Karen Gambaro

The Genomic Landscape of TP53 and p53 Annotated High Grade Ovarian Serous Carcinomas from a Defined Founder Population Associated with Patient Outcome

High-grade ovarian serous carcinomas (HGSC) are characterized by TP53 mutations and non-random patterns of chromosomal anomalies, where the nature of the TP53 mutation may correlate with clinical outcome. However, the frequency of common somatic genomic events occurring in HGSCs from demographically defined populations has not been explored. Whole genome SNP array, and TP53 mutation, gene and protein expression analyses were assessed in 87 confirmed HGSC samples with clinical correlates from French Canadians, a population exhibiting strong founder effects, and results were compared with independent reports describing similar analyses from unselected populations. TP53 mutations were identified in 91% of HGSCs. Anomalies observed in more than 50% of TP53 mutation-positive HGSCs involved gains of 3q, 8q and 20q, and losses of 4q, 5q, 6q, 8p, 13q, 16q, 17p, 17q, 22q and Xp. Nearly 400 regions of non-overlapping amplification or deletion were identified, where 178 amplifications and 98 deletions involved known genes. The subgroup expressing mutant p53 protein exhibited significantly prolonged overall and disease-free survival as compared with the p53 protein null subgroup. Interestingly, a comparative analysis of genomic landscapes revealed a significant enrichment of gains involving 1q, 8q, and 12p intervals in the subgroup expressing mutant p53 protein as compared with the p53 protein null subgroup. Although the findings show that the frequency of TP53 mutations and the genomic landscapes observed in French Canadian samples were similar to those reported for samples from unselected populations, there were differences in the magnitude of global gains/losses of specific chromosomal arms and in the spectrum of amplifications and deletions involving focal regions in individual samples. The findings from our comparative genomic analyses also support the notion that there may be biological differences between HGSCs that could be related to the nature of the TP53 mutation.

Small cell ovarian carcinoma: genomic stability and responsiveness to therapeutics

BackgroundThe biology of small cell ovarian carcinoma of the hypercalcemic type (SCCOHT), which is a rare and aggressive form of ovarian cancer, is poorly understood. Tumourigenicity, in vitro growth characteristics, genetic and genomic anomalies, and sensitivity to standard and novel chemotherapeutic treatments were investigated in the unique SCCOHT cell line, BIN-67, to provide further insight in the biology of this rare type of ovarian cancer.MethodThe tumourigenic potential of BIN-67 cells was determined and the tumours formed in a xenograft model was compared to human SCCOHT. DNA sequencing, spectral karyotyping and high density SNP array analysis was performed. The sensitivity of the BIN-67 cells to standard chemotherapeutic agents and to vesicular stomatitis virus (VSV) and the JX-594 vaccinia virus was tested.ResultsBIN-67 cells were capable of forming spheroids in hanging drop cultures. When xenografted into immunodeficient mice, BIN-67 cells developed into tumours that reflected the hypercalcemia and histology of human SCCOHT, notably intense expression of WT-1 and vimentin, and lack of expression of inhibin. Somatic mutations in TP53 and the most common activating mutations in KRAS and BRAF were not found in BIN-67 cells by DNA sequencing. Spectral karyotyping revealed a largely normal diploid karyotype (in greater than 95% of cells) with a visibly shorter chromosome 20 contig. High density SNP array analysis also revealed few genomic anomalies in BIN-67 cells, which included loss of heterozygosity of an estimated 16.7 Mb interval on chromosome 20. SNP array analyses of four SCCOHT samples also indicated a low frequency of genomic anomalies in the majority of cases. Although resistant to platinum chemotherapeutic drugs, BIN-67 cell viability in vitro was reduced by >75% after infection with oncolytic viruses.ConclusionsThese results show that SCCOHT differs from high-grade serous carcinomas by exhibiting few chromosomal anomalies and lacking TP53 mutations. Although BIN-67 cells are resistant to standard chemotherapeutic agents, their sensitivity to oncolytic viruses suggests that their therapeutic use in SCCOHT should be considered.

