Puberty-specific promotion of mammary tumorigenesis by a high animal fat diet

Abstract

Introduction: Increased animal fat consumption is associated with increased premenopausal breast cancer risk in normal weight, but not overweight, women. This agrees with our previous findings in obesity-resistant BALB/c mice, in which exposure to a high saturated animal fat diet (HFD) from peripuberty through adulthood promoted mammary tumorigenesis. Epidemiologic and animal studies support the importance of puberty as a life stage when diet and environmental exposures affect adult breast cancer risk. In this study, we identified the effects of peripubertal exposure to HFD and investigated its mechanism of enhancing tumorigenesis.

Methods: Three-week-old BALB/c mice fed a low-fat diet (LFD) or HFD were subjected to 7,12-dimethylbenz[a]anthracene (DMBA)-induced carcinogenesis. At 9 weeks of age, half the mice on LFD were switched to HFD (LFD-HFD group) and half the mice on HFD were switched to LFD (HFD-LFD group). Tumor gene expression was evaluated in association with diet and tumor latency.

Results: The peripubertal HFD reduced the latency of DMBA-induced mammary tumors and was associated with tumor characteristics similar to those in mice fed a continuous HFD. Notably, short-latency tumors in both groups shared gene expression characteristics and were more likely to have adenosquamous histology. Both HFD-LFD and continuous HFD tumors showed similar gene expression patterns and early latency. Adult switch from HFD to LFD did not reverse peripubertal HFD tumor promotion. Increased proliferation, hyperplasia, and macrophages were present in mammary glands before tumor development, implicating these as possible effectors of tumor promotion. Despite a significant interaction between pubertal diet and carcinogens in tumor promotion, peripubertal HFD by itself produced persistent macrophage recruitment to mammary glands.

Conclusions: In obesity-resistant mice, peripubertal HFD is sufficient to irreversibly promote carcinogen-induced tumorigenesis. Increased macrophage recruitment is likely a contributing factor. These results underscore the importance of early life exposures to increased adult cancer risk and are consistent with findings that an HFD in normal weight premenopausal women leads to increased breast cancer risk. Notably, short-latency tumors occurring after peripubertal HFD had characteristics similar to human basal-like breast cancers that predominantly develop in younger women.

Fig. 1

Experimental design and Kaplan-Meier plot of 7,12-dimethylbenz[a]anthracene (DMBA)-induced mammary tumors. a BALB/c mice were started on a high saturated animal fat diet (HFD) or a low-fat diet (LFD) at 3 weeks old. Mice were switched from HFD to LFD and from LFD to HFD at 9 weeks old. DMBA was administered weekly from ages 5 to 8 weeks. Tumor incidence was observed. b–e Kaplan-Meier plots of tumor incidence. Time = number of days after last DMBA treatment (HFD mice, n = 101; LFD mice, n = 90; HFD-LFD mice, n = 42; LFD-HFD mice, n = 45). b HFD-LFD vs. HFD. c LFD-HFD vs. LFD. d LFD-HFD vs. HFD-LFD. e LFD vs. HFD

Fig. 2

Tumor characteristics based on histopathology. a Mice that were on the high saturated animal fat diet (HFD) and were switched to the low-fat diet (HFD-LFD) had an increased proportion of adenosquamous tumors compared with mice on the HFD that were switched to LFD (LFD-HFD) (HFD-LFD 55 % vs. LFD-HFD 0 %). *p < 0.05. Conversely, LFD-HFD-fed mice had an increased proportion of nonadenosquamous tumors compared with HFD-LFD-fed mice (LFD-HFD 100 % vs. HFD-LFD 45 %). *p < 0.05. b Adenosquamous tumors (n = 6) in HFD-LFD-fed mice had reduced latency compared with other tumor types (n = 5). *p < 0.05. c Both adenosquamous and other mammary tumor types had increased proliferation in HFD-LFD-fed mice compared with tumors in LFD-HFD-fed mice (n = 6). *p < 0.05. b and c Bars represent mean ± standard error of the mean (SEM). d HFD-fed (n = 29 tumors) and LFD-fed (n = 16 tumors) mice had a similar incidence of all tumor types. e HFD-fed mice had reduced tumor latency for both adenosquamous (n = 12) and other tumor types (n = 17) compared with total tumors in LFD-fed mice (adenosquamous, n = 5; other, n = 11). *p < 0.05. f Adenosquamous tumors from HFD-fed mice had increased proliferation compared with all other tumors in HFD- and LFD-fed mice. *p < 0.05. b, c, e, and f Bars represent mean ± SEM. g Incidence of estrogen receptor– and progesterone receptor–negative (ER − PR−) tumors was increased among HFD-early (n = 11 of 12) and HFD-LFD-early (n = 6 of 6) tumors compared with LFD (n = 4 of 16), HFD-late (n = 2 of 5), LFD-HFD (n = 0 of 6), and HFD-LFD-late (n = 2 of 5) tumors. *p < 0.05. BrdU 5-bromo-2′-deoxyuridine, DMBA 7,12-dimethylbenz[a]anthracene

Fig. 3

Overall tumor characteristics in HFD-, LFD-, HFD-LFD-, and LFD-HFD-fed mice. a Blood vessel density (CD31 staining) was increased in LFD-HFD-fed murine mammary tumors (#p = 0.07) and significantly increased in HFD- and HFD-LFD-fed murine mammary tumors compared with LFD-fed murine mammary tumors. *p < 0.05. b Macrophage (Macs; F4/80 staining) recruitment was increased within the stroma of tumors from mice fed a HFD-LFD vs. LFD-HFD. *p < 0.05. Bars represent mean ± standard error of the mean; n = 4–10 tumors per diet treatment. Arg1 arginase 1, HFD high saturated animal fat diet, HFD-LFD mice on high saturated animal fat diet switched to low-fat diet, LFD low-fat diet, LFD-HFD mice on low-fat diet switched to high saturated animal fat diet

Fig. 4

Microarray heat map cluster analysis. The heat map and dendrogram represent a two-class significance analysis of genes differentially expressed between early vs. late tumor onset and high saturated animal fat diet (HFD) vs. a low-fat diet (LFD). Two gene clusters were identified: one enriched for genes upregulated among early-occurring tumors (gray bar) and one enriched for genes downregulated among early-occurring tumors (black bar). Early-occurring tumors from HFD- and HFD-LFD-fed mice cluster together. Diet group, early vs. late tumor onset, and adenosquamous vs. other histologies are noted. HFD-LFD mice on high saturated animal fat diet switched to low-fat diet, LFD-HFD mice on low-fat diet switched to high saturated animal fat diet

Fig. 5

Effect of various diets on mammary glands at 13 weeks of age and before tumor development. a More hyperplastic lesions (hyperplasias) developed in HFD- and HFD-LFD-fed, 7,12-dimethylbenz[a]anthracene-treated mice. *p < 0.05. b Mice fed HFD or HFD-LFD exhibited increased cellular proliferation in both normal epithelium and hyperplastic lesions (hyperplasia) compared with those fed LFD or LFD-HFD, respectively, as measured by 5-bromo-2′-deoxyuridine incorporation. *p < 0.05. c Macrophage (F4/80-stained cells) recruitment was increased adjacent to normal ducts and hyperplastic lesions (hyperplasia) (*p < 0.05) in mammary glands of HFD-, HFD-LFD-, and LFD-HFD-fed mice compared with those of LFD-fed mice. d Blood vessel density (CD31 staining) was significantly increased adjacent to normal mammary gland structures and hyperplastic lesions (hyperplasia) (*p < 0.05) in HFD-fed compared with LFD-, HFD-LFD-, or LFD-HFD-fed mice. Bars represent mean ± standard error of the mean (n = 5 mice per diet group for each assay). HFD high saturated animal fat diet, HFD-LFD mice on high saturated animal fat diet switched to low-fat diet, LFD low-fat diet, LFD-HFD mice on low-fat diet switched to high saturated animal fat diet

Fig. 6

β-catenin regulation by diet treatment in mammary glands at 13 weeks of age. Before tumor development, β-catenin levels were measured based on immunofluorescence. β-catenin levels were increased by HFD (p = 0.08) and HFD-LFD. *p < 0.05. Bars represent mean ± standard error of the mean (n = 5 mice per diet treatment). HFD high saturated animal fat diet, HFD-LFD mice on high saturated animal fat diet switched to low-fat diet, LFD low-fat diet, LFD-HFD mice on low-fat diet switched to high saturated animal fat diet

Fig. 7

β-catenin regulation by diet treatment in tumors. β-catenin levels were measured based on immunofluorescence intensity or nuclear localization in adenosquamous (ADSQ) and nonadenosquamous (Other) tumors. a β-catenin levels were increased in nonadenosquamous tumors from HFD-fed mice (HFD Other) compared with those from LFD-mice (LFD Other). *p < 0.05. b These same tumors from HFD-fed mice (HFD Other) showed a trend toward increased nuclear β-catenin compared with nonadenosquamous tumors from LFD-fed mice (LFD Other) (p = 0.09). c Tumors from HFD-LFD- and LFD-HFD-fed mice had similar β-catenin levels. Bars represent mean ± standard error of the mean (n = 4–8 tumors per diet treatment). HFD high saturated animal fat diet, HFD-LFD mice on high saturated animal fat diet switched to low-fat diet, LFD low-fat diet, LFD-HFD mice on low-fat diet switched to high saturated animal fat diet

Fig. 8

Effects of diet with or without 7,12-dimethylbenz[a]anthracene (DMBA) treatment in the mammary gland at 13 weeks of age. a Proliferation in mammary glands from mice fed HFD, LFD, HFD-LFD, or LFD-HFD was measured by 5-bromo-2′-deoxyuridine incorporation. Only diet treatments with DMBA increased proliferation (*p < 0.05). b Only HFD with DMBA increased blood vessel density (CD31 staining) compared with all other diet treatments (*p < 0.05). c HFD with or without DMBA and HFD-LFD without DMBA increased macrophage recruitment (F4/80 staining) compared with LFD and LFD-HFD without DMBA. *p < 0.05. Bars represent mean ± standard error of the mean (n = 5 mice per diet and DMBA treatment). HFD high saturated animal fat diet, HFD-LFD mice on high saturated animal fat diet switched to low-fat diet, LFD low-fat diet, LFD-HFD mice on low-fat diet switched to high saturated animal fat diet

Fig. 9

Effects of high saturated animal fat diet (HFD) with and without 7,12-dimethylbenz[a]anthracene (DMBA) on mammary gland morphology. Mammary gland morphology was assessed from whole mounts of mice fed HFD or a low-fat diet (LFD) with or without DMBA from 3 to 13 weeks of age. Mammary glands from either DMBA-treated HFD- or LFD-fed mice retained a pubertal ductal organization with numerous terminal end buds present (indicated by arrows). In the absence of DMBA, HFD- and LFD-fed mice exhibited a mature morphology indicated by extensive ductal branching and presence of ductal side branches (indicated by arrowheads). Representative photomicrographs of five mice per diet and DMBA treatment are shown. Scale bar = 1 mm