Mario Medvedovic

The transcriptional signature of dioxin in human hepatoma HepG2 cells

We have used a high density microarray hybridization approach to characterize the transcriptional response of human hepatoma HepG2 cells to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). We find that exposure to 10 nM TCDD for 8 hr alters by at least a factor of 2.1 the expression of 310 known genes and of an equivalent number of expressed sequence tags. Treatment with TCDD in the presence of 20 microg/mL of cycloheximide blocked the effect on 202 of these genes, allowing us to distinguish between primary effects of TCDD exposure, which take place whether cycloheximide is present or not, and secondary effects, which are blocked by inhibition of protein synthesis. Of the 310 known genes affected by TCDD, 30 are up-regulated and 78 are down-regulated regardless of cycloheximide treatment, and 84 are up-regulated and 118 are down-regulated only when protein synthesis is not inhibited. Functional clustering of genes regulated by TCDD reveals many potential physiological interactions that might shed light on the multiple biological effects of this compound. Our results, however, suggest that arriving at a sound understanding of the molecular mechanisms governing the biological outcome of TCDD exposure promises to be orders of magnitude more complicated than might have been previously imagined.

Bayesian infinite mixture model based clustering of gene expression profiles

MOTIVATION The biologic significance of results obtained through cluster analyses of gene expression data generated in microarray experiments have been demonstrated in many studies. In this article we focus on the development of a clustering procedure based on the concept of Bayesian model-averaging and a precise statistical model of expression data. RESULTS We developed a clustering procedure based on the Bayesian infinite mixture model and applied it to clustering gene expression profiles. Clusters of genes with similar expression patterns are identified from the posterior distribution of clusterings defined implicitly by the stochastic data-generation model. The posterior distribution of clusterings is estimated by a Gibbs sampler. We summarized the posterior distribution of clusterings by calculating posterior pairwise probabilities of co-expression and used the complete linkage principle to create clusters. This approach has several advantages over usual clustering procedures. The analysis allows for incorporation of a reasonable probabilistic model for generating data. The method does not require specifying the number of clusters and resulting optimal clustering is obtained by averaging over models with all possible numbers of clusters. Expression profiles that are not similar to any other profile are automatically detected, the method incorporates experimental replicates, and it can be extended to accommodate missing data. This approach represents a qualitative shift in the model-based cluster analysis of expression data because it allows for incorporation of uncertainties involved in the model selection in the final assessment of confidence in similarities of expression profiles. We also demonstrated the importance of incorporating the information on experimental variability into the clustering model. AVAILABILITY The MS Windows(TM) based program implementing the Gibbs sampler and supplemental material is available at http://homepages.uc.edu/~medvedm/BioinformaticsSupplement.htm CONTACT medvedm@email.uc.edu

MicroRNA-494 Targeting Both Proapoptotic and Antiapoptotic Proteins Protects Against Ischemia/Reperfusion-Induced Cardiac Injury

Background— MicroRNAs (miRs) participate in many cardiac pathophysiological processes, including ischemia/reperfusion (I/R)-induced cardiac injury. Recently, we and others observed that miR-494 was downregulated in murine I/R-injured and human infarcted hearts. However, the functional consequence of miR-494 in response to I/R remains unknown. Methods and Results— We generated a mouse model with cardiac-specific overexpression of miR-494. Transgenic hearts and wild-type hearts from multiple lines were subjected to global no-flow I/R with the Langendorff system. Transgenic hearts exhibited improved recovery of contractile performance over the reperfusion period. This improvement was accompanied by remarkable decreases in both lactate dehydrogenase release and the extent of apoptosis in transgenic hearts compared with wild-type hearts. In addition, myocardial infarction size was significantly reduced in transgenic hearts on I/R in vivo compared with wild-type hearts. Similarly, short-term overexpression of miR-494 in cultured adult cardiomyocytes demonstrated an inhibition of caspase-3 activity and reduced cell death on simulated I/R. In vivo treatment with antisense oligonucleotide miR-494 increased I/R-triggered cardiac injury relative to the administration of mutant antisense oligonucleotide miR-494 and saline controls. We further identified that 3 proapoptotic proteins (PTEN, ROCK1, and CaMKII&dgr;) and 2 antiapoptotic proteins (FGFR2 and LIF) were authentic targets for miR-494. Importantly, the Akt-mitochondrial signaling pathway was activated in miR-494–overexpressing myocytes. Conclusions— Our findings suggest that although miR-494 targets both proapoptotic and antiapoptotic proteins, the ultimate consequence is activation of the Akt pathway, leading to cardioprotective effects against I/R-induced injury. Thus, miR-494 may constitute a new therapeutic agent for the treatment of ischemic heart disease.

