Gene expression abnormalities in histologically normal breast epithelium of breast cancer patients

Abstract

Normal‐appearing epithelium of cancer patients can harbor occult genetic abnormalities. Data comprehensively comparing gene expression between histologically normal breast epithelium of breast cancer patients and cancer‐free controls are limited. The present study compares global gene expression between these groups. We performed microarrays using RNA from microdissected histologically normal terminal ductal‐lobular units (TDLU) from 2 groups: (i) cancer normal (CN) (TDLUs adjacent to untreated ER+ breast cancers (n = 14)) and (ii) reduction mammoplasty (RM) (TDLUs of age‐matched women without breast disease (n = 15)). Cyber‐T identified differentially expressed genes. Quantitative RT‐PCR (qRT‐PCR), immunohistochemistry (IHC), and comparison to independent microarray data including 6 carcinomas in situ (CIS), validated the results. Gene ontology (GO), UniProt and published literature evaluated gene function. About 127 probesets, corresponding to 105 genes, were differentially expressed between CN and RM (p < 0.0009, corresponding to FDR <0.10). 104/127 (82%) probesets were also differentially expressed between CIS and RM, nearly always (102/104 (98%)) in the same direction as in CN vs. RM. Two‐thirds of the 105 genes were implicated previously in carcinogenesis. Overrepresented functional groups included transcription, G‐protein coupled and chemokine receptor activity, the MAPK cascade and immediate early genes. Most genes in these categories were under‐expressed in CN vs. RM. We conclude that global gene expression abnormalities exist in normal epithelium of breast cancer patients and are also present in early cancers. Thus, cancer‐related pathways may be perturbed in normal epithelium. These abnormalities could be markers of disease risk, occult disease, or the tissue's response to an existing tumor. © 2007 Wiley‐Liss, Inc.

Gene expression in histologically normal epithelium from breast cancer patients and from cancer-free prophylactic mastectomy patients shares a similar profile

“…Of the 98 probe sets, 66 (67%) had a higher gene expression in RM samples, and 32 of the 98 (33%) had a higher gene expression in HN samples. In all, 36 of the 98 probe sets (37%) were among the 127 probe sets identified in our initial report, which compared NlEpi between RM and HN samples only from ER þ breast cancer patients (Tripathi et al, 2008). A clustering analysis of RM and HN samples using these 98 probe sets shows that the two groups generally separate, although five HN samples cluster with RMs (see Supplementary Figure S2).…”

“…Tissue preparation, microdissection, RNA extraction and amplification, array hybridisation, and data normalisation were performed as described earlier (Tripathi et al, 2008). Briefly, tissues were snap frozen, embedded in optimal cutting temperature embedding medium, sectioned at 10 mm, stained with diluted haematoxylin and eosin (H&E) (see Figure 1), and then NlEpi -both TDLUs and ducts -was microdissected (see Supplementary Figure S1).…”

“…Controls and cases were tightly age matched -no pair differed by more than 2 years -to adjust for age-associated changes and generate age-independent data. Data from 11 RM and 5 HN samples were reported earlier (Tripathi et al, 2008). The third, the high-risk group, consisted of women with a personal history or with a strong family history of breast cancer, or who were known BRCA mutation carriers, who were undergoing PM of a cancer-free breast (PM ¼ 6).…”

“…Much less is known about the gene expression in premalignant breast tissue, and studies that focused on HN epithelium (NlEpi) are particularly limited (Finak et al, 2006;Grigoriadis et al, 2006;Tripathi et al, 2008;Chen et al, 2009). This is partly because of the difficulty in obtaining homogeneous epithelial cell populations from fresh tissue.…”

“…Some evidence already exists suggesting that alterations could have clinical significance (Yang et al, 2005;Chen et al, 2009). In our pilot study, using a different statistical approach and study design, we identified the differences in gene expression between microdissected NlEpi of controls (women undergoing reduction mammoplasty (RM)) and those of cases (women with oestrogen receptor (ER) þ breast cancers) (Tripathi et al, 2008). That study, however, could not address whether these differences were an effect of the existing tumour, a marker of increased cancer risk, or a profile of early carcinogenesis.…”