A maternal diet high in n − 6 polyunsaturated fats alters mammary gland development, puberty onset, and breast cancer risk among female rat offspring

In Utero Manipulations.

Three-month-old female Sprague–Dawley rats (Charles River Breeding Laboratories) were fed either a high or low n − 6 PUFA diet, the feeding beginning 5 days before mating and continuing throughout pregnancy. The animals were returned to a Purina laboratory chow 5001, which contains 12% calories from fat (saturated and unsaturated), immediately after delivery. On postnatal day 2, the offspring were crossfostered and kept with nursing dams that were always fed the Purina chow. Crossfostering was done to control for possible differences in the milk content of mothers fed with the high or low PUFA diet during pregnancy. The mean litter-size in each group also was balanced to 8–10 offspring.

The experimental diets, prepared commercially by Food-Tek (Morris Plains, NJ) and Bioserv (Frenchtown, NJ), contained either 43% or 46% of calories from corn oil (high n − 6 PUFA), or 12% or 16% calories from corn oil (low n − 6 PUFA) (20). These percentages of calories from fat are comparable to those consumed in the American diet (≈37%) (36), and those achievable in low-fat (15–20%) human dietary intervention studies (37). The diet we address as low n − 6 PUFA contains optimal levels of this fatty acid as well as n − 3 PUFA for rodent needs (38). However, because the 12–16% calories from fat diet is comparable to a human low-fat diet, the rat diet also was called a low n − 6 PUFA diet. Formulation of the semipurified animal feeds (AIN-76A) were based on the recommendations of the American Institute of Nutrition (39). The protein, mineral, and vitamin contents of the high and low n − 6 PUFA diets were similar (19–20% kcal from protein, 50 g of mineral mix, and 35 g of vitamin mix per 1 kg of feed). To achieve the same caloric intake, the high n − 6 PUFA diets were made isocaloric with the low n − 6 PUFA diets by adjusting their caloric contents with carbohydrates (corn starch, maltose, and sucrose) and fiber (alphacel). The caloric density of the diets was 3.7 kcal/g feed. Thus, the high n − 6 PUFA diets contained less carbohydrates (38% versus 64% kcal from carbohydrates) and more fiber than the low n − 6 PUFA diets. In a separate set of experiments, 3-month-old pregnant rats were injected subcutaneously with 20 ng of E2 (Sigma) and/or oil vehicle on days 14–20 of gestation in the absence of dietary manipulations.

Serum Estradiol.

Total plasma E2 levels were measured using a double-antibody kit (Diagnostic Products, Los Angeles, or ICN) according to the manufacturers’ specifications. The blood was collected by cardiac puncture from pregnant rats kept on the high or low n − 6 PUFA diets on gestation day 14, and from their offspring on postnatal days 8 and 60. In the offspring of E2-treated mothers, the blood was collected on postnatal day 64. The estrus cycle was matched among the offspring, most of them being in estrus at the time of blood collection.

Mammary Parenchymal Patterns.

We dissected the fourth abdominal mammary glands from 4-, 7-, and 11-week-old rats that were exposed to a high or low n − 6 PUFA diet in utero, and from 4- and 7-week-old rats exposed to E2 in utero. Tissues were processed for morphological analysis as whole mounts. By week 4, the occurrence of terminal end buds (TEBs) is high (40). By week 7, the offspring were exposed to DMBA to assess effects on susceptibility to transformation. By week 11, the typical adult pattern of epithelial cell growth can be seen.

Analysis of mammary epithelial structures was made through visual evaluation and computer-assisted image analysis. We have developed and characterized a visual scale to assess the growth patterns of mammary epithelial cells (41). In our scale method, the stained mammary whole mounts were examined under an Olympus dissecting scope. The following characteristics of the coded mammary glands were evaluated double-blind, using a 4-point scale (0 = absent, 1 = few, 2 = moderate, 3 = numerous): the relative density of (i) epithelial ducts; (ii) TEBs; and (iii) alveolar buds (ABs). The inter-individual correlation was 0.7–0.9 for various glandular structures. These epithelial structures are well characterized by Russo and Russo (40). Measurement of the density of epithelial structures also was performed using the Optimas Image Analysis system. Evaluation of mammary gland morphology was based on images of the glands taken from a camera-affixed, Olympus stereo microscope, using Optimas software. At the age of 4 weeks, total epithelial area was used for density measurements with the image analysis (only two glands per group were suitable for the image analysis at this age). At the ages of 7 and 11 weeks, an area of 3 × 3 mm, obtained from the same location of the gland for each animal, was used.

Mammary Tumorigenesis.

Mammary tumors were induced by administering 10 mg of DMBA (Sigma) to the 55-day-old offspring by oral gavage. The animals were checked regularly for mammary tumors by palpation once per week. Tumor growth rates were measured by recording the tumor diameters with a caliper and determining the length of the longest axis and the width perpendicular to the longest axis. Tumor doubling time was determined only for proliferating tumors, tumor volume and tumor doubling time being estimated as described by Rygaard and Spang-Thomsen (42). The animals were sacrificed when detectable tumor burden approximated 10% of total body weight. The surviving animals and animals that did not develop mammary tumors were sacrificed on week 18 after the DMBA administration.

Puberty Onset: Vaginal Opening.

Sexual maturation in rodents can be determined by establishing the age when vaginal opening occurs, the first estrus occurring within a few days of this event (43). The animals were examined daily starting on postnatal day 33.

