Constitutive activation of the RON gene, known to code for the tyrosine-kinase receptor for Macrophage Stimulating Protein (also known as Scatter Factor 2), has been shown to induce invasive-metastatic phenotype in vitro. As yet, nothing is known about the expression of this novel member of the MET-oncogene family in spontaneously occurring human cancers. Here we report that Ron is expressed at abnormally high levels in about 50% primary breast carcinomas (35/74 patients). Among these, the expression is increased more than 20-fold in 12 cases and the overexpressed protein is constitutively phosphorylated on tyrosine residues. Notably, Ron is only barely detectable in epithelial cells of the mammary gland, and its expression remains unchanged in benign breast lesions (including adenomas and papillomas). Overexpression was observed in di erent histotypic variants of carcinomas; it is associated with the disease at any stage and correlates with the post-menopausal status. In breast carcinoma cells grown in vitro, activation of the Ron receptor resulted in proliferation, migration and invasion through reconstituted basement membranes. Altogether, these data suggest a role for the RON gene in progression of human breast carcinomas to the invasive-metastatic phenotype.

Summary Monoclonal antibody SP-2 to the tumour-associated antigen 90K was generated by immunisation with conditioned medium of human breast cancer cells. We investigated whether circulating levels of 90K can influence the prognosis of patients with breast cancer. Serum samples were obtained from 425 patients with histologically proven breast cancer with no clinical evidence of disease after surgery (NED) and in 310 patients with metastatic disease. Serum 90K was determined by a new immunoradiometric assay (IRMA). Antigen levels in NED patients were elevated in 18.5% of cases, mean levels being higher than in healthy controls (P = 0.001). Among 375 evaluable patients, the 75-month overall survival for 90K-negative ( < 11 U ml-') and 90K-positive (>11 U ml-) patients was 78% and 53% respectively (P = 0.004 (Suter et al., 1989). Biotinylation of SP-2 was carried out according to Guesdon et al. (1979). After coating, the beads were washed extensively with 0.9% sodium chloride solution and incubated with biotinylated SP-2 (5fig ml-') at room temperature for 18 h. Coated beads were treated with an overcoating solution of bovine serum albumin (BSA) (2 mg ml-') for 1 h at room temperature, washed with distilled water and stored at room temperature until used. Beads treated in this fashion were stable for at least 6 months.With each assay, 200-jil aliquots of appropriately diluted samples or standards were incubated with SP-2-coated beads for 1 h at 37°C. The beads were washed with distilled water followed by the addition of 100ll of '25I-labelled SP-2 (approximately 50,000c.p.m.; specific activity 10 CifLg-1) in PBS, pH7.4, containing 5% BSA, 0.1mgml-' normal mouse IgG and 0.1 % sodium azide for an additional hour at 37'C. Beads were washed with distilled water and counted in a gamma-counter. The amount of 90K was calculated by reference to the amount present in standard preparations made from a pool of sera from breast cancer patients and titred to contain 40, 20, 10 and 5 arbitrary units per ml. The simultaneous assay of 120 sera from breast cancer patients using the IRMA and the previously developed ELISA (lacobelli et al., 1988) gave a correlation coefficient of 0.91

The Ku70/80 heterodimer is the regulatory subunit of the DNA-dependent protein kinase (DNA-PK) and its DNAbinding activity mediates DNA double-strand breaks repair. Although Ku80 was recently proposed as a caretaker gene involved in the control of genome integrity, no data are available on Ku70/80 DNAbinding activity in human tumors. Heterodimer DNAbinding activity and protein expression were assayed by electrophoretic-mobility-shift-assay (EMSA) and Western blot analysis, in nuclear and cytoplasmic extracts from eight breast, seven bladder primary tumors and three metastatic nodes from breast cancers. Corresponding normal tissues of the same patients were used as controls. Ten out of 15 tumors showed nuclear Kubinding activity 3 ± 10 times higher than in the normal tissues, irrespective of bladder or breast origin. Conversely, in 5/15 primary tumors and in all the metastatic nodes analysed, nuclear Ku-activity was 1.5 ± 4.5-fold lower than in the corresponding normal tissues. Cytoplasmic heterodimer activity signi®cantly diered between tumor and normal tissues, displaying a 2 ± 10-fold increase in neoplastic tissues. Three dierent patterns combining both Ku expression and activity with tumor characteristics were identi®ed. In low aggressive breast tumors p70/p80 proteins were expressed in tumor but not in normal tissues. The heterodimer bindingactivity matched the protein levels. In non-invasive bladder carcinomas no signi®cant dierences in protein expression between tumor and the corresponding normal tissues were found, however heterodimer binding-activity was increased in tumor samples. In breast and bladder tumors, at the advanced stage and in node metastases, the binding activity was strongly reduced in tumor biopsies, however no dierences were demonstrated between normal and tumor protein levels. Our results suggest a dierent modulation of Ku70/80 DNA-binding activity in human neoplastic tissues, possibly related to tumor progression. Findings provide further data on tissue-speci®c protein expression and post-translational regulation of heterodimer activity. Oncogene (2001) 20, 739 ± 747.

The androgen receptor (AR) is a transcription factor mediating the action of androgens. The AR gene is localized on chromosome X and it contains a series of CAG trinucleotide repeats. The length of the CAG repeats varies among individuals and this polymorphism is believed to be related to AR transcriptional activity. Studies have shown that fewer CAG repeats are associated with an increased risk as well as more aggressive forms of prostate cancer. Although AR is expressed in breast cancer and the impact of androgen and AR on breast cancer has been recognized, the role of the CAG repeats in breast cancer remains unknown. In this study, we measured the CAG repeats in breast cancer tissue using a PCR-based method. Of the 133 patients with primary breast cancer, 102 were heterozygous and 31 were homozygous. The mean CAG repeat number for homozygous women was 21; for heterozygous women the repeat number mean was 20 for the short allele and 24 for the long allele. The length of CAG repeats either in one allele or in both alleles was inversely correlated with the histological grade of breast cancer (r = -0.23 or -0.26, respectively, p < 0.05). An association between positive lymph nodes and fewer CAG repeats in both alleles was also suggested (p = 0.06). Furthermore, survival analysis indicated that the total number of CAG repeats in both alleles was associated with patient overall survival. With every CAG repeat increase, there was a 6% reduction in the risk of death (RR = 0.94, p = 0.03). The association remained significant after controlling for the homozygous and heterozygous status (RR = 0.92, p = 0.01). The association became no longer significant when clinical and pathological variables were adjusted in the analysis but this could be due to the reduction of sample size in the multivariate analysis. CAG heterozygosity and difference in number of CAG repeats between the two alleles were not associated with either disease features or patient survival. Our results suggest that longer CAG repeats may occur more frequently in less aggressive cancer and that the CAG repeats may play a role in breast cancer progression.

The combination of tamoxifen with ovarian suppression seems to be safe and to yield comparable results relative to standard CMF.