Exposure to bisphenol A correlates with early-onset prostate cancer and promotes centrosome amplification and anchorage-independent growth in vitro

Abstract

Human exposure to bisphenol A (BPA) is ubiquitous. Animal studies found that BPA contributes to development of prostate cancer, but human data are scarce. Our study examined the association between urinary BPA levels and Prostate cancer and assessed the effects of BPA on induction of centrosome abnormalities as an underlying mechanism promoting prostate carcinogenesis. The study, involving 60 urology patients, found higher levels of urinary BPA (creatinine-adjusted) in Prostate cancer patients (5.74 µg/g [95% CI; 2.63, 12.51]) than in non-Prostate cancer patients (1.43 µg/g [95% CI; 0.70, 2.88]) (p = 0.012). The difference was even more significant in patients <65 years old. A trend toward a negative association between urinary BPA and serum PSA was observed in Prostate cancer patients but not in non-Prostate cancer patients. In vitro studies examined centrosomal abnormalities, microtubule nucleation, and anchorage-independent growth in four Prostate cancer cell lines (LNCaP, C4-2, 22Rv1, PC-3) and two immortalized normal prostate epithelial cell lines (NPrEC and RWPE-1). Exposure to low doses (0.01-100 nM) of BPA increased the percentage of cells with centrosome amplification two- to eight-fold. Dose responses either peaked or reached the plateaus with 0.1 nM BPA exposure. This low dose also promoted microtubule nucleation and regrowth at centrosomes in RWPE-1 and enhanced anchorage-independent growth in C4-2. These findings suggest that urinary BPA level is an independent prognostic marker in Prostate cancer and that BPA exposure may lower serum PSA levels in Prostate cancer patients. Moreover, disruption of the centrosome duplication cycle by low-dose BPA may contribute to neoplastic transformation of the prostate.

Figure 1. Scatter plots of LnBPA.

Urine BPA levels are associated with PCa. The log-transformed BPA is referred to as LnBPA. Values in graph are mean ± SD of LnBPA. (A) Urine BPA levels are higher in PCa patients than in non-PCa patients. Means of LnBPA  = 1.75±1.97 in PCa (blue, n = 27) vs. 0.35±2.14 in non-PCa (red, n = 33), p = 0.012. (B) LnBPA in PCa vs. LnBPA in non-PCa, stratified by age = 65. Urine BPA levels are significantly higher in young PCa patients than in the respective non-PCa patients only in the age group <65 years old; p = 0.006. (C) Linear regression analyses of Serum PSA vs. LnBPA in patients <65 years old only (n = 30). Blue solid squares represent PCa patients; red inverse-circles represent non-PCa patients. Blue and red solid lines represent their regression lines, respectively. (D) Comparison of the geometric mean of BPA in PCa and non-PCa groups. The geometric mean (Geo) is defined as the exponential of the mean of LnBPA. Values are geometric means (95% CI) of BPA in unit of µg/g creatinine.

Figure 2. Low doses of BPA have an adverse effect on centrosome numbers in prostate cancer cells.

The cell lines NPrEC, RWPE1, LNCaP, C4-2, 22Rv1, and PC3 were treated with medium containing 10% CSS plus 0, 0.01 nM, 0.1 nM, 1 nM, 10 nM and 100 nM BPA for 72 h. Cells were fixed with 100% cold methanol and immunostained for centrosomes and nuclei. The number of centrosomes per cell was scored by fluorescence microscopy. The results are shown as an average determined from five separate experiments. The scatter plot was generated of the percentage of cells with an abnormal number of centrosomes in response to BPA. Analyses was performed using a fixed effect model for each cell line. Post hoc comparisons of means were adjusted using Bonferroni's tests. The fold change is the percentage of cells with abnormal centrosomes at 0.1 nM BPA/the percentage of cells with abnormal centrosomes at 0 nM BPA.

Figure 3. An increase in centrosome numbers is seen in prostate cancer cells exposed to BPA.

(I) An increase in centrosome numbers. The cell lines NPrEC, RWPE1, LNCaP, C4-2, 22Rv1 and PC3 were treated with medium containing 10% CSS plus 0 or 0.1 nM BPA for 72 h. Cells were fixed with 100% cold methanol and immunostained for centrosomes (anti-γ-tubulin, red) and nucleus (DAPI, blue). The cells were examined by fluorescence microscopy. Arrows point to the positions of centrosomes, and panels on the right show magnified images of the indicated areas. Scale bar, 10 µm. (II) Centrosome amplification in the presence of BPA is not due to centriole separation. RWPE-1 cells were treated with 0.1 nM BPA for 3 days. Cells were fixed and immunostained for centrosomes (anti-γ-tubulin, red), centrioles (anti-centrin, green), and nucleus (DAPI, blue). Arrows point to the positions of centrosomes. Panels on right show magnified images of the indicated areas. Scale bar, 10 µm.

Figure 4. BPA enhances centrosomal aster formation.

The microtubule aster formation assay was performed 2-immunostained for centrosomes (anti-γ-tubulin, red) and MTs (anti-α-tubulin, green). The centrosomal aster formation was assessed as positive if centrosomes had an MT aster with more than 15 MTs. The results shown in (A) are the average ± standard error (SE) from three experiments. For each experiment, >200 cells were examined. Significance was calculated using Student's t-test vs. 0 pM. *p≤0.00002.

Figure 5. Cells grown in the absence and presence of 0.1-independent growth.

Representative pictures of colonies after 2 weeks of incubation in agar. C4-2 cells in the presence of 0.1 nM BPA formed larger colonies (B, B′, 100–1200 µm diameter) compared with those grown in the absence of BPA (A, A′, 50–400 µm diameter).