Dietary tomato and lycopene impact androgen signaling- and carcinogenesis-related gene expression during early TRAMP prostate carcinogenesis

Abstract

Consumption of tomato products containing the carotenoid lycopene is associated with a reduced risk of prostate cancer. To identify gene expression patterns associated with early testosterone-driven prostate carcinogenesis, which are impacted by dietary tomato and lycopene, wild-type (WT) and transgenic adenocarcinoma of the mouse prostate (TRAMP) mice were fed control or tomato- or lycopene-containing diets from 4 to 10 weeks of age. Eight-week-old mice underwent sham surgery, castration, or castration followed by testosterone repletion (2.5 mg/kg/d initiated 1 week after castration). Ten-week-old intact TRAMP mice exhibit early multifocal prostatic intraepithelial neoplasia. Of the 200 prostate cancer-related genes measured by quantitative NanoString, 189 are detectable, 164 significantly differ by genotype, 179 by testosterone status, and 30 by diet type (P < 0.05). In TRAMP, expression of Birc5, Mki67, Aurkb, Ccnb2, Foxm1, and Ccne2 is greater compared with WT and is decreased by castration. In parallel, castration reduces Ki67-positive staining (P < 0.0001) compared with intact and testosterone-repleted TRAMP mice. Expression of genes involved in androgen metabolism/signaling pathways is reduced by lycopene feeding (Srd5a1) and by tomato feeding (Srd5a2, Pxn, and Srebf1). In addition, tomato feeding significantly reduced expression of genes associated with stem cell features, Aldh1a and Ly6a, whereas lycopene feeding significantly reduced expression of neuroendocrine differentiation-related genes, Ngfr and Syp. Collectively, these studies demonstrate a profile of testosterone-regulated genes associated with early prostate carcinogenesis that are potential mechanistic targets of dietary tomato components. Future studies on androgen signaling/metabolism, stem cell features, and neuroendocrine differentiation pathways may elucidate the mechanisms by which dietary tomato and lycopene impact prostate cancer risk.

Figure 1

Representative histological images (magnification = 200x) of H&E stained 10-wk-old mouse anterior prostate lobes from WT or TRAMP, after either sham surgery (Intact), castration surgery (Castration), or castration surgery followed by testosterone repletion (Repletion). Arrow indicates an early PIN lesion (A-F).

Figure 2

Hierarchical clustering of 189 detectable prostatic genes in response to genotype and testosterone status. I, intact; C; castration; R, castration followed by testosterone repletion (A). Venn diagram depicting the numbers of prostatic genes for which expression was significantly (P≤0.05) impacted by either the main effects of genotype, testosterone, or diet (non-overlapping portions) or a combination of main effects (overlapping portions) (B).

Figure 3

Gene expression was strongly up-regulated (A-D) or down-regulated (E&F) in TRAMP compared to WT mice with a significant main effect of genotype (P<0.05), in intact, castration, or castration followed by testosterone repletion (repletion) groups, and was significantly decreased in castration compared to intact and testosterone repletion group (P<0.0001; not denoted on figures). Fold-change of expression is relative to the intact, WT group. Data are expressed as mean ± SE (n=3-5/group). *** Denotes a significant between-genotype, within surgical treatment difference (P<0.0001). Castration significantly decreased Ki67 protein expression of AP in WT (G) and TRAMP (H). *** Denotes a significant differences in castration compared to intact and repletion in both WT and TRAMP, regardless of diet type.

Figure 4

Prostatic gene expression significantly impacted by androgen status (A-F) and a significant genotype × testosterone status interaction (G&H). *** Indicates significant differences between androgen status groups, within genotype (P<0.0001 compared to intact and repletion groups). Fold change of expression is relative to intact, WT. Data are expressed as mean ± SE (n=3-5/group).

Figure 5

The top 10 canonical pathways significantly impacted by genotype (A) and castration (B), as calculated from gene expression data by IPA.

Figure 6

Prostatic gene expression significantly impacted by an overall main effect of diet (P<0.05). Bars represent gene expression in tomato- or lycopene-fed mice, relative to control-fed mice, in which equal expression is represented as “1” (n=21-22/group). ^ Indicates significant differences between tomato- vs. control-fed and * indicates significant differences between lycopene- vs. control-fed mice (P<0.05). (A). Gene expression of testosterone metabolism and signaling-related genes down-regulated with tomato or lycopene-feeding compared to control-feeding. Fold change is calculated as a ratio of gene expression to control group and expressed as mean ± SE (n=21-22/group) (B-E). * Denotes a significant between-diet group difference, relative to control-feeding, (P<0.05).