Citations

Abstract A better understanding of gene-specific risks for development of breast cancer will lead to improved screening, prevention, and therapeutic strategies for individuals identified to carry germline mutations. We performed targeted massively-parallel sequencing to identify mutations and large genomic rearrangements in 26 known or proposed breast cancer susceptibility genes in 2134 BRCA-negative women with familial breast cancer (FBC). A case-control analysis was performed comparing the frequency of internally classified mutations identified in FBC cases to that in non-Finnish European controls from the Exome Aggregation Consortium (ExAC) excluding samples from The Cancer Genome Atlas. Including large genomic rearrangements, mutations were identified in 8.2% of FBC cases compared to 6.2% of ExAC controls, including mutations in high-penetrance genes (0.6% in cases vs. 0.1% in controls), moderate-penetrance genes (3.7% vs 1.7%), and seven cases with two mutations (0.3%). The remainder of FBC cases and ExAC controls had mutations in proposed breast cancer genes (1.6% of cases vs 2.4% of controls), Lynch syndrome genes (0.5% vs. 0.5%) or were heterozygous MUTYH carriers (1.5% vs. 1.5%). Case-control analysis demonstrated significant associations with FBC for ATM, PALB2, and TP53 mutations (OR>3.0, p<10-4), BARD1 mutations (OR=3.2, p=0.012), and CHEK2 truncating mutations (OR=1.6, p=0.041). Our results therefore demonstrate that only approximately 4% of BRCA1/2 negative FBC patients have mutations in genes definitively associated with breast cancer at this time. Large case-control studies are needed to fully evaluate the breast cancer risks associated with moderate penetrance and proposed breast cancer susceptibility genes. Citation Format: Kara N. Maxwell, Thomas Paul Slavin, Jenna M. Lilyquist, Joseph Vijai, Susan L. Neuhausen, Steven N. Hart, Vignesh Ravichandran, Tinu Thomas, Ann Maria, Kasmintan A. Schrader, Raymond Moore, Chunling Hu, Brad Wubbenhorst, Brandon M. Wenz, Kurt D'Andrea, Susan M. Domchek, Mark E. Robson, Paulo Peterlongo, Paolo Radice, James M. Ford, Judy E. Garber, Csilla Szabo, Kenneth Offit, Katherine L. Nathanson, Fergus J. Couch, Jeffrey N. Weitzel. Risks of familial breast cancer associated with known and proposed breast cancer susceptibility genes [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 4265. doi:10.1158/1538-7445.AM2017-4265

Abstract Benign breast disease, a known risk factor for subsequent breast cancer, includes a histological spectrum of lesions that can be subdivided based on the overall impression into non-proliferative lesions, proliferative lesions without atypia, and atypical hyperplasia. Yet little is known about the molecular characteristics of benign breast tissue from African American women, who suffer from disproportionately high breast cancer mortality rates. 72 breast samples from African American women were obtained. RNA was extracted from archival formalin-fixed, paraffin-embedded (FFPE) blocks. Gene expression profiling was performed on these samples using the Affymetrix HTA2.0 exon array chip. A fit linear model using generalized least squares identified probe sets with differential expression in benign-in situ and benign-invasive breast cancer paired lesions. 36 benign and 36 breast cancer paired lesions were included in the study. In the benign-in situ set (n=20), 978 probe sets were differentially expressed (threshold: fold-change, 1.50; p < 0.05). Probe sets mapped to 125 annotated genes, of which 69 (55.2%) were involved in cancer. The benign-invasive set (n=52) identified 351 differentially expressed probe sets that mapped to 84 annotated genes, of which 61 (72.6%) had known cancer significance. Genes involved in cellular growth and proliferation were significantly over-expressed in both benign-in situ and benign-invasive lesions (p-values < 0.0104). In the population of 36 paired samples, differentially expressed genes were found to regulate the estrogen receptor (ESR1) through distinct network associations. Many genes are differentially expressed by benign and breast cancer lesions, and in situ/invasive histology contribute to the distinct molecular characteristics and pathways of breast malignancy. Our findings highlight the first study to assess expression profiling of paired benign and breast cancer lesions among African American women. Citation Format: Andreana Natalie Holowatyj, Julie J. Ruterbusch, Rouba Ali-Fehmi, Sudeshna Bandyopadhyay, Christine Mehner, Melody Stallings Mann, Gregory Dyson, Derek Radisky, Michele L. Cote. Expression profiling of paired benign and breast cancer lesions in African American women [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 4253. doi:10.1158/1538-7445.AM2017-4253

