Analysis of estrogen-responsive finger protein expression in benign and malignant human breast.

The estrogen-responsive finger protein (EFP) gene was originally identified in a screen of genomic DNA for genes containing estrogen-response elements (EREs), and its expression was subsequently shown to be estrogen-regulated and correlated with estrogen receptor (ER)alpha-positive tissues in mice. Human chromosomal mapping localized it to 17q23.1, close to BRCA1, in a region frequently lost in breast cancers. Structurally related proteins have been implicated in a variety of important cellular processes, including carcinogenisis. Given that ER is over-expressed in a large proportion of breast cancers, we reasoned that EFP may play a role in mediating the estrogen-dependent progression of breast cancer. We raised anti-sera to EFP and show that EFP is present in the cytoplasm in mammary cell lines and epithelial cells of normal breast tissue. Furthermore, EFP is present in cell culture medium, suggesting that it may be secreted. Immunohistochemistry of paraffin-embedded breast biopsy specimens showed significantly greater levels of EFP in lactating breast and fibroadenomata compared to normal breast (p<0.001 and p = 0.001, respectively), which is likely to be a result of estrogen responsiveness. Levels were reduced in breast cancer (p = 0.02), where no correlation was seen with other immunohistochemical, histopathological or clinical data. The lack of correlation between EFP and ER status of tumors could indicate escape from estrogenic control, pointing to new models of tumor pathogenesis. Increased levels of EFP in lactating breast and the reduction in malignancy suggest a role for EFP in promoting mammary gland differentiation.

An important role for BRCA1 in breast cancer progression is indicated by its loss in a large proportion of non-familial breast cancers.

The presence of BRCA1 protein was determined immunohistochemically in normal and benign breast biopsies, non-familial breast carcinomas and breast carcinomas from one or more individuals from 8 BRCA1 families. Strikingly, little staining was detected in breast carcinomas from BRCA1 families, regardless of the position or type of mutation, whereas strong immunostaining was observed in 28/28 of non-malignant breast biopsies. Furthermore, BRCA1 staining was reduced in non-familial breast carcinomas, since loss of nuclear BRCA1 staining was evident in 19% of non-familial breast carcinomas whilst a similar proportion (20%) showed absence of either cytoplasmic or nuclear BRCA1 staining. Statistical analysis indicates that breast cancer is characterised by a reduction in levels of nuclear BRCA1 in familial (p < 0.001) and non-familial breast cancer (p = 0.001). In non-familial breast cancer absence of nuclear BRCA1, but not cytoplasmic BRCA1, is more common in high grade breast carcinomas (p = 0.03) and in patients with evidence of lymph node involvement (p = 0.05). Correlation between the absence of BRCA1 protein with high grade is consistent with previous findings of a correlation between mutations in the BRCA1 gene and high grade. Our findings provide new evidence in support of BRCA1 as a tumour suppressor protein in non-familial breast cancer.

Modulation of elongation factor-1 delta (EF-1 delta) expression by oncogenes in human epithelial cells.

A cDNA clone covering part of the C-terminal domain of human EF-1 delta was isolated from mammary cancer cells by subtractive hybridisation. The higher expression of EF-1 delta in the tumours suggested that malignant transformation in vivo requires an increase in translation factor mRNA and protein synthesis for entry into and transition through the cell cycle. To explore the relation between cell division and EF-1 delta expression, MCF-7 cells were treated with dexamethasone, an inducer of differentiation. There was no change in the mRNA levels of EF-1 delta in the dexamethasone-treated cells. To explore the relation between oncogenes and EF-1 delta expression, a variety of oncogenes were introduced into human mammary epithelial cells (MCF-7) and human keratinocytes (HaCaT). Despite high oncogene mRNA expression, there was no significant change in the EF-1 delta mRNA level by v-src, c-erbB (EGF Receptor), c-erbB-2, v-myc and v-fos oncogenes. However, overexpression of v-Ha-ras in HaCaT cells resulted in a three to five-fold decrease in the steady-state mRNA level of EF-1 delta. Taken together, the data provides further support on the interaction of translation factors and oncogenic transformation.

Histochemistry to detect Helix pomatia lectin binding in breast cancer: methodology makes a difference.

A number of studies have shown that altered cellular glycosylation, as detected by binding of Helix pomatia lectin to paraffin sections, is associated with metastatic disease and consequent poor patient prognosis in breast and other cancers. In a 24-year retrospective study, sections of 373 primary breast cancers were stained for binding of the lectin using two different histochemical techniques: a direct method (using peroxidase-conjugated lectin) and an indirect method (using native, unconjugated lectin). Similar percentages of cases were positive (79%) and negative (21%) for lectin binding with either technique, but there was enormous inconsistency when individual cases were examined. A total of 38/373 (10.2%) cases that were negative by the indirect method were positive by the direct method, and 37/373 (9.9%) cases that were negative by the direct method were positive by the indirect method. Life tables calculated for lectin staining vs nonstaining cases showed a very strong correlation between lectin binding and long-term survival (p < 0.0001) when staining was performed by the indirect method, but only very weak correlation with prognosis (p < 0.03, borderline significance) when the direct technique was employed. SDS-PAGE revealed that there were differences in breast cancer glycoproteins recognized by native lectin and peroxidase-conjugated lectin immobilized on Sepharose 4B affinity beads. Helix pomatia lectin binding appears to be an intriguing and potentially valuable marker of biological behavior in breast cancer. This study emphasizes the importance of selecting an appropriate immunohistochemical technique for its visualization.

Subtraction hybridization cloning of RNA amplified from different cell populations microdissected from cryostat tissue sections.

We describe a generally applicable and easily reproducible method for the cloning of differentially expressed RNA, amplified from small numbers of enriched cell populations, obtained by microdissection from single cryostat sections. The procedure involves homopolymeric A tailing of cDNA synthesized from released RNA using an anchored (NN)T12 primer. Subsequent entire cDNA population polymerase chain reaction amplification was carried out using a biotinylated (X)nT16 primer-adaptor in the presence of biotin-dATP. This biotinylated driver cDNA was then twice hybridized in 50-fold excess to heterologous target cDNA made with nonbiotinylated (Y)nT16 primer; common hybrids and excess driver cDNA were magnetically removed following the addition of streptavidin-coated magnetospheres which bound biotinylated strands, leaving enriched target population sequences. These were then directly amplified through the tails using a primer containing only the target-specific (Y)n sequence. Insertion into a lambda-phage vector was facilitated by means of an EcoR1 site incorporated in the (Y)n primer. Subsequent packaging and transformation into Escherichia coli NM522 resulted in cDNA libraries containing approximately 5 x 10(3)-10(4) pfu. Screening of these primary libraries with cDNA derived from the starting populations yielded a large number of differentially hybridizing clones which are currently under analysis.