Clustering gene-expression data with repeated measurements

Clustering is a common methodology for the analysis of array data, and many research laboratories are generating array data with repeated measurements. We evaluated several clustering algorithms that incorporate repeated measurements, and show that algorithms that take advantage of repeated measurements yield more accurate and more stable clusters. In particular, we show that the infinite mixture model-based approach with a built-in error model produces superior results.

Persistent Hypomethylation in the Promoter of Nucleosomal Binding Protein 1 (Nsbp1) Correlates with Overexpression of Nsbp1 in Mouse Uteri Neonatally Exposed to Diethylstilbestrol or Genistein

Neonatal exposure of CD-1 mice to diethylstilbestrol (DES) or genistein (GEN) induces uterine adenocarcinoma in aging animals. Uterine carcinogenesis in this model is ovarian dependent because its evolution is blocked by prepubertal ovariectomy. This study seeks to discover novel uterine genes whose expression is altered by such early endocrine disruption via an epigenetic mechanism. Neonatal mice were treated with 1 or 1000 microg/kg DES, 50 mg/kg GEN, or oil (control) on d 1-5. One group of treated mice was killed before puberty on d 19. Others were ovariectomized or left intact, and killed at 6 and 18 months of age. Methylation-sensitive restriction fingerprinting was performed to identify differentially methylated sequences associated with neonatal exposure to DES/GEN. Among 14 candidates, nucleosomal binding protein 1 (Nsbp1), the gene for a nucleosome-core-particle binding protein, was selected for further study because of its central role in chromatin remodeling. In uteri of immature control mice, Nsbp1 promoter CpG island (CGI) was minimally methylated. Once control mice reached puberty, the Nsbp1 CGI became hypermethylated, and gene expression declined further. In contrast, in neonatal DES/GEN-treated mice, the Nsbp1 CGI stayed anomalously hypomethylated, and the gene exhibited persistent overexpression throughout life. However, if neonatal DES/GEN-treated mice were ovariectomized before puberty, the CGI remained minimally to moderately methylated, and gene expression was subdued except in the group treated with 1000 microg/kg DES. Thus, the life reprogramming of uterine Nsbp1 expression by neonatal DES/GEN exposure appears to be mediated by an epigenetic mechanism that interacts with ovarian hormones in adulthood.

LRpath: a logistic regression approach for identifying enriched biological groups in gene expression data.

MOTIVATION The elucidation of biological pathways enriched with differentially expressed genes has become an integral part of the analysis and interpretation of microarray data. Several statistical methods are commonly used in this context, but the question of the optimal approach has still not been resolved. RESULTS We present a logistic regression-based method (LRpath) for identifying predefined sets of biologically related genes enriched with (or depleted of) differentially expressed transcripts in microarray experiments. We functionally relate the odds of gene set membership with the significance of differential expression, and calculate adjusted P-values as a measure of statistical significance. The new approach is compared with Fishers exact test and other relevant methods in a simulation study and in the analysis of two breast cancer datasets. Overall results were concordant between the simulation study and the experimental data analysis, and provide useful information to investigators seeking to choose the appropriate method. LRpath displayed robust behavior and improved statistical power compared with tested alternatives. It is applicable in experiments involving two or more sample types, and accepts significance statistics of the investigators choice as input.

Conditional Activation of RET/PTC3 and BRAFV600E in Thyroid Cells Is Associated with Gene Expression Profiles that Predict a Preferential Role of BRAF in Extracellular Matrix Remodeling

Papillary thyroid cancers (PTC) are associated with nonoverlapping mutations of genes coding for mitogen-activated protein kinase signaling effectors (i.e., the TK receptors RET or NTRK and the signaling proteins RAS and BRAF). We examined the pattern of gene expression after activation of these oncoproteins in thyroid PCCL3 cells, with the goal of identifying pathways or gene subsets that may account for the phenotypic differences observed in human cancers. We hybridized cDNA from cells treated with or without doxycycline to induce expression of BRAF(V600E), RET/PTC3, or RET/PTC3 with small interfering RNA-mediated knockdown of BRAF, respectively, to slides arrayed with a rat 70-mer oligonucleotide library consisting of 27,342 oligos. Among the RET/PTC3-induced genes, 2,552 did not require BRAF as they were similarly regulated by RET/PTC3 with or without BRAF knockdown and not by expression of BRAF(V600E). Immune response and IFN-related genes were highly represented in this group. About 24% of RET/PTC3-regulated genes were BRAF dependent, as they were similarly modified by RET/PTC3 and BRAF(V600E) but not in cells expressing RET/PTC3 with knockdown of BRAF. A gene cluster coding for components of the mitochondrial electron transport chain pathway was down-regulated in this group, potentially altering regulation of cell viability. Metalloproteinases were also preferentially induced by BRAF, particularly matrix metalloproteinase 3 (MMP3), MMP9, and MMP13. Accordingly, conditional expression of BRAF was associated with markedly increased invasion into Matrigel compared with cells expressing RET/PTC3. The preferential induction of MMPs by BRAF could explain in part the more invasive behavior of thyroid cancers with BRAF mutations.

