Gene expression profiles of human breast cancer progression

Abstract

Although distinct pathological stages of breast cancer have been described, the molecular differences among these stages are largely unknown. Here, through the combined use of laser capture microdissection and DNA microarrays, we have generated in situ gene expression profiles of the premalignant, preinvasive, and invasive stages of human breast cancer. Our data reveal extensive similarities at the transcriptome level among the distinct stages of progression and suggest that gene expression alterations conferring the potential for invasive growth are already present in the preinvasive stages. In contrast to tumor stage, different tumor grades are associated with distinct gene expression signatures. Furthermore, a subset of genes associated with high tumor grade is quantitatively correlated with the transition from preinvasive to invasive growth.

Figure 1

LCM. Phenotypically normal breast epithelium (white arrows) and phenotypically abnormal epithelium (black arrows) from ADH, DCIS, and IDC from a single breast specimen (case 79) were captured from hematoxylin and eosin-stained sections (8 μm). Images of precapture (lane a), postcapture (lane b), and the captured epithelial compartments (lane c) are shown.

Figure 2

Expression profiles of breast cancer progression. (A) Two-dimensional hierarchical clustering of the data matrix consisting of 1,940 genes by 61 samples of different pathological stages (Table 4). Rows represent genes and columns represent samples, which are color-coded by tumor grade (blue, green, and red correspond to grades I, II, and III, respectively). Color scale is shown at bottom left. (B) In situ hybridization of CRIP1 mRNA. DIG-labeled RNA probes from both the antisense and sense (negative control) strands of CRIP1 transcript were hybridized to sections of normal and IDC components of case 179. Hybridization signals were visualized by alkaline phosphatase-conjugated anti-DIG antibody using fast red as substrate.

Figure 3

Two-dimensional clustering of 61 samples and the top 200 genes correlating with tumor grade. Genes (rows) and samples (columns) were clustered independently by hierarchical clustering. Three main clusters are highlighted by color bars. See Fig. 2A for color scale and designations. The original data are in Table 5.

Figure 4

Genes with increased expression in IDC relative to DCIS. (A) Cluster of 29 genes showing consistent up-regulation in IDC. Expression values are expressed as log2-ratios of expression in IDC to that in patient-matched DCIS. Color scheme is shown at bottom left; see Fig. 2A for sample color designations. The original data are in Table 5. (B) Confirmation by QRT-PCR of increased expression in IDC for CKS2, RRM2, and UBE2C. Fold changes from DCIS to IDC and associated standard errors are plotted. Data shown are averages of triplicate QRT-PCR measurements. Values outside the scale in the y axis are marked by a star. The patient case numbers from left to right are: 14, 25, 79, 102, 173, 180, and 193 for grade I; 41, 43, 121, 130, 131, 148, 169, 170, and 198 for grade II; and 8, 30, 44, 65, 88, 96, 112, 133, 178, and 179 for grade III.