Search

Effective identification of genomic variations is a crucial step to understand the relationship between genotype and phenotype and it can yield important insights into rare diseases and cancer biology. Exome and whole-genome sequencing are now routinely used in clinics for detection of disease causing variants in Mendelian disorders. Despite great success achieved using these methods, the diagnostic rate for identifying the right variant is ~25-50%. Recent studies have shown that using RNA sequencing (RNA-seq) the diagnostic rate can be improved and can be used as a complementary method for clinical diagnostic. In this study, we have developed a pipeline to detect germline variants from RNA-seq data. The pipeline steps include: pre-processing, alignment, GATK best practices for RNA-seq and variant filtering. The pre-processing step includes base and adapter trimming and removal of contamination reads from rRNA, tRNA, mitochondrial DNA and repeat regions. The read alignment of the pre-processed reads is performed using STAR/HiSAT. After this we used GATK best practices for the RNA-seq dataset to call germline variants. We benchmarked our pipeline on NA12878 RNA-seq data downloaded from SRA (SRR1258218). After variant calling, the quality passed variants were compared against the gold standard variants provided by GIAB consortium. Of the total ~3.6 million high quality variants reported as gold standard variants for this sample (considering whole genome), our pipeline identified ~58,104 variants to be expressed in RNA-seq. Our pipeline achieved more than 99% of sensitivity in detection of germline variants.

Familial Adenomatous Polyposis (FAP) is characterized by the manifestation of adenomatous polyps in the colon and rectum at an early age (mean age of 16), which if left untreated, leads to aggressive and fatal tumors by the age of 40 years. FAP is caused by autosomal dominant inheritance of germ line mutations in the Adenomatous polyposis coli (APC) gene, a well characterized tumor suppressor gene. Over fifty percent of FAP affected individuals with germline mutations eventually develop colon cancer and therapeutic intervention to prevent cancer progression remains a major unmet medical need. In the last five years, therapies aimed at restoring or enhancing the host’s immune response to treat cancers has gained momentum. Cancer immunotherapy triggers a patient’s immune system to destroy tumor cells (apoptosis) by recognizing tumor-derived neoantigens, presented on the tumor cell surface as peptides bound to class I and II major histocompatibility complex (MHC). Our approach in this study was to determine if FAP could be targeted by immunotherapeutic approaches to reduce polyp numbers, thereby limiting the risk of progression to colorectal cancer. In this study, we identified a novel germline mutation in the APC gene in 10/26 members of a FAP-affected family. To find out if peptides derived from the novel APC mutation could induce a cytotoxic T cell response, peptides harboring the variant amino acids were first interrogated in silico for their immunogenicity using a proprietary neoepitope prioritization pipeline, OncoPeptVAC. A single 9-mer peptide was predicted to be immunogenic. Remarkably, CD8+ T cells isolated from either a FAP+/ APCmut individual, or from a FAP-/ APCmut individual, failed to respond to the peptide, whereas those from either an unaffected family member (FAP-/ APCwt) or from healthy unrelated donors with same HLA type, showed a robust response. We conclude that CD8+ T cells from affected individuals carrying this germline APC mutation have been tolerized against the mutation. Additionally, in silico analyses showed that of the 996 previously reported APC gene mutations in FAP, 42% are potentially immunogenic. These immunogenic mutations could provide novel opportunities to treat FAP patients and to delay their progression to colorectal cancer.

Cancer immunotherapy provides durable response to a small subset of treated patients. A variety of approaches are being developed to increase the long term benefit of checkpoint blockade. These include radiation and cytotoxic therapies and use of cancer vaccines among others. Preclinical and clinical studies have demonstrated that cancer vaccines evoke strong anti-tumor immune response by mobilizing CD8+ T-cells. A challenge in the field of cancer vaccines is identifying mutations that are T-cell activating (neoepitopes). Advances in next generation sequencing permit accurate detection of cancer mutations, even when present at a low frequency. However, neoepitope prediction involves a large number of steps many of which cannot be accurately modeled. In humans, class-I peptides, 9-11–mer in length are presented by HLA – A, B and C alleles and activate CD8+ T-cells. Class-II peptides are 14-17-mer, presented by DPA, DPB, DQA, DQB, DRA and DRB alleles and activate CD4+ T-cells.The current method to identify epitopes (peptides) depends primarily on HLA binding prediction algorithm. Our analysis of 9mer peptides from IEDB database showed that there is no difference in binding affinity of peptide that can activate (immunogenic) and that cannot activate (non-immunogenic) the T-cells. The specificity to identify immunogenic peptide using HLA binding based method is only 27.59%. In this study, we present a novel approach using machine learning technique that can predict whether the peptide will be immunogenic or not. We have generated the model using features generated from amino-acid composition, HLA-binding, structural features, peptide processing and peptide transport. Our model achieved an overall accuracy of 77.30% with a specificity of 92.24% on unseen dataset.

Neoepitopes are tumor-derived immunogenic peptides that arise from intracellular proteolytic processing of somatic mutation in protein coding genes. These peptides bind HLA Class I proteins and are presented on the surface by antigen presenting cells. Productive engagement of HLA Class I-bound peptide with T-cell receptor (TCR) activates CD8+ T-cells to generate cytotoxic T cells, which mediate lysis of the neoantigen-expressing tumor cells. Neoepitopes can be used as cancer vaccines to prime CD8+ T-cells against tumor cells. As tumors accumulate hundreds of mutations during cancer development and only a small subset of these are immunogenic, identifying the neoepitopes requires accurate modeling of the steps involved in peptide production, presentation as well as TCR binding. Most current pipelines prioritize neoepitopes based on the expression of mutant proteins and their HLA binding affinity. However, presentation of peptides on the surface is not sufficient to activate T-cells. To improve the predictive power of our neoepitope prioritization pipeline and to circumvent selection biases that are inherent with using HLA binding as a proxy for neoepitope prediction, we developed a novel algorithm that prioritizes peptide interactions with the TCR. OncoPeptVAC algorithm uses features selected by analyzing crystal structures of TCR and HLA-peptide complex present in the Protein Data Bank. A neural network model was built to derive a composite score that includes, besides TCR binding, other features associated with the peptide, such as level of expression of the mutant allele, affinity of HLA binding, affinity of TAP binding and sensitivity to proteasomal processing. We validated our neural network prediction model on known immunogenic and non-immunogenic peptides and achieved superior accuracy, sensitivity and specificity of prediction compared to using the standard HLA binding affinity of ≤500 nM. By applying OncoPeptVAC neoepitope prioritization solution to 2.2 million unique somatic mutations, we identified ~700 immunogenic peptides derived from recurrent somatic mutations in all cancers. Several of these peptides were validated on a CD8+ T cell-activation assay. Our in silico prioritization platform combined with the cell-based validation method is a powerful tool to identify therapeutic vaccines for personalized cancer immunotherapy applications.

