation of S6 kinase. We also found that MmuPV1 E7 potentiates this signaling through increased sensitivity to epidermal growth factor stimulation. Our collective data show that MmuPV1 E7 promotes several phenotypes associated with "high-risk" HPV infection and cancers.

Lung adenocarcinoma (LUAD) exhibits differences between the sexes in incidence, prognosis, and therapy, suggesting underexplored molecular mechanisms. We conducted an integrative multi-omics analysis using the Clinical Proteomic Tumor Analysis Consortium (CPTAC) and The Cancer Genome Atlas (TCGA) datasets to contrast transcriptomes and proteomes between sexes. We used TIGER to analyze TCGA-LUAD expression data and found sex-biased activity of transcription factors (TFs); we used PTM-SEA with CPTAC-LUAD proteomics data and found sex-biased kinase activity. We combined these to construct a kinase-TF signaling network and discovered druggable pathways linked to cancer-related processes. We also found significant sex biases in clinically relevant TFs and kinases, including NR3C1, AR, and AURKA. Using the PRISM drug screening database, we identified potential sex-specific drugs, such as glucocorticoid receptor agonists and aurora kinase inhibitors. Our findings emphasize the importance of considering sex and using multi-omics network methods to discover personalized cancer therapies.

HCMV genes UL135 and UL138 play opposing roles regulating latency and reactivation in CD34+ human progenitor cells (HPCs). Using the THP-1 cell line model for latency and reactivation, we designed an RNA sequencing study to compare the transcriptional profile of HCMV infection in the presence and absence of these genes. The loss of UL138 results in elevated levels of viral gene expression and increased differentiation of cell populations that support HCMV gene expression and genome synthesis. The loss of UL135 results in diminished viral gene expression during an initial burst that occurs as latency is established and no expression of eleven viral genes from the ULb' region even following stimulation for differentiation and reactivation. Transcriptional network analysis revealed host transcription factors with potential to regulate the ULb' genes in coordination with pUL135. These results reveal roles for UL135 and UL138 in regulation of viral gene expression and potentially hematopoietic differentiation.

Transcriptional regulation plays a crucial role in determining cell fate and disease, yet inferring the key regulators from gene expression data remains a significant challenge. Existing methods for estimating transcription factor (TF) activity often rely on static TF-gene interaction databases and cannot adapt to changes in regulatory mechanisms across different cell types and disease conditions. Here, we present a new algorithm - Transcriptional Inference using Gene Expression and Regulatory data (TIGER) - that overcomes these limitations by flexibly modeling activation and inhibition events, up-weighting essential edges, shrinking irrelevant edges towards zero through a sparse Bayesian prior, and simultaneously estimating both TF activity levels and changes in the underlying regulatory network. When applied to yeast and cancer TF knock-out datasets, TIGER outperforms comparable methods in terms of prediction accuracy. Moreover, our application of TIGER to tissue- and cell-type-specific RNA-seq data demonstrates its ability to uncover differences in regulatory mechanisms. Collectively, our findings highlight the utility of modeling context-specific regulation when inferring transcription factor activities.

Abstract African Americans (AAs) have among the highest incidence and mortality rates of colorectal cancer (CRC) in the US. They present with more right-sided, microsatellite stable (MSS) tumors and are diagnosed at earlier ages than non-Hispanic Whites (NHW). While DNA methylation changes and their significance have been previously described in CRC, much less is known about the unique DNA methylation changes that occur in AA CRCs compared to NHW CRCs. In the current study, we analyzed DNA methylation changes in AA patients and compared those changes to data from NHWs. AA patients diagnosed with a MSS colorectal tumor at the Yale New Health System were included in the study. A pathologist reviewed archival slides, and selected areas from CRCs (n=160) and adjacent normal tissue (n=42) were cored (median age=59; 82 females, 78 males). DNA was bisulfite converted and analysis performed using EPIC arrays. The 450K methylation array from NHW CRCs was also analyzed for 217 CRCs and 20 normal tissues obtained from TCGA (median age=68; 103 females, 114 males). Differentially methylated regions (DMRs) with |Δβ|>0.1, FDR corrected p-value <0.05 between tumor and normal tissue were identified using SeSAMe. We identified 4881 DMRs (57% hyper- and 44% hypomethylated) between AA CRC and normal tissue. Of these, 46% were in CpG Islands, 15% in CpG Shores, 4% in CpG shelves, and 36% in open sea. Most DMRs (71%) were in promoter regions, with 18% in gene bodies, 5% in enhancer regions, and 6% in intergenic regions. Next, we performed additional DMR analysis using overlapping probes from both arrays to compare NHW with AA data. We identified 2239 DMRs between AA CRC and normal tissue compared to 4138 in NHW. Of the 2239 AA DMRs, 1419 (66%) were common with NHW DMRs, 790 (34%) were unique to AAs. There were many overlapping pathways and genes; however, the DMRs impacting these pathways differed between AA and NHW CRCs. Hypermethylation of LCK, a tyrosine kinase involved in T-cell receptor signaling, was found in 60% of AA CRCs compared to 5.9% of NHW CRCs. LCK hypermethylation leads to CD4/8 T-cell inactivation and linked to poor outcomes in lung cancer. We observed AA-specific hypermethylation of SSTR1 in 57.2% of AA CRCs compared to 27.6% of NHW CRCs. Somatostasin (SST) suppresses tumor growth by regulating certain factors involved in cell proliferation. Hypermethylation of SSTR1 has been associated with various malignancies including gastric cancer. Lastly, promoter hypermethylation of 3 protocadherin genes was found in 65.1% of AAs compared to 33.1% NHWs. Protocadherins regulate WNT signaling and hypermethylation is associated with poor outcomes in multiple cancers. Taken together, these results show AA-specific CRC methylation changes compared with NHW CRCs. Further work is warranted to investigate the potential functional and clinical implications of these methylation alterations, some of which may represent therapeutic targets and biomarkers. Citation Format: Seeta Rajpara, David N. Buckley, Rosa Xicola, Reger Mikaeel, Mary K. Yagle, Baris Kerimoglu, Curtis Thorne, Dante A. Bellomo, Megha Padi, Xavier Llor, Nathan Ellis, Bodour Salhia. DNA Methylation analysis of African American colorectal cancers reveal race-specific alterations [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 7004.

