The ß-Catenin Pathway is Overexpressed in Focal Nodular Hyperplasia but not in Cirrhotic FNH-like Nodules

Focal nodular hyperplasias (FNHs) are benign liver lesions considered to be a hyperplastic response to increased blood flow in otherwise normal liver. In contrast, FNH-like nodules occur in cirrhotic liver but share similar histopathological features. To better understand the pathophysiology of FNH, we performed a transcriptomic analysis. Methods: Affymetrix and cDNA microarrays were used to compare gene expression in eight FNHs with that in tissue from six normal livers. Selected genes were validated with quantitative RT-PCR in 70 benign liver tumors including adenomas and cirrhotic and FNH-like lesions. Results: Among the deregulated genes in FNHs, 19 were physiologically zonated in the normal liver lobule. All six periveinous genes were up-regulated in FNH, whereas 13 genes normally expressed in the periportal area were down-regulated. Immunohistochemistry revealed that glutamine synthetase was markedly overexpressed, forming anastomosed areas usually centered on visible veins. ß-catenin mRNA was slightly but significantly overexpressed, as were several known ß-catenin target genes. Moreover, activated hypophosphorylated ß-catenin protein accumulated in FNH in the absence of activating mutations. These results suggest zonated activation of the ß-catenin pathway specifically in FNH, whereas the other benign hepatocellular tumors, including FNH-like lesions, demonstrated an entirely different pattern of ß-catenin expression. Conclusions: In FNH, increased expression of the ß-catenin pathway was restricted to enlarged periveinous areas, which may explain the slight polyclonal over-proliferation of hepatocytes at the origin of the lesion. FNH-like nodules may have a different pathogenetic origin. Keywords: Disease state analysis In a first experiment, microarray analyses were applied to 7 Focal nodular hyperplasias (FNHs) and 8 non related non-tumor livers using cDNA in-house manufactured arrays following a randomized and blinded unbalanced design. RNA labelling, hybridization and analysis of fluorescence was carried out, as described in [Graudens et al., Genome Biol 2006, 7: R19; PMID: 16542501], considering that uneven numbers of samples were randomly allocated to each of the engineers who were not aware of sample phenotypes. To assess data reproducibility and minimize dye bias effects, each of the samples was measured four times, twice with Cy3 and twice with Cy5. To ensure robustness and flexibility in data analysis, a reference design was used with a universal reference sample (Stratagene, USA) serving as a baseline for the comparisons of tumor samples. Hybridizations were performed onto an 11K human array (11K_VJF-ARRAY, GPL3282), which provides a genome-wide coverage of functional pathways. Raw data were obtained using the ArrayVision™ 7.0 software (Imaging Research Inc., USA); the resulting hybridization data points collected from 30 arrays were stored in a MIAME-compliant database and pre-processed for normalization and filtering as described in [Graudens et al., Genome Biol 2006, 7: R19; PMID: 16542501]. In a second experiment, HG-U133A Affymetrix GeneChipTM arrays were used to compare the expression profiles of 4 Focal nodular hyperplasias (FNHs) and 4 non related non-tumor livers. RNA labelling, hybridization and analysis were carried out following the manufacturer’s instructions (Affymetrix, Santa Clara, CA). Raw data were obtained by using Microarray Suite 5.0 (MAS5) software, embedded in the Affymetrix GeneChip Operating Software (Santa Clara, USA); the resulting raw numerical data (CEL files) - available as supplementary files - collected from 8 Affymetrix GeneChips were pre-processed for normalization and filtering as described in [Rebouissou et al., J Biol Chem 2007, 282(19):14437-46, PMID: 17379603; and Rebuissou et al., in preparation]. A platform-dependant statistical comparison was performed considering Focal nodular hyperplasias (FNHs) versus non tumor liver tissues, using multiple testing procedures to evaluate statistical significance for differentially expressed genes, as described in [Graudens et al., Genome Biol 2006, 7: R19; PMID: 16542501] and in the supplemental experimental procedures in [Rebouissou et al., J Biol Chem 2007, 282(19):14437-46, PMID: 17379603; and Rebuissou et al., in preparation]. Thus, a combined cross-platform analysis was done considering the UGCluster Identifiants as a common primary key (Unigene Build184-June05). Additional information on the procedures is provided in the supplemental experimental procedures in [Rebouissou et al., J Biol Chem 2007, 282(19):14437-46, PMID: 17379603; and Rebuissou et al., in preparation].