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common malignancy globally, but ranks first in India because of the extensive use of tobacco, betel-quid and alcohol. Despite, the availability of aggressive treatments, the survival for HNSCC patients remains relatively poor. High-throughput sequencing technologies have identified several mutant genes and pathways in cancer genomes but the transcription patterns of these genetic alterations remain unclear. We performed an integrated analysis of whole exome and RNA sequencing to gain a better understanding of the transcriptional consequences of these genetic mutations, thus paving the way to identifying biomarkers and neo-epitopes. To this end, we conducted a prospective study on patients diagnosed with HNSCC from Gujarat, India. Detailed clinical data was collected from a total of 541 consented patients. Additionally, DNA and RNA was extracted from a subset of nine buccal cancer patients from both tumor and blood (Age 44 ± 9: M: 7, F: 2). Whole exome and RNA sequencing was performed on an Illumina platform followed by data analysis using bioinformatics tools. Normalized expression of all coding variants were compiled and variants with Alt allele depth in RNA-seq >= 1 was prioritized for HLA binding and neo-epitope prediction using our proprietary pipeline OncoPeptVAC. As previously reported, DNA sequencing results revealed recurrent frameshift/nonsense and missense mutations in the TP53, PIK3CA, CASP8, FAT1, TTN, FSIP2 and CDKN2A genes. Additionally, unique mutations were observed in the CA6, FGD3, FLG2, COX 4l1, CD58, MAPK1, HLA-A and HLA-DQB1genes, previously not associated with head and neck cancers. A positive correlation between the mutant allele frequency in Exome and RNA-seq data confirmed the expression of both missense and frameshift mutant genes. Our results show that the majority of the mutation derived peptides were not predicted to be immunogenic for the HLA types considered. Peptides derived from variants in TP53, ANKRD12, EPHB2, ZNF200, TRPM4, CHD4, PLEC, KMT2C, GLTSCR1 were predicted to have neo-epitopes properties in buccal cancer. These findings provide a platform to predict immunogenic neo-epitopes which could be used in the development of personalized vaccines for patients with buccal cancer.

Cancer immunotherapy is now established as a major therapeutic modality, and 70% of all cancer patients are estimated to receive some form of an immunotherapy treatment as a part of their disease control by 2025. Several different tumor cell-intrinsic and extrinsic features including the tumor microenvironment, driver gene mutations, host genetics, microbiome and environmental factors modulate response to immune checkpoint inhibitors. The tumor microenvironment plays the most significant role harbouring various components that influence the malignancy of a disease. Clinical investigations on tumor infiltrating immune cells have established the roles of cytotoxic T cells (CTLs) and tumor-associated macrophages (TAMs) in several cancers. It is now strongly evident that cancer outcome and response to therapy is guided by diverse immune cell activity in tumors. We have developed an in silico platform to profile immune cell type for tumor samples from gene expression data. Using an integrated knowledge based method on pure cell type gene expression samples, we have created gene signatures for 8 different immune cell types (Bcell, CD4, CD8, Monocyte, Treg, Macrophage, Neutrophil, NK) summing up to a total of 42 genes. The identified signature was validated on pure cell samples from RNA-seq, microarray and IHC. Comparison of our immune signatures with some of the published signatures showed better performance of our signatures. We applied our in silico platform on the TCGA cancer samples. Our analysis showed that patients with higher CD8 score have better survival in bladder carcinoma, endocervical adenocarcinoma, head & neck, sarcoma, melanoma, thyroid, whereas, patients with higher monocyte score have poor survival in endocervical adenocarcinoma, kidney renal clear cell carcinoma, kidney renal papillary cell carcinoma, lower grade glioma and liver hepatocellular carcinoma.

Hemophagocytic lymphohistiocytosis (HLH) is a rare disorder of uncontrolled cytotoxic T-lymphocyte and macrophage activation associated with extreme inflammation, which if left untreated is invariably fatal. HLH can be classified either as primary (familial) or secondary (acquired). Genetic mutations underlie primary HLH which typically occurs in young children, while secondary HLH is associated with infections, metabolic disease, and malignancies and manifests later in life. Both categories of HLH affect the function of T-lymphocytes and natural killer (NK) cells. Homozygous null mutations in several genes including PRF1, UNC13D (MUNC13-4), STX11, and STXBP2 (MUNC18-2) have been described in cases of primary HLH, where consanguinity has been shown to play a role. These mutations contribute to an uncontrolled inflammatory response, over production of interferon gamma (IFNγ) and pro-inflammatory cytokines leading to macrophage activation and eventually to tissue and organ damage. While no known genetic mutations have been described in patients with acquired HLH, an underlying genetic predisposition has been suggested. In an effort to better understand the genetic and clinical correlates of this disease, we performed retrospective analyses of 60 suspected HLH cases that were submitted by physicians for genomic analysis to our diagnostic laboratory. High throughput exome sequencing of the 60 HLH samples was carried out on an Illumina HiSeq 2500 platform, followed by bioinformatics analysis. Our data showed that 1) 34% of the patients had homozygous mutations in known HLH-associated genes, the majority of which were in the Perforin1 (PRF1) gene (70%) followed by mutations in the UNC13D (20%) and STXBP2 (10%) genes. Most patients with homozygous PRFI mutations were products of consanguineous marriages. The parents of these patients while being carriers, however, remained unaffected. 2) 18% of our patients had mutations in other congenital immune deficiency syndromes that are known to lead to HLH. 3) The remaining 48% of the patients lacked any known HLH-associated mutation. Our data suggest that heterozygous mutations in novel genes may contribute to predisposing patients to secondary HLH, particularly since most of these genes are associated with T-lymphocyte activation, which plays a vital role in the etiology of this devastating disease.