Abstract Background: The incidence of early-onset colorectal cancer (EO-CRC) has been alarmingly increasing, particularly among non-Hispanic Whites (NHWs), resulting in similar rates to ones in African Americans (AAs). Overall, the median age of diagnosis has decreased from 72 to 66 years. The molecular features that could potentially distinguish this phenotype are still not well-established. This study aimed to identify genetic, epigenetic, and transcriptomic factors associated with EO-CRC. Methods: 510 patients with microsatellite-stable tumors analyzed for somatic mutations, targeting 20 cancer driver genes. We included a cohort of 175 AAs from the Yale New Haven Health System, a previously studied cohort of 51 AAs from the Chicago Colorectal Cancer Consortium, and the publicly available molecular data from TCGA COAD-READ PanCancer cohort (50 AAs and 229 NHWs). We analyzed the transcriptomic and methylation profiles from the Yale cohort using Tempo-SeqTM and Illumina EPIC array, respectively. Results: Of the 510 patients, 18% were diagnosed with EO-CRC. Using a multiple variant analysis, adjusting for sex, race, tumor location, and stage, EO-CRC patients were less likely to be male (OR = 0.59, 95% CI: 0.37-0.94, p-value = 0.03), their tumors lacked mutations in APC (OR = 0.37, 95% CI: 0.22-0.63, p-value = 0.0002), and they were more likely to have mutations in FBXW7 (OR = 3.1, 95% CI: 1.60-6.00, p-value = 0.0007) and BCL9L (OR = 3.1, 95% CI: 1.00-8.72, p-value = 0.04). We compared tumors with and without mutations in FBXW7 or BCL9L in two independently analyses. Combining transcriptomics and methylation data, we identified 10 candidate genes that showed significant gene downregulation with hypermethylation or upregulation with hypomethylation. Among FBXW7 mutant tumors, we found the Cartilage oligomeric matrix protein (COMP) gene was significantly downregulated and hypermethylated. In CRC, COMP has been associated with a crosstalk between the TGF-β pathway and immune cell infiltration. Moreover, inactivation of FBXW7 resulted in the accumulation of phosphorylated TGIF1 molecules and repression of TGFβ-dependent transcription in cancer cell lines. Deficiency of FBXW7 is a key element in the phosphorylation-dependent ubiquitination and subsequent proteasome degradation of oncoproteins, such as c-MYC, NOTCH, and Cyclin E. Among BCL9L mutant tumors, keratin 20 (KRT20), the intermediate filament protein and marker of intestinal differentiation in CRC, was downregulated and hypermethylated. BCL9L plays an important role in tumorigenesis induced by aberrant Wnt signaling or aneuploidy tolerance. Conclusion: Our findings further demonstrate the importance of specific etiologic mechanisms in EO-CRC at the level of somatic alterations, providing new ways to understand the biology of EO-CRC and prospects for the development of targeted therapies. Citation Format: Reger Mikaeel, Dante Bellomo, Seeta Rajpara, David Buckley, Mary Yagle, Baris Kerimoglu, Curtis Thorne, Megha Padi, Bodour Salhia, Nathan A. Ellis, Xavier Llor, Rosa Xicola. FBXW7 and BCL9L altered pathways as hallmarks of early-onset colorectal cancer [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 4409.

