University of Pittsburgh Annual Progress Report: 2008 Formula Grant

Transcription

1 University of Pittsburgh Annual Progress Report: 2008 Formula Grant Reporting Period July 1, 2012 December 31, 2012 Formula Grant Overview The University of Pittsburgh received $8,975,120 in formula funds for the grant award period January 1, 2009 through December 31, Accomplishments for the reporting period are described below. Research Project 1: Project Title and Purpose Small Molecule Inhibitors of HIV Nef Signaling - The project focuses on discovery of small molecules to target a unique human immunodeficiency virus type 1 (HIV-1) protein, called Nef, which is essential for progression from a state of HIV infection to full-blown acquired immune deficiency syndrome (AIDS). Using in vitro and cell-based screens, the investigators aim to identify two classes of drug-like compounds with potent activity against HIV replication. These studies could eventually provide new tools for understanding Nef interactions with host cell proteins and identify novel lead compounds for anti-hiv therapeutics. Duration of Project 1/1/2009-3/31/2011 This project ended during a prior state fiscal year. For additional information, please refer to the Commonwealth Universal Research Enhancement C.U.R.E. Annual Reports on the Department's Tobacco Settlement/Act 77 web page at Research Project 2: Project Title and Purpose Alterations in Bioenergetics and Mitochondrial Function in Tumor Cells - Cells generate the energy they need to function and divide through two different metabolic pathways, mitochondrial oxidative phosphorylation and glycolysis. Due to their rapid growth, tumor cells have unique energy demands. However, the mechanism by which tumor cells regulate energy production is poorly understood. This project will examine the balance between these two energy producing pathways in tumor cells. The cells to be examined are associated with specific genetic alterations in critical genes (oncogenes, like Myc and Ras) or include critical genes that University of Pittsburgh 2008 Formula Grant Page 1

2 have been mutated (tumor suppressor genes, like p53). This project also addresses how shifts in energy demands can increase genome instabilities like mutations. Understanding bioenergetic shifts of specific tumor types may improve therapeutic approaches. Duration of Project 1/1/ /31/2010 This project ended during a prior state fiscal year. For additional information, please refer to the Commonwealth Universal Research Enhancement C.U.R.E. Annual Reports on the Department's Tobacco Settlement/Act 77 web page at Research Project 3: Project Title and Purpose Clinical Trials in Melanoma - The University of Pittsburgh Cancer Institute s (UPCI) Clinical Research Service (CRS) provides valuable resources for researchers and clinicians seeking to improve clinical research as a means to elevate patient care standards and treatment efficacy. This project will evaluate the clinical response of melanoma patients at various disease stages to novel therapeutic regimens designed to be specific to their treatment needs. Duration of Project 1/1/ /31/2010 This project ended during a prior state fiscal year. For additional information, please refer to the Commonwealth Universal Research Enhancement C.U.R.E. Annual Reports on the Department's Tobacco Settlement/Act 77 web page at Research Project 4: Project Title and Purpose Therapeutic Vaccine Trials in Recurrent Prostate Cancer - The University of Pittsburgh Cancer Institute s (UPCI) Clinical Research Service (CRS) provides valuable resources for researchers and clinicians seeking to improve clinical research to elevate patient care standards and treatment efficacy. This project will conduct two novel vaccine trials in patients with recurrent prostate cancer: (1) a safety and efficacy study of a new vaccine targeted against mucin-1 (MUC-1) in conjunction with poly-iclc (polyinosinic-polycytidylic acid stabilized with polylysine and carboxymethylcellulose) in patients with recurrent and/or advanced prostate cancer and (2) a safety and efficacy evaluation of alpha-type-1 dendritic cell (DC)-based vaccines loaded with allogeneic prostate tumor cells in combination with androgen ablation in patients with recurrent prostate cancer. University of Pittsburgh 2008 Formula Grant Page 2