VGLL3 expression is associated with a tumor suppressor phenotype in epithelial ovarian cancer

Previous studies have implicated vestigial like 3 (VGLL3), a chromosome 3p12.3 gene that encodes a putative transcription co‐factor, as a candidate tumor suppressor gene (TSG) in high‐grade serous ovarian carcinomas (HGSC), the most common type of epithelial ovarian cancer. A complementation analysis based on microcell‐mediated chromosome transfer (MMCT) using a centric fragment of chromosome 3 (der3p12‐q12.1) into the OV‐90 ovarian cancer cell line haploinsufficient for 3p and lacking VGLL3 expression was performed to assess the effect on tumorigenic potential and growth characteristics. Genetic characterization of the derived MMCT hybrids revealed that only the hybrid that contained an intact VGLL3 locus exhibited alterations of tumorigenic potential in a nude mouse xenograft model and various in vitro growth characteristics. Only stable OV‐90 transfectant clones expressing low levels of VGLL3 were derived. These clones exhibited an altered cytoplasmic morphology characterized by numerous single membrane bound multivesicular‐bodies (MVB) that were not attributed to autophagy. Overexpression of VGLL3 in OV‐90 was achieved using a lentivirus‐based tetracycline inducible gene expression system, which also resulted in MVB formation in the infected cell population. Though there was no significant differences in various in vitro and in vivo growth characteristics in a comparison of VGLL3‐expressing clones with empty vector transfectant controls, loss of VGLL3 expression was observed in tumors derived from mouse xenograft models. VGLL3 gene and protein expression was significantly reduced in HGSC samples (>98%, p < 0.05) relative to either normal ovarian surface epithelial cells or epithelial cells of the fallopian tube, possible tissues of origin of HGSC. Also, there appeared to be to be more cases with higher staining levels in stromal tissue component from HGSC cases that had a prolonged disease‐free survival. The results taken together suggest that VGLL3 is involved in tumor suppressor pathways, a feature that is characterized by the absence of VGLL3 expression in HGSC samples.

Low levels of IGFBP7 expression in high-grade serous ovarian carcinoma is associated with patient outcome.

BackgroundInsulin-like growth factor binding protein 7 (IGFBP7) has been suggested to act as a tumour suppressor gene in various human cancers, yet its role in epithelial ovarian cancer (EOC) has not yet been investigated. We previously observed that IGFBP7 was one of several genes found significantly upregulated in an EOC cell line model rendered non-tumourigenic as consequence of genetic manipulation. The aim of the present study was to investigate the role of IGFBP7 in high-grade serous ovarian carcinomas (HGSC), the most common type of EOC.MethodsWe analysed IGFBP7 gene expression in 11 normal ovarian surface epithelial cells (NOSE), 79 high-grade serous ovarian carcinomas (HGSC), and seven EOC cell lines using a custom gene expression array platform. IGFBP7 mRNA expression profiles were also extracted from publicly available databases. Protein expression was assessed by immunohistochemistry of 175 HGSC and 10 normal fallopian tube samples using tissue microarray and related to disease outcome. We used EOC cells to investigate possible mechanisms of gene inactivation and describe various in vitro growth effects of exposing EOC cell lines to human recombinant IGFBP7 protein and conditioned media.ResultsAll HGSCs exhibited IGFBP7 expression levels that were significantly (p = 0.001) lower than the mean of the expression value of NOSE samples and that of a whole ovary sample. IGFBP7 gene and protein expression were lower in tumourigenic EOC cell lines relative to a non-tumourigenic EOC cell line. None of the EOC cell lines harboured a somatic mutation in IGFBP7, although loss of heterozygosity (LOH) of the IGFBP7 locus and epigenetic methylation silencing of the IGFBP7 promoter was observed in two of the cell lines exhibiting loss of gene/protein expression. In vitro functional assays revealed an alteration of the EOC cell migration capacity. Protein expression analysis of HGSC samples revealed that the large majority of tumour cores (72.6%) showed low or absence of IGFBP7 staining and revealed a significant correlation between IGFBP7 protein expression and a prolonged overall survival (p = 0.044).ConclusionThe low levels of IGFPB7 in HGSC relative to normal tissues, and association with survival are consistent with a purported role in tumour suppressor pathways.

Overexpressing the CCL2 chemokine in an epithelial ovarian cancer cell line results in latency of in vivo tumourigenicity

The frequent loss of heterozygosity of chromosome (Chr) 17 in epithelial ovarian cancer (EOC), particularly high-grade ovarian serous carcinomas (HGOSCs), has been attributed to the disruption of known tumour suppressor genes, such as TP53 (17p13), as well as other genes on this chromosome that alone or in combination have a role in EOC. In a transcriptome analysis of Chr17 genes, we observed significant underexpression of the chemokine CCL2 (17q12) in a small set of HGOSC samples relative to normal ovarian surface epithelial cells and a significant upregulation of CCL2 in the TP53-mutated OV-90 EOC cell line rendered non-tumourigenic as a consequence of genetic manipulation. Here, we report that overexpressing CCL2 in OV-90 resulted in latency of tumour formation at intraperitoneal (i.p.) but not subcutaneous sites in a mouse xenograft model. Overexpressing CCL2 affected cell morphology and exerted modest, but not significant effects on cell viability, colony formation and cell migration. We report significant underexpression of CCL2 by transcriptome analysis (P=0.015) and by immunohistochemistry in 77% of HGOSC samples (n=65). Absent or a very low level of protein expression by immunohistochemistry was also observed in 71% of additional HGOSC samples (n=122). However, CCL2 protein expression did not significantly correlate with overall or disease-free survival. The epithelial cells of normal fallopian tubes, a purported origin of HGOSC, exhibited expression of CCL2 protein by immunohistochemistry. Our results affirm that CCL2 underexpression is a significant feature of HGOSC samples, and that CCL2 overexpression in an EOC cell line model affects tumourigenic potential in the i.p. setting.

