Use of gene expression and pathway signatures to characterize the complexity of human melanoma

Abstract

A defining characteristic of most human cancers is heterogeneity, resulting from the somatic acquisition of a complex array of genetic and genomic alterations. Dissecting this heterogeneity is critical to developing an understanding of the underlying mechanisms of disease and to paving the way toward personalized treatments of the disease. We used gene expression data sets from the analysis of primary and metastatic melanomas to develop a molecular description of the heterogeneity that characterizes this disease. Unsupervised hierarchical clustering, gene set enrichment analyses, and pathway activity analyses were used to describe the genetic heterogeneity of melanomas. Patterns of gene expression that revealed two distinct classes of primary melanoma, two distinct classes of in-transit melanoma, and at least three subgroups of metastatic melanoma were identified. Expression signatures developed to predict the status of oncogenic signaling pathways were used to explore the biological basis underlying these differential patterns of expression. This analysis of activities revealed unique pathways that distinguished the primary and metastatic subgroups of melanoma. Distinct patterns of gene expression across primary, in-transit, and metastatic melanomas underline the genetic heterogeneity of this disease. This heterogeneity can be described in terms of deregulation of signaling pathways, thus increasing the knowledge of the biological features underlying individual melanomas and potentially directing therapeutic opportunities to individual patients with melanoma.

Figure 1

Unsupervised hierarchical clustering analysis of melanoma expression data. MAS5.0 gene expression data were log transformed and normalized, using BFRM. Expression data were filtered in Cluster (see Materials and Methods) to 1000 genes that are represented in the data matrix. Within the data matrix, columns represent samples; rows, genes; red, higher expression values; and green, lower expression values. The color bar below the data matrix defines the sample in each column. Subgroups of samples exhibiting similar gene expression patterns, as defined by the nodes of the array tree, are color coded within the array tree.

Figure 2

The status of oncogenic signaling pathway activities analyzed in the benign nevi, primary melanoma, and metastatic melanoma samples. Within the heat map, columns represent samples; and rows, the oncogenic signaling pathway analyzed. Pathway activities on a low to high continuum are represented by a blue to red continuum, respectively. The color bar below the heatmap defines the sample in each column. The samples are ordered according to the defined subgroups, as shown in Figure 1 (the array tree above the heat map defines the subgroups of melanoma samples exhibiting similar gene expression patterns). CAT indicates catenin; EGFR, epidermal growth factor receptor; ER, estrogen receptor; HER, 2 human epidermal growth factor receptor 2; PI3k, phosphatidylinositol 3-kinase; PR, progesterone receptor.

Figure 3

Statistically significant differences in the distribution of activities of particular oncogenic signaling pathways identified between subgroups of melanoma samples. A: Oncogenic signaling pathways, as indicated, determined to exhibit statistically significantly distinct levels of activity between the two primary melanoma subgroups (P1 and P2). B: Oncogenic signaling pathways, as indicated, determined to exhibit statistically significantly distinct levels of activity among the three metastatic melanoma subgroups (M1, M2, and M3). Relevant sections of the heat map from Figure 2 are reproduced. The differences in the distribution of activities of particular oncogenic signaling pathways within samples composing one subgroup of melanomas compared with another subgroup of melanomas, as indicated, were quantified with a Mann-Whitney U-test. CAT indicates catenin; EGFR, epidermal growth factor receptor; ER, estrogen receptor; STAT3, signal transducer and activator of transcription.

Figure 4

Unsupervised hierarchical clustering analysis of in-transit melanoma expression data. MAS5.0 gene expression data were log transformed and normalized, using BFRM. Expression data were filtered in Cluster (see Materials and Methods) to 1000 genes that are represented in the data matrix. Within the data matrix, columns represent samples; rows, genes; red, higher expression values; and green, lower expression values. The color bar below the data matrix defines the sample in each column.