Transcriptomic Features Research Articles

Cross-modality mapping using image varifolds to align tissue-scale atlases to molecular-scale measures with application to 2D brain sections

This paper explicates a solution to building correspondences between molecular-scale transcriptomics and tissue-scale atlases. This problem arises in atlas construction and cross-specimen/technology alignment where specimens per emerging technology remain sparse and conventional image representations cannot efficiently model the high dimensions from subcellular detection of thousands of genes. We address these challenges by representing spatial transcriptomics data as generalized functions encoding position and high-dimensional feature (gene, cell type) identity. We map onto low-dimensional atlas ontologies by modeling regions as homogeneous random fields with unknown transcriptomic feature distribution. We solve simultaneously for the minimizing geodesic diffeomorphism of coordinates through LDDMM and for these latent feature densities. We map tissue-scale mouse brain atlases to gene-based and cell-based transcriptomics data from MERFISH and BARseq technologies and to histopathology and cross-species atlases to illustrate integration of diverse molecular and cellular datasets into a single coordinate system as a means of comparison and further atlas construction.

Antipsychotic-induced epigenomic reorganization in frontal cortex of individuals with schizophrenia

Genome-wide association studies have revealed >270 loci associated with schizophrenia risk, yet these genetic factors do not seem to be sufficient to fully explain the molecular determinants behind this psychiatric condition. Epigenetic marks such as post-translational histone modifications remain largely plastic during development and adulthood, allowing a dynamic impact of environmental factors, including antipsychotic medications, on access to genes and regulatory elements. However, few studies so far have profiled cell-specific genome-wide histone modifications in postmortem brain samples from schizophrenia subjects, or the effect of antipsychotic treatment on such epigenetic marks. Here, we conducted ChIP-seq analyses focusing on histone marks indicative of active enhancers (H3K27ac) and active promoters (H3K4me3), alongside RNA-seq, using frontal cortex samples from antipsychotic-free (AF) and antipsychotic-treated (AT) individuals with schizophrenia, as well as individually matched controls (n=58). Schizophrenia subjects exhibited thousands of neuronal and non-neuronal epigenetic differences at regions that included several susceptibility genetic loci, such as NRG1, DISC1, and DRD3. By analyzing the AF and AT cohorts separately, we identified schizophrenia-associated alterations in specific transcription factors, their regulatees, and epigenomic and transcriptomic features that were reversed by antipsychotic treatment; as well as those that represented a consequence of antipsychotic medication rather than a hallmark of schizophrenia in postmortem human brain samples. Notably, we also found that the effect of age on epigenomic landscapes was more pronounced in frontal cortex of AT-schizophrenics, as compared to AF-schizophrenics and controls. Together, these data provide important evidence of epigenetic alterations in the frontal cortex of individuals with schizophrenia, and remark for the first time on the impact of age and antipsychotic treatment on chromatin organization.

Functional specialization of hippocampal somatostatin-expressing interneurons

Hippocampal somatostatin-expressing (Sst) GABAergic interneurons (INs) exhibit considerable anatomical and functional heterogeneity. Recent single-cell transcriptome analyses have provided a comprehensive Sst-IN subpopulations census, a plausible molecular ground truth of neuronal identity whose links to specific functionality remain incomplete. Here, we designed an approach to identify and access subpopulations of Sst-INs based on transcriptomic features. Four mouse models based on single or combinatorial Cre- and Flp- expression differentiated functionally distinct subpopulations of CA1 hippocampal Sst-INs that largely tiled the morpho-functional parameter space of the Sst-INs superfamily. Notably, the Sst;;Tac1 intersection revealed a population of bistratified INs that preferentially synapsed onto fast-spiking interneurons (FS-INs) and were sufficient to interrupt their firing. In contrast, the Ndnf;;Nkx2-1 intersection identified a population of oriens lacunosum-moleculare INs that predominantly targeted CA1 pyramidal neurons, avoiding FS-INs. Overall, our results provide a framework to translate neuronal transcriptomic identity into discrete functional subtypes that capture the diverse specializations of hippocampal Sst-INs.

Transcriptomic and proteomic features of a mouse model of sperm DNA damage induced by benzo(a)pyrene

This study replicated a mouse model of sperm DNA damage induced by benzo(a)pyrene (BaP), and the transcriptomic and proteomic features of the model were examined to clarify the pathways related to BaP-induced damage to sperm DNA. Male mice in the BaP group were subjected to BaP at a dosage of 100 mg/kg/d or an equivalent quantity of saline solution in the control group for 60 days. Subsequently, the DNA fragmentation index (DFI) in sperm was assessed using a sperm chromatin structure assay (SCSA). RNA-seq and data-independent acquisition (DIA) were used to identify the mRNA and protein expression patterns in the testis. The sperm DFI significantly increased in the BaP group. Compared to the control group, the BaP group exhibited differential expression of 240 genes (referred to as DEGs) and 616 proteins (referred to as DEPs). These molecules included Aldh1a1, Cyb5r3, Fads1, Oxsm, Rcn3, and Prss45. Pathways in cancer, the PI3K-Akt signaling pathway, metabolic pathways, and the MAPK signaling pathway were the primary areas where these genes showed enrichment. BaP can damage the DNA of sperm and affect metabolism, the PI3K-Akt pathway, and pathways associated with cancer signaling.

