Metabolomics,Unknown,Transcriptomics,Genomics,Proteomics

Dataset Information

H3K27me3 in four squamous cell carcinoma cell lines

ABSTRACT: ChIP-Seq for H3K27 trimethylation was performed for two HPV-positive and two HPV-negative squamous cell carcinoma cell lines. The data served two purposes. First, the data were used as an example implementation of our novel ChIP-Seq Peak Prioritization pipeline, PePr. We have developed the PePr pipeline, a ChIP-Seq Peak Prioritization pipeline that accounts for the variation among replicates and peak location relative to a gene. We show, using a transcription factor dataset (which exhibited small variation among samples), that PePr performs favorably compared to commonly used peak callers and that it achieves balanced sensitivity and specificity. We also show, using histone modification data (which exhibited larger variation among samples), that PePr can improve the detection of differential H3K27me3 regions compared with a common current approach. Using data from ChIP-Seq and gene expression experiments performed in parallel on the same samples, we show that the incorporation of functional annotations can improve the prioritization of functional sites. Secondly, the data were used to assess real differences in the genome-wide H3K27me3 profiles between HPV-positive and HPV-negative carcinoma cell lines. Careful analysis and integration of the data with DNA methylation and gene expression data performed on the same cell lines demonstrated striking differences exist. ChIP-Seq for H3K27 trimethylation was performed for two HPV-positive and two HPV-negative squamous cell carcinoma (SCC) cell lines. Input DNA was also sequenced for each sample to serve as a control. The goal was to determine overall differences in H3K27me3 patterns observed between the HPV-positive and HPV-negative SCC cell lines.

ORGANISM(S): Homo sapiens

SUBMITTER: Maureen Sartor

PROVIDER: E-GEOD-38629 | biostudies-arrayexpress |

REPOSITORIES: biostudies-arrayexpress

ACCESS DATA

Similar Datasets

Project description:Bilateral retinoblastoma has served as a critical model for studying genetic predisposition to cancer, however international incidence of unilateral retinoblastoma is variable across populations suggesting the involvement of environmental factors. DNA from Human Papilloma Virus (HPV) has been found in human retinoblastoma tissue through polymerase chain reaction based methods in several independent reports. To search for HPV oncoproteins and characterize transcriptional profiles in retinoblastoma, we analyzed tumor tissue obtained after enucleation and cultured for one week from 13 previously untreated Mexican children with retinoblastoma. We identified the presence of E7 HPV oncoproteins using western blot analysis in HPV positive tumors, and using non-supervised methods of classification on gene expression microarrays data, tumors clustered according to the presence or absence of HPV DNA identified by PCR. These methods were unable to discriminate retinoblastoma tissues by laterality. 2,404 genes were differentially expressed between HPV positive and HPV negative retinoblastoma, including a group of genes involved in a transcriptional response to viral infection. This molecular evidence supports the hypothesis that HPV is involved in the molecular pathogenesis of these tumors. The 13 children and tumors examined here, are incident cases participating in a larger case-control study at two Pediatric Hospitals of third level of medical attention in Mexico City, that form part of the public health system. Nine different tumors positive and four negative to HPV by PCR were included in the experimental design. From these, seven are unilateral and six are bilateral and are considered biological replicas. Two unilateral and two bilateral cases were chosen to make technical replicas. From the surgical specimens, two tumor portions were obtained, one was frozen, DNA was extracted from it, and HPV status was assigned using MY09 and MY11 primers by PCR. Another portion of tumor tissue was used to grow in culture, separating viable cells from non viable cells through a ficoll gradient. In the eye, retinoblastoma usually grows into the vitreous which has no cells, and thus the tissue obtained is predominantly formed by tumor cells and few endothelial cells from capillaries in the tumor. Retinoblastoma grow in culture as non adherent cells, and when adherent cells were also present, cell suspension was moved to a new dish to keep retinoblastoma cells as the main or only cellular type as possible. We used a two channel platform with in house spotted microarrays. Since retinoblastoma has known classes: uni and bilateral, we sought to verify these classes or discover new, we used a class discovery approach and a reference experimental design. As the reference sample we used RNA form exponentially growing Saccharomyces cerevisiae because major signaling pathways and basic cellular processes among Saccharomyces cerevisiae and mammalian cells including cell cycle control and DNA repair mechanisms have a high degree of conservation. Also because yeast is easy to grow, and thus a reliable and robust source of reference RNA.

Dataset Information

H3K27me3 in four squamous cell carcinoma cell lines

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure