Dataset Information

Widespread molecular patterns associated with drug sensitivity in breast cancer cell lines, with implications for human tumors.

ABSTRACT:

Background

Recent landmark studies have profiled cancer cell lines for molecular features, along with measuring the corresponding growth inhibitory effects for specific drug compounds. These data present a tool for determining which subsets of human cancer might be more responsive to particular drugs. To this end, the NCI-DREAM-sponsored DREAM7: Drug Sensitivity Prediction Challenge (sub-challenge 1) set out to predict the sensitivities of 18 breast cancer cell lines to 31 previously untested compounds, on the basis of molecular profiling data and a training subset of cell lines.

Methods and results

With 47 teams submitting blinded predictions, team Creighton scored third in terms of overall accuracy. Team Creighton's method was simple and straightforward, incorporated multiple expression data types (RNA-seq, gene array, RPPA), and incorporated all profiled features (not only the "best" predictive ones). As an extension of the approach, cell line data, from public datasets of expression profiling coupled with drug sensitivities (Barretina, Garnett, Heiser) were used to "predict" the drug sensitivities in human breast tumors (using data from The Cancer Genome Atlas). Drug sensitivity correlations within human breast tumors showed differences by expression-based subtype, with many associations in line with the expected (e.g. Lapatinib sensitivity in HER2-enriched cancers) and others inviting further study (e.g. relative resistance to PI3K inhibitors in basal-like cancers).

Conclusions

Molecular patterns associated with drug sensitivity are widespread, with potentially hundreds of genes that could be incorporated into making predictions, as well as offering biological clues as to the mechanisms involved. Applying the cell line patterns to human tumor data may help generate hypotheses on what tumor subsets might be more responsive to therapies, where multiple cell line datasets representing various drugs may be used, in order to assess consistency of patterns.

SUBMITTER: Creighton CJ

PROVIDER: S-EPMC3873899 | biostudies-literature | 2013

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Widespread molecular patterns associated with drug sensitivity in breast cancer cell lines, with implications for human tumors.

Creighton Chad J CJ

PloS one 20131227 12

<h4>Background</h4>Recent landmark studies have profiled cancer cell lines for molecular features, along with measuring the corresponding growth inhibitory effects for specific drug compounds. These data present a tool for determining which subsets of human cancer might be more responsive to particular drugs. To this end, the NCI-DREAM-sponsored DREAM7: Drug Sensitivity Prediction Challenge (sub-challenge 1) set out to predict the sensitivities of 18 breast cancer cell lines to 31 previously unt ...[more]

PMID: 24386073

Similar Datasets

Project description:MicroRNAs (miRNAs) regulate gene expression post-transcriptionally. In this way they might influence whether a cell is sensitive or resistant to a certain drug. So far, only a limited number of relatively small scale studies comprising few cell lines and/or drugs have been performed. To obtain a broader view on miRNAs and their association with drug response, we investigated the expression levels of 411 miRNAs in relation to drug sensitivity in 36 breast cancer cell lines. For this purpose IC50 values of a drug screen involving 34 drugs were associated with miRNA expression data of the same breast cancer cell lines. Since molecular subtype of the breast cancer cell lines is considered a confounding factor in drug association studies, multivariate analysis taking subtype into account was performed on significant miRNA-drug associations which retained 13 associations. These associations consisted of 11 different miRNAs and eight different drugs (among which Paclitaxel, Docetaxel and Veliparib). The taxanes, Paclitaxel and Docetaxel, were the only drugs having miRNAs in common: hsa-miR-187-5p and hsa-miR-106a-3p indicative of drug resistance while Paclitaxel sensitivity alone associated with hsa-miR-556-5p. Tivantinib was associated with hsa-let-7d-5p and hsa-miR-18a-5p for sensitivity and hsa-miR-637 for resistance. Drug sensitivity was associated with hsa-let-7a-5p for Bortezomib, hsa-miR-135a-3p for JNJ-707 and hsa-miR-185-3p for Panobinostat. Drug resistance was associated with hsa-miR-182-5p for Veliparib and hsa-miR-629-5p for Tipifarnib. Pathway analysis for significant miRNAs was performed to reveal biological roles, aiding to find a potential mechanistic link for the observed associations with drug response. By doing so hsa-miR-187-5p was linked to the cell cycle G2-M checkpoint in line with this checkpoint being the target of taxanes. In conclusion, our study shows that miRNAs could potentially serve as biomarkers for intrinsic drug resistance and that pathway analyses can provide additional information in this context.