Abstract LB-231: Genomic profiling in serial metastatic colorectal tumors identifies copy number alterations and spatio temporal intra-patient heterogeneity profiles associated with clinical response. Q-CROC-01: NCT00984048

Introduction: Colorectal cancer (CRC) is the third leading cause of cancer related deaths primarily due to its resistance to current treatments. Studies aiming at understanding mechanisms of resistance have largely investigated the genomic landscape of primary tumors at diagnosis. However, selective pressures during therapy can lead to the expansion of resistant clones and tumor heterogeneity. This highlights the need to characterize the molecular changes of metastasis over time of treatment and response to decipher tumor evolution and therapeutic resistance mechanisms. Methods: Metastatic liver tissue samples were collected at baseline (pre-biopsies) and at the time of resistance (post-biopsies) in responder and non-responder CRC patients undergoing the same first-line treatment. Paired pre/post biopsies were collected from 14 patients including 4 patients with multiple post-biopsies to assess temporal and spatio-temporal tumor heterogeneity following treatment exposure. Biopsies were profiled using exome and transcriptome sequencing as well as high-density Single-Nucleotide Polymorphism (SNP) array analysis to capture chromosomal anomalies, loss of heterozygosity and copy number (CN) variations. Results: Profiling of 45 samples with both high-density SNP array and exome sequencing revealed 97.4% similarity between both technologies in the identification of genes targeted by copy number changes. Using chemo-naive biopsies, we identified 120 CN gains and 47 CN loss that were significantly associated with patient progression free survival. Integrative analysis with transcriptome data revealed that only 10% of the genomic CN gains and 17% of the CN loss correlated with their gene expression levels. Based on CN variants comparison between paired pre/post treatment samples, we found high temporal intra-patient heterogeneity over time of treatment. Interestingly, we observed a relationship between heterogeneity and tumor response; showing that acquired resistant tumors have the highest temporal variations. Conclusion: This study, using a multi-omic approach to profile serial liver metastatic samples in CRC patients, highlights the genomic changes in tumor composition after treatment exposure and constitutes an innovative approach to identify clinical biomarkers and molecular signatures of resistance. Citation Format: Mathilde Couetoux du Tertre, Maud Marques, Karen Gambaro, Michael Witcher, Benoit Samson, Bernard Lesperance, Yoo-Joung Ko, Richard Dalfen, Eve St-Hilaire, Lucas Sideris, Felix Couture, Sabine Tejpar, Ronald Burkes, Mohammed Harb, Errol Camlioglu, Adrian Gologan, Vincent Pelsser, Andre Constantin, Suzan McNamara, Petr Kavan, Claudia Kleinman, Gerald Batist. Genomic profiling in serial metastatic colorectal tumors identifies copy number alterations and spatio temporal intra-patient heterogeneity profiles associated with clinical response. Q-CROC-01: NCT00984048 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr LB-231.

Abstract 2795: Integrating multiomics discovery approaches to identify biomarkers of therapeutic resistance in metastatic colorectal cancer through analyses of multiple sequential tumor and liquid biopsies; Qcroc01: NCT00984048

