A molecular signature of normal breast epithelial and stromal cells from Li-Fraumeni syndrome mutation carriers

Abstract

Specific changes in gene expression during cancer initiation should enable discovery of biomarkers for risk assessment, early detection and targets for chemoprevention. It has been previously demonstrated that altered mRNA and proteome signatures of morphologically normal cells bearing a single inherited "hit" in a tumor suppressor gene parallel many changes observed in the corresponding sporadic cancer. Here, we report on the global gene expression profile of morphologically normal, cultured primary breast epithelial and stromal cells from Li-Fraumeni syndrome (LFS) TP53 mutation carriers. Our analyses identified multiple changes in gene expression in both morphologically normal breast epithelial and stromal cells associated with TP53 haploinsufficiency, as well as interlocking pathways. Notably, a dysregulated p53 signaling pathway was readily detectable. Pharmacological intervention with the p53 rescue compounds CP-31398 and PRIMA-1 provided further evidence in support of the central role of p53 in affecting these changes in LFS cells and treatment for this cancer. Because loss of signaling mediated by TP53 is associated with the development and survival of many human tumors, identification of gene expression profiles in morphologically normal cells that carry "one-hit" p53 mutations may reveal novel biomarkers, enabling the discovery of potential targets for chemoprevention of sporadic tumors as well.

Figure 1. Gene expression patterns between TP53 heterozygous and WT breast epithelial and stromal cells.

A) Principal component analyses (PCA) of samples. Spheres depict breast epithelial samples (BR), while triangles depict stromal samples. Red objects represent LFS patient 50 samples, blue objects represent samples derived from IUSM-LFS patient, and green objects represent normal/WT samples. B) PCA of LFS-50 compared to LFS-IUSM samples. C) Supervised heat-map with gene expression patterns of the top 100 genes noting clusters differentiating diseased vs. non-diseased samples. The bars above the panel depict sample clusters; top: tissue/sample type (left to right: BR, breast epithelial; STR, stromal); middle: ID/cell line (left to right: NA, normal/WT; 50, patient 50; IUSM, cells derived from IUSM patient); bottom: genotype/phenotype (left to right: N, normal/WT; LF, Li-Fraumeni syndrome). The different colored bars on the left of the panel represent different clusters of biological processes. Gene expression variation is depicted by color (red, up-regulated; blue, down-regulated; gray, no significant change). The genes and Gene Ontology of Biological Processes are listed in Supplemental Data.

Figure 2. Top gene networks generated from IPA and significantly modulated (log p-value) between both of the LFS cells and WT cells.

A) Top ten networks significantly modulated between LFS and WT breast epithelial cells. B) Top ten networks significantly modulated between LFS and WT breast stromal cells. Molecules associated within these pathways are listed in Table 4.

Figure 3. Ingenuity pathway analysis of genes differentially regulated in LFS vs. WT epithelial cells.

Functional pathway analysis by IPA of ERK (A) and IL1B/p300/BIRC3 (B) genes and their interaction nodes in LFS breast epithelial cells relative to WT breast epithelial cells. Gene expression variation by at least 2-fold is depicted by color (red, up-regulated; green, down-regulated; gray, no significant change).

Figure 4. Ingenuity pathway analysis of genes differentially regulated in LFS vs. WT stromal cells.

Functional pathway analysis by IPA of IL1B/CDK2 (A) and TP53 (B) gene pathways and their interaction nodes in LFS breast stromal cells relative to WT breast stromal cells.

Figure 5. Ingenuity Pathway Analysis (IPA) for LFS-50 vs. WT Epithelial Cells and the Comparison of LFS-IUSM vs. WT Epithelial Cells.

Functional pathway analysis by IPA of LFS-50 vs. WT Epithelial Cells (A) and LFS-IUSM vs. WT Epithelial Cells (B) gene comparisons and their interaction nodes.

Figure 6. TP53 rescue agents restore expression of dysregulated genes in LFS breast epithelial cells.

A) Effects on LFS-50 breast epithelial cell growth by the p53 rescue agents PRIMA-1 and CP-31398, compared to untreated control (normalized to 100%). B) Analysis of top genes (via qRT-PCR) from LFS microarray/p53 network of LFS epithelial cells (LFS-50 breast epithelial cells) treated with 10 μM PRIMA-1, CP-31398, or combination of both. Data is average of at least two independent experiments, with three replicates per treatment group, plus standard error. Statistical significance was determined by a two-tailed Students' t-test (MS Excel) where P<0.05 was considered significant (*, P<0.05; **, P<0.01; ***, P<0.001).