3 Duration of Project 1/1/ /31/2010 This project ended during a prior state fiscal year. For additional information, please refer to the Commonwealth Universal Research Enhancement C.U.R.E. Annual Reports on the Department's Tobacco Settlement/Act 77 web page at Research Project 5: Project Title and Purpose Research Infrastructure: Bridgeside Point II Building Renovations - Bridgeside Point II (BPII), the University of Pittsburgh s newest research facility, is nearing completion. This project will address final construction and laboratory fit-out in the building. By establishing this additional space, the Department of Orthopaedic Surgery, the Department of Microbiology and Molecular Genetics, and the McGowan Institute for Regenerative Medicine will be able to expand their research programs, recruit additional faculty members, and accommodate researchers changing needs. Duration of Project 12/17/2009 9/30/2010 This project ended during a prior state fiscal year. For additional information, please refer to the Commonwealth Universal Research Enhancement C.U.R.E. Annual Reports on the Department's Tobacco Settlement/Act 77 web page at Research Project 6: Project Title and Purpose Estrogen Receptor Genetics in Breast Cancer - Breast cancer remains a major health issue in the United States, with approximately 210,000 new cases and 40,000 deaths each year. Breast cancer can be subdivided into estrogen receptor (ER) positive and negative disease. Breast cancers that express ER are often less aggressive and respond to anti-hormone therapy, but a significant fraction of patients who are ER positive do not respond to therapy. New sequencing tools have recently allowed an unprecedented view of ER action in breast cancer. This project will investigate how genetic alterations in ER and the ER pathway affect response to the hormone and whether the ER itself is involved in generating genetic changes in breast cancer. Better knowledge of ER action should translate into improved treatment of ER positive breast cancer. University of Pittsburgh 2008 Formula Grant Page 3

4 Duration of Project 3/30/ /31/2012 Project Overview ER is a major determinant of breast cancer initiation, progression, and response to anti-hormone therapy. Recent advances in molecular techniques have allowed an unprecedented view into how this hormone receptor orchestrates its biological response in normal cells and tumors. This project is based on the use of novel next-generation sequencing techniques, which have provided a unique insight into genetic changes in ER in human breast cancers and new ideas about how ER may effect genetic changes in breast cancer. This project will use an array of new technologies to comprehensively interrogate ER in breast cancer cell lines. In the first aim, we will study the effect of germline single nucleotide polymorphisms (SNPs) and cancer-specific somatic mutations on ER function. Preliminary studies using next-generation sequencing in human breast tumors have identified numerous genetic changes in ER, and we will examine these changes with specific regard to their effects on estrogen-mediated cancer phenotypes, such as proliferation and invasion. In the second aim, we will examine how alterations in the deoxyribonucleic acid (DNA) sequence that ER binds might affect subsequent hormone response. Target sites will be identified by chromatin immunoprecipitation and sequencing; and changes will be characterized, especially related to hormone response. In the third aim, we will examine whether ER binding sites are associated with regions of DNA translocations in breast cancer cells and whether ER can itself direct genomic translocation in cells with reduced DNA repair activity. This project will provide significant new information regarding ER action in breast cancer and will likely have a major impact in improving prediction and response to hormone therapy. Principal Investigator Adrian V. Lee, PhD University of Pittsburgh Magee-Womens Research Institute 204 Craft Avenue Room B705 Pittsburgh, PA Other Participating Researchers Steffi Oesterreich, PhD - employed by University of Pittsburgh Expected Research Outcomes and Benefits Two-thirds (~130,000) of breast cancers detected in the United States each year are ER-positive and are treated with anti-hormone therapy. However, not all tumors respond to this therapy, and a majority will eventually develop resistance. This project will investigate genetic changes in ER University of Pittsburgh 2008 Formula Grant Page 4

5 and their role in causing differential responses to anti-hormone therapy. In Aim 1, we expect to find that somatic mutations in ER render the tumor resistant to anti-hormone therapy, indicating the need for alternative therapies. Similarly, germline polymorphisms in ER are likely to modulate the response to anti-hormone therapy; and some may actually render tumors hypersensitive to therapy, in which case the tumors may require anti-hormone therapy alone, thereby sparing the patient unnecessary toxic therapy. In Aim 2, we predict that genetic changes in ER binding sites will correlate with sensitivity or resistance to hormone therapy. Finally, we expect Aim 3 to produce novel data regarding how ER affects genomic translocations. The discovery of specific ER-directed genomic translocations will have major implications in breast cancer, as they may produce novel gene products that can act as tumor-specific diagnostic and therapeutic targets (similar to the oncogene fusion BCR-Abl in chronic myelogenous leukemia). In summary, this project will provide novel and critical information regarding ER function in breast cancer and is likely to have a major impact on hormone treatment of breast cancers. Early studies of ER1 using Sanger sequencing and conformational assays reported few if any mutations in ER-positive tumors, although more recently a few somatic mutations have been described in tumors and cell lines. We set out to validate the authenticity of these reported somatic mutations by analyzing ER DNA sequence variants (DSVs) in 66 ER+ breast tumors and 39 breast cancer cell lines. Genomic DNA from primary breast tumors, part of the Baylor College of Medicine Breast Tumor DNA Bank was used. We identified 12 DSVs in ESR1 exons, five of which were previously described SNPs (rs , rs746432, rs , rs , rs ). One DSV was previously described as a mutation (H6Y), and six are potential novel DSVs. For sequence confirmation we used a Maldi-Tof mass spectrometer from Sequenom. This technology works with small amounts of degraded DNA, its multiplexing capacity makes it highly efficient, and a variety of studies have demonstrated that the sensitivity of mass spectrometric methods exceeds that of traditional Sanger sequencing and is highly concordant with Sanger sequencing, pyrosequencing, and allele-specific PCR. This analysis was carried out in collaboration with the National Surgical Adjuvant Breast and Bowel Project (NSABP) in Pittsburgh (Patrick Gavin). We developed mass-spectrometry assays, including controls, for these DSVs and included 10 DSVs that had previously been reported as somatic mutations in breast (and other) cancers (Figure 1). Using the Sequenom-based analysis, we were able to confirm previously reported SNPs and one previously reported mutation (H6Y). This mutation was found in one human breast tumor and one cell line. None of the other reported somatic mutations could be confirmed in tumors or cell lines. This result clearly indicates that ESR1 mutations are very rare in primary human breast tumors. This study is the largest ER sequencing study done to date, and although negative, provides very critical information. A poster summarizing these data was shown at the San Antonio Breast Cancer Symposium (12/2012, Oesterreich et al Lack of Frequent ER Mutations in Primary Breast Cancer ), and we are currently preparing the manuscript for submission. University of Pittsburgh 2008 Formula Grant Page 5

6 We set out to identify the biological importance of the ESR1H6Y DSV. This was accomplished through ERE-tk-luciferase reporter assays (Figure 2). We introduced the H6Y (C16T) mutation into an HA-tagged ESR1 construct using site-directed mutagenesis (Quick Change Site- Directed Mutagenesis Kit from Stratagene, La Jolla, CA). After confirming the correct sequence of the ESR1 coding region, we determined the effect of the mutation on ESR1 transcriptional activity using transient reporter assays (ERE-Tk-Luc). We also measured the proportion of cells in S-phase (through incorporation of BrdU) and analyzed cell cycle distribution in the presence of ER wildtype or H6Y to determine the effect of the mutation on cell cycle (data not shown). These functional assays on the ESR1H6Y DSV failed to identify increased or altered activity of the mutant. The lack of a selective phenotype and the infrequent occurrence of this DSV (1/60 tumors) suggest that ESR1H6Y does not play a major role in breast cancer. As stated above, this analysis suggests that ESR1 mutations are a rare event in ER+ primary breast cancer. Interestingly, while studying the degradation of ER in a panel of breast cancer cell lines, we noticed that invasive lobular cancer (ILC) cell lines express very high levels of ER protein compared to invasive ductal cancer (IDC) cell lines, whereas there are no differences at the RNA level (data not shown). We subsequently found that these cell lines are inherently resistant to antiestrogen, and preliminary data suggest that the ER in these cell lines is also resistant to hormone-mediated degradation. We recently published a review summarizing data on hormone response in ILC (Sikora et al, Steroids 2012) and are continuing an in-depth analysis of this phenomenon. We have begun to study the interaction between ER and repressive histone modifying enzymes. Briefly, we performed a screen for epigenetic changes (DNA promoter methylation) in endocrine resistant cell lines (C4-12 and LTED). In this unbiased screen, we found that the estrogenrepressed gene HOXC10 was hypermethylated in endocrine resistant cells, and consistent with this observation HOXC10 mrna and protein were down-regulated. Intriguingly, we observed that there was a step-wise epigenetic reprogramming of the HOXC10 promoter during the loss of hormone response, including an ER-EZH2 interaction and bivalent chromatin marks, followed by DNA methylation. These studies were submitted for publication and are currently under review (Science Translational Medicine). We performed ChIP-seq studies in a number of cell lines, including estrogen-responsive osteoblastoma-like U2OS cells, the ER+ invasive ductal breast cancer cell line MCF-7, and the invasive lobular cancer cell line MDA-MB-134VI. The studies in U2OS cells led to the identification of estrogen regulation of a prostaglandin transporter, PGE2, which had not previously been implicated in estrogen-mediated tumor growth or metastasis. We have therefore begun a collaboration with Dr. Pawel Kalinksi (University of Pittsburgh Cancer Institute [UPCI], Pittsburgh, PA), an expert in the study of PGE2, to follow up on these studies. These studies are ongoing. The ER binding sites in the ILC MDA-MB-134VI cells are currently being sequenced, and we expect to obtain these results in the near future. Validity of our approach was confirmed by candidate gene PCR using the ChIP DNA (Figure 3). University of Pittsburgh 2008 Formula Grant Page 6

7 We performed gene expression array studies in MDA-MB-134VI cells (-/+ estrogen, 6-hour and 24-hour time points) (these data were presented in poster form at the San Antonio Breast Cancer Symposium, 12/2012). The UPCI Biostatistics Facility is currently analyzing the gene expression results. We also requested tumor samples from the University of Pittsburgh Health Sciences Tissue Bank, and we are currently isolating DNA and RNA to confirm DSVs and correlate gene expression changes in clinical samples. In collaboration with Dr. Panayiotis V. Benos and Dr. Xinghua Lu (Center for Translational Bioinformatics, University of Pittsburgh), we downloaded publicly available ChIP-seq datasets of ESR1 and identified DSVs. For example, we currently have access to nine ER ChIP-seq data sets for the MCF-7 cell line. Focusing on the overlapping (i.e., high confidence) estrogeninduced ESR1 binding sites, we identified ~ 5,200 binding sites. Cross-referencing these sites with dbsnp and the 1,000 Genomes Project data showed that approximately 85% of these sites harbor known SNPs. The remaining 15% are either sequencing artifacts (unlikely given the Genome Analysis Toolkit [GATK] workflow stringency criteria applied to our analysis), they are low frequency SNPs not yet deposited in public databases, or they might be somatic mutations in ER binding sites. To address this question, we decreased the number of unknown DSVs by developing and applying bioinformatics models that predict the effect of the DSV on ER binding. A further analysis of these binding sites has led to the discovery that there are fundamental differences in ER and cofactor recruitment between estrogen and repressed genes, and this finding will need to be considered for functional analyses of the DSVs in binding sites. These novel observations have been submitted to Genome Biology and are currently under review (Hatice Ulku Osmanbeyoglu, Roger Day, Steffi Oesterreich, Panayiotis V. Benos, Claudia Coronnello, Xinghua Lu. Estrogen represses gene expression through reconfiguring chromatin structures). As expected, these studies have led to increased interaction and collaboration with our bioinformatics colleagues, especially Drs. Panayiotis V. Benos and Xinghua Lu (University of Pittsburgh). Working together, we developed and tested an algorithm that identifies DSVs in mirna binding sites. The description of this model is published (Coronello, et al, PLoS Computational Biology, (12)1-13). We examined statistical enrichment for epigenomic marks and transcription factor binding sites (such as ER binding site) in the dataset of genomic translocations (Figure 4). We used a total of 259 epigenomic feature tracks from the following sources: 148 transcription factor binding tracks generated by the ENCODE project, 66 transcription factor binding tracks and 25 other epigenomic marks in breast cancer from the Cistrome database, and 20 normal epigenomic tracks from the Human Epigenome Atlas Release 2. We computed enrichment scores for each of the 259 epigenomic features for both germline and putatively somatic breakpoints using a 50,000 base pair radius around each breakpoint. A total of 107 (out of 259) features showed significantly different patterns of enrichment (Mann-Whitney-Wilcoxon test, p-value<0.05) between the two types of breakpoints. The enrichment scores for discriminating features were used to cluster the samples and the features (Figure 4). University of Pittsburgh 2008 Formula Grant Page 7

8 Published Samples Among the 74 features enriched around somatic breakpoints are ESR1 (eight datasets), FoxA1 (two datasets), PGR, and SRC-3. This finding supports our previous observation of ER-directed translocation at the PLAC1 promoter. We observed strong enrichment around somatic translocations for 49 transcription factors profiled by the ENCODE project, including GATA- 1/2, c-jun, CEBPB, FoxA1, ERα, and p300 (p<0.02). To examine possible interactions between ESR1 and other transcription factors, we identified somatic breakpoints within 50kb of an ESR1 binding site for each ESR1 experiment and identified genes within 10kb of such breakpoints. A total of 40 genes were identified by this method in at least 3/21 cancer samples. Analysis of the 40 genes by GSEA (gene set enrichment analysis) indicated enrichment for ETS1 (p-value =3.23x10-4 ) and p53 (p-value=3.71x10-4 ) binding in promoter regions of these genes. The enrichment patterns suggest that transcription factors either individually or, as part of larger complexes, may promote genomic instability. A manuscript describing this work and the associated bioinformatic platform is in preparation for submission. Published SNVs S10S H6Y* SNPs R243R P325P R394L P399S P399L Q499L AF-1 DBD Hinge LBD AF-2 T553A H550Y T594T H6Y 1 M264I 1 S341L 3 G400V 6 L489V 7 N532Y* 7 K303R 2 E352V 4 M490I 7 Y537N 8 V364E 5 D538G 7 Breast Cancer M543V 9 GBM Lung Ovarian 1) Chanock, SJ et al. Breast Cancer Res (2007) 9:R5. 2) Fuqua et al. Cancer Res (2000) 60: ) Parsons, DW et al. Science (2008) 321: ) Karnik, PS et al. Cancer Res (1994) 54:349. 5) McInerney, EM et al. Mol Endocrinol (1996) 10:1519 (functional but not in clinical samples) 6) Tora, L et al. EMBO J (1989) 8:1981 (functional, but not in clinical samples) 7) Kan, Z et al. Nature (2010) 466:869. 8) Kim, JH et al. J Mol Endo (2005) 35:449. 9) Nichols, M et al. Breast Cancer Res Treat (2010) 120: 761. Figure 1. DSVs tested through a mass-spectrometric approach (Sequenom). University of Pittsburgh 2008 Formula Grant Page 8

9 Input% Figure 2: ESR1H6Y has transcriptional activity similar to wild-type ESR1. Cell lines (C4-12 and HeLa) were transiently transfected with ERE-Tk-reporter and with ESR1 wildtype (ER-HA), ESR1 mutation H6Y (C16T), or empty plasmid (vector) as a control. Cells were treated with vehicle (EtOH), 10-8M estradiol (E2), 10-6M 4-OH-tamoxifen (Tam), or a combination of 10-8M estradiol and 10-6M 4-OH-tamoxifen (E2/Tam). Luciferase values were determined using Luciferase reporter kit (Promega), and a Luminometer ER IgG ER IgG ER IgG ER IgG V E NFERE IGFBP4 Ps2 GREB1 Figure 3: ChIP results in MDA-MB-134VI cells. Cells were incubated in 5% CSS and treated with 10-8M E2 for 45 mins. ChIP assays and qpcr were performed, analyzing three candidate genes (IGFBP4, ps2, and GREB1). A nonfunctional ERE (NFERE) served as negative control. University of Pittsburgh 2008 Formula Grant Page 9

10 Figure 4 (next page). Hierarchical clustering of breakpoint sets based on enrichment for transcription factor binding and epigenomic features. Note the distinct patterns of enrichment for germline breakpoints (columns to the left) and putatively somatic breakpoints (columns to the right). University of Pittsburgh 2008 Formula Grant Page 10