ELAPOR1 induces the classical/progenitor subtype and contributes to reduced disease aggressiveness through metabolic reprogramming in pancreatic cancer.

Pancreatic ductal adenocarcinoma (PDAC) is a heterogeneous disease with distinct molecular subtypes described as classical/progenitor and basal-like/squamous PDAC. We hypothesized that integrative transcriptome and metabolome approaches can identify candidate genes whose inactivation contributes to the development of the aggressive basal-like/squamous subtype. Using our integrated approach, we identified endosome-lysosome associated apoptosis and autophagy regulator 1 (ELAPOR1/KIAA1324) as a candidate tumor suppressor in both our NCI-UMD-German cohort and additional validation cohorts. Diminished ELAPOR1 expression was linked to high histological grade, advanced disease stage, the basal-like/squamous subtype, and reduced patient survival in PDAC. In vitro experiments demonstrated that ELAPOR1 transgene expression not only inhibited the migration and invasion of PDAC cells but also induced gene expression characteristics associated with the classical/progenitor subtype. Metabolome analysis of patient tumors and PDAC cells revealed a metabolic program associated with both upregulated ELAPOR1 and the classical/progenitor subtype, encompassing upregulated lipogenesis and downregulated amino acid metabolism. 1-Methylnicotinamide, a known oncometabolite derived from S-adenosylmethionine, was inversely associated with ELAPOR1 expression and promoted migration and invasion of PDAC cells in vitro. Taken together, our data suggest that enhanced ELAPOR1 expression promotes transcriptome and metabolome characteristics that are indicative of the classical/progenitor subtype, whereas its reduction associates with basal-like/squamous tumors with increased disease aggressiveness in PDAC patients. These findings position ELAPOR1 as a promising candidate for diagnostic and therapeutic targeting in PDAC.

Transcriptomic features of programmed and inflammatory cell death in gingival tissues.

This study used an experimental periodontitis model to detail the gingival transcriptome related to cell death processes of pyroptosis, necroptosis, ferroptosis, and cuproptosis. Healthy Macaca mulatta primates stratified by age, ≤3 years (young), 7-12 years (adolescent), 12-15 years (adult), and 17-23 years (aged), provided gingival tissue biopsies for microarray analysis focused on 257 genes representative of the four cell death processes and bacterial plaque samples for 16S rRNA gene analysis. Age differences in the profiles of gene expression in healthy tissues were noted for cuproptosis, ferroptosis, necroptosis, and pyroptosis. Major differences were then observed with disease initiation, progression, and resolution also related to the age of the animals. Distinct bacterial families/consortia of species were significantly related to the gene expression differences for the cell death pathways. These results emphasized age-associated differences in the gingival tissue molecular response to changes in the quality and quantity of bacteria accumulating with the disease process reflected in regulated cell death pathways that are both physiological and pathophysiological.

Deciphering age-specific molecular features in cervical cancer and constructing an angio-immune prognostic model.

Cancer incidence is increasingly seen in younger individuals. Molecular distinctions between young and elderly patients at onset are understudied. This study used public databases to explore genomic, transcriptomic, and immune-related features across age groups in cervical cancer. Additionally, it aims to create a prognostic model applicable across diverse age cohorts, enabling precise patient stratification, and personalized therapies. Gene mutations, expression data, and clinicopathological information were obtained from 317 cervical cancer patients. These patients were divided into a young group and an old group based on the median age of onset. The characteristics of differential gene mutation, gene expression, and immune cells analysis were analyzed by R software. Finally, the prognostic model was constructed by univariate Cox, least absolute shrinkage and selection operator, and multivariate Cox regression analyses of angiogenic and immune gene sets. Its validity was further confirmed using an additional 300 cervical squamous cell carcinoma and endocervical adenocarcinoma tissues. Cervical cancer patients at elderly onset age exhibit a significantly higher frequency of NOTCH1 and TP53 driver mutations compared to young patients, along with a notably higher tumor mutational burden. However, there were no significant differences between the 2 groups in terms of genomic instability and age-related mutational signatures. Differential gene expression analysis revealed that the young group significantly upregulated interferon-alpha and gamma responses and exhibited significantly higher activity in multiple metabolic pathways. Immune microenvironment analysis indicated enrichment of dendritic cells and natural killer cells in the young group, while transforming growth factor-β signature was enriched in the elderly group, indicating a higher degree of immune exclusion. A multigene prognostic model based on angiogenesis and T cell immune gene sets showed excellent prognostic performance independent of clinical factors such as age. High-risk groups identified by the model exhibit significant activation of tumor-promoting processes, such as metastasis and angiogenesis. Our study reveals distinct patterns in cancer-driving mechanisms, biological processes, and immune system status between young and elderly patients at onset with cervical cancer. These findings shed light on the age-specific underlying mechanisms of carcinogenesis. Furthermore, an independent molecular prognostic model is constructed to provide valuable references for patient stratification and the development of potential drug targets.