Abstract Prostate cancer (PCa) stands as the second leading cause of cancer death in American men, with overtreatment being common given the difficulty in distinguishing between indolent and aggressive cases. Unlike many cancers, PCa lacks hallmark mutations in key oncogenes and tumor-suppressor genes. Instead, PCa exhibit extensive genomic rearrangements and chromosomal instability (CIN). CIN occurs during PCa progression, producing ETS gene family fusions, PTEN loss and androgen receptor amplification. Importantly, the mechanism underlying CIN in PCa remains unclear. Previously, we found that cell within early-grade human primary prostate adenocarcinomas (PRAD) frequently lack centrosomes, and this frequency correlates with tumor grade. We demonstrated that transient removal of centrosomes within non-tumorigenic human prostate epithelial cells (hPrEC) induces CIN and was, strikingly, sufficient to transform subpopulations of cells capable of producing xenograft tumors in mice. To unravel the molecular mechanisms underlying this path to tumorigenesis, we isolated DNA and RNA from parental hPrECs, clonal lines subjected to transient centrosome loss, and xenograft tumor cells and performed whole genome sequencing (WGS) and bulk RNA-seq. This allowed us to characterize the genomic profiles induced by centrosome loss and to identify an associated mutational signature. We used MUTECT2 to identify single nucleotide variants. Kataegis loci were observed on multiple chromosomes in all transient centrosome removal samples. Copy number variations were also detected in these samples using FACETS and SEQUENZA. Next, we extracted the copy number (CN) signature with SigProfiler and Sigminer, comparing it to COSMIC CN signatures and WGS data from PRAD patients. Our results revealed that the transient centrosome loss signature bears similarity to CN signatures associated with chromothripsis, loss of heterozygosity, and homologous recombination repair deficiency. Patients with PRAD displaying the centrosome loss CN signature had a poorer prognosis. Additionally, our CN analysis discovered that centrosome loss induced mosaic loss of chromosome Y in our samples. To detect structural variations, we employed DELLY, MANTA, SVABA, revealing a high occurrence of both non-clustered and clustered translocations in transient centrosome loss samples. These translocations were validated in our RNA-seq data using STAR-Fusion. Ongoing analyses include inferring CNV from PCa scRNA-seq data, and determining cell populations with the centrosome loss CN-signature and their associated transcriptome features. This work unveils a comprehensive genomic profile stemming from centrosome loss and demonstrates its role in driving oncogenesis in PCa. These findings have the potential to establish centrosome loss as a hallmark in PCa stratification, and lead to a significant advancement in prostate tumor treatment. Citation Format: Jiawen Yang, Diogo de Oliveira Pessoa, John M. Ryniawec, Emily Loertscher, Anne E. Cress, Megha Padi, Gregory C. Rogers. Characterization of a centrosome loss-induced tumorigenic signature in prostate epithelial cells [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 4349.

Natural Killer (NK) cells can target and destroy cancer cells, yet tumor microenvironments typically suppress NK cell recruitment and cytotoxicity. The epidermal growth factor receptor (EGFR) is a potent oncogene that can activate survival, migration, and proliferation pathways, and clinical data suggests it may also play an immunomodulating role in cancers. Recent work has demonstrated a novel role for nuclear EGFR (nEGFR) in regulating transcriptional events unique from the kinase domain. Using a novel peptide therapeutic (cSNX1.3) that inhibits retrograde trafficking of EGFR and an EGFR nuclear localization mutant, we discovered that nEGFR suppresses NK cell recruitment and cytotoxicity. RNA-Seq analysis of breast cancer cells treated with cSNX1.3 or modified to lack a nuclear localization sequence (EGFRΔNLS) revealed the EGF-dependent induction of NK activating receptor ligands, while kinase inhibition by erlotinib did not impact these genes. NanoString analysis of tumor-bearing WAP-TGFα transgenic mice treated with cSNX1.3 demonstrated an increase in immune cell populations and activating genes. Additionally, immunohistochemistry confirmed an increase in NK cells upon cSNX1.3 treatment. Finally, cSNX1.3 treatment was found to enhance NK cell recruitment and cytotoxicity in vitro. Together, the data demonstrate a unique immunomodulatory role for nEGFR.

African Americans (AAs) have among the highest incidence and mortality rates of colorectal cancer (CRC) in the US. They present with more right-sided, microsatellite stable tumors and are diagnosed at earlier ages than non-Hispanic Whites (NHW). While DNA methylation changes and their significance have been previously described in CRC, much less is known about the unique DNA methylation changes that occur in AA CRCs compared to NHW CRCs. In the current study, we analyzed DNA methylation changes in AA patients and compared those changes to data from NHWs. AA patients diagnosed with a microsatellite-stable colorectal tumor at the Yale New Health System were included in the study. A pathologist reviewed archival slides, and selected areas from tumor (n=160) and adjacent normal tissue (n=42) were cored. The median age of patients was 59 from 82 females and 78 males. Extracted DNA was bisulfite converted and tested using EPIC arrays. DNA methylation (450K) from NHW CRCs was also analyzed for 217 CRCs and 20 normal tissues obtained from TCGA (median age=70; 114 males, 103 females). Differentially methylated regions (DMRs) with |Db|>0.1, FDR corrected p-value <0.05 between tumor and normal tissue were identified using SeSAMe. We identified 4881 DMRs (57% hyper- and 44% hypomethylated) between AA CRC and normal tissue. Of these, 46% were in CpG Islands, 15% in CpG Shores, 4% in CpG shelves, and 36% in open sea. Most DMRs (71%) were in promoter regions, compared with 18% in gene bodies, 5% in enhancer regions, and 6% in intergenic regions. Next, we performed an additional DMR analysis using overlapping probes from the 450K and EPIC arrays to compare TCGA NHW with Yale AA data. We identified 2290 DMRs between AA CRC and normal tissue compared to 4657 in NHW. Of the 2290 AA DMRs, 1405 (61%) were common with NHW DMRs, 885 (39%) were unique to AAs. There were many overlapping pathways and genes involving metabolic functions, DNA damage response, immune response, and WNT signaling. However, the DMRs impacting these pathways differed between AA and NHW CRCs. Several AA-specific DMRs were associated with immune response and included LCK and GPSM3 genes. Hypermethylation of LCK, a tyrosine kinase involved in T-cell receptor signaling, was found in 60% of AA CRCs compared to 8% of NHW CRCs. LCK hypermethylation leads to CD4/8 T-cell inactivation and is linked to poor outcomes in lung cancer. We also observed AA-specific hypermethylation of GPSM3 in 60% of AA CRCs compared to 18% of NHW CRCs. GPSM3 is associated with CD4/8 T-cell infiltration. Lastly, promoter hypermethylation of 6 protocadherin genes was found in 59% of AAs compared to 20% NHWs. Protocadherins play a role in regulating WNT signaling and hypermethylation has been associated with poor outcomes in multiple cancers. Taken together, these results show AA-specific CRC methylation changes compared with NHW CRCs. Further work is warranted to investigate the potential functional and clinical implications of these methylation alterations, some of which may represent therapeutic targets and biomarkers. Citation Format: David N. Buckley, Rosa Xicola, Mikaeel Reger, Mary K. Yagle, Baris Kerimoglu, Curtis Thorne, Dante Bellomo, Megha Padi, Xavier Llor, Nathan Ellis, Bodour Salhia. DNA Methylation analysis of African American colorectal cancers reveal race-specific alterations [abstract]. In: Proceedings of the 16th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2023 Sep 29-Oct 2;Orlando, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2023;32(12 Suppl):Abstract nr C060.

Table S11: RPPA differential abundance in AZD1208 resistant compared to DMSO treated HSB-2 cells. Table S12: RPPA differential abundance in AZD-1µM (AZD1208 sensitive) compared to AZDR1 HSB-2 cells. Table S13: RPPA differential abundance in LGB321 resistant compared to DMSO treated HSB-2 cells. Table S14: RPPA differential abundance in LGB-1µM (LGB321 sensitive) compared to LGBR2 HSB-2 cells.

Table S4: GSEA curated canonical pathway enrichment (FDR<0.25) Table S5: GSEA oncogenic signature enrichment of AZDR1 compared to Naïve HSB-2 (FDR<0.25)

Salivary gland hypofunction is an adverse side effect associated with radiotherapy for head and neck cancer patients. This study delineated metabolic changes at acute, intermediate, and chronic radiation damage response stages in mouse salivary glands following a single 5 Gy dose. Ultra-high performance liquid chromatography-mass spectrometry was performed on parotid salivary gland tissue collected at 3, 14, and 30 days following radiation (IR). Pathway enrichment analysis, network analysis based on metabolite structural similarity, and network analysis based on metabolite abundance correlations were used to incorporate both metabolite levels and structural annotation. The greatest number of enriched pathways are observed at 3 days and the lowest at 30 days following radiation. Amino acid metabolism pathways, glutathione metabolism, and central carbon metabolism in cancer are enriched at all radiation time points across different analytical methods. This study suggests that glutathione and central carbon metabolism in cancer may be important pathways in the unresolved effect of radiation treatment.

Table S2: GO terms enriched from hierarchical clustering with optimal cut tree on differentially expressed (FDR<0.01) genes in AZDR1 compared to Naïve HSB-2. Table S3: GO terms enriched from hierarchical clustering with cut tree at h=2 on differentially expressed (FDR<0.01) genes in AZDR1 compared to Naïve HSB-2.