Influence of atrazine administration and reduction of calorie intake on prostate carcinogenesis in probasin/SV40 T antigen transgenic rats

Abstract

Atrazine, which has been used worldwide as a pesticide, is now known to exert endocrine disrupting (antiandrogenic) effects in mammals. In this study, modifying effects of dietary feeding of 500 and 1000 p.p.m. atrazine on the development of androgen‐dependent prostate cancer were investigated using male probasin/SV40 T antigen transgenic (TG) rats. As administration of atrazine has now been identified as causing a decrease in bodyweight, a dietary‐restricted TG rat group was also included in order to elucidate the influence of reduction of calorie intake per se on the development of prostate cancer. At week 13, almost the entire lobes of the prostate were occupied with tumor lesions, with no clear intergroup differences in the incidences and multiplicities. Therefore, morphometrical assessment ratios of the prostate epithelial area to the whole prostate tissue area were evaluated. The ratio in the lateral lobe of the 1 000 p.p.m. atrazine‐treated group was significantly decreased, and there was a tendency to decrease in the ratios in the dorsal lobe of the atrazine‐treated groups. However, dietary restriction itself without atrazine treatment caused the same reduction to a similar or greater extent. Testosterone levels were not affected by atrazine administration or dietary restriction. Our results indicate that the observed atrazine‐related suppression of prostate carcinogenesis was probably caused by the decrease in calorie intake, rather than by atrazine‐related endocrine disruption. (Cancer Sci 2005; 96: 221–226)

A number of chemicals in the environment have the potential to disturb the endocrine system. These are classified as endocrine disruptors, and many of them are known to have an influence on reproductive potential. ¹ , ² , ³ Atrazine, once one of the most widely used agricultural pesticides all over the world, is now recognized to have disrupting effects on the reproductive systems of mammals. ⁴ , ⁵ , ⁶ , ⁷ , ⁸ Although agricultural use of atrazine is prohibited in many countries at present, a certain amount of the compound remains in soil and water. ⁹ , ¹⁰ Therefore, humans and wildlife are at risk of exposure and it remains important to assess the potential affects on the reproductive systems in laboratory animals. Both the female ⁴ , ⁵ and male ⁶ , ⁷ , ⁸ reproductive systems are affected in rats. Friedmann ⁽⁷⁾ reported that atrazine acts as an endocrine disrupter in male rats by inhibiting testosterone production in the Leydig cells directly. Therefore, we examined the effects of atrazine exposure on development of androgen dependent prostate carcinomas, using male transgenic (TG) rats having a probasin promoter/simian virus 40 (SV40) T antigen construct. ¹¹ , ¹² These rats, established in our laboratory, developed well‐differentiated adenocarcinomas in the prostate at 100% incidence before 15 weeks of age. ¹¹ , ¹² Castration at 20 weeks of age caused complete regression or involution of carcinomas in these rats; and castration at 5 weeks of age completely inhibited the prostate carcinoma development. ⁽¹²⁾ Since the tumors are considered to be completely androgen dependent, we hypothesized that exposure of atrazine to these rats would cause inhibition of testosterone production, resulting in reduced carcinoma development. To test this, we examined the prostate of animals histopathologically after dietary administration of 500 or 1000 p.p.m. atrazine for 13 weeks. We also examined serum testosterone levels at the time of necropsy. As it has been reported that administration of high doses of atrazine causes a decrease of bodyweight in rats, ⁽⁸⁾ a group of rats given a restricted amount of basal diet was included with adjustment of the mean bodyweight of this group to be similar to that of the 1000 p.p.m. atrazine treatment group. We also measured levels of atrazine and its metabolites in the serum at the end of the treatments in order to confirm absorption of the dietary administered compound.

Materials and methods

Animals, chemical, and diet.  Male heterozygous TG rats established in our laboratory with a Sprague‐Dawley genetic background were used in the present study. ¹¹ , ¹² At 6 weeks of age, animals were housed two or three per cage on wood‐chip bedding in an air‐conditioned animal room at 23 ± 2°C and 50 ± 10% humidity and provided with powdered diet with or without atrazine and tap water ad libitum, except in the dietary‐restricted group. Atrazine was from Wako Pure Chemical Industries Ltd. (Osaka, Japan).