Redoxhigh phenotype mediated by KEAP1/STK11/SMARCA4/NRF2 mutations diminishes tissue-resident memory CD8+ T cells and attenuates the efficacy of immunotherapy in lung adenocarcinoma

ABSTRACT Metabolism reprogramming within the tumor microenvironment (TME) can have a profound impact on immune cells. Identifying the association between metabolic phenotypes and immune cells in lung adenocarcinoma (LUAD) may reveal mechanisms of resistance to immune checkpoint inhibitors (ICIs). Metabolic phenotypes were classified by expression of metabolic genes. Somatic mutations and transcriptomic features were compared across the different metabolic phenotypes. The metabolic phenotype of LUAD is predominantly determined by reductase-oxidative activity and is divided into two categories: redoxhigh LUAD and redoxlow LUAD. Genetically, redoxhigh LUAD is mainly driven by mutations in KEAP1, STK11, NRF2, or SMARCA4. These mutations are more prevalent in redoxhigh LUAD (72.5%) compared to redoxlow LUAD (17.4%), whereas EGFR mutations are more common in redoxlow LUAD (19.0% vs. 0.7%). Single-cell RNA profiling of pre-treatment and post-treatment samples from patients receiving neoadjuvant chemoimmunotherapy revealed that tissue-resident memory CD8+ T cells are responders to ICIs. However, these cells are significantly reduced in redoxhigh LUAD. The redoxhigh phenotype is primarily attributed to tumor cells and is positively associated with mTORC1 signaling. LUAD with the redoxhigh phenotype demonstrates a lower response rate (39.1% vs. 70.8%, p = 0.001), shorter progression-free survival (3.3 vs. 14.6 months, p = 0.004), and overall survival (12.1 vs. 31.2 months, p = 0.022) when treated with ICIs. The redoxhigh phenotype in LUAD is predominantly driven by mutations in KEAP1, STK11, NRF2, and SMARCA4. This phenotype diminishes the number of tissue-resident memory CD8+ T cells and attenuates the efficacy of ICIs.

Single-cell RNA sequencing reveals aberrant sphingolipid metabolism in non-small cell lung cancer impacts tumor-associated macrophages and stimulates angiogenesis via macrophage inhibitory factor signaling.

Sphingolipids not only serve as structural components for maintaining cell membrane fluidity but also function as bioactive molecules involved in cell signaling and the regulation of various biological processes. Their pivotal role in cancer cell development, encompassing cancer cell proliferation, migration, angiogenesis, and metastasis, has been a focal point for decades. However, the contribution of sphingolipids to the complexity of tumor microenvironment promoting cancer progression has been rarely investigated. Through the integration of publicly available bulk RNA-seq and single-cell RNA-seq data, we conducted a comprehensive analysis to compare the transcriptomic features between tumors and adjacent normal tissues, thus elucidating the intricacies of the tumor microenvironment (TME). Disparities in sphingolipid metabolism (SLM)-associated genes were observed between normal and cancerous tissues, with the TME characterized by the enrichment of sphingolipid signaling in macrophages. Cellular interaction analysis revealed robust communication between macrophages and cancer cells exhibiting low SLM, identifying the crucial ligand-receptor pair, macrophage inhibitory factor (MIF)-CD74. Pseudo-time analysis unveiled the involvement of SLM in modulating macrophage polarization towards either M1 or M2 phenotypes. Categorizing macrophages into six subclusters based on gene expression patterns and function, the SPP1+ cluster, RGS1+ cluster, and CXCL10+ cluster were likely implicated in sphingolipid-induced M2 macrophage polarization. Additionally, the CXCL10+, AGER+, and FABP4+ clusters were likely to be involved in angiogenesis through their interaction with endothelial cells. Based on multiple scRNA-seq datasets, we propose that a MIF-targeted strategy could potentially impede the polarization from M1 to M2 and impair tumor angiogenesis in low-SLM non-small cell lung cancer (NSCLC), demonstrating its potent antitumor efficacy.

Abstract LB242: Prediction of patient response to neoadjuvant chemotherapy in breast cancer from their deconvolved tumor microenvironment transcriptome