Pregnancy Outcome.

Pregnancy was initiated and maintained with a similar success rate in both groups consuming a high- or low-fat diet. Food consumption and body weight gain of the pregnant rats, and the number of pups born and their body weights, did not differ between the two groups (Table 1). Among the pregnant rats exposed to E2 or vehicle, pregnancy- and nursing-related variables also were similar in the two groups (not shown).

Circulating E2 Levels.

The plasma levels of E2 were approximately 30% higher in pregnant females consuming the highest-fat feed (46% calories from corn oil) than in pregnant females consuming the lowest-fat feed (12% calories from corn oil) (t = 3.37, df = 9, P < 0.01; Student’s t test) (Fig. 1). The plasma E2 levels were not altered in the 8-day-old offspring of the dietary-manipulated mothers, who from the birth onward were exposed to the Purina rodent chow (Fig. 1). This is consistent with the clearance of estrogens from neonates during the first few days postpartum. There were no differences in serum E2 levels between the 2-month-old female offspring of mothers fed with a high and low n − 6 PUFA diet, or between the offspring of mothers injected with E2 or oil vehicle during pregnancy (data not shown). Thus, an in utero exposure to a high-fat diet does not appear to significantly alter adult E2 levels.

Mammary Parenchymal Patterns.

High-fat diet. The total mammary fat pad area, as measured by a caliper, was larger in the female rats exposed to a high n − 6 PUFA diet in utero than in the female rats exposed to a low-fat diet (F(1, 12) = 5.7, P < 0.03; two-way ANOVA) (not shown). This finding is consistent with the observation that early postnatal exposure to estrogens stimulates growth of the mammary fat pad (44). The visual image analysis indicated that, on week 4, the total epithelial cell density was clearly higher in the high-fat (5, 582 ± 100 mm²) than in the low-fat group (188 ± 30 mm²). In addition, in the 7- and 11-week-old offspring of high-fat-fed mothers, the representative epithelial cell area was significantly more dense than that of the offspring of the low-fat fed mothers (F(1, 8) = 6.6, P < 0.03) (Figs. 2A and 3). The scale-method produced comparable results. The epithelial tree was more dense, as defined by the visual score, in the female rats exposed to a high-fat diet in utero than in the rats exposed to a low-fat diet (F(1, 21) = 13.43, P < 0.001) (Fig. 2A).

The mammary epithelial structures were significantly different between the female rats exposed to a high and low n − 6 PUFA diets through their pregnant mother (Fig. 2B). The relative density of TEBs, the sites of carcinogen-induced malignant transformation (40), was higher on weeks 4, 7, and 11 in the high-fat animals, when compared with the low-fat animals (F(1, 21) = 15.0, P < 0.0009). The relative density of ABs was not significantly different between the groups on week 7. However, when examined on week 11, the mammary glands in the offspring of female rats fed with a high-fat diet during pregnancy had significantly more ABs than the low-fat animals (F(1, 11) = 5.7, P < 0.04) (Fig. 2B).

Mammary Tumorigenesis.

High-fat diet. The number of animals in each group with one or more mammary tumors (tumor incidence) was higher among the rats who were exposed in utero through a maternal diet high in n − 6 PUFA, when compared with the offspring of rats fed the low n − 6 PUFA diet (z = 4.03, P < 0.001; Gehan–Wilcoxan test) (Fig. 5A). On week 18 after DMBA administration, 60% of the female offspring of mothers exposed to a high n − 6 PUFA diet during pregnancy had developed mammary tumors, whereas 30% of the offspring of low n − 6 PUFA mothers had mammary tumors. Tumor multiplicity (the mean number of tumors per female) was similar between the two groups. The latency for tumor appearance was significantly shorter in the high-fat diet group (t = 3.4, df = 51, P < 0.002) (Table 2). The proportions of histotypes (77% malignant adenocarcinomas, 6% squamous cell carcinomas, and 17% nonmalignant adenomas) were comparable to those previously reported with DMBA (40) and were not affected by the in utero dietary manipulations.

Tumor growth rate and size upon first detection did not differ between the offspring of mothers fed a high- or low-fat diet during pregnancy (Table 2). These endpoints reflect tumor promotion/progression events, i.e., rate of growth of initiated cells. Thus, the effects of a fetal exposure to a high-fat diet through mother on mammary tumorigenesis are more closely associated with factors influencing breast cancer risk.

Vaginal Opening.

High-fat diet. Fig. 6A shows that vaginal opening occurred at a younger age in the offspring exposed to a diet high in n − 6 PUFA in utero (z value = 3.9, P < 0.0004; Gehan–Wilcoxan test). These data suggest that critical events may occur in utero that influence the “priming” of the hypothalamic-gonadal axis for the timing of the onset of puberty. These events may be mediated by estrogens and appear sensitive to dietary influences.

Estrus cycling and uterine wet weights were studied to assess the endocrinologic/estrogenic environment at the time when carcinogen was administered (between 50–60 days of age). The length of the estrus cycle was normal (4–5 days) in 80% of both the rats exposed to a high- and low n − 6 PUFA diet in utero via the maternal diet. Uterine wet weights were similar in the 55-day-old offspring of the high- and low-fat groups. Thus, the reproductive hormonal environment at the time of DMBA exposure is unlikely to have influenced mammary tumor incidence between females exposed to a high- or low-fat diet in utero.