Abstract By regulating gene expression networks that mediate neoplastic behavior, epigenetic protein complexes are the ultimate effectors of oncogenic pathways. This hierarchical position of epigenetic pathways within the mechanisms that regulate cancer makes them ideal candidates for therapeutic targets. The current work was designed to characterize the role of the K9H3 Histone Methyltransferase (HMT) pathway in mediating oncogenic signals downstream of Aurora Kinase A (AurkA), with the goal of designing efficient combinatorial therapies against PDAC. Affinity protein purification, combined with mass spectrometry, demonstrates that HP1γ isolated from mitotic cells interacts with AurkA and the HMTs, G9a and GLP. We also find that this complex is critical for mitotic progression and cell proliferation. Congruently, treatment of PDAC cells with individual drugs against AurkA or HP1-HMT inhibits cell growth by inducing senescence, a cytostatic response. However, the combination of these agents has a synergistic effect of reducing cell growth in both, monolayer and spheroid cultures to result in a cytotoxic effect. Confocal and electron microscopy, along with cell cycle analysis, demonstrate that the cytotoxic effect of this combination is due to induction of mitotic catastrophe, a distinct form of cell death that occurs during mitosis. Molecularly, the combination of AurkA and HP1-HMT inhibition bypasses G2/M arrest with downregulation of the Chk1-Cdc25c pathway. In vivo treatment of PDAC orthotopic xenografts with the HP1-HMT inhibitor alone demonstrated reduced PDAC growth, and increased efficacy in PDAC growth reduction was observed in combination with the AurkA inhibitor over either individual treatment. Thus, our data demonstrate a novel AurkA-HP1-HMT mitotic pathway that holds promise for potential pharmacological targeting in combinatorial therapy for this malignancy. Citation Format: Angela Mathison, Ann Salmonson, Mckenna Missfeldt, Thiago de Assuncao, Jennifer Bintz, Trace Christensen, Sarah Kossak, Asha Nair, Juan Iovanna, Robert Huebert, Gwen Lomberk. Pharmacological targeting of the Aurora A and Histone 3 lysine 9 methyltransferase pathways in pancreatic cancer induces mitotic catastrophe [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 4060. doi:10.1158/1538-7445.AM2017-4060

Abstract Pancreatic ductular adenocarcinoma (PDAC) ranks fourth as a cause of cancer death in the USA and is almost universally fatal, with the annual number of deaths equivalent to the number of newly diagnosed cases. Valuable research in the field has revealed genetic aberrations that contribute to PDAC development and progression, with KRAS being one of the most frequent mutations in more than 90% of patient samples. However, to date, any efforts to directly target KRAS have failed in the clinic. Thus, there is indisputably an urgent need to further improve our understanding of molecular mechanisms underlying PDAC development as to identify novel therapeutic targets, including druggable important downstream targets and nodes orchestrated by oncogenic KRAS. In particular, we are interested in epigenetic pathways involved in PDAC development and progression due to the potential reversibility of any alteration, unlike genetic mutation. In the current study, using a cell model that allows inducible expression of mutant KRASG12D, we find that protein levels of the dimethyl-K9H3 histone methyl transferase (HMT), G9a, and its complex partners are increased in response to activation of the oncogenic Kras pathway. Furthermore, the activation of this oncogenic pathway results in the formation of the G9a-GLP-Wiz trimer complex, as determined by affinity protein purification, combined with mass spectrometry. In vivo experiments involving the cross of the Pdx1-CRE/LSL-KRASG12D mice with G9afl/fl animals demonstrate that a loss of the H3K9Me2 mark in the nucleus of exocrine cells is accompanied by a significantly reduced number of PanIN lesions. RNA-Seq experiments from these animals reveal that these mice have reduced levels of typical molecular markers of PanINs. In addition, these experiments show changes in the levels in several genes, which have been previously been shown to synergize with Kras to mediate pancreatic cancer initiation. Congruently, pharmacological inhibition of G9a using BRD4770 displays an inhibitory effect on KRASG12D-induced cell proliferation. Combined, these data provide evidence for a key role of the meK9H3-G9a pathway as a mediator of the oncogenic Kras response and defines a novel point of potential therapeutic intervention for PDAC. Citation Format: Angela Mathison, Ann Salmonson, Brooke Paradise, Mckenna Missfeldt, Juan Iovanna, Daniel Billadeau, Raul Urrutia, Gwen Lomberk. The epigenetic regulator, G9a, is a KRAS-inducible protein and its inactivation inhibits PanIN formation by this oncogene [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 1391. doi:10.1158/1538-7445.AM2017-1391

