Powerful New Genetic Tools May Unlock Old Mysteries

Coordinate Expression of Regulatory Genes Differentiates Embryonic and Perinatal Forms of Biliary Atresia. Zhang D, Sabla G, Shivakumar P, Tiao G, Sokol R, Mack C, Shneider B, Aronow B, Bezerra J. Hepatology 2004;39:954–962. Summary: The authors report a novel description of transcriptional differences between the two clinical forms of biliary atresia (BA) based on analysis of gene chip microarrays. Patients with BA who had undergone liver wedge biopsies as part of their diagnostic evaluation for BA were classified as having either an embryonic form, in which BA is associated with major congenital nonhepatic malformations and early onset conjugated hyperbilirubinemia (n = 5), or a perinatal form with later onset jaundice and no major associated congenital malformations (n = 6). Microarray analysis was performed in the liver samples of these patients against 44,760 probe sets corresponding to a high fraction of the genes of the human genome. Analysis of the data by hierarchical clustering and supervised data filtering according to clinical form identified 230 genes with significant differences between the two groups and generated an expression profile that clearly differentiated the embryonic from the perinatal form of BA. Regulatory genes (whose functions include DNA and RNA processing, imprinting, laterality, signal transduction, transcription regulation and cell cycle control) were predominantly overexpressed in the embryonic form of BA (35%), whereas regulators of metabolic function were overexpressed primarily in the perinatal form (40%). Hepatic expression of imprinted genes was uniformly increased in the embryonic form of BA 1.8- to 2.9-fold above levels of infants with the perinatal form, and 1.8- to 3.5-fold above levels of the infants with neonatal cholestasis. Of the 45 genes that function to regulate laterality, only 4 showed increased expression in the embryonic versus the perinatal form. Livers of infants with the embryonic form had a mild increase (1.4–1.8-fold) in expression of six genes that function as regulators of cell cycle progression. Testing for functional differentiation of helper T cells consistent with prior reports of increased levels of proinflammatory cytokines in BA failed to reveal significant differences in the pattern of expression of immunity/inflammation genes between the different forms of BA. Additional quantitative analysis of cytokines by real-time polymerase chain reaction did not reveal a preferential activation of proinflammatory cytokines in livers of infants with the embryonic form of BA. Comment: The powerful technology of microarray analysis has allowed researchers to obtain unbiased surveys of gene expression in tissue samples on a genome-wide scale and has been especially useful in clinical oncology, where different biologic subtypes of cancer are accompanied by differences in transcriptional programs (1). Expression microarrays contain tens of thousands of short fragments of genomic DNA tethered to a flat surface in an orderly array. Each short segment of DNA corresponds to a small gene segment. RNA that is produced from the genes in that DNA segment will bind to the DNA in a specific location on the array. The quantity of RNA then bound to each particular location can be analyzed and compared. By this method, relative expression of numerous genes can be quantified from RNA extracted from tissue samples. The microarray results are quantified by image analysis of the chip using software that extracts measures of RNA bound to specific locations on the chip and processes the output through filters and an analytic algorithm to produce descriptions of the pattern of binding. The investigators in this study applied a standard filter to the expression data such that only those genes whose expression actually varied between the clinical types of BA were included in the analysis. They then sought to identify a core group of genes whose differential expression could be used to search for molecular signatures highly associated with each clinical form. This strategy revealed no preferential clustering of differentially expressed genes in specific chromosomal segments. This core group of genes was further classified according to their biologic function. There was expression of regulatory genes segregating toward the embryonic form of BA, which was interpreted by the authors as revealing a core molecular footprint orchestrating several biologic processes. The authors assert that within these processes, regulators of imprinting, laterality, and cell cycle progression may be relevant to understanding the phenotype of BA. Among these genes, only regulators of laterality have been associated with BA (2). Interestingly, the laterality genes that significantly differed between the groups have not been associated with defects in laterality in humans. The authors speculate that this may represent adaptive changes from a not-yet-defined genetic defect. Clearly, analysis of these genes in a larger group of subjects is necessary to address how this increased expression may influence laterality in infants with BA. A very interesting finding seen in the infants with the embryonic form of BA is the increase in the expression of five imprinted genes, as well as genes involved in chromatin structure/histone deacetylation. Genomic imprinting is a form of gene silencing that is an epigenetic modification of a specific parental allele of a gene leading to differential expression of the two alleles of the gene in somatic cells of the offspring. Loss of imprinting simply means loss of preferential parental origin-specific gene expression. This can involve abnormal expression of the normally silent allele and lead to bi-allelic expression. Alternatively, silencing of the normally expressed allele can result, leading to epigenetic silencing of the locus. This is found in abnormal imprinting in certain cancers, leading to activation of normally silent growth-promoting genes, such as the insulin growth factor II (IGF2) gene, whose expression is increased in colorectal cancer and embryonal tumors, including Wilms (3). This finding of increased expression of IGF2, Smarca-1, Hda3, and Rybp, raises the possibility that stable alterations in gene expression that arise during development and cell proliferation may indeed modulate the phenotypic features of infants with the embryonic form of BA. Clearly, validation of this requires additional studies of index genes implicated in loss of imprinting and DNA methylation in affected infants. Microarray analysis is a powerful research tool. It allows the clinical investigation of small numbers of patients, yielding a high density of data that can uncover robust linkages between clinical phenotypes and molecular signatures of gene expression. This information may reveal hints at dominant biologic processes involved in disease pathogenesis. It is a process that, applied to hypothesis-driven analysis, will lead to a rapid accumulation of new information. The challenge will remain in the translation of this information into an understanding of pathophysiology and advancement of clinical treatments. D. Ekong Karan Emerick Northwestern University Medical School, Chicago, Illinois, U.S.A.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.