Figure - available from: Nature
This content is subject to copyright. Terms and conditions apply.
HLA loss as a mechanism of immune escape a, Left, distribution over cells of chromosome arm 6p BAF in scRNA-seq data with ranking by median 6p BAF per cell type. Right, allelic imbalance in 6p BAF across cancer cell clusters. White vertical lines indicate the median. Chr., chromosome. b, Left, percentage of cancer cells with 6p LOH per patient. Right, site- and clone-specific percentage of cancer cells with 6p LOH. Het., heterozygous. c, Percentage of cancer cells with 6p LOH per sample as a function of mutational signature. Pie charts show the fraction of samples with heterozygous, subclonal LOH and clonal LOH 6p status. d, Percentage of patients with LOH of any HLA class I gene in the MSK-IMPACT HGSOC cohort (n = 1,298 patients) for BRCA1-, BRCA2- and CDK12-mutant and CCNE1-amplified tumours, mapping to HRD-Dup, HRD-Del, TD and FBI signatures, respectively. Error bars, 95% binomial confidence intervals. e, Percentage of cancer cells with 6p LOH per sample as a function of anatomical site. Pie charts show the fraction of samples by 6p status. f, UMAP plots of cancer cells from representative HRD-Dup and FBI cases. Density plots show site-specific 6p BAF. g, Fraction of naive and dysfunctional T cells in CD45⁺ samples as a function of the 6p LOH clonality of cancer cells in matched CD45⁻ samples. *P < 0.05; brackets indicate two-sided Wilcoxon pairwise comparisons. In b, c, e and g, 6p LOH status is defined as follows: heterozygous, percentage 6p LOH ≤ 20%; subclonal LOH, 20% < percentage 6p LOH ≤ 80%; clonal LOH, percentage 6p LOH > 80%. In c, e and g, box plots and violin plots show the median, top and bottom quartiles; whiskers correspond to 1.5× IQR. In a–e, only BAF estimates from cells with ≥10 reads aligning to 6p were considered and allelic imbalance states were assigned on the basis of the mean 6p BAF per cell as follows: balanced, BAF ≥ 0.35; imbalanced, 0.15 ≤ BAF < 0.35; LOH, BAF < 0.15 (Methods).
Source publication
High-grade serous ovarian cancer (HGSOC) is an archetypal cancer of genomic instability1–4 patterned by distinct mutational processes5,6, tumour heterogeneity7–9 and intraperitoneal spread7,8,10. Immunotherapies have had limited efficacy in HGSOC11–13, highlighting an unmet need to assess how mutational processes and the anatomical sites of tumour...
Similar publications
Objective：Here, we explored the expression of SYNM by means of Gene Expression Omnibus (GEO) and investigated its prognostic significance as well as potential functions in gastric cancer (GC).
Methods：Toward this goal, differential gene expression analysis, univariate Cox regression, Lasso regression, best subset regression, Gene Set Enrichment Ana...
Citations
... Tumourinfiltrating CD8-positive T cells secrete interferon-γ, which further upregulates PD-L1 expression on OC cells and macrophages [124][125][126]. Additionally, the impact of mutational profile and anatomic sites has been suggested to contribute to the immunosuppressed and hostile TME [127]. However, a recent press release from sponsors of the DUO-O trial suggest significant PFS benefit with the addition of durvalumab to platinum-based chemotherapy and bevacizumab followed by durvalumab/olaparib/bevacizumab maintenance in newly diagnosed advanced OC [128]. ...
Ovarian cancer is the leading cause of gynecological cancer death. Improved understanding of the biologic pathways and introduction of poly (ADP-ribose) polymerase inhibitors (PARPi) during the last decade have changed the treatment landscape. This has improved outcomes, but unfortunately half the women with ovarian cancer still succumb to the disease within 5 years of diagnosis. Pathways of resistance to PARPi and chemotherapy have been studied extensively, but there is an unmet need to overcome treatment failure and improve outcome. Major mechanisms of PARPi resistance include restoration of homologous recombination repair activity, alteration of PARP function, stabilization of the replication fork, drug efflux, and activation of alternate pathways. These resistant mechanisms can be targeted to sensitize the resistant ovarian cancer cells either by rechallenging with PARPi, overcoming resistance mechanism or bypassing resistance pathways. Augmenting the PARPi activity by combining it with other targets in the DNA damage response pathway, antiangiogenic agents and immune checkpoint inhibitors can potentially overcome the resistance mechanisms. Methods to bypass resistance include targeting non-cross-resistant pathways acting independent of homologous recombination repair (HRR), modulating tumour microenvironment, and enhancing drug delivery systems such as antibody drug conjugates. In this review, we will discuss the first-line management of ovarian cancer, resistance mechanisms and potential strategies to overcome these.
... More broadly, RNAseq of matched pairs has previously identified changes in immune-related genes 46 . However, transcriptional subtypes of HGSC 47,48 derived from bulk analyses largely reflect immune cell composition and abundance of fibroblasts rather than intrinsic difference in malignant cells 49,50 . Overall, detailed single-cell analyses of matched pairs will be required to elucidate critical changes in copy number and gene transcription at relapse. ...
The drivers of recurrence and resistance in ovarian high grade serous carcinoma remain unclear. We investigate the acquisition of resistance by collecting tumour biopsies from a cohort of 276 women with relapsed ovarian high grade serous carcinoma in the BriTROC-1 study. Panel sequencing shows close concordance between diagnosis and relapse, with only four discordant cases. There is also very strong concordance in copy number between diagnosis and relapse, with no significant difference in purity, ploidy or focal somatic copy number alterations, even when stratified by platinum sensitivity or prior chemotherapy lines. Copy number signatures are strongly correlated with immune cell infiltration, whilst diagnosis samples from patients with primary platinum resistance have increased rates of CCNE1 and KRAS amplification and copy number signature 1 exposure. Our data show that the ovarian high grade serous carcinoma genome is remarkably stable between diagnosis and relapse and acquired chemotherapy resistance does not select for common copy number drivers.
... GeneVector summarizes co-expression of genes as mutual information between the probability distribution of read counts across cells. We showcase GeneVector on four scRNA datasets produced from a diverse set of experiments: peripheral blood mononuclear cells (PBMCs) subjected to interferon beta stimulation 19 , the Tumor Immune Cell Atlas (TICA) 20 , treatment naive multi-site samples from High Grade Serous Ovarian Cancers (HGSOC) 21 and a time series of cisplatin treatment in patient-derived xenografts (PDX) of triple negative breast cancer (TNBC) 22 . We first confirm GeneVector's ability to identify putatively co-regulated gene pairs from sparse single cell expression measurements using the TICA and PBMC datasets. ...
... GeneVector provides a framework for asking such questions in the form of latent space arithmetic. By defining the difference between two sites as a vector, where the direction defines transcriptional change, we identify metagenes associated with expression loss and gain between primary and metastasis sites from six patients in the Memorial Sloan Kettering Cancer Center SPECTRUM cohort of patients with high-grade serous ovarian cancer (HGSOC) 21 . ...
... A set of 270,833 cells quality control filtered cells from adnexa (primary) and bowel (metastasis) samples were processed with Gene-Vector (Fig. 5A). The unnormalized counts were subset to 2000 highly variable genes were used as input to GeneVector and cells were classified to one of six cell types using gene markers curated for HGSOC and two markers for cancer cells (EPCAM and CD24) 21 . We performed cell type classification (Methods: Cell Type Assignment) and compared GeneVector accuracy to the original annotations generated from Cel-lAssign for each cell type in a confusion matrix (Fig. 5B). ...
Deciphering individual cell phenotypes from cell-specific transcriptional processes requires high dimensional single cell RNA sequencing. However, current dimensionality reduction methods aggregate sparse gene information across cells, without directly measuring the relationships that exist between genes. By performing dimensionality reduction with respect to gene co-expression, low-dimensional features can model these gene-specific relationships and leverage shared signal to overcome sparsity. We describe GeneVector, a scalable framework for dimensionality reduction implemented as a vector space model using mutual information between gene expression. Unlike other methods, including principal component analysis and variational autoencoders, GeneVector uses latent space arithmetic in a lower dimensional gene embedding to identify transcriptional programs and classify cell types. In this work, we show in four single cell RNA-seq datasets that GeneVector was able to capture phenotype-specific pathways, perform batch effect correction, interactively annotate cell types, and identify pathway variation with treatment over time.
... These two pathway activities were correlated with gene expression patterns characteristic of tumor cells from treatment-naive HGSOC patients: Cancer.3 and Cancer.6, respectively 13 . PROGENy signatures and associated gene expression patterns were localized to distinct cell subsets within the epithelial cell embedding ( Fig. 4i and Supplementary Fig. 11a-d). ...
... ; https://doi.org/10.1101/2023.07.17.549422 doi: bioRxiv preprint detected differential ELF3 transcript structures associated with chr1:202011127 A>C in exon 2 for which flanking retained introns were observed among REF reads (1,130 UMIs in 574 cells); whereas ALT reads (2,962 UMIs in 1,013 cells) exhibited normal splicing (Fig. 5d). ELF3 is a transcription factor strongly expressed in epithelial tissue and has been shown to inhibit the epithelial-to-mesenchymal transition 13,38 while supporting angiogenesis 39 . Another example of differential transcript structures linked to an SNV (chr2:190975811 C>A) was observed with the transcription factor STAT1, which exhibited distinct allele-specific events (REF = 2,501 UMIs in 525 cells and ALT = 1,254 UMIs in 353 cells) that spanned both CDS and UTR/Intron (Fig. 5e). ...
Single-cell RNA sequencing predominantly employs short-read sequencing to characterize cell types, states and dynamics; however, it is inadequate for comprehensive characterization of RNA isoforms. Long-read sequencing technologies enable single-cell RNA isoform detection but are hampered by lower throughput and unintended sequencing of artifacts. Here we developed Single-cell Targeted Isoform Long-Read Sequencing (scTaILoR-seq), a hybridization capture method which targets over a thousand genes of interest, improving the median number of unique transcripts per cell by 29-fold. We used scTaILoR-seq to identify and quantify RNA isoforms from ovarian cancer cell lines and primary tumors, yielding 10,796 single-cell transcriptomes. Using long-read variant calling we revealed associations of expressed single nucleotide variants (SNVs) with alternative transcript structures. In addition, phasing of SNVs across transcripts facilitated measurement of allelic imbalance within distinct cell populations. Overall, scTaILoR-seq is a long-read targeted RNA sequencing method and analytical framework for exploring transcriptional variation at single-cell resolution.
... In contrast, mistakes made by other segmentation methods potentially lead to implausible outcomes (Schmidt et al., 2018). An example of StarDist application to HMTI data is segmenting high-grade serous ovarian cancer (HGSOC) images leading to immune recognition and evasion investigations (Vázquez-García et al., 2022). ...
Since its introduction into the field of oncology, deep learning (DL) has impacted clinical discoveries and biomarker predictions. DL-driven discoveries and predictions in oncology are based on a variety of biological data such as genomics, proteomics, and imaging data. DL-based computational frameworks can predict genetic variant effects on gene expression, as well as protein structures based on amino acid sequences. Furthermore, DL algorithms can capture valuable mechanistic biological information from several spatial “omics” technologies, such as spatial transcriptomics and spatial proteomics. Here, we review the impact that the combination of artificial intelligence (AI) with spatial omics technologies has had on oncology, focusing on DL and its applications in biomedical image analysis, encompassing cell segmentation, cell phenotype identification, cancer prognostication, and therapy prediction. We highlight the advantages of using highly multiplexed images (spatial proteomics data) compared to single-stained, conventional histopathological (“simple”) images, as the former can provide deep mechanistic insights that cannot be obtained by the latter, even with the aid of explainable AI. Furthermore, we provide the reader with the advantages/disadvantages of DL-based pipelines used in preprocessing highly multiplexed images (cell segmentation, cell type annotation). Therefore, this review also guides the reader to choose the DL-based pipeline that best fits their data. In conclusion, DL continues to be established as an essential tool in discovering novel biological mechanisms when combined with technologies such as highly multiplexed tissue imaging data. In balance with conventional medical data, its role in clinical routine will become more important, supporting diagnosis and prognosis in oncology, enhancing clinical decision-making, and improving the quality of care for patients.
... The intrinsic heterogeneity of HGSOC may justify our findings. In fact, immune response is an evolutionary process that depends on multiple factors, such as the anatomical site, tumor progression and mutational status [38]. ...
Early stages are under-represented in studies on the molecular and immune features of high-grade serous ovarian carcinoma (HGSOC), and specific studies focused on early-stage HGSOC are required for a better prognostic stratification and to personalize chemotherapy. The aim of this study was to determine the prognostic significance of CD8+ and CD4+ tumor-infiltrating lymphocytes (TILs), tumoral cell PD-L1 expression, BRCA mutational status and tumor mutation burden (TMB) in early-stage HGSOC. A retrospective study was performed on stage I and II HGSOC from the Molecular Reclassification of Early Stages of Ovarian Cancer (RECLAMO) cohort from the Spanish Group of Ovarian Cancer Research (GEICO). Centralized histological typing was performed based on morphological and immunohistochemical features. Intraepithelial (i) and stromal (s) CD8+ and CD4+ T cells and PD-L1 were evaluated on tissue microarrays by immunohistochemistry. BRCA1 and BRCA2 mutation status and TMB were analyzed in tumor DNA using next-generation sequencing. The study included 124 tumors. High iCD8+ (>20 TILs/core), low/intermediate CD4+ (<20 TILs/core) and high CD8+/CD4+ ratio (>35/core) were associated with favorable outcomes. Tumor cell PD-L1 expression (TPS ≥ 1) was present in only 8% of tumors. In total, 11 (16%) and 6 (9%) out of 69 HGSOC tested carried pathogenic or likely pathogenic BRCA1 or BRCA2 mutations, respectively. Median TMB of 40 tumors analyzed was 5.04 mutations/Mb and only 6 tumors had 10 or more mutations/Mb. BRCA status and TMB were not associated with TILs or prognosis. When compared with studies on advanced HGSOC, our results suggested that prognostic variables differed according to stage and that more studies focused on early stages of HGSOC are needed to better stratify these tumors.
... 30)) and methods specifically designed for calling mutations in single-cell data (Monovar 31 and SCReadCounts 32 ). To this end, we used the matched genome sequencing and scRNA-seq data from epithelial cells from seven out of the eight cSCC tumors described above 26 , and from nine kidney and fourteen ovarian tumors 33,34 . In total, we considered 416 mutations detected in WES or whole-genome sequencing (WGS) data from 30 tumors with sufficient coverage in scRNA-seq data for benchmarking. ...
... Similarly, the mutational signatures detected in the ovarian cancer data set using the mutations called in scRNA-seq by SComatic were highly concordant with the WGS data, in stark contrast to the results obtained with the other algorithms ( Fig. 3e and Supplementary Figs. 9 and 10). Moreover, we found an enrichment of mutations attributable to mutational signature 3 (SBS3) in ovarian tumors with homologous repair deficiency (n = 22) compared to tumors with homologous repair proficiency (n = 13), which is consistent with previous studies of homologous repair deficient tumors 33,35,36 , highlighting the power of SComatic to detect clinically relevant mutational processes in scRNA-seq data ( Fig. 3f and Supplementary Fig. 10). ...