Abstract Introduction: The tumor microenvironment (TME) is a complex and dynamic ecosystem that plays critical roles in tumor development and clinical outcome. While the multifarious cellular interactions within the TME have been extensively studied with a significant emphasis on immunotherapy, understanding its role in chemotherapy outcome remains less explored. To this end, we present a novel generic computational framework named DECODEM (DEcoupling Cell-type-specific Outcomes using DEconvolution and Machine learning). DECODEM facilitates the exploration of the association of cell-type-specific gene expression profiles with the effectiveness of neoadjuvant chemotherapy (NAC) in treating breast cancer (BC). This approach holds promise for elucidating the nuanced ways in which the TME can modulate therapeutic efficacy, potentially paving the way for more personalized and effective cancer treatments. Methods: DECODEM employs cellular deconvolution of bulk transcriptomics to extract cell-type-specific gene expression profiles within the TME, and subsequently harnesses machine learning to construct cell-type-specific predictors of clinical outcomes. Utilizing DECODEM, we analyzed three publicly available cohorts of HER2-negative BC patients who received NAC. We assessed the predictive capabilities of the individual cell types as well as their collective influence, benchmarking against a state-of-the-art predictor that employs both clinical and transcriptomic features. We corroborated our findings in a single-cell (SC) transcriptomics cohort of patients with triple negative breast cancer (TNBC) undergoing chemotherapy. Further, we expanded DECODEM to DECODEMi, which accounts for cell-cell interactions (CCIs) among the predominant cell types, offering potential insights into the mechanistic processes governing clinical responses. Results: Our analysis indicated that the heterogeneous composition of the TME is crucial for chemotherapy response prediction in HER2-negative BC, beyond the influence of malignant cells alone. The key insights from our study are as follows:1. Gene expression of immune cells (myeloid, plasmablasts, B-cells) and stromal cells (endothelial, normal epithelial, cancer-associated fibroblasts) displayed comparable-to-superior predictive powers over both bulk expression and the state-of-the-art predictor.2. The predictive accuracy is further augmented by multi-cell-type ensembles, with a triad ensemble comprising endothelial, myeloid, and plasmablast cells achieving the highest predictive accuracy across multiple cohorts.3. DECODEMi can effectively capture the regulatory impact of CCIs (validated in a SC dataset sourced from the SC-TNBC cohort) and has identified novel CCIs likely to influence chemotherapy response in HER2-negative BC. Conclusion: We provide an unprecedented computational methodology to quantitatively assess the cell-type-specific influences of the TME on clinical outcome. Our analysis offers insights into the complex interactions among TME cells and their correlation with chemotherapy efficacy in HER2-negative BC, thereby enriching the knowledge base of BC treatment effectiveness. Citation Format: Saugato Rahman Dhruba, Sahil Sahni, Binbin Wang, Di Wu, Yael Schmidt, Eldad Shulman, Sanju Sinha, Stephen-John Sammut, Carlos Caldas, Kun Wang, Eytan Ruppin. Prediction of patient response to neoadjuvant chemotherapy in breast cancer from their deconvolved tumor microenvironment transcriptome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB242.

GTDrift: a resource for exploring the interplay between genetic drift, genomic and transcriptomic characteristics in eukaryotes.

We present GTDrift, a comprehensive data resource that enables explorations of genomic and transcriptomic characteristics alongside proxies of the intensity of genetic drift in individual species. This resource encompasses data for 1506 eukaryotic species, including 1413 animals and 93 green plants, and is organized in three components. The first two components contain approximations of the effective population size, which serve as indicators of the extent of random genetic drift within each species. In the first component, we meticulously investigated public databases to assemble data on life history traits such as longevity, adult body length and body mass for a set of 979 species. The second component includes estimations of the ratio between the rate of non-synonymous substitutions and the rate of synonymous substitutions (dN/dS) in protein-coding sequences for 1324 species. This ratio provides an estimate of the efficiency of natural selection in purging deleterious substitutions. Additionally, we present polymorphism-derived N e estimates for 66 species. The third component encompasses various genomic and transcriptomic characteristics. With this component, we aim to facilitate comparative transcriptomics analyses across species, by providing easy-to-use processed data for more than 16 000 RNA-seq samples across 491 species. These data include intron-centered alternative splicing frequencies, gene expression levels and sequencing depth statistics for each species, obtained with a homogeneous analysis protocol. To enable cross-species comparisons, we provide orthology predictions for conserved single-copy genes based on BUSCO gene sets. To illustrate the possible uses of this database, we identify the most frequently used introns for each gene and we assess how the sequencing depth available for each species affects our power to identify major and minor splice variants.

Association between the lung microbiome and perioperative prognosis in lung transplant recipients.

Studies have confirmed that the lung microbiome of lung transplant recipients is altered and serves as a prognostic indicator for long-term mortality. Other studies reported that the lung microbiome affects host immunity and the transcriptome. However, the lung microbiome composition at the early post-transplant period following lung transplantation is unclear, and the relationship of the lung microbiome with pulmonary immunity and the host transcriptome is also not well understood. We hypothesize that changes in the lung microbiome composition in the early post-transplant period may have a predictive value for perioperative outcomes following lung transplantation and that the lung microbiome is correlated with pulmonary immunity and the host transcriptome. Thus, this prospective study aimed at observing the lung microbiome composition in the early post-transplant period and the impact of the lung microbiome on pulmonary cytokines and the host transcriptome. Our findings will help us gain a comprehensive understanding of the distribution and significance of the lung microbiome in the early post-transplant period. An observational study was conducted to identify the lung microbiome and the host transcriptome characteristics using next-generation sequencing. Luminex was employed for quantifying alveolar cytokines. Spearman's correlation analysis was utilized to assess the impact of the lung microbiome on pulmonary immunity and differentially expressed genes in patients who died perioperatively after lung transplantation. Patients with poor perioperative outcomes showed an increase in Mycoplasma and Arcobacter, a decrease of Gemella, and increased interleukin (IL)-10, IL-1β, and tumor necrosis factor (TNF)-α concentration. The lung microbiome correlates with lung immunity in lung transplant recipients. In the death group, the function of differentially expressed genes is associated with cell apoptosis, and promoting TNF production is upregulated. The lung microbiome is related to differentially expressed genes between the two groups. The lung microbiome and cytokines can be considered as potential biomarkers for early prognosis in lung transplant recipients. The lung microbiome is associated with both lung immunity and differentially expressed genes in lung transplant recipients.

