CAV1 has been reported to interact with various intracellular signalling pathways and is thought to suppress tumour growth in breast cancer cell lines (Lee et al, 1998
; Razani et al, 2001
; Fiucci et al, 2002
). Prior studies have suggested that oncogenic transformation results in reduced cellular levels of caveolin (Glenney and Soppet, 1992
; Sager et al, 1994
; Koleske et al, 1995
), and that this reduction probably contributes to a loss of caveolae (Koleske et al, 1995
). Lee et al (1998)
found that the CAV1 levels were inversely correlated to breast cancer progression in vitro
and the overexpression of CAV1 resulted in substantial growth inhibition of breast tumour cells, which normally had no endogenous caveolin expression.
Our study confirmed that mRNA level of CAV1 and CAV2 were significantly downregulated in human breast cancer tissues compared to corresponding normal tissues (P<0.001), and that CAV1 and CAV2 mRNA levels were significantly correlated with each other in breast cancer cell lines and tissues. Hurlstone et al reported that CAV1 was expressed in the normal breast tissue by myoepithelial cells but not by ductal epithelial cells. This is in agreement with our present study that observed much more CAV1 and CAV2 expression in myoepithelial cells than in ductal epithelial cells in normal breast tissue by immunohistochemistry.
CAV1-null mice show a striking increase in the frequency and size of multifocal dysplastic lesions in mammary grands, with the nuclei and the nuclei of mammary grand cells showing anaplastic characteristics with increased mitotic figures (Williams et al, 2003
). This suggests that CAV1-mediated interactions between myoepithelial cells and the rest of mammary gland may play an important role in oncogenesis. In addition, our study found that breast cancer patients with tumours that expressed low levels of CAV1 mRNA tended to have larger tumour sizes, which supported the hypothesis that CAV1 acts as a growth suppressor in breast tumours (Lee et al, 1998
There was a significant correlation between high CAV1 and CAV2 mRNA level and positive hormonal receptor status. Razandi et al
reported that E2 stimulates the synthesis CAV1 and CAV2 proteins and activated a CAV1 promoter/luciferase reporter construct transfected into in smooth muscle cells. CAV1 also stimulated ER translocation to the cell membrane in MCF-7 cells, inhibiting E2-induced ERK (MAPK) activation, required for DNA synthesis and cell survival (Razandi et al, 2002
). Thus, our results support the hypothesis that CAV1 and CAV2 play an important role in the process of hormonal translocation.
Our present results also suggested that HER2/neu level and hormonal receptor expression was inversely correlated. This finding is supported by the Southwest Oncology Group Study and Konecney Report (Elledge et al, 1998
; Ferrero-Pous et al, 2000
; Konecny et al, 2003
). While previous studies have reported HER2/neu to be a poor prognostic factor (Slamon et al, 1989
; Quenel et al, 1995
), neither CAV1, CAV2, nor HER2/neu appeared to be a predictor of disease outcome in our study. This may be due to the average follow-up period in our study being only 1141 days, such that few disease recurrences were observed. Thus, future studies may require longer observation periods and higher patient numbers for analysis.
Several mechanisms could be involved in the suppression of CAV gene expression observed in our study, including loss of heterozygosity (LOH) (Tatarelli et al, 2000
; Zenklusen et al, 2001
; Lee et al, 2002
), point mutations (Hayashi et al, 2001
) and methylation (Engelman et al, 1999
). As for LOH, several LOH markers are located near the CAV1 gene locus on human chromosome 7q31. Loss of heterozygosity is frequently encountered in this region in a variety of human neoplasias, indicating the presence of a tumour-suppressor gene. With regard to mutation, Hayashi et al (2001)
reported CAV1 gene mutations in approximately 16% of human breast cancers, and that these mutations may play a role in malignant progression. Although CAV1 expression is increased in prostate cancer (Yang et al, 1999
), oesophageal cancer (Kato et al, 2002
) and ovarian cancer (Davidson et al, 2001
), future studies should include assessment of CAV1 mutation. As for methylation, Engelman et al (1999)
reported that the first and second exons of the CAV1 and CAV2 genes are embedded within CpG islands, such that it is possible that caveolin gene expression is controlled, at least in part, by methylation of these CpG regions.
Zihui et al
reported that growth factor receptor stimulation activated the phosphatidylinositol 3′-kinase and β
-catenin pathways in mammary epithelial cells. β
-Catenin activates cyclin D1 and c-Myc transcription, and subsequently, c-Myc suppressing caveolin-1 transcription (Xie et al, 2003
). Furthermore, some studies have shown that activation of HER2/neu downregulates CAV1 expression in vitro
(Couet et al, 1997
; Kim et al, 2000
), while other studies have reported the negative regulation of expression and signal transduction between CAV1 and HER2/neu (Engelman et al, 1998
; Zhang et al, 2000
). However, our present study revealed no such correlations in vivo
. This may be due to a discrepancy between experimental and clinical studies.
In summery, our results indicated that CAV1 suppression correlated closely with that of CAV2 in breast cancer, that CAV1 level was inversely correlated with tumour size, and that CAV1 and CAV2 levels were correlated with hormonal receptor status. Therefore, CAV1 and CAV2 play an important role in tumour progression in breast cancer patients.