Expression of SGK-1, a PI3-kinase-dependent kinase implicated in cell survival signaling, in primary human breast cancers Free

3743

Activation of SGK-1 signaling has been shown to inhibit breast epithelial cell apoptosis in response to a variety of environmental stressors including growth factor deprivation and UV irradiation. Using biochemical fractionation, we have shown that SGK-1 localizes predominantly to the endoplasmic reticulum (ER)-containing microsomal fraction and to the cytoplasm. More recently, immunofluorescent analysis of epithelial cells has confirmed the predominant ER and cytoplasmic localization of SGK-1; a smaller proportion of SGK-1 was also seen in the mitochondria and nuclei. Interestingly, SGK-1 can form a complex with the stress-associated C-terminus of Hsc70 interacting protein (CHIP), a known E3 ligase involved in ER-associated protein degradation. In our studies, CHIP was also found to be required for both SGK-1 ubiquitin modification and its rapid proteasomal degradation. Furthermore, CHIP and SGK-1 co-localize in association with the ER.

Because CHIP-mediated regulation of SGK-1 has only been examined in cell lines, we wished to determine whether or not an association exists between CHIP and SGK-1 expression in primary human breast tissues. To this end, we examined SGK-1 and CHIP expression by immunohistochemistry in breast tissue microarrays representing unaffected breast epithelium (n=75), DCIS (n=56), invasive breast cancer (n=213), and metastatic breast cancer (n=31) from 255 patients. Cytoplasmic SGK-1 expression was significantly increased in 44% of tumors versus only 18% of normal breast tissues (P<.05). In addition, we found that among samples with high CHIP expression, SGK-1 had significantly lower expression levels in normal tissues compared with tumor samples (P=0.031). These findings suggest that CHIP and its associated degradation of SGK-1 may be deregulated in breast cancer.

In addition, we have discovered two low molecular weight SGK-1 protein products that appear to be less susceptible to proteasome-mediated degradation when compared to full-length SGK-1. Because these variants are detected with an anti-C-terminus- but not an anti-N-terminus-SGK-1 antibody, they likely lack the N-terminal region of SGK-1 known to be required for rapid proteasomal degradation. Future experiments will determine the relative expression and subcellular localization of these variants in tumor cells versus normal breast epithelium. Such variants could contribute to the differences in tumor versus normal cell SGK-1 expression that we have detected in this study.