Altered levels of plasma chemokines in breast cancer and their association with clinical and pathological characteristics.

Chemokines are a family of small, structurally related cytokines with chemoattractant and activation properties. In breast cancer, both epithelial cancer cells and cells within the microenvironment secrete chemokines with either tumor-promoting or anti-malignant potential. The equilibrium between these two chemokine activities plays a key role in the biology of the developing tumor, including its ability to metastasize. Here we evaluated the expression of chemokines in breast tumors and the plasma of breast cancer patients before treatment in order to identify a blood-based signature that could distinguish between malignant and non-malignant processes. We screened the mRNA expression of chemokine genes using cDNA microarray on homogenous, laser-capture microdissected breast cancer specimens. Further, using a protein array approach, we determined the levels of selected chemokines in the plasma of patients with breast cancer, benign breast tumors and healthy women. Finally, we analyzed the association between the levels of chemokines in breast and blood samples with the pathological characteristics of the disease. At mRNA level, 27 chemokines and 11 chemokine receptors were differentially expressed in cancers when compared with normal breast tissue. When compared to benign tumors, the only chemokine significantly upregulated in cancers was CXCL10. At protein level, with the exception of CXCL13, nine out of the ten selected chemokines (CCL2, CCL7, CCL18, CCL22, CXCL8, CXCL9, CXCL10, CXCL11 and osteoprotegerin) were significantly overexpressed in the plasma of breast cancers patients compared to healthy controls. After grouping, CXCL8, CXCL9 and CCL22 proved to be significant predictors for breast cancers as compared to healthy controls in a model of logistic regression. We found upregulation of CXCL8, CXCL11 and CXCL9 in triple negative carcinomas, CXCL9 in low proliferative carcinomas, and CXCL10, CCL7 and osteoprotegerin in poorly differentiated carcinomas. Furthermore, CXCL9 was overexpressed in lymph node negative tumors, whereas CXCL8 and CCL18 were higher in advanced stage carcinomas. We identified a panel of chemokines dysregulated in breast cancer that could be further investigated as prospective novel diagnostic markers or for therapeutic and prognostic applications.

Interaction between estrogens and androgen receptor genes microsatellites, prostate-specific antigen and androgen receptor expressions in breast cancer.

The role of estrogen and androgen receptors signaling in breast cancer is widely accepted, but the interrelations between them are not well understood. It was suggested that PSA could be a marker of endogenous balance between androgens and estrogens. In this context, we intended to investigate the potential of relationship between polymorphic tandem repeats (CAG, TA and CA) in AR (androgen receptor), ERalpha (estrogen receptor alpha) and ERbeta (estrogen receptor beta) genes and the immunoexpression of PSA and AR proteins. We assessed also the possible influences of CAG, TA, and CA variables and other available prognostic factors (ER, PR, AR, HER2/neu, PSA expression, and nodal status) on disease-free survival. We assessed the polymorphic tandem repeats lengths by genotyping, followed by high-resolution denaturing polyacrylamide gel electrophoresis in 163 breast cancers. Immunohistochemistry was performed to assess the expressions of AR, PSA, ER, PR and HER2/neu proteins. Our results showed that PSA was correlated with the length of CA repeats in the 3'-untranslated region of ERbeta, shorter CA repeats being correlated with PSA expression (p=0.03). AR immunoexpression was correlated with CAG repeats on AR gene, higher number of repeats being linked to a higher AR immunoexpression (p=0.04). Performing logistic regression to investigate relationships with prognosis, we observed that PSA immunoexpression (p=0.004), the nodal status (p-<0.001) and marginally, longer TA repeats (p=0.05) were correlated with increased disease-free survival. AR expression presented a low statistical value (p=0.054) in predicting evolution and was not entered into the multivariate regression analysis. Altogether, our findings supports the hypothesis that estrogens, through both alpha and beta-receptors variants are mediating the AR signaling pathway.