The goal of the present study is to characterize the transcriptomic regulatory dynamics operational in the regenerating rat liver affected by chronic alcohol intake. While transcriptomic data on the gene expression changes during normal liver regeneration is increasingly becoming available from multiple studies, the effect of chronic alcohol consumption on these changes is less clear. Such an alcohol-mediated disruption of repair mechanisms presumably contributes to the onset of liver damage in alcoholic liver disease.
We take an integrated experimental and computational approach to investigate the dynamics of the system-wide transcriptional changes, use sensitive bioinformatics methods to infer regulatory modules comprising of transcription factors (TFs) and target genes, and validate these for changes in the activity of TFs and their binding at target gene promoters. In our experimental approach, rats were fed a liquid diet containing 36% of total calories derived from ethanol; corresponding pair-fed calorie-matched controls received liquid diets in which ethanol calories were replaced either by carbohydrate or by fat. After five weeks, rats were subjected to 2/3d PHx by surgical excision of left-lateral and medial (LLM) lobes. Remnant liver tissues were harvested after 1, 6 and 24h to represent the priming phase, early G1, and G1/S transition, respectively. LLM tissues obtained at t=0 served as within-animal controls. We used Affymetrix Rat Gene 1.0 ST arrays with ~25000 probe sets to obtain global gene expression data from each liver sample (n=4 replicate animals, 72 arrays total). Analysis of the gene expression data using a 3-way ANOVA revealed a total of 6893 transcript clusters as responsive to PHx in at least one of the diet groups at either 1, 6 or 24h post PHx. Clustering the sample replicates indicated that the gene expression response to PHx is the primary factor separating the groups.
We developed a novel dynamic response pattern analysis approach in which the average differential gene expression data was discretized to three levels (+1, 0, -1) based on a fold change threshold of 1.5 up or down regulation. Within each diet group, this discretization yielded a dynamic response pattern for each gene, encoded by one of 27 possible ordered sets of the three levels +1, 0, and -1. Pairs of diet groups were compared to count the number of genes that follow each of the possible 27 * 27 (=729) comparative dynamic response patterns. Only a select set of the possible 729 patterns contained more than 25 genes indicating that the ethanol effects are mediated through specific mechanisms. The largest set of genes with different patterns between the diet groups were those with lack of differential regulation in ethanol group. This was followed by another gene set with de novo differential expression in ethanol group but not in that of the control diets. Other patterns involved alterations of dynamic gene expression from a transient response at one of the time points in control groups to a more persistent differential gene expression in ethanol group, and vice versa. Notably, ethanol group showed a marked lack of up regulation of cell cycle related gene expression at 6 and 24h.
In complementary experiments, we detected genome-wide NF-kB binding targets using Roche NimbleGen ChIP-chip microarray platform using Chromatin Immunoprecipitated (ChIP) samples. Our results indicate that PHx induced significant increase in NF-kB genome-wide binding at 1h, more so in chronic ethanol samples (5300 genes) than in controls (4000 genes), with several common targets (~3300 genes). At 6h post PHx, only the control samples showed a further increase in NF-kB promoter binding (~5500 genes). Pathway analysis revealed that NF-kB target genes specific to the chronic ethanol group participate in key processes such as DNA methylation, cell death, proteolysis, histone modification, and regulation of cell cycle. Our dynamic response pattern analysis of NF-kB promoter binding revealed similar alterations seen in gene expression response patterns, notably the conversion of persistent response to a transient response and vice versa in the ethanol group as compared to the control diet groups.
Taken together, our results provide a first system-wide view of the regulatory network dynamics in liver regeneration as it is affected by alcohol, and implicate combinatorial regulatory modules as participating in the adverse liver regeneration response. The above results indicate a phenomenon in which compensatory mechanisms that lead to an adaptive state in alcoholic rats play a role in suppressing the sensitivity of the system to certain perturbations, affecting the ability of the liver to initiate effective repair responses in the face of additional environmental challenges.
Supported by: NIH AA008714, AA014986, AA017261, AA016919, and AA018873.