Low-fat, low-glycemic load diet and gene expression in human prostate epithelium: a feasibility study of using cDNA microarrays to assess the response to dietary intervention in target tissues.

PURPOSE We examined the feasibility of using gene expression changes in human prostate epithelium as a measure of response to a dietary intervention. MATERIALS AND METHODS Eight men with newly diagnosed prostate cancer were randomized to a low-fat/low-glycemic load intervention arm (<20% energy from fat and total daily glycemic load <100) or a "standard American" control arm (approximately 35% energy from fat and total daily glycemic load >200). Prostate tissue was collected before randomization and approximately 6 weeks later, at the time of radical prostatectomy. Epithelium was acquired by laser capture microdissection, and transcript abundance levels were measured by cDNA microarray hybridization and confirmed by quantitative reverse transcription-PCR. RESULTS Men in the intervention arm consumed 39% less total energy (P = 0.004) and the difference in weight change between intervention and control arms was -6.1 kg (P = 0.02). In the intervention arm, 23 (0.46%) of 5,711 cDNAs with measurable expression were significantly altered (P < 0.05; false discovery rate, <or=10%). In the control arm, there were no significant changes in transcript expression, even when using a false discovery rate as high as 50%. CONCLUSIONS A 6-week, low-fat/low-glycemic load diet was associated with significant gene expression changes in human prostate epithelium. These results show the feasibility of using prostate tissues collected at diagnosis and at surgery to study the effects of dietary manipulation on prostate tissue, which may give insight into the molecular mechanisms underlying the associations of diet and obesity with the development or progression of prostate cancer.