Geographic origin shapes the adaptive divergences of Rotaria rotatoria (Rotifera, Bdelloidea) to thermal stress: Insights from ecology and transcriptomics.

Global warming has raised concerns regarding the potential impact on aquatic biosafety and health. To illuminate the adaptive mechanisms of bdelloid rotifers in response to global warming, the ecological and transcriptomic characteristics of two strains (HX and ZJ) of Rotaria rotatoria were investigated at 25°C and 35°C. Our results showed an obvious genetic divergence between the two geographic populations. Thermal stress significantly reduced the average lifespan of R. rotatoria in both strains, but increased the offspring production in the ZJ strain. Furthermore, the expression levels of genes Hsp70 were significantly upregulated in the HX strain, while GSTo1 and Cu/Zn-SOD were on the contrary. In the ZJ strain, the expression levels of genes Hsp70, CAT2, and GSTo1 were upregulated under thermal stress. Conversely, a significant decrease in the expression level of the Mn-SOD gene was observed in the ZJ strain under thermal stress. Transcriptomic profiling analysis revealed a total of 105 and 5288 differentially expressed genes (DEGs) in the HX and ZJ strains under thermal stress, respectively. The PCA results showed clear differences in gene expression pattern between HX and ZJ strains under thermal stress. Interestingly, compared to the HX strain, numerous downregulated DEGs in the ZJ strain were enriched into pathways related to metabolism under thermal stress, suggesting that rotifers from the ZJ strain prioritize resource allocation to reproduction by suppressing costly metabolic processes. This finding is consistent with the life table results. This study provides new insights into the adaptive evolution of aquatic animals in the context of global climate change.

Targeting Group 3 Medulloblastoma by the Anti-PRUNE-1 and Anti-LSD1/KDM1A Epigenetic Molecules.

Medulloblastoma (MB) is a highly malignant childhood brain tumor. Group 3 MB (Gr3 MB) is considered to have the most metastatic potential, and tailored therapies for Gr3 MB are currently lacking. Gr3 MB is driven by PRUNE-1 amplification or overexpression. In this paper, we found that PRUNE-1 was transcriptionally regulated by lysine demethylase LSD1/KDM1A. This study aimed to investigate the therapeutic potential of inhibiting both PRUNE-1 and LSD1/KDM1A with the selective inhibitors AA7.1 and SP-2577, respectively. We found that the pharmacological inhibition had a substantial efficacy on targeting the metastatic axis driven by PRUNE-1 (PRUNE-1-OTX2-TGFβ-PTEN) in Gr3 MB. Using RNA seq transcriptomic feature data in Gr3 MB primary cells, we provide evidence that the combination of AA7.1 and SP-2577 positively affects neuronal commitment, confirmed by glial fibrillary acidic protein (GFAP)-positive differentiation and the inhibition of the cytotoxic components of the tumor microenvironment and the epithelial-mesenchymal transition (EMT) by the down-regulation of N-Cadherin protein expression. We also identified an impairing action on the mitochondrial metabolism and, consequently, oxidative phosphorylation, thus depriving tumors cells of an important source of energy. Furthermore, by overlapping the genomic mutational signatures through WES sequence analyses with RNA seq transcriptomic feature data, we propose in this paper that the combination of these two small molecules can be used in a second-line treatment in advanced therapeutics against Gr3 MB. Our study demonstrates that the usage of PRUNE-1 and LSD1/KDM1A inhibitors in combination represents a novel therapeutic approach for these highly aggressive metastatic MB tumors.

Quantification and molecular correlates of tertiary lymphoid structures in primary prostate cancer.