... For fitting of COSMIC signatures, we used only the mutational processes known to be operative in each sample type analyzed 7,42 : (1) SBS1, SBS5, SBS6, SBS10a, SBS10b, SBS14, SBS15, SBS17a, SBS17b, SBS18, SBS21, SBS26, SBS28, SBS37, SBS40 and SBS44 for colorectal cancer samples; (2) SBS1, SBS2, SBS5, SBS7a, SBS7b, SBS7c, SBS7d, SBS13, SBS32 and SBS40 for cSCC samples; (3) SBS1, SBS2, SBS3, SBS5, SBS8, SBS9, SBS13, SBS18, SBS26, SBS40 and SBS44 for ovarian cancer samples 33 ; and (4) SBS1, SBS2, SBS4, SBS5, SBS7a, SBS7b, SBS13, SBS16, SBS17b, SBS18, SBS22, SBS23, SBS32, SBS40, SBS41 and SBS88 for MPNs and non-neoplastic samples. We also included SBS6, SBS8, SBS19, SBS32, SBS35, SBS39 and SBS44 when analyzing heart samples 46 . ...
Characterization of somatic mutations at single-cell resolution is essential to study cancer evolution, clonal mosaicism and cell plasticity. Here, we describe SComatic, an algorithm designed for the detection of somatic mutations in single-cell transcriptomic and ATAC-seq (assay for transposase-accessible chromatin sequence) data sets directly without requiring matched bulk or single-cell DNA sequencing data. SComatic distinguishes somatic mutations from polymorphisms, RNA-editing events and artefacts using filters and statistical tests parameterized on non-neoplastic samples. Using >2.6 million single cells from 688 single-cell RNA-seq (scRNA-seq) and single-cell ATAC-seq (scATAC-seq) data sets spanning cancer and non-neoplastic samples, we show that SComatic detects mutations in single cells accurately, even in differentiated cells from polyclonal tissues that are not amenable to mutation detection using existing methods. Validated against matched genome sequencing and scRNA-seq data, SComatic achieves F1 scores between 0.6 and 0.7 across diverse data sets, in comparison to 0.2–0.4 for the second-best performing method. In summary, SComatic permits de novo mutational signature analysis, and the study of clonal heterogeneity and mutational burdens at single-cell resolution.
... HGSOC deconvolution results are concordant across datasets, reference profiles, and platforms samples from Tothill et al. [10] (Table 1). We performed analyses in parallel using different reference profiles: HGSOC-Penn/Utah data [29,30], and data sampled from Vázquez-García et al [31]. Each reference profile comprised gene expression signatures from scRNA-seq data annotated by cell type (see Methods). ...
... In turn, Tothill had a higher average content of most other cell types, particularly endothelial cells. To assess whether deconvolution results were affected by the reference profile used, we performed parallel analyses using a representative 10% sample of single-cell data from Vázquez-García et al [31]. This reference profile was grouped into fewer cell types by the original authors than the HGSOC-Penn/Utah reference profile, meaning an exact comparison was not possible, but the deconvolution estimates were qualitatively similar in broad profiles ( Fig. 1D-F). ...
... We also used data from Vázquez-García et al [31] (Gene Expression Omnibus (GEO) ID GSE180661) as an independent reference profile. This dataset was too large to feasibly analyze with BayesPrism, which is an accurate but computationally expensive deconvolution algorithm, in its entirety. ...
Ovarian cancer is a deadly disease with few effective therapies. The most common form is high-grade serous ovarian cancer (HGSOC). Transcriptomic subtypes of HGSOC have shown promise in characterizing tumor heterogeneity and are associated with survival. Gene expression signatures for the subtypes suggest variation in stromal cell types in the tumor microenvironment (TME). Here, we characterize the TME composition of HGSOC on a population scale by performing deconvolution on bulk transcriptomic data. We use comprehensive cell type profiles from 164 HGSOC tumor samples from two independent reference datasets, in order to compare cell type proportions across and within bulk transcriptomic datasets, and assess their alignment to the subtypes proposed by The Cancer Genome Atlas. We also assess the relationship between tumor composition and clinical outcomes. Our results suggest that HGSOC transcriptomic subtypes are driven by TME composition, specifically fibroblast and immune cell content, and we propose a modified HGSOC subtype model informed by cell composition.