Abstract Background: Large-scale genome-wide association studies (GWAS) have identified ~35 loci associated with epithelial ovarian cancer risk. However, these genetic loci together explain only a small portion of the heritability of this malignancy. The large majority of the GWAS-identified variants are located in non-coding regions, thus possible causal genes in these loci remain largely unknown. We performed a transcriptome-wide association study (TWAS) to search for novel genetic loci for ovarian cancer risk and plausible causal genes at GWAS-identified loci. Method: Transcriptome data from normal ovarian tissue samples (n=68) and all tissue samples (n = 369), along with their high-density genotyping data, obtained from only European descendants included in the Genotype-Tissue Expression Project (GTEx), were used to build ovarian linear prediction models and cross-tissue models (to increase statistical power) using the elastic net method. Based on model performance, we evaluated 17,121 genes for their cis-predicted gene expressions in relation to ovarian cancer risk using summary statistics data generated in GWAS of ovarian cancer from 25,509 cases and 40,941 controls. MetaXcan was used to integrate gene expression prediction model with summary statistics. Results: We identified 35 genes with predicted expression levels associated with ovarian cancer risk at P value < 2.2 × 10-6, the Bonferroni corrected significance level for multiple comparisons. Of these, 12 genes at 4 genetic loci are located ≥500kb away from risk SNPs previously reported in GWAS, representing potential novel genetic loci for ovarian cancer risk. The remaining 23 genes at 12 loci are located within known ovarian cancer risk loci. Fifteen of these genes at the 12 loci have not been reported in previous studies. Analyses by ovarian cancer histological subtypes showed that the majority of these 35 genes are associated with serous invasive carcinoma. Several new associations were identified in histological subtype analyses. Conclusion: In this TWAS we identified multiple genes with predicted expressions related to ovarian cancer risk and provide substantial new information to enhance the understanding of ovarian cancer biology and genetics. Citation Format: Yingchang Lu, Joellen M. Schildkraut, Thomas A. Sellers, Lang Wu, Xingyi Guo, Bingshan Li, Y. Ann Chen, Jennifer B. Doherty, Simon Gayther, Ellen L. Goode, Hae Kyung Im, Siddhartha Kar, Kate Lawrenson, Ani W. Manichaikul, Jennifer B. Permuth, Brett M. Reid, Jamie K. Teer, Paul Pharoah, Wei Zheng, Jirong Long. Transcriptome-wide association study among 66,450 women to identify candidate susceptible genes for ovarian cancer risk [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 1308. doi:10.1158/1538-7445.AM2017-1308

