Exposure of cells to anticancer agents can result in altered gene expression. In general, only a single gene or a small number of genes are studied. To improve the efficiency and value of such experiments, we are interested in using DNA array technology to identify, on a genome-wide scale, drug-responsive genes which might serve both as surrogate markers for drug action and as future targets for drug development. As a means to establish this approach in our Centre we have carried out two exploratory projects. In the first project, we examined gene expression in p53 wild-type and mutant ovarian cell lines established from two patients with ovarian cancer who achieved a complete remission following treatment. Neither line can grow in vitro and were passaged as subcutaneous xenografts in nude mice. In the xenograft model both tumours were curable following multiple dosing with cisplatin. Growth-delay experiments following a single dose of 6 mg/kg cisplatin demonstrated that the p53 mutant PXN65 line was more sensitive to cisplatin than the wild-type PXN100 line. Northern blot, RNase protection and western blot demonstrated the induction of p21, MDM-2 and Bax expression in PXN100, but not PXN65. This was consistent with the p53 status of these tumours. These studies were extended to two commercial cDNA arrays, one encompassing genes involved in signal transduction, cell cycle and apoptosis regulation and the other with ESTs of known identity. There was very little difference in the overall constitutive expression pattern between the two cell lines. Following cisplatin treatment the expression of a number of p53-responsive genes was increased in the wild-type PXN100 cell line. There were also changes in the expression of a number of cell-cycle regulators and regulators of the insulin-like growth/survival factor pathway. In contrast, no major changes in gene expression were detected in the p53 mutant PXN65 line. In the second project, we have looked at the effects of a novel agent, the benzoquinone ansamycin 17-allylamino geldanamycin (17-AAG), in two colon adenocarcinoma cancer lines in vitro. This agent inhibits hsp-90 activity, a protein required to maintain the stability of a number of signalling proteins, including c-Raf-1, and is about to enter clinical trial in our centre and elsewhere. Treatment of HT29 and HCT116 cells rapidly reduced the levels of c-Raf-1 detected by western blot and inhibited the high levels of constitutive MAP kinase phosphorylation detected in HCT116 cells. This occurred as early as 24 hours and resulted in death by apoptosis 72 hours after treatment. Data detailing changes in gene expression in response to 17-AAG treatment of both HT29 and HCT 116 cells will be presented.