Colorectal cancer (CRC) is the 2nd leading cause of cancer related-death in Canada. Clinical responses of metastatic (m)CRC to first-line treatment range from 35 to 60%, but even responders inevitably develop therapeutic resistance. Studies aiming at understanding mechanisms of resistance have largely investigated primary tumors. However, selective pressures during therapy can lead to the expansion of resistant clones and tumor heterogeneity. This highlights the need to characterize the molecular changes of metastasis and plasma over time of treatment and response to decipher tumor evolution and therapeutic resistance mechanisms. In this multicenter study, 52 tissue samples from liver metastasis were collected at baseline (pre-biopsies) and at the time of resistance (post-biopsies) in responder and non-responder mCRC patients (n=44) undergoing the same standard first-line treatments. Multiple post-biopsies also have been harvested in 4 patients, to allow the assessment of tumor heterogeneity and as well as the evolution of the genomic complexity after treatment exposure. Analyses were carried out across multiple omic platforms to identify resistant signatures and characterize molecular changes during treatment. Biopsies were profiled using exome and transcriptome sequencing as well as high-density SNP array analysis to capture chromosomal anomalies, loss of heterozygosity (LOH) and copy number variations (CNV). Additionally, serial blood samples were collected for proteomic, ctDNA and cytokine analysis. Our preliminary analysis of transcriptomes performed on serial biopsies from a set of 11 patients identified genes consistently overexpressed at resistance. Cytogenetics analysis showed similar genomic profiles of matched pre- and post-biopsies and allowed the establishment of LOH and CNV catalogues of liver metastasis, while exome sequencing revealed cumulative somatic mutations over time of treatment, which suggests subclonal and acquired “driver” mutations of resistance. Plasma-derived ctDNA analysis was performed to investigate the mutational status during treatment and whether they correlate with their relative levels in biopsies. Immune gene expression analysis of a test set of 27 metastases revealed strong clustering of 7 metastases due to overexpression of transcripts related to active immune response, allowing to define novel subgroups of patients based on immune response status. Our study, using a multi-omic strategy and integration of independent molecular platforms to profile liver metastasis samples of responder and non-responder mCRC patients, constitutes an innovative approach to identify clinical biomarkers and molecular signature of resistance, which may enhance individualization of cancer medicine and customized therapy. Citation Format: Karen Gambaro, Maud Marques, Ryan Morin, Claudia Kleinman, Michael Witcher, Simon Turcotte, Benoit Samson, Bernard Lesperance, Yoo-Joung Ko, Richard Dalfen, Eve St-Hilaire, Lucas Sideris, Felix Couture, Sabine Tejpar, Ronald Burkes, Mohammed Harb, Francine Aubin, Thierry Alcindor, Errol Camlioglu, Adriana Aguilar, Mathilde Couetoux du Tertre1, Suzan McNamara, Adrian Gologan, Petr Kavan, Gerald Batist. Integrating multiomics discovery approaches to identify biomarkers of therapeutic resistance in metastatic colorectal cancer through analyses of multiple sequential tumor and liquid biopsies; Qcroc01: NCT00984048 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2795. doi:10.1158/1538-7445.AM2017-2795

Abstract 3059: CCL2 overexpression in an epithelial ovarian cancer cell line results in reduced tumorigenic potential

Proceedings: AACR 102nd Annual Meeting 2011‐‐ Apr 2‐6, 2011; Orlando, FL Chromosome 17 abnormalities are frequently observed in epithelial ovarian cancer (EOC), suggesting that one or more genes mapping to this chromosome contribute to EOC development and/or progression. We have previously performed a comparative transcriptome analysis of primary cultures of normal ovarian surface epithelial (NOSE) cells and malignant, serous subtype ovarian tumor (TOV) samples in order to identify differentially expressed chromosome 17 genes. CCL2, at 17q12, was significantly differentially underexpressed at least 3-fold in all TOV samples. We have confirmed underexpression in an independent series of 99 serous EOC tumor samples. Interestingly, CCL2 was identified as a significantly up-regulated gene occurring as a consequence of tumor suppression in the characterization of a novel EOC cell line model developed by transferring chromosome 3 fragments into the tumorigenic OV-90 EOC cell line. We have further characterized CCL2 expression in a series of EOC cell lines that differ in their tumorigenic potential, and have shown the lowest levels of gene expression occur in tumorigenic cell lines. DNA sequencing did not reveal any evidence of mutations. To determine if overexpression affects tumorigenicity or growth characteristics, we have generated stable overexpressing OV-90 clones using a commercially available pDream: CCL2 expression vector. Overexpression was verified by ELISA. In cell culture, the CCL2 overexpressing clones appeared less tightly packed and more elongated as compared with OV-90 cells. The cell viability of the CCL2 clones appears to be reduced when compared to the parental OV-90 cell line in XTT cell proliferation assays. The ability of the CCL2 clones to form colonies was also reduced as compared to OV-90 in a colony formation assay. There is also evidence of an increase in the percentage of senescent cells in the CCL2 clones based on a beta-galactosidase staining assay. Both CCL2 overexpressing clones exhibit latency in tumor formation in intraperitoneal tumorigenicity assays involving SCID mice as compared with the parental OV-90 cell line. This is the first report of stable CCL2 transfection into an EOC cell line. Our results suggest that CCL2 expression may reduce the tumorigenic potential of EOC cells, and draws attention to the role of the innate immune system and the tumor microenvironment in EOC development and progression. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3059. doi:10.1158/1538-7445.AM2011-3059