... Beyond the already-known proteins, such as Ku70/ 80, DNA-PKcs, Artemis, DNA pol λ/μ, DNA ligase IV-XRCC4, and XLF, new proteins are involved in the NHEJ, namely PAXX, MRI/CYREN, TARDBP of TDP-43, IFFO1, ERCC6L2, and RNase H2. Among them, MRI/CYREN has a dual role, as it stimulates NHEJ in the G1 phase of the cell cycle and inhibits the pathway in the S and G2 phases [43] . In addition to this genomic complexity, a challenging clinical feature of HGSOC is the presence of widespread intraperitoneal disease at the time of diagnosis. ...
Epithelial ovarian cancer (EOC) is the most lethal gynaecological malignancy, and despite advancements in therapeutics, most women unfortunately still succumb to their disease. Immunotherapies, in particular immune checkpoint inhibitors (ICI), have been therapeutically transformative in many tumour types, including gynaecological malignancies such as cervical and endometrial cancer. Unfortunately, these therapeutic successes have not been mirrored in ovarian cancer clinical studies. This review provides an overview of the ovarian tumour microenvironment (TME), particularly factors associated with survival, and explores current research into immunotherapeutic strategies in EOC, with an exploratory focus on novel therapeutics in navigating drug resistance.
... Traditionally, ovarian cancer is considered a typical immunosuppressive tumor characterized by lower immune infiltration and a low response rate to immunotherapy. Although several studies have shown that presence of tumorinfiltrating lymphocytes (TILs) is associated with improved clinical out-comes in ovarian cancer patients, human ovarian cancer remains poorly responsive to immunotherapy [6,7]. The expression of multiple coinhibitory receptors on T cell infil-trates as well as redundant immune suppressive mechanisms are actually the culprits to limiting the antitumor efficacy of immune checkpoint inhibitors in ovarian cancer [8]. ...
Background:: A plateau has been reached for conventional therapies in ovarian cancer as there is no definitive increase in overall survival. It is necessary to distinguish molecular subtypes based on multi-omics integration to take more targeted strategies to improve prognosis and increase therapeutic efficacy. Methods:: A total of 202 patients with mRNA, miRNA, DNA methylation, somatic mutation, and Reverse Phase Protein Array (RPPA) data were assembled into one multi-omics dataset and then developed a model via MOVICS. A total of 1028 OV samples with raw microarray data and clinicopathological information were obtained from the Gene Expression Omnibus (GEO) as a validation cohort. Then, a series of immune infiltration analyses were performed by GSVA, TIDE, EaSIeR, and ESTIMATE package. Lastly, Metascape was conducted to unravel activated signaling pathways, and the ChAMP package was used to identify potential biomarkers. Results:: 202 OV samples retrieved from the multi-omics TCGA cohort were categorized into two subgroups by MOVICS. Drug sensitivity analysis revealed that targeted therapy and immunotherapy are possibly considered as preferable approaches to prolong the survival of such patients as C2 while platinum-based drugs to C1. Furthermore, immune infiltration analysis revealed that there is a negative correlation between myeloid and neutrophil cells (MDSC) and prognosis and response to immunotherapy in C1. Immune-associated pathways were mainly enriched in C1 and metabolic-related pathways in C2. CRISPLD2, SLC12A8, STARD8, and DCHS1 were considered as potential targets to possibly improve the outcome by disrupting the immunosuppressive immunocytes by integration of DNA methylation and transcriptome expression. Conclusion:: The molecular subtypes based on integrated multi-omics provided novel insights into the molecular basis of immune recognition and immune regulation of cancer cells for predicting the prognosis of ovarian cancer and potential clinical therapeutic targets for OV.