To morphologically describe tertiary lymphoid structures (TLS) in prostatectomy specimens and correlate them with clinical and transcriptomic features. A total of 72 consecutive cases of entirely submitted radical prostatectomy (RP) patients tested with the Decipher Genomic Classifier were included in the study. Images were manually annotated using QuPath tools to denote tumor regions and each cluster of TLS. Clusters of lymphocytes that were surrounded on all four sides by tumor were defined as intra-tumor TLS (IT-TLS). Clusters of lymphocytes at the leading edge of carcinoma with either the prostatic pseudocapsule or benign parenchyma at one end were defined as peri-tumor TLS (PT-TLS). A classification algorithm to distinguish lymphocytes from non-lymphocytic cells using a supervised machine learning model was used. The associations between TLS formation and 265 gene expression-based signatures were examined. The magnitude of total TLS correlations with primary tumor gene expression signatures was moderate (~0.35-0.5) with several HLA, T-cell and B-cell Cluster signatures, showing positive correlation with various metrics for quantification of TLS. On the other hand, immune suppressive signatures (Treg, MDSC) were negatively correlated. While signatures for macrophages, NK cells and other immune cell types were uncorrelated for the most part. PT-TLS was associated with MHC signatures while IT TLS correlated with MHC andT-cell signatures. Clusters of inflammatory cells in the RP specimen can be divided spatially into PT TLS and IT-TLS, each with its unique molecular correlates of tumor immune microenvironment. The presence of TLS is positively correlated with MHC signatures, T- cell and B-cell cluster signatures but, negatively correlated with immune suppressive signatures. A subset of prostate cancer demonstrate a robust inflammatory response, and warrant further characterization in larger cohorts.

Abstract 620: Screening of small-molecule MAPK pathway inhibitors on a panel of 130 human cancer cell lines identifies determinants of response in RAS-mutant subgroups

Abstract Oncogenic mutations in members of the RAS family, HRAS, NRAS, and KRAS, are frequently identified in various cancers and often lead to increased signaling of the mitogen-activated protein kinase (MAPK) pathway. KRAS is the most frequently mutated RAS isoform, with a mutation incidence of >10% across all cancers, and up to 70% and 20% in pancreatic and non-small cell lung cancer (NSCLC), respectively. RAS has long been considered to be undruggable, and direct inhibition of RAS with small molecule inhibitors has only recently been achieved, with the approval of the KRAS G12C-mutant selective inhibitors sotorasib and adagrasib for treatment of NSCLC. Indirect targeting of RAS-mutated cancers by inhibitors of downstream components such as MEK and ERK has shown variable responses. Novel insights into MAPK signaling have renewed interest into targeting the MAPK pathway in RAS-mutant cancer, with pan-RAF dimer inhibitors and dual RAF/MEK inhibitors reaching clinical trials. To improve the chances of success, it is important to determine which patient populations are most likely to respond to these inhibitors. To gain more insight into the cellular targeting of MAPK pathway inhibitors, eighteen inhibitors of the MAPK components KRAS, RAF, MEK, and ERK were profiled on a panel of 130 cell lines in cell viability assays (the Oncolines® panel). The cell lines represented a diverse range of tumor tissues and had different genomic backgrounds. Many of the cell lines had mutations in KRAS, NRAS, or BRAF. Drug responses were determined by exposing the cell lines to nine-point duplicate dose ranges of the inhibitors. Using bioinformatics, cell line responses to the inhibitors were related to genomic and transcriptomic features, to identify determinants of drug response in both tissue-agnostic and tissue-specific contexts. The results of our analyses revealed variable responses of KRAS-mutant cell lines to the inhibitors, reflecting clinical responses. Tissue-focused analyses showed that the different responses could be traced back to the different KRAS amino acid substitutions. For instance, in pancreatic cancer cell lines, the KRAS Q61H mutation predicted increased sensitivity to MAPK pathway inhibitors compared to other KRAS amino acid substitutions. Additionally, in tissue groups with a more heterogenous mutation profile, co-mutations explained the decreased effectivity of MAPK pathway inhibitors in KRAS-mutant cell lines, such as PIK3CA mutations in colorectal cancer cell lines. Lastly, a gene expression-based signature of MAPK activity predicted response to MAPK pathway inhibitors, independent of mutation status. Our findings have uncovered determinants of response to MAPK pathway inhibitors in RAS-mutant cell lines. These determinants can serve as markers to select patients who are most likely to benefit from MAPK inhibitors. Citation Format: Jeffrey J. Kooijman, Awan Al Koerdi, Daphne J. Kluitmans, Esmee van den Bossche, Jelle Dylus, Jeroen A. de Roos, Janneke J. Melis, Guido J. Zaman. Screening of small-molecule MAPK pathway inhibitors on a panel of 130 human cancer cell lines identifies determinants of response in RAS-mutant subgroups [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 620.

Abstract 1846: Integrative histopathologic, proteomic, phosphoproteomic, genomic, and transcriptomic characterization of renal cell carcinoma