Rationale for dose selection.  Friedmann ⁽⁷⁾ reported that 50 mg/kg per day atrazine given by oral gavage acted as an endocrine disrupter in male rats by inhibiting testosterone production. Therefore, atrazine concentrations in the basal diet were set at 500 and 1000 p.p.m. because mean atrazine intake in the 1000 p.p.m. group was estimated to be around 50 mg/kg/day, considering a predicted daily food consumption of TG rats.

Experimental procedure.  A total of 40 heterozygous male TG rats were divided into four groups, each containing 10 animals. Animals in the non‐treatment control group were fed basal diet (soybean‐free modified NIH‐07; Oriental yeast Co., Tokyo, Japan) and animals in atrazine treatment groups received the same diet containing 500 or 1000 p.p.m. atrazine. Animals in the dietary‐restricted group were given an amount of basal diet restricted (15–20 g). The amount was determined on each day when bodyweights were measured, so as to adjust mean bodyweight of the dietary‐restricted group to that of the 1000 p.p.m. atrazine treatment group. Mean food consumption in the dietary‐restricted group was about 70% of that in the control group. Bodyweights for all the animals and food consumptions for all the cages were measured at least once in each treatment week. At the end of the 13‐week treatment period, all rats were killed by exsanguination via the thoracic and abdominal aorta under light ether anesthesia after blood collection from the abdominal aorta. The testes and the accessory sex organs were removed and fixed in 10% neutral buffered formalin except for the testes of two animals in each group, which were fixed in formalin‐sucrose‐acetic acid (FSA) composed of 37% formalin 5.0, 5% sucrose solution 15.0, and acetic acid 0.8 in a volume ratio, praised for its suitability in evaluation of testis morphology. ⁽¹³⁾ The testis weights were measured before fixation. The ventral prostate, the dorsolateral prostate including the urethra, and the anterior prostate/seminal vesicle complex were removed after fixation, and each was weighed. Relative organ weights were calculated as percentages of body weights at the time of necropsy.

Testosterone and atrazine levels in serum, and atrazine concentrations in diets.  The collected blood samples at the time of necropsy were centrifuged at 1000 r.p.m. for 10 min to obtain serum. Testosterone levels in the serum were analyzed for every animal in each group using radioimmunoassays by a commercial laboratory (SRL, Tokyo, Japan). Serum levels of atrazine and its metabolites were measured for three animals in each group using gas chromatography‐mass spectrometry according to the method of Brzezicki et al. ⁽¹⁴⁾ The concentrations of atrazine in the remaining diets at the time of necropsy were measured for each group using liquid chromatography‐mass spectrometry. The measurement of atrazine in the serum and the diets was performed at Japanese Food Research Laboratories (Tokyo, Japan).

Histopathology.  For histopathological observation, one transversal slice of each side of the testes, the caput, and cauda regions of each side of the epididymides, two sagittal slices from the ventral prostate, two transverse slices from the dorsolateral prostate including the urethra, and four transverse slices from each side of the seminal vesicles including the anterior prostate were embedded in paraffin, sectioned, stained with hematoxylin‐eosin (HE), and examined by light microscopy. Histopathological evaluation of the prostate, including morphometry, was performed for each prostate lobe: the ventral, dorsal, lateral, and anterior lobes.

Immunohistochemistry.  Immunohistochemistry for SV40 T antigens and androgen receptors (AR) were performed using avidin‐biotin‐peroxidase complex (ABC) method. The sections were treated with mouse anti‐SV40 T antigen (Pharmingen, San Diego, CA, USA) and rabbit anti‐AR (Affinity Bioreagents, Golden, CO, USA). Binding was visualized with a Vectastain Elite ABC kit (Vector Laboratory, Burlingame, CA, USA) followed by light hematoxylin counterstaining to facilitate microscopic examination.

Morphometry for prostate epithelium area.  As the entire prostate lobes were occupied with tumor lesions, differences in the incidences and multiplicities of tumors were not observed clearly among the groups (see Results). Therefore, ratios of epithelial areas including tumors to whole prostate tissue areas were evaluated morphometrically. The prostate epithelial areas including prostatic intraepithelial neoplasias (PIN) and adenocarcinomas for each section were quantitatively measured using the H&E sections and an image processor for analytical pathology (IPAP) (Sumika Technos, Osaka, Japan). The whole prostate tissue area including lumen for each section was also measured in the same way to allow calculation of the ratio. All of the H&E slides were thus evaluated for the ventral, dorsal, and lateral lobes of the prostate. For the anterior lobe, the left sides only were measured.

Statistical evaluation.  The data of bodyweights, serum testosterone levels, organ weights, and morphometrical ratios were analyzed statistically as follows. For the statistical evaluation between the control group and the atrazine treatment groups, first, the data were tested by Bartlett's test ⁽¹⁵⁾ for homogeneity of variance. When the variances were homogeneous, Williams’ test ⁽¹⁶⁾ was performed assuming a dose‐related trend. When the variances were heterogeneous, the Shirley‐Williams test ¹⁷ , ¹⁸ was performed assuming a dose‐related trend. For the statistical evaluation between the control group and the dietary‐restricted group, the data were tested by the F‐test ⁽¹⁵⁾ for homogeneity of variance between the groups. When the variances were homogeneous, Student's t‐test ⁽¹⁵⁾ was used, and when the variances were heterogeneous, the Aspin & Welch t‐test ⁽¹⁵⁾ was performed to compare the mean in the control group with that in the dietary‐restricted group. Bartlett's test was conducted at the significance level of 0.05 and the Williams and Shirley‐Williams tests were conducted at the two‐tailed significance level of 0.05. The F‐test was conducted at the significance level of 0.20, and the Student's t‐test and the Aspin & Welch t‐test were conducted at the two‐tailed significance levels of 0.05 and 0.01. The Williams test and the Shirley‐Williams test were performed using the SAS function PROBMC. ⁽¹⁹⁾

Results

General observation.  No deaths were observed in any group during the treatment period. Bodyweights in the atrazine treatment groups were significantly lower than in the control group, almost throughout the treatment period and with dose‐dependence. Bodyweights of the dietary‐restricted and 1000 p.p.m. atrazine groups were similar, and significantly lower than the control values throughout the treatment period (Fig. 1). Food consumption in the atrazine treatment groups was lower than in the control group throughout the treatment period with dose‐dependence (Fig. 2). Mean atrazine intakes were calculated from the targeted concentrations of the substance in the prepared diets (500 p.p.m. and 1000 p.p.m.), the food consumption values, and the bodyweights. The results were 33.0 mg/kg/day for the 500 p.p.m. group and 61.6 mg/kg/day for the 1000 p.p.m. group.

Figure 1.

Body weight curves of all four groups over the treatment period (n = 10).

Figure 2.

Food consumption curves of all four groups over the treatment period (n = 10).

Organ weights.  A significant decrease in relative weight of the ventral prostate was noted only in the dietary‐restricted group (Table 1). No other treatment‐related significant change was observed in any group. Although statistically significant increases in relative weights of testes were observed in the atrazine treatment groups and in the dietary‐restricted group, these were considered to be superficial changes because absolute weights did not significantly differ (data not shown). They were rather due to decrease in bodyweight gain.

Table 1.

Final body weights and relative organ weights in TG rats (n = 10)

	Body weight (g)	Relative organ weight (%)
		Testes	Vetral prostate	Dorsolateral prostate	Anterior prostate + Seminal vesicle
Control	475.2 ± 26.8	0.721 ± 0.031	0.076 ± 0.016	0.185 ± 0.076	0.500 ± 0.106
Atrazine (500 p.p.m.)	423.7 ± 28.6*	0.766 ± 0.226*	0.078 ± 0.014	0.165 ± 0.014	0.554 ± 0.093
Atrazine (1000 p.p.m.)	385.0 ± 22.8*	0.894 ± 0.057*	0.081 ± 0.013	0.168 ± 0.012	0.578 ± 0.081
Food restriction	389.3 ± 18.4 †	0.836 ± 0.102 †	0.057 ± 0.008 †	0.155 ± 0.017	0.432 ± 0.071

^* Significantly different from the control group, P < 0.05 (Williams test),

^† Significantly different from the control group, P < 0.01 (paired comparison).

Testosterone levels in serum.  Serum testosterone levels at the end of the treatment period were not significantly different in the atrazine treatment groups and in the dietary‐restricted group compared to the control group value (Fig. 3).

Figure 3.

Serum testosterone levels in transgenic rats at the end of the treatment period (n = 10). Data are mean±SD. No significant difference from the control group was observed.