Project description:BACKGROUND:Proper cell models for breast cancer primary tumors have long been the focal point in the cancer's research. The genomic comparison between cell lines and tumors can investigate the similarity and dissimilarity and help to select right cell model to mimic tumor tissues to properly evaluate the drug reaction in vitro. In this paper, a comprehensive comparison in copy number variation (CNV), mutation, mRNA expression and protein expression between 68 breast cancer cell lines and 1375 primary breast tumors is conducted and presented. RESULTS:Using whole genome expression arrays, strong correlations were observed between cells and tumors. PAM50 gene expression differentiated them into four major breast cancer subtypes: Luminal A and B, HER2amp, and Basal-like in both cells and tumors partially. Genomic CNVs patterns were observed between tumors and cells across chromosomes in general. High C?>?T and C?>?G trans-version rates were observed in both cells and tumors, while the cells had slightly higher somatic mutation rates than tumors. Clustering analysis on protein expression data can reasonably recover the breast cancer subtypes in cell lines and tumors. Although the drug-targeted proteins ER/PR and interesting mTOR/GSK3/TS2/PDK1/ER_P118 cluster had shown the consistent patterns between cells and tumor, low protein-based correlations were observed between cells and tumors. The expression consistency of mRNA verse protein between cell line and tumors reaches 0.7076. These important drug targets in breast cancer, ESR1, PGR, HER2, EGFR and AR have a high similarity in mRNA and protein variation in both tumors and cell lines. GATA3 and RP56KB1 are two promising drug targets for breast cancer. A total score developed from the four correlations among four molecular profiles suggests that cell lines, BT483, T47D and MDAMB453 have the highest similarity with tumors. CONCLUSIONS:The integrated data from across these multiple platforms demonstrates the existence of the similarity and dissimilarity of molecular features between breast cancer tumors and cell lines. The cell lines only mirror some but not all of the molecular properties of primary tumors. The study results add more evidence in selecting cell line models for breast cancer research.

Project description:DNA methylation is an epigenetic event that involves adding a methyl group to the cytosine (C) site, especially the one that pairs with a guanine (G) site (ie, CG or CpG site), in a human genome. This event plays an important role in both cancerous and normal cell development. Previous studies often assume symmetric methylation on both DNA strands. However, asymmetric methylation, or hemimethylation (methylation that occurs only on 1 DNA strand), does exist and has been reported in several studies. Due to the limitation of previous DNA methylation sequencing technologies, researchers could only study hemimethylation on specific genes, but the overall genomic hemimethylation landscape remains relatively unexplored. With the development of advanced next-generation sequencing techniques, it is now possible to measure methylation levels on both forward and reverse strands at all CpG sites in an entire genome. Analyzing hemimethylation patterns may potentially reveal regions related to undergoing tumor growth. For our research, we first identify hemimethylated CpG sites in breast cancer cell lines using Wilcoxon signed rank tests. We then identify hemimethylation patterns by grouping consecutive hemimethylated CpG sites based on their methylation states, methylation "M" or unmethylation "U." These patterns include regular (or consecutive) hemimethylation clusters (eg, "MMM" on one strand and "UUU" on another strand) and polarity (or reverse) clusters (eg, "MU" on one strand and "UM" on another strand). Our results reveal that most hemimethylation clusters are the polarity type, and hemimethylation does occur across the entire genome with notably higher numbers in the breast cancer cell lines. The lengths or sizes of most hemimethylation clusters are very short, often less than 50?base pairs. After mapping hemimethylation clusters and sites to corresponding genes, we study the functions of these genes and find that several of the highly hemimethylated genes may influence tumor growth or suppression. These genes may also indicate a progressing transition to a new tumor stage.