Esophageal squamous cell carcinoma (ESCC) is one of the most common cancers with high mortality rate. Smoking is one of the established risk factors of ESCC. However, there is limited data on molecular alterations associated with cigarette smoke exposure in esophageal cells. Understanding the effects of cigarette smoke on esophageal squamous epithelial cells at a molecular level would lead to a better understanding of the pathobiology of ESCC which has implications for identification of early biomarkers and therapeutic targets. To investigate the effect of cigarette smoke exposure, we developed a cell line model where Het1A cells (non-neoplastic human esophageal epithelial cells) were chronically treated with cigarette smoke condensate (CSC) for 2 months, 4 months, 6 months and 8 months. We carried out comparative proteomic, phosphoproteomic and whole exome sequencing analyses on CSC treated and untreated cells. Increased cell proliferation, invasion and anchorage independent growth of Het1A cells was observed after chronic exposure to cigarette smoke. Using quantitative proteomic and phosphoproteomic analyses, we identified 35 proteins and 118 phosphoproteins that showed differential expression. Bioinformatics analysis of differentially expressed proteins and phosphoproteins showed enrichment of molecules involved in DNA damage response pathway. To further understand the mutational burden associated with cigarette smoke, we did whole exome sequencing of CSC treated and untreated cells which also revealed mutations and copy number alterations in genes associated with DNA damage response. By correlating WES, proteomic and phosphoproteomic results, we observed potential loss of function in HMGN2 and MED1 that were reported as potential tumor suppressors and are known to play important role in DNA damage response. We also observed decreased expression of HMGN2 in tissue section of ESCC. Overexpression of HMGN2 and MED1 lead to decreased proliferative and invasive ability of CSC treated cells. These findings suggest that cigarette smoke affects genes and proteins associated with DNA damage response pathways which might play a vital role in development of ESCC.

Tobacco usage is a known risk factor associated with development of oral cancer. It is mainly consumed in two different forms (smoking and chewing) that vary in their composition and methods of intake. Despite being the leading cause of oral cancer, molecular alterations induced by tobacco are poorly understood. To investigate the adverse effects of cigarette smoke/chewing tobacco exposure in oral keratinocytes, we developed two cellular models where normal oral keratinocytes were chronically exposed to cigarette smoke and chewing tobacco for a period of 8 months. Cellular assays reveal that OKF6/TERT1 cells acquire an oncogenic phenotype after chronic exposure to cigarette smoke/chewing tobacco. We employed both whole exome sequencing (WES) and quantitative proteomics approaches to investigate the molecular alterations in oral keratinocytes (OKF6/TERT1) chronically exposed to smoke and chewing tobacco. Exome sequencing revealed a much higher rate of C>A transversions in smoke exposed cells in conjunction with previous studies. In contrast, C>G transversions were observed to be higher in chewing tobacco exposed cells. Diverse mutations in both treated cells further highlight the distinct effects of each exposure. Distinct proteomic alterations were observed in smoke and chewing tobacco exposed cells compared to parental cells. In addition, we observe enrichment of different signaling cascades in transformed oral cells upon chronic exposure to either cigarette smoke or chewing tobacco. Current analysis defines a clear distinction in the molecular dysregulation in oral cells in response to different tobacco-based insults. Future studies are needed to validate some of the genetic and proteomic alterations unique to each form of tobacco exposure. This study can serve as a reference for fundamental damage on oral cells as a consequence of exposure to different forms of tobacco.

Colorectal cancer (CRC) is an aging related disease with 72% of newly diagnosed cases aged 65 years or older (Cancer Registration Statistics, 2012). Past few decades of research from the West have identified aberrantly activated canonical Wnt/β-catenin signaling and microsatellite instability (MSI) as the major pathways driving CRC tumorigenesis. In India however majority of cases appear to belong to a younger age group with preponderance of rectal cancer. Molecular genetic screening of early-onset sporadic rectal cancer (EOSRC) performed earlier from our laboratory identified a significant proportion of EOSRC to be driven neither by aberrant Wnt signaling nor MSI. We performed whole exome sequencing of 27 tumor/normal pairs obtained from surgically resected rectal adenocarcinomas (microsatellite stable with no aberrant Wnt signaling) using the Illumina HiSeq 2500 platform. We identified recurrently mutated genes in EOSRC including known tumor suppressors (TSG) and oncogenes such as TP53, APC, KRAS, SMAD4 and PIK3CA. More importantly, we discovered mutations in uncharacterized TSGs and oncogenes such as ARID2, FAT3, FAT4, ZEB2 and TRPC7. These included putative oncogenic mutations in the zinc finger domains of ZEB2, a DNA-binding transcriptional repressor that promotes epithelial to mesenchymal transition in tumors. Functional work is underway to characterize the effect of these mutations on CRC tumor progression. APC mutations detected in this study were present in major population of cells, yet were not driving β-catenin to the nucleus. This observation suggests β-catenin degradation independent tumor suppressor function of APC, which is being validated.

Epidermal growth factor receptor (EGFR) is overexpressed in 90% of head and neck squamous cell carcinomas (HNSCC). However, most clinical trials with tyrosine kinase inhibitors (TKIs) have shown modest response rates due to development of acquired resistance. We performed whole exome sequencing (WES) of an isogenic pair of erlotinib-sensitive (SCC-S) and resistant (SCC-R) HNSCC cell lines to elucidate the molecular mechanisms that govern acquired resistance to erlotinib. Exome sequencing resulted in identification of 148 non-synonymous single nucleotide variants (SNVs) in 139 genes and copy number alterations (CNA) (≥2 fold) affecting 339 genes in SCC-R cells compared to SCC-S cells. Comparison of SNVs from SCC-R against post-translational modification databases resulted in identification of loss of ubiquitinylation site at p.K57E in dual specificity mitogen-activated protein kinase kinase 1 (MAP2K1) which was validated using in-house high-throughput proteomic data. Substitution mutation K57N in MAP2K1 is shown to result in its constitutive activation and subsequent gefitinib resistance in lung adenocarcinoma. We also identified a well-known driver mutation p.G13R in Harvey rat sarcoma viral oncogene homolog (HRAS) (AllF: 20.41%). In addition, we also observed CNA in other genes of this pathway including RAC-beta serine/threonine-protein kinase (AKT2), glycogen synthase kinase-3 alpha (GSK3A), Rho guanine nucleotide exchange factor 1 (ARHGEF1) amongst others. Corresponding protein expression changes of these genes were also observed in proteomics data. Quantitative phosphoproteomics revealed hyperphosphorylation of other proteins involved in MAPK pathway such as serine/threonine-protein kinase B-raf (BRAF), MAP2K2, mitogen-activated protein kinases such as MAPK1 and MAPK3. Integrative multi-omic analysis revealed constitutive activation of key intermediates of MAPK pathway in SCC-R cells compared to SCC-S cells which may be essential in the development of acquired resistance to erlotinib in these cells. We hypothesize that combinatorial treatment regime involving inhibition of putative targets such as MAP2K1 with erlotinib therapy may aid in tackling acquired erlotinib resistance in HNSCC patients.