Abstract Epithelial ovarian cancer (EOC) is a rare but deadly disease for which there is notably poorer survival in women of African Ancestry (AA) compared to women of European Ancestry (EA). Previous EA-based genome-wide association studies (GWAS) have identified 30 common, low penetrant EOC susceptibility alleles. Using the custom-designed 533,631 SNP Illumina OncoArray and imputation to ~12 million genetic variants in the 1000 Genomes Phase 3, we conducted a GWAS in 755 AA EOC cases, including 537 high-grade serous ovarian cancer (HGSOC) cases, and 1,235 AA controls from the Ovarian Cancer Association Consortium (OCAC). We identified novel susceptibility loci with suggestive evidence of association with EOC, based on a threshold of P<1x10-6, at three distinct loci including 10p15.1 (lead SNP rs4525119, intronic to AKR1C3, P=4.9 x 10-7, minor allele frequency [MAF]=0.33), 3p25.3 (lead SNP rs7643459, intronic to LOC101927394, P=8.4 x 10-7, MAF=0.36), and 4q13.3 (lead SNP rs4286604, 15 kb 3’ of UGT2A1, P=8.5x10-7, MAF=0.27). In analysis of HGSOC, we identified suggestive evidence of association at six distinct loci including 5q11.2 (lead SNP rs37792, 132 kb 5’ of FST, P=6.0x10-8, MAF=0.34), Xq27.2 (lead SNP rs57403204, 81 kb 3' of MAGEC1, P=1.7x10-7, MAF=0.06), 10p15.1 (lead SNP rs79079890, LOC105376360 intronic, P=3.0x10-7, MAF=0.03), 17p25.1 (lead SNP rs66459581, 5 kb 5’ of PRPSAP1, P=5.1x10-7, MAF=0.23), 15p12 (lead SNP rs116046250, GABRG3 intronic, P=8.7x10-7, MAF=0.05), and 4q21.21 (lead SNP rs192876988, 30 kb 3’ of GK2, P=9.2x10-7, MAF=0.05). These SNPs showed no evidence of association with risk in the OCAC EA GWAS including up to 23,543 EOC cases and 29,444 controls. Of the SNPs identified with suggestive evidence of association in our GWAS, two are located near genes known to regulate hormones and previously reported in relation to diseases of the ovary (AKR1C3 and FST), two have been linked to cancer (AKR1C3 and MAGEC1) and one represents testis-specific expression (GK2). Of the 30 SNPs identified previously in a EA EOC GWAS, we observed nominally significant associations (P<0.05) with consistent direction of effect for eight SNPs in AA women. Our investigation presents evidence of (1) variants for EOC shared among EA and AA women and (2) novel EOC risk loci of importance for AA women. Citation Format: Ani Manichaikul, Lauren C. Peres, Xin-Qun Wang, Kate Lawrenson, Sarah Abbott, Ann G. Schwartz, Anna H. Wu, Edward Peters, Patricia G. Moorman, Michele L. Cote, Melissa Bondy, Linda Kelemen, Ellen L. Goode, Jill Barnholtz-Sloan, Simon A. Gather, Andrew Berchuck, Jennifer A. Doherty, Paul Pharoah, Joellen Schildkraut. Identification of novel epithelial ovarian cancer loci in women of African ancestry from the Ovarian Cancer Association Consortium [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 1307. doi:10.1158/1538-7445.AM2017-1307

Abstract Mismatch repair (MMR) is essential for temozolomide (TMZ) sensitivity, and MMR deficiency causes TMZ resistance and disease recurrence. However, the role MMR in TMZ response has been linked with Msh2/Msh6 (MutSα), while role of Msh3, a component of MutSβ remains unclear. To delineate role of MutSα and MutSβ in TMZ response, we used siRNA to disrupt MSH6 and MSH3 expression. As expected, silencing of MSH6 led to resistance in a TMZ-sensitive U251 cell line. In contrast, silencing of MSH3 enhanced TMZ sensitivity as compared to the control cells. These results support established role of MutSα in TMZ response via replication stress and DNA double strand breaks (DSB), and reveal that MutSβ may potentially influence TMZ response by modulating DSB repair. Consistently, host cell reactivation (HCR) assays in U251 expressing DRGFP reporter showed decreased homologous recombination (HR) after MSH3 knockdown (KD), while Msh6 KD had no effect on HR efficiency. Supporting the concept of synthetic lethality of HR with PARP inhibition, clinically relevant dose of veliparib (1 µM) showed a modest but consistent increase in TMZ sensitization measured in terms of potentiation factor (PF, ratio of relative growth with TMZ alone/ TMZ and veliparib) in MSH3 KD (PF 1.6±0.2) as compared to control cells (PF 1.1±0.2). Sensitizing effect of 1 µM veliparib in MSH6 KD (PF 1.7±0.5) was comparable to that in MSH3 KD cells. Interestingly, supra-therapeutic dose of veliparib (10 µM) induced more pronounced sensitization in MSH6 KD cells (PF 2.6±0.3) as compared to MSH3 KD (PF 2.4±0.3) or control cells (PF 1.6±0.2). Consistently, veliparib (10 µM) alone had no impact on DNA damage signaling, while TMZ induced prominent increase in phosphorylation of KAP1, Chk1 and Chk2 in control and MSH3 KD cells, but not in MSH6 KD cells; veliparib/TMZ co-treatment led a robust signaling in all 3 types of cells. Increased sensitization by veliparib after MSH3 KD was also examined in a TMZ-resistant (GBM22TMZ) PDX model carrying resistance due to a homozygous mutation in MSH6, and a 4-fold increase in HR efficiency. In clonogenic assays silencing of MSH3 or BRCA1 had no effect on TMZ sensitivity in GBM22TMZ cells. However, veliparib alone (10 µM) decreased colony formation with relative colony formation 0.54±0.03 and 0.32±0.06 in MSH3 and BRCA1 KD cells, respectively as compared to 0.82±0.04 in controls. TMZ/veliparib co-treatment in GBM22TMZ led to further decrease in colony formation in MSH3 KD and BRCA1 KD cells with relative colony formation of 0.30±0.04 and 0.19±0.02, respectively versus 0.46±0.03 in control. Taken together, our results suggest that MutSα and MutSβ differentially regulate cytotoxic effects of TMZ. Decreased HR efficiency and enhanced sensitizing effects of veliparib in MSH3 KD cells partially define synthetic lethality of MMR deficiency with PARP, and may help development of PARP inhibitors as sensitizer of TMZ therapy in GBM. Citation Format: Shiv K. Gupta, Emily J. Smith, Ann M. Tuma, Brett L. Carlson, Mark A. Schroeder, Gasper J. Kitange, Jann N. Sarkaria. MutSα and MutSβ mismatch repair complexes differentially regulate temozolomide efficacy & sensitizing effects of veliparib in glioblastoma cells [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 1109. doi:10.1158/1538-7445.AM2017-1109