Project description:BackgroundIncreased focus surrounds identifying patients with advanced non-small cell lung cancer (NSCLC) who will benefit from treatment with epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI). EGFR mutation, gene copy number, coexpression of ErbB proteins and ligands, and epithelial to mesenchymal transition markers all correlate with EGFR TKI sensitivity, and while prediction of sensitivity using any one of the markers does identify responders, individual markers do not encompass all potential responders due to high levels of inter-patient and inter-tumor variability. We hypothesized that a multivariate predictor of EGFR TKI sensitivity based on gene expression data would offer a clinically useful method of accounting for the increased variability inherent in predicting response to EGFR TKI and for elucidation of mechanisms of aberrant EGFR signalling. Furthermore, we anticipated that this methodology would result in improved predictions compared to single parameters alone both in vitro and in vivo.ResultsGene expression data derived from cell lines that demonstrate differential sensitivity to EGFR TKI, such as erlotinib, were used to generate models for a priori prediction of response. The gene expression signature of EGFR TKI sensitivity displays significant biological relevance in lung cancer biology in that pertinent signalling molecules and downstream effector molecules are present in the signature. Diagonal linear discriminant analysis using this gene signature was highly effective in classifying out-of-sample cancer cell lines by sensitivity to EGFR inhibition, and was more accurate than classifying by mutational status alone. Using the same predictor, we classified human lung adenocarcinomas and captured the majority of tumors with high levels of EGFR activation as well as those harbouring activating mutations in the kinase domain. We have demonstrated that predictive models of EGFR TKI sensitivity can classify both out-of-sample cell lines and lung adenocarcinomas.ConclusionThese data suggest that multivariate predictors of response to EGFR TKI have potential for clinical use and likely provide a robust and accurate predictor of EGFR TKI sensitivity that is not achieved with single biomarkers or clinical characteristics in non-small cell lung cancers.

Project description:PurposeBluePrint (BP) is an 80-gene molecular subtyping test that classifies early-stage breast cancer (EBC) into Basal, Luminal, and HER2 subtypes. In most cases, breast tumors have one dominant subtype, representative of a single activated pathway. However, some tumors show a statistically equal representation of more than one subtype, referred to as dual subtype. This study aims to identify and examine dual subtype tumors by BP to understand their biology and possible implications for treatment guidance.MethodsThe BP scores of over 15,000 tumor samples from EBC patients were analyzed, and the differences between the highest and the lowest scoring subtypes were calculated. Based upon the distribution of the differences between BP scores, a threshold was determined for each subtype to identify dual versus single subtypes.ResultsApproximately 97% of samples had one single activated BluePrint molecular subtype, whereas ~ 3% of samples were classified as BP dual subtype. The most frequently occurring dual subtypes were the Luminal-Basal-type and Luminal-HER2-type. Luminal-Basal-type displays a distinct biology from the Luminal single type and Basal single type. Burstein's classification of the single and dual Basal samples showed that the Luminal-Basal-type is mostly classified as 'luminal androgen receptor' and 'mesenchymal' subtypes, supporting molecular evidence of AR activation in the Luminal-Basal-type tumors. Tumors classified as Luminal-HER2-type resemble features of both Luminal-single-type and HER2-single-type. However, patients with dual Luminal-HER2-type have a lower pathological complete response after receiving HER2-targeted therapies in addition to chemotherapy in comparison with patients with a HER2-single-type.ConclusionThis study demonstrates that BP identifies tumors with two active functional pathways (dual subtype) with specific transcriptional characteristics and highlights the added value of distinguishing BP dual from single subtypes as evidenced by distinct treatment response rates.

Dataset Information

Widespread molecular patterns associated with drug sensitivity in breast cancer cell lines, with implications for human tumors.

Background

Methods and results

Conclusions

Publications

Widespread molecular patterns associated with drug sensitivity in breast cancer cell lines, with implications for human tumors.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets