Metastasis is the primary cause of death of patients with breast cancer. A xenograft model of human breast cancer progression was used to identify genes potentially involved with metastasis, comparing gene expression patterns in the poorly metastatic GI101A human breast cancer cell line and a highly metastatic variant GILM2. cDNA microarray analyses of these isogenic variants were performed using 16K Operon 70-mer oligonucleotide microarray slides. Differentially expressed genes were identified by Analysis of Variance, and differences of 2.5-fold or greater in expression were found for 106 genes. The differences in expression were confirmed by independent assays using RNA and/or protein samples from the two cell lines for ten out of twelve genes. Three of the markers identified, the heat shock protein HSP-70, the chemokine CXCL-1, and the serine protease inhibitor SLPI, were studied further using large cohort breast cancer tissue microarrays, with a novel method of automated, quantitative analysis (AQUA™). The system used cytokeratin to define pixels as breast cancer within the tissue array spot, and then measured intensity of marker expression using Cy5-conjugated antibodies within the defined areas. Scores were correlated with clinical and pathological variables. High HSP-70 and high nuclear CXCL-1 expression in primary human breast cancers were both associated with decreased survival (P = 0.05 and P = 0.026, respectively). Expression of each marker was strongly associated with lymph node involvement (HSP-70, P= 0.0002; nuclear CXCL-1, P= 0.008, cytoplasmic CXCL-1, P= 0.0012; SLPI, P= 0.012). The identification of genes associated with breast cancer metastasis may have clinical implications for the management of breast cancer, as these genes might be useful as prognostic markers or molecular targets for novel therapies. The highly metastatic GILM2 cells show constitutive Akt activity that can be inhibited with the phosphatidylinositol 3-kinase (PI3K) inhibitor LY294002. We tested whether genes differentially expressed between GI101A and GILM2, which were identified in the cDNA microarrays, were regulated through PI3K-Akt signaling. GILM2 cells were treated with 10μM LY294002, and RNA and culture supernatants collected at 6h and 24 h, respectively. The expression of SLPI was reduced by 56% (measured by quantitative RT-PCR) and expression of CXCL-1 was reduced by 78% (measured by ELISA) in LY294002-treated cells. These data suggest that inhibition of the PI3K-Akt signaling pathway may inhibit expression of genes associated with breast cancer progression and metastasis. The GILM2 cell line is a suitable experimental model for testing this potential therapeutic strategy.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]