Levels of atrazine in serum and concentrations of atrazine in diets.  Data for serum levels of atrazine and its metabolites in each group are summarized in Table 2. Diaminochlorotriazine (DACT) was the major metabolite, with desethylatrazine (DE‐ATRA) and desisopropylatrazine (DI‐ATRA) as minor metabolites. Their concentrations were dose‐dependent. Atrazine and its metabolites were essentially not detectable in the serum in the control and dietary‐restricted groups. Actual concentrations of atrazine in the prepared diet were 300 p.p.m. (60%) for the 500 p.p.m. group and 910 p.p.m. (91%) for the 1000 p.p.m. group.

Table 2.

Serum atrazine and its metabolites in TG rats (n = 3)

	ATRA* (µg/mL)	DE‐ATRA † (µg/mL)	DI‐ATRA ‡ (µg/mL)	DACT § (µg/mL)
Control	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.04 ± 0.04
Atrazine (500 p.p.m.)	0.00 ± 0.00	0.08 ± 0.01	0.23 ± 0.04	3.63 ± 0.67
Atrazine (1000 p.p.m.)	0.02 ± 0.01	0.15 ± 0.03	0.38 ± 0.03	9.27 ± 2.05
Food restriction	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.04 ± 0.05

^* atrazine,

^† desethylatrazine,

^‡ desisopropylatrazine,

^§ diaminochlorotriazine.

Histopathology and immunohistochemistry.  The histopathological lesions of the prostate were diagnosed according to the criteria described previously. ¹¹ , ¹² Marked epithelial proliferation was observed with formation of irregular glands and luminal bridging to give a cribriform pattern in focal populations. The nuclei demonstrated enlargement and severe atypia. These lesions appeared compatible with well‐differentiated adenocarcinomas in human cases and were therefore diagnosed as adenocarcinomas (Fig. 4a,b). They were observed in the ventral, lateral, and anterior lobes of the prostate in almost all rats. In the dorsal lobe, adenocarcinomas were observed in two or three rats of each group. The incidences of adenocarcinomas in any treatment groups were not dramatically different from those of the control group (Table 3). Foci within glands with less proliferation were also observed. These exhibited crowding of stratified epithelial cells with irregular spacing and occasional luminal bridging. Although nuclear atypia were severe, basic structures of the prostate glands were mostly maintained, comparable with PIN of human cases (Fig. 4a,b). Such lesions were observed in all lobes of the prostate in all rats (Table 3) with no significant intergroup differences. SV40 Tag and androgen receptor expression were detected immunohistochemically in almost all cells of the prostate adenocarcinomas and PIN (Fig. 4c,d). No treatment‐related changes in the testes, epididymides, and seminal vesicles were observed. Marked atrophy of the seminiferous tubules in the testes was observed in a rat in the 500 p.p.m. atrazine group and in another rat in the dietary‐restricted group. Decreased sperm in the corresponding epididymides were observed in those two rats. However, they were considered to be spontaneous changes, because the incidences were sporadic and the other rats showed no remarkable changes in the testes and epididymides. In addition, similar changes were often observed in non‐treated rats.

Figure 4.

Representative histopathology and immunohistochemical analysis of ventral prostates in transgenic rats. (a) Control. Atypical glands diagnosed as adenocarcinomas are evident in the upper, and prostatic intraepithelial neoplasia (PIN) (marked with *) is also apparent. (b) 1000 p.p.m. atrazine. Adenocarcinomas are evident in the center, and PIN (marked with *) is also apparent. (c) Most atypical cells have positive reactions for SV40 T antigen (same region as a). (d) Most atypical cells have positive reactions for androgen receptors (same region as a).

Table 3.

Histopathological findings of neoplastic lesions in the prostate of TG rats

	Number of rats	Number of rats with prostatic lesions
		Ventral		Lateral		Dorsal		Anterior
		PIN	AC	PIN	AC	PIN	AC	PIN	AC
Control	10	10	10	10	9	10	3	10	10
Atrazine (500 p.p.m.)	10	10	10	10	9	10	3	10	10
Atrazine (1000 p.p.m.)	10	10	10	10	10	10	3	10	9
Food restriction	10	10	10	10	9	10	2	10	9

PIN, prostatic intraepithelial neoplasia; AC, adenocarcinoma.

Morphometry for evaluation of the prostate cancer development.  Results of morphometry for the prostate epithelium including atypical and cancer cells are summarized in Fig. 5. Significant decreases in the ratios of the epithelial areas to whole prostate tissue areas were observed in the ventral, dorsal, and lateral lobes in the dietary‐restricted group, and in the lateral lobe in the 1000 p.p.m. atrazine group. The values in the dorsal lobe in the atrazine treatment groups also showed a tendency to decrease in a dose‐dependent manner, albeit without statistical significance. No significant effect was observed in the anterior lobe in any treatment groups.

Figure 5.

Results of morphometry for the prostate epithelium in transgenic rats (n = 10). Data are mean±SD. Each longitudinal axis represents a percentage ratio of the epithelial area to the whole prostate tissue area. W, significantly different from the control group, P < 0.05 (Williams test). **, significantly different from the control group, P < 0.01 (paired comparison).

Discussion

In the present study, dietary administration of atrazine for 13 weeks showed slight suppressive effects on the development of prostate cancer in the probasin‐SV40 T antigen transgenic rats, even though the atrazine treatment did not affect the serum testosterone levels. The morphometrically assessed ratio of prostate epithelial lesion area to the whole tissue area was significantly decreased in the lateral lobe of the 1000 p.p.m. atrazine group. A similar tendency for decrease was also observed in the dorsal lobe. However, equal or greater suppression of prostate neoplastic lesion development was also observed in the dietary‐restricted group. Therefore, the influence observed in the atrazine treated groups was probably caused by the decrease in calorie intake, rather than endocrine effects.

Bodyweights in the dietary‐restricted group were similar to those in the 1000 p.p.m. atrazine group. Diet is considered to be a major and important environmental factor contributing to cancers of hormonal tissues, including the prostate. ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ Snyder et al. ⁽²³⁾ and Pollard et al. ²⁴ , ²⁵ reported that dietary restriction reduced the frequency and delayed the occurrence of spontaneous prostate adenocarcinoma in Lobund‐Wistar rats. Mukherjee et al. ⁽²⁶⁾ found dietary restriction to reduce prostate cancer growth in transplantable prostate cancer models (i.e. the Dunning R3327‐H adenocarcinoma in rats). Interestingly, they revealed that tumor microvessel density and vascular endothelial growth factor (VEGF) expression were reduced by dietary restriction, and they hypothesized that dietary restriction reduces prostate cancer growth by inhibiting tumor angiogenesis.

We couldn’t detect histopathological differences between the control group and the treated groups in terms of mitotic figures and apoptotic body formation. In addition, even though we performed immunohistochemistry for single‐stranded DNA, we couldn’t find noticeable differences (data not shown). Therefore, detailed mechanisms for inhibition of prostate carcinogenesis observed in this study were not clear.

Several reports have documented the effects of atrazine on the male reproductive systems in rats. ⁶ , ⁷ , ⁸ Friedmann ⁽⁷⁾ reported that atrazine given by oral gavage acted as an endocrine disrupter in male rats by inhibiting testosterone production in the Leydig cells. However, in the present study, serum testosterone levels were not affected by dietary administration of the compound, despite the fact that serum levels of atrazine, including its metabolites, were consistent with values reported previously. ¹⁴ , ²⁷ This clearly indicated that atrazine was certainly absorbed in the TG rats. The reason for the discrepancy between the present study and the previous reports remained unclear, but gavage administration might conceivably exert different effects from dietary administration. There have been reports in which atrazine was administered by food to rats and expected effects were not observed. ²⁸ , ²⁹ Son et al. ⁽²⁸⁾ reported that dietary administration of atrazine did not influence thyroid carcinogenesis in the rat thyroid two‐stage carcinogenesis model applying N‐bis(2‐hydroxypropyl)nitrosamine (DHPN) as an initiator, even though atrazine given by oral gavage was documented to affect rat thyroid histopathology along with thyroid hormones in an earlier article. ⁽³⁰⁾ Tanaka et al. ⁽²⁹⁾ described dietary administration of atrazine to lack influence on the occurrence of ovarian adenocarcinoma in a rat model initiated with 7,12‐dimethyl‐benz(a)anthracene (DMBA), even though it was reported that atrazine given by oral gavage affected rat ovaries in terms of weights and estrous cyclicity. ⁽³¹⁾

Actual concentrations of atrazine in the prepared diets were found here to be 60–91% of the target values. The reason for the discrepancy is not clear, although homogeneity of the diet preparation or duration between the preparation and the measurement (5 or 6 months) might have affected the results.

In conclusion, our results indicate that high dose atrazine in the diet is associated with slight inhibition of prostate carcinogenesis in probasin‐SV40 T antigen transgenic rats, but this effect is probably caused by suppression of calorie intake, rather than by atrazine‐related endocrine disruption.