Progression of BRAF -induced thyroid cancer is associated with epithelial–mesenchymal transition requiring concomitant MAP kinase and TGFβ signaling

Mice with thyroid-specific expression of oncogenic BRAF (Tg-Braf) develop papillary thyroid cancers (PTCs) that are locally invasive and have well-defined foci of poorly differentiated thyroid carcinoma (PDTC). To investigate the PTC–PDTC progression, we performed a microarray analysis using RNA from paired samples of PDTC and PTC collected from the same animals by laser capture microdissection. Analysis of eight paired samples revealed a profound deregulation of genes involved in cell adhesion and intracellular junctions, with changes consistent with an epithelial–mesenchymal transition (EMT). This was confirmed by immunohistochemistry, as vimentin expression was increased and E-cadherin lost in PDTC compared with adjacent PTC. Moreover, PDTC stained positively for phospho-Smad2, suggesting a role for transforming growth factor (TGF)β in mediating this process. Accordingly, TGFβ-induced EMT in primary cultures of thyroid cells from Tg-Braf mice, whereas wild-type thyroid cells retained their epithelial features. TGFβ-induced Smad2 phosphorylation, transcriptional activity and induction of EMT required mitogen-activated protein kinase (MAPK) pathway activation in Tg-Braf thyrocytes. Hence, tumor initiation by oncogenic BRAF renders thyroid cells susceptible to TGFβ-induced EMT, through a MAPK-dependent process.

From co-expression to co-regulation: how many microarray experiments do we need?

BackgroundCluster analysis is often used to infer regulatory modules or biological function by associating unknown genes with other genes that have similar expression patterns and known regulatory elements or functions. However, clustering results may not have any biological relevance.ResultsWe applied various clustering algorithms to microarray datasets with different sizes, and we evaluated the clustering results by determining the fraction of gene pairs from the same clusters that share at least one known common transcription factor. We used both yeast transcription factor databases (SCPD, YPD) and chromatin immunoprecipitation (ChIP) data to evaluate our clustering results. We showed that the ability to identify co-regulated genes from clustering results is strongly dependent on the number of microarray experiments used in cluster analysis and the accuracy of these associations plateaus at between 50 and 100 experiments on yeast data. Moreover, the model-based clustering algorithm MCLUST consistently outperforms more traditional methods in accurately assigning co-regulated genes to the same clusters on standardized data.ConclusionsOur results are consistent with respect to independent evaluation criteria that strengthen our confidence in our results. However, when one compares ChIP data to YPD, the false-negative rate is approximately 80% using the recommended p-value of 0.001. In addition, we showed that even with large numbers of experiments, the false-positive rate may exceed the true-positive rate. In particular, even when all experiments are included, the best results produce clusters with only a 28% true-positive rate using known gene transcription factor interactions.

Control of Nutrient Stress-Induced Metabolic Reprogramming by PKCζ in Tumorigenesis

Tumor cells have high-energetic and anabolic needs and are known to adapt their metabolism to be able to survive and keep proliferating under conditions of nutrient stress. We show that PKCζ deficiency promotes the plasticity necessary for cancer cells to reprogram their metabolism to utilize glutamine through the serine biosynthetic pathway in the absence of glucose. PKCζ represses the expression of two key enzymes of the pathway, PHGDH and PSAT1, and phosphorylates PHGDH at key residues to inhibit its enzymatic activity. Interestingly, the loss of PKCζ in mice results in enhanced intestinal tumorigenesis and increased levels of these two metabolic enzymes, whereas patients with low levels of PKCζ have a poor prognosis. Furthermore, PKCζ and caspase-3 activities are correlated with PHGDH levels in human intestinal tumors. Taken together, this demonstrates that PKCζ is a critical metabolic tumor suppressor in mouse and human cancer.