|
|
|
|
|
|
|||||||
|
 |
|||||||
|
Contents
*To download a copy
|
Application of Integrative Functional Genomics To Decode Cancer Signatures
Orthologous human and mouse genes from both data sets were selected, and the gene expression data were integrated after standardizing the relative expression levels for both species. In hierarchical clustering analysis of integrated data, gene expression patterns of HCC from Myc, E2f1, and Myc/E2f1 mice were most similar to those of the better survival group of human HCC, whereas the expression patterns of Myc/Tgfa and diethylnitrosamine (DENA)–induced mouse HCC were most similar to those of the poorer survival group of human HCC. These results suggest that these two classes of mouse models might closely recapitulate the molecular patterns of the two subclasses of human HCC. The similarity of gene expression profiles between human and mouse models are in good agreement with the phenotypic characteristics of the tumors (Figure 1). The human tumors with increased proliferation, decreased apoptosis, and worse prognosis are paired with the mouse models that have the same characteristics. The gene expression–based prediction of mouse models is highly concordant with the phenotypes of mice. Myc/Tgfa mice have a typically poor prognosis phenotype, such as an earlier and higher incident rate of HCC development and higher mortality, genomic instability, and expression of poor prognostic marker. Myc and Myc/E2f1 mice have a relatively higher mutation frequency regarding β-catenin as well as a higher nuclear accumulation of the protein, which in human HCC are indicative of lower genomic instability and better prognosis. The fact that these findings were first uncovered by using unsupervised methods and validated later using supervised methods indicates that the underlying principles in gene expression changes are conserved between mouse and human HCC. Figure 1. Phenotypic similarities between hepatocellular carcinomas (HCCs) generated in transgenic mouse models and subclasses A and B of human HCC. The best-fit HCC mouse models can be used to test hypotheses on tumor progression that are generated by analysis of cross-species gene expression patterns or from other experimental data. These models should also be extremely valuable for testing both potential therapeutic targets identified in human studies and preclinical trials of drugs. H, high; L, low; M, medium. In our second study, we extended this comparative functional genomic approach to address the issue pertaining to the cell(s) of origin for tumors. It is axiomatic that cancer cells evolve from normal cells after accumulation of genetic and epigenetic alterations. Also, it has been shown that the gene expression patterns in cancer cells reflect these alterations. Nevertheless, a considerable fraction of the gene expression program of cancer cells is characteristic of the non-transformed cellular lineages from which they originated. Furthermore, analysis of gene expression profiles of cancer cell lines indicated that neither physiological adaptation in vivo nor experimental adaptation in vitro is sufficient to abolish the gene expression programs acquired during development. These data suggest that the global gene expression profiles of tumors might provide critical information on the cellular origin of tumors. Because HCC could originate from both adult hepatocytes and hepatic progenitor cells, we decided to test whether global gene expression analysis of human HCC could identify subtypes of HCC derived from these different cell types. The experimental strategy involved the generation of gene expression data from multiple species suitable for integration and cross-comparison. We integrated, using only orthologous genes, gene expression data from rat fetal hepatoblasts and adult rat hepatocytes with HCC gene expression data from human and mouse models. By applying hierarchical clustering analysis of gene expression patterns from human HCC, mouse HCC, rat fetal hepatoblasts, and adult rat hepatocytes, we identified a new prognostic subtype of HCC that shares gene expression patterns with fetal hepatoblasts. The hepatoblast (HB) subtype is distinguished from other types of HCC by the differential expression of hundreds of genes, and the robustness of this gene expression signature in the HB subtype was validated in an independent cohort of HCC patients. HCC patients who shared a gene expression pattern with fetal hepatoblasts had a poor prognosis. The gene expression program that distinguished this subtype from other types of HCC included markers of hepatic oval cells, suggesting that HCC of this subtype arises from hepatic progenitor cells. Application of network-based pathway analyses of gene expression provided important insights into the pathogenesis of the HB subtype of HCC (Figure 2). Enrichment of predicted JUN and FOS activity in the HB subtype led us to hypothesize that the activator protein 1 (AP-1) complex was the major driving force in tumorigenesis of the HB subtype. Previous studies have shown that Jun is essential for normal liver development, and it could also be crucial for the initiation of HCC development in mice. Furthermore, higher expression of JUN target genes involved in invasive phenotypes (e.g., MMP1, PLAUR, TIMP1, CD44, and VIL2) was observed in the HB subtype of HCC, indicating the cellular origin of these tumors and accounting for the poor prognosis of the affected individuals. Figure 2. Gene networks of activator protein 1 (AP-1) transcription factors in the HCC subtype with liver progenitor cell signature. Upregulated and downregulated genes in the HCC subtype, with progenitor signature, are indicated in red and green, respectively. Genes in gray color are not on the list but are associated with the regulated genes. Gray lines and arrows represent the direction of transcriptional regulation, and plus and minus signs indicate positive and negative regulation of gene expression. Purple lines represent known physical interactions between connected genes. The success of the new experimental and analytical approaches presented here strongly suggests that more integration of independent data sets will enhance our ability to identify key regulatory elements in cancer development. It is, therefore, reasonable to expect that the clinical inference from transcriptome analyses will be significantly strengthened when gene expression data are integrated with diverse genomics information obtained from DNA sequence, array-based comparative genomic hybridization (CGH), and noncoding gene (i.e., microRNA) expression analyses.
|
|||||||
|
 |
he
success of comparative sequence analysis in identifying and characterizing genomic
regulatory regions with important functional roles is well recognized. Indeed,
the neutral theory of molecular evolution provides a framework for the identification
of functional DNA sequences in genomes of different species. The theory predicts
that functionally important elements in genome sequences tend to evolve at a
slower rate than do less important elements. This difference has permitted the
identification of both protein-coding and functional non-coding sequences in
the genome. Although many of the functional genomic elements are protein-coding
sequences, a large number of conserved sequences are regulatory elements known
to modulate gene expression. These observations constituted the basis for our
hypothesis that, if regulatory elements of evolutionarily related species are
conserved, then gene expression signatures reflecting similar phenotypes in
the species would also be conserved. In our first study, we tested this hypothesis
by comparing global expression patterns of orthologous genes in human and mouse
hepatocellular carcinomas (HCCs) and asked whether similar tumor phenotypes
could be recognized and, thus, allow the identification of the best-fit mouse
models for human HCC. We considered it important to use the mouse cancer models
because the extent to which these models reproduce features observed in the
corresponding human conditions is uncertain.