Abstract Although histologically defined subtypes exist, renal cell carcinoma (RCC) is a heterogenous disease, resulting in various treatment outcomes and prognosis. Here, we conduct comprehensive profiling of RCC with histopathologic, proteomic, phosphoproteomic, genomic, and transcriptomic analyses on tumor and paired adjacent normal tissues from 113 patients. RCC tumors are distinctly separated from corresponding normal tissues through multi-omics clustering, a distinction further supported by pathway analysis. In addition, various hallmarks of cancer were differentially observed at the individual patient level, highlighting the intertumoral heterogeneity of RCC. Proteomic subtyping classifies RCC tumors into four subtypes (C1-4) with distinct phosphoproteomic, genomic, and transcriptomic characteristics with prognostic and predictive implications. Among these subtypes, C1-3 were almost composed of clear cell RCC (ccRCC; 96.8%), whereas C4 was composed of non-clear cell RCC (nccRCC) or other types of cancer (100.0%). Among each subtype, C1 (N=37) exhibits the strongest adaptive immune response and angiogenesis with favorable prognosis after surgical resection. C2 (N=39) was characterized with enrichment in DNA repair pathways and alteration in VHL (100.0%) along with BAP1 (24.0%) genes. C3 (N=19) was associated with enrichment in inflammatory response with dismal prognosis and therapeutic resistance to tyrosine kinase inhibitors targeting vascular endothelial growth factor receptors. C4 (N=18) was exclusively enriched with nccRCC with active mitochondrial and oxidation phosphorylation. Transglutaminase 2 (TGM2), a crosslinking enzyme, is identified as a potential marker with therapeutic implication for ccRCC. Collectively, this study reports a large-scale multi-omics-based approach of RCC, providing insights into biologic underpinning and evidence for rational treatment selection as well as linking the multi-omics-derived phenotypes to clinical outcomes of ccRCC. Citation Format: Woo Sun Kwon, Chang Gon Kim, Harim Koo, Woo Kyun Bae, Eu Chang Hwang, Won Sik Ham, Nam Hoon Cho, Jason K. Sa, Jong Bae Park, Sun Young Rha. Integrative histopathologic, proteomic, phosphoproteomic, genomic, and transcriptomic characterization of renal cell 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 1846.

Abstract 5887: Identifying new immunotherapy targets using machine learning and ex vivo validation

Abstract Overview: Immunotherapy has revolutionized cancer treatment. However, existing immune checkpoint inhibitors (CPI) yield low response rates in most cancers, highlighting the need for new therapeutic options. Traditional immune-oncology (IO) target discovery relies on preclinical models, which struggle to recapitulate human tumor complexity, limiting translation potential. Attention is thus directed at harnessing multimodal patient molecular data with modern machine learning (ML) techniques to identify new IO targets which may have higher clinical viability. Approach: We posed IO target discovery as a binary classification task, training ML systems on known candidate drug targets that have progressed to stage I or higher clinical trials. We constructed a rich knowledge graph database to support model development. Graph nodes (n=11,919) comprised genes with edges linking genes involved in n=99,275 protein-protein interactions (PPIs; PMID: 28936969). Genes were labelled with n=6,387 gene-disease associations alongside bulk exome and transcriptome (RNAseq) features from n=1,317 CPI-treated patients (PMID: 33508232). We considered the role of different immune subsets by incorporating cell type-specific PPIs from single cell RNAseq atlases of n=350 samples (PMID: 31786210). To capture how antigen processing affects immunotherapy response, we also examined the immunopeptidome of n=60 patients (PMID: 30556813). Causal data on immune responses to genetic perturbation stemmed from n=7 publicly available CRISPR tumor-T cell co-cultures and n=15,442 SNP-phenotype links. Lastly, we used interpretability analysis to understand drivers of model predictions and elucidate critical genes that influence multiple IO target pathways. Results: Firstly, we developed an ensemble ML approach which achieved test ROC-AUC>0.75. Next, we used a graph-based ML framework which yielded superior performance (test ROC-AUC>0.90). Orthogonal validation confirmed these models can discriminate known targets by trial phase (p<0.001), predict patient response in new CPI trials (p<0.05), and identify genes that rank highly in unseen genome-wide CRISPR screens. Targets from both methods have entered experimental validation in patient-derived explants and organoid-immune co-cultures. Already, we have identified 2 novel targets predicted to relate to macrophage activity. Early data suggest that perturbing them leads to macrophage repolarization. Further validation experiments are ongoing. Conclusions: We will reveal candidate targets, showing that our method is effective at uncovering new IO targets and can identify critical nodes in biological networks that might be attractive hits. In addition, deciphering data types that drive model predictions provides a broader immunobiological understanding of anti-tumor immune responses. Thus, our results endorse using ML and multimodal data for novel IO target discovery. Citation Format: Marcellus Augustine, Nuno Rocha Nene, Krupa Thakkar, Danwen Qian, Evelyn Fitzsimons, Benjamin S. Simpson, Chris Watkins, Charles Swanton, Kevin Litchfield. Identifying new immunotherapy targets using machine learning and ex vivo validation [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 5887.

Abstract 2486: Dissecting tumor-immune interaction in response and resistance to immune checkpoint blockade in metastatic melanoma

