Comparative analysis of the human hepatic and adipose tissue transcriptomes during LPS-induced inflammation leads to the identification of differential biological pathways and candidate biomarkers.

BACKGROUND: Insulin resistance (IR) is accompanied by chronic low grade systemic inflammation, obesity, and deregulation of total body energy homeostasis. We induced inflammation in adipose and liver tissues in vitro in order to mimic inflammation in vivo with the aim to identify tissue-specific processes implicated in IR and to find biomarkers indicative for tissue-specific IR. METHODS: Human adipose and liver tissues were cultured in the absence or presence of LPS and DNA Microarray Technology was applied for their transcriptome analysis. Gene Ontology (GO), gene functional analysis, and prediction of genes encoding for secretome were performed using publicly available bioinformatics tools (DAVID, STRING, SecretomeP). The transcriptome data were validated by proteomics analysis of the inflamed adipose tissue secretome. RESULTS: LPS treatment significantly affected 667 and 483 genes in adipose and liver tissues respectively. The GO analysis revealed that during inflammation adipose tissue, compared to liver tissue, had more significantly upregulated genes, GO terms, and functional clusters related to inflammation and angiogenesis. The secretome prediction led to identification of 399 and 236 genes in adipose and liver tissue respectively. The secretomes of both tissues shared 66 genes and the remaining genes were the differential candidate biomarkers indicative for inflamed adipose or liver tissue. The transcriptome data of the inflamed adipose tissue secretome showed excellent correlation with the proteomics data. CONCLUSIONS: The higher number of altered proinflammatory genes, GO processes, and genes encoding for secretome during inflammation in adipose tissue compared to liver tissue, suggests that adipose tissue is the major organ contributing to the development of systemic inflammation observed in IR. The identified tissue-specific functional clusters and biomarkers might be used in a strategy for the development of tissue-targeted treatment of insulin resistance in patients.

Resistin is more abundant in liver than adipose tissue and is not up-regulated by lipopolysaccharide.

CONTEXT: Resistin is an adipokine correlated with inflammatory markers and is predictive for cardiovascular diseases. There is evidence that serum resistin levels are elevated in obese patients; however, the role of resistin in insulin resistance and type 2 diabetes remains controversial. OBJECTIVE: We addressed the question of whether inflammation may induce expression of resistin in organs involved in regulation of total body energy metabolism, such as liver and adipose tissue (AT). METHODS: Human liver tissue, sc AT, and omentum were cultured in the absence/presence of lipopolysaccharide (LPS). The resistin and cytokine mRNA and protein expression levels were determined by real-time PCR, ELISA, and Multiplex Technology, respectively. The localization of resistin in human liver was analyzed by immunohistochemistry. RESULTS: Resistin gene and protein expression was significantly higher in liver than in AT. Exposure of human AT and liver tissue in culture to LPS did not alter resistin concentration; however, concentrations of IL-1beta, IL-6, and TNFalpha were significantly increased in these tissues. In liver, resistin colocalizes with markers for Kupffer cells, for a subset of endothelial and fibroblast-like cells. CONCLUSIONS: High level of resistin gene and protein expression in liver compared to AT implies that resistin should not be considered only as an adipokine in humans. LPS-induced inflammation does not affect resistin protein synthesis in human liver and AT. This suggests that elevated serum resistin levels are not indicative for inflammation of AT or liver in a manner similar to known inflammatory markers such as IL-1beta, IL-6, or TNFalpha.

Analysis of bile acid-induced regulation of FXR target genes in human liver slices.

Information about the role of nuclear receptors has rapidly increased over the last decade. However, details about their role in human are lacking. Owing to species differences, a powerful human in vitro system is needed. This study uses for the first time precision-cut human liver slices in the nuclear receptor field. The farnesoid X receptor (FXR) was chosen as a model. We were able to demonstrate that human liver slices efficiently take up bile acids and show a stable expression of a wide variety of genes relevant for bile acid metabolism, including bile acid transporters, cytochrome P450 enzymes and transcription factors. Treatment with chenodeoxycholate induced small heterodimer partner, bile salt export pump and p-glycoprotein, ABCB4 and repressed cholesterol 7alpha hydroxylase, hepatocyte nuclear factor (HNF)1, HNF4 and organic anion transporting peptide (OATP)1B1. OATP1B3, FXR, HNF3beta and cytochrome P450 enzyme remained relatively constant. In contrast to what has been observed in mice and rat studies, SHP induction did not result in repression of sodium-dependent bile acid cotransporter expression. Further, regulation of genes seemed to be dependent on concentration and time. Taken together, the study shows that the use of liver slices is a powerful technique that enables to study nuclear receptors in the human liver.

LPS-induced downregulation of MRP2 and BSEP in human liver is due to a posttranscriptional process.

Endotoxin-induced cholestasis in rodents is caused by hepatic downregulation of transporters, including the basolateral Na+-dependent taurocholate transporter (ntcp) and the canalicular bile salt export pump (bsep) and multidrug resistance-associated protein 2 (mrp2). Details about the regulation of the human transporter proteins during this process are lacking. We used precision-cut human and rat liver slices to study the regulation of transporter expression during LPS-induced cholestasis. We investigated the effect of LPS on nitrate/nitrite and cytokine production in relation to the expression of inducible nitric oxide synthase, NTCP, BSEP, and MRP2 both at the level of mRNA with RT-PCR and protein using immunofluorescence microscopy. In liver slices from both species, LPS-induced expression of inducible nitric oxide synthase was detected within 1-3 h and remained increased over 24 h. In rat liver slices, this was accompanied by a significant decrease of rat ntcp and mrp2 mRNA levels, whereas bsep levels were not affected. These results are in line with previous in vivo studies and validate our liver slice technique. In LPS-treated human liver slices, NTCP mRNA was downregulated and showed an inverse correlation with the amounts of TNF-alpha and Il-1beta produced. In contrast, MRP2 and BSEP mRNA levels were not affected under these conditions. However, after 24-h LPS challenge, both proteins were virtually absent in human liver slices, whereas marker proteins remained detectable. In conclusion, we show that posttranscriptional mechanisms play a more prominent role in LPS-induced regulation of human MRP2 and BSEP compared with the rat transporter proteins.