A 28‐year‐old female with PIK3CA‐related segmental overgrowth presented with headaches. She also had a unilateral vestibular schwannoma (VS), as well as 3 small (<2 cm) meningiomas, which according to the Manchester consensus diagnostic criteria for neurofibromatosis 2 (NF2) is sufficient for a clinical diagnosis. Analysis of blood revealed a mosaic PIK3CA c.2740G>A (p.Gly914Arg) mutation, confirming the diagnosis of PIK3CA‐related overgrowth, but no mutations in NF2 were detected. Although VS has not previously been reported in PIK3CA‐related segmental overgrowth, meningiomas have, raising the question of whether this patient's VS and meningiomas represent coincidental NF2 or phenotypic extension of her overgrowth syndrome. Genetic analysis of the VS revealed a heterozygous NF2 mutation c.784C>T (p.Arg262Ter) and loss of a portion of 22q, including NF2, SMARCB1, and LZTR1 genes. These results suggest that the patient has 2 different mosaic disorders, NF2 and PIK3CA‐related overgrowth. The PIK3CA mutation was also present in the VS. Confirmation of the clinical diagnosis of mosaic NF2 in this patient has implications for monitoring and highlights the possibility of co‐occurrence of mosaicism for multiple rare disorders in a single patient.

Abstract Purpose: Receptor tyrosine kinases (RTKs) are frequently deregulated in leukemia, yet the biological consequences of this deregulation remain elusive. The mechanisms underlying aberrant methylation, a hallmark of leukemia, are not fully understood. Here we investigated the role of RTKs in methylation abnormalities and characterized the hypomethylating activities of RTK inhibitors. Experimental Design: Whether and how RTKs regulate expression of DNA methyltransferases (DNMTs), tumor suppressor genes (TSGs) as well as global and gene-specific DNA methylation were examined. The pharmacologic activities and mechanisms of actions of RTK inhibitors in vitro, ex vivo, in mice, and in nilotinib-treated leukemia patients were determined. Results: Upregulation of RTKs paralleled DNMT overexpression in leukemia cell lines and patient blasts. Knockdown of RTKs disrupted, whereas enforced expression increased DNMT expression and DNA methylation. Treatment with the RTK inhibitor, nilotinib, resulted in a reduction of Sp1-dependent DNMT1 expression, the diminution of global DNA methylation, and the upregulation of the p15INK4B gene through promoter hypomethylation in AML cell lines and patient blasts. This led to disruption of AML cell clonogenicity and promotion of cellular apoptosis without obvious changes in cell cycle. Importantly, nilotinib administration in mice and human patients with AML impaired expression of DNMTs followed by DNA hypomethylation, TSG re-expression, and leukemia regression. Conclusions: Our findings demonstrate RTKs as novel regulators of DNMT-dependent DNA methylation and define DNA methylation status in AML cells as a pharmacodynamic marker for their response to RTK-based therapy, providing new therapeutic avenues for RTK inhibitors in overcoming epigenetic abnormalities in leukemia. Clin Cancer Res; 23(20); 6254–66. ©2017 AACR.