Abstract Background: Immune checkpoint blockade (ICB) has revolutionized the treatment and prognosis of various cancers, including melanoma. Nevertheless, intrinsic or acquired resistance is common, and the interactions between tumor-intrinsic features and the microenvironment underlying these resistances remain unclear for most patients. Methods: In this study, we developed a framework to analyze response and resistance, both intrinsic and acquired, via tumor-intrinsic programs, immune features, and tumor-stromal-immune interactions in a standardized, uniformly processed, and deeply clinically annotated cohort of metastatic melanoma patients (n=61) treated with ICB as part of the NCI funded human tumor atlas network (HTAN) initiative. From the same tumor samples (pre-treatment (n=33) and post-progression (n=28)), we conducted bulk whole-exome sequencing (WES), single-nucleus RNA sequencing, and for a subset of the samples high-resolution spatial imaging (including protein mIHC, CODEX, and transcriptomics MERFISH). Standardized processing and data pipelines allowed for integration of genomic, transcriptomic, and spatial features to elucidate shared tumor and microenvironmental states and their relationships with resistance. Results: Our preliminary analysis on a subset of 28 samples identified an enrichment of B cells, plasma cells, and T follicular helper-like cells in the tumor microenvironment (TME) of patients sensitive to treatment. The TME of non-responders was predominantly characterized by macrophage/monocyte cell populations. Post-treatment responder samples showed an enrichment of CD4+ T cells. Genomic heterogeneity, assessed through WES, was confirmed as a feature of intrinsic resistance across different ICB treatment settings, while ploidy was associated with treatment response in pre-treatment, ipilimumab-naïve patients treated with PD-1 inhibitors. Comparing the immune compartments of samples with high and low heterogeneity revealed NK cell enrichment in samples with a high proportion of subclonal mutations. Conclusions and Future Directions: Overall, our work provides a high-resolution understanding of the tumor-stromal-immune interaction in metastatic melanoma, shedding light on the factors contributing to therapy response and resistance, and contributing a repository of data from deeply clinically annotated, multimodally characterized and uniformly processed patient samples. Ongoing spatial analysis will further enhance our understanding of these compartments at a spatial level. The results of our study have the potential to guide more personalized and effective treatment strategies for metastatic melanoma in the future. Citation Format: Giuseppe Tarantino, Yiwen He, Priyanka Solanky, Aaron Thorner, Tyler Aprati, Ryan Sullivan, Emily Robitschek, Timothy Blosser, Xiaowei Zhuang, Xiaowei Zhuang, Elliot Boblitt, Allison Frangieh, Marta Holovatska, Aleigha Lawless, Michael Manos, Kathleen Pfaff, Karla Helvie, Tatyana Sharova, Dennie Frederick, James Liam Fahey, Diego Villamarin, Sami Farhi, Scott Rodig, Bruce Johnson, Alex K. Shalek, Eliezer Van Allen, Stephen Hodi, Genevieve M. Boland, David Liu. Dissecting tumor-immune interaction in response and resistance to immune checkpoint blockade in metastatic melanoma [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 2486.

Abstract 4907: RanBALL: Identifying B-cell acute lymphoblastic leukemia subtypes based on an ensemble random projection model

Abstract As the most common pediatric malignancy, B-cell acute lymphoblastic leukemia (B-ALL) has multiple distinct subtypes characterized by recurrent and sporadic somatic and germline genetic alterations like chromosomal alteration, transcription factor rearrangement or kinase inhibition. The treatment of B-ALL patients is personalized based on specific subtypes, as the treatment responses for different B-ALL subtypes may vary considerably. Identification of B-ALL subtypes can facilitate risk stratification and enable tailored therapeutic approaches. Existing methods for B-ALL subtyping primarily depend on immunophenotypic, cytogenetic and genomic analyses, which would be costly, complicated, and laborious in clinical practice applications. To overcome these challenges, we present RanBALL (an Ensemble Random Projection-Based Model for Identifying B-Cell Acute Lymphoblastic Leukemia Subtypes), an accurate and cost-effective model for B-ALL subtype identification based on transcriptomic profiling only. RanBALL leverages random projection (RP) to construct an ensemble of dimension-reduced multi-class classifiers for B-ALL subtyping. Specifically, the transcriptomic profiling features were projected onto low-dimensional spaces by random projection matrices whose elements conform to a distribution characterized by zero mean and unit variance. To ensure reliable and robust performance, we selected 20 subspace dimensions ranging from 600 to 2500, with intervals of 100. The transformed low dimensional data matrix was used for training an ensemble of multi-class support vector machine (SVM) classifiers, each corresponding to one of the RP matrices of various dimensions. The predicted probabilistic scores of each B-ALL subtype were integrated for determining the final decision. Results based on 10 times 10-fold cross validation tests for >1700 B-ALL patients demonstrated that the proposed model achieved an accuracy of 93.7%, indicating promising prediction capabilities of RanBALL for B-ALL subtyping. Furthermore, the 30% held-out tests suggested that the model was robust and consistent to maintain high confidence levels for accurate predictions. The high accuracies of RanBALL suggested that our model could effectively capture underlying patterns of transcriptomic profiling for accurate B-ALL subtype identification. To extend the impact of RanBALL, we have established a free and publicly available python package for RanBALL available at https://github.com/wan-mlab/RanBALL. We believe RanBALL will facilitate the discovery of B-ALL subtype-specific marker genes and therapeutic targets, and eventually have consequential positive impacts on downstream risk stratification and tailored treatment design. Citation Format: Lusheng Li, Hanyu Xiao, Joseph D. Khoury, Jieqiong Wang, Shibiao Wan. RanBALL: Identifying B-cell acute lymphoblastic leukemia subtypes based on an ensemble random projection model [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 4907.