Pubertal high fat diet: effects on mammary cancer development

Abstract

Introduction: Epidemiological studies linking dietary fat intake and obesity to breast cancer risk have produced inconsistent results. This may be due to the difficulty of dissociating fat intake from obesity, and/or the lack of defined periods of exposure in these studies. The pubertal mammary gland is highly sensitive to cancer-causing agents. We assessed how high fat diet (HFD) affects inflammation, proliferative, and developmental events in the pubertal gland, since dysregulation of these can promote mammary tumorigenesis. To test the effect of HFD initiated during puberty on tumorigenesis, we utilized BALB/c mice, for which HFD neither induces obesity nor metabolic syndrome, allowing dissociation of HFD effects from other conditions associated with HFD.

Methods: Pubertal BALB/c mice were fed a low fat diet (12% kcal fat) or a HFD (60% kcal fat), and subjected to carcinogen 7,12-dimethylbenz[a]anthracene (DMBA)-induced tumorigenesis.

Results: HFD elevated mammary gland expression of inflammatory and growth factor genes at 3 and 4 weeks of diet. Receptor activator of nuclear factor kappa-B ligand (RANKL), robustly induced at 4 weeks, has direct mitogenic activity in mammary epithelial cells and, as a potent inducer of NF-κB activity, may induce inflammatory genes. Three weeks of HFD induced a transient influx of eosinophils into the mammary gland, consistent with elevated inflammatory factors. At 10 weeks, prior to the appearance of palpable tumors, there were increased numbers of abnormal mammary epithelial lesions, enhanced cellular proliferation, increased growth factors, chemokines associated with immune-suppressive regulatory T cells, increased vascularization, and elevated M2 macrophages. HFD dramatically reduced tumor latency. Early developing tumors were more proliferative and were associated with increased levels of tumor-related growth factors, including increased plasma levels of HGF in tumor-bearing animals. Early HFD tumors also had increased vascularization, and more intra-tumor and stromal M2 macrophages.

Conclusions: Taken together in this non-obesogenic context, HFD promotion of inflammatory processes, as well as local and systemically increased growth factor expression, are likely responsible for the enhanced tumorigenesis. It is noteworthy that although DMBA mutagenesis is virtually random in its targeting of genes in tumorigenesis, the short latency tumors arising in animals on HFD showed a unique gene expression profile, highlighting the potent overarching influence of HFD.

Figure 1

Characteristics of tumor development in high fat diet- versus low fat diet-fed mice. (A) Kaplan-Meier plot of all tumors developing in high fat diet (HFD)- and low fat diet (LFD)-fed mice. Time = number of days post last 7,12-dimethylbenz[a]anthracene (DMBA) treatment (HFD mice, n = 95; LFD mice, n = 87) (B) Time-course of epithelial proliferative lesion development. More hyperplastic and precancerous lesions developed in HFD-fed DMBA-treated mice at 8 and 14 weeks post first DMBA treatment. Bars represent mean ± standard error of the mean of lesions per mouse; n = 5 mice at each time for HFD and LFD. ^*P = 0.05; ^**P = 0.003. (C) Epithelial proliferative lesions comprised (i) terminal duct hyperplasia, (ii) lobular hyperplasia, and (iii) mixed dysplasia. Scale bar = 1 mm.

Figure 2

Proliferation and angiogenesis in mammary glands, tumors, and tumor microenvironments in mice fed on high fat and low fat diets. (A) Proliferation: fold increases in proliferation in normal mammary epithelium at 4 and 10 weeks, hyperplastic foci at 10 weeks, and in tumor epithelium from mice on a high fat diet (HFD) versus low fat diet (LFD). At 4 weeks, mice fed HFD exhibited increased cellular proliferation as measured by proliferating cell nuclear antigen (PCNA); ^*P <0.05. Note, the 4-week data are a re-analysis of data published in Olson et al. (2008). At 10 weeks, mice fed HFD exhibited increased cellular proliferation in both normal epithelium and hyperplastic foci, as measured by 5-bromo-2’-deoxyuridine (BrdU) incorporation; ^*P <0.05. Early developing tumors on HFD (HFD-E) also exhibited increased cellular proliferation, as measured BrdU incorporation; ^*P <0.05. (B) Angiogenesis: blood vessel density was measured, as described in Materials and Methods, by the area occupied by CD31-positive vessels near normal mammary epithelium at 3, 4, and 10 weeks, and in tumor epithelium from mice on HFD versus LFD. At 10 weeks, CD31-stained vessels were significantly increased adjacent to normal mammary gland structures (^*P = 0.01), hyperplastic foci (^**P = 0.04) and microscopic tumors (tumors versus hyperplasia; ^***P = 0.02) in HFD-fed compared to low fat diet (LFD)-fed mice. CD31 staining was also greater in HFD-E tumors (^#P = 0.01) compared to LFD tumors. (C) Insets show increased CD31 staining of (ii) an HFD-E tumor compared with (i) an LFD tumor. Scale bars = 50 mm.

Figure 3

Inflammatory cell recruitment in mammary gland, tumors, and tumor microenvironments in high fat diet- and low fat diet-fed mice. (A and B) BALB/c mice were started on diets at 3 weeks of age and terminated after 3 or 4 weeks for analysis of eosinophil (A) and mast cell (B) recruitment to mammary gland epithelial structures, as described in the Materials and Methods. At 3 weeks on diet, eosinophil recruitment (A) for all mammary structures and mast cell recruitment (B) for large ducts was significantly increased in high fat diet (HFD) compared to low fat diet (LFD)-fed mice. ^*P = 0.0001; ^**P = 0.03. (C, D, E, and F) Sections from mice terminated at 3 weeks on diet (C), 4 weeks on diet (D), 10 weeks on diet (E) and from HFD-E and LFD tumors (F) were double immunofluorescently stained with anti-F4/80 and anti-Arg1 antibodies, as described in Materials and Methods, and then analyzed for macrophage recruitment. At 10 weeks, total macrophage (F4/80) and M2 macrophage (Arg1 + F4/80) recruitment (E) was increased adjacent to small ducts (^*P = 0.01) and large ducts (^**P = 0.05) in mammary glands of HFD-fed mice. The increase in F4/80 and Arg1 + F4/80 staining in HFD versus LFD hyperplasia was not significant (P = 0.16). Tumor-associated macrophages (F) were quantified based on their location at the tumor edge, within the tumor (intra-tumor), in the tumor stroma (stroma), and combined for total tumor-associated macrophages. ^*P = 0.05 that there were more F4/80 and Arg1 + F4/80 labeled macrophages within HFD-E tumors. ^#P = 0.01 that there were more Arg1 + F4/80 macrophages (total) in HFD-E tumors and within HFD-E stroma.

Figure 4

Increase in significant genes identified by targeted pathway PCR arrays between 4 and 10 weeks on diet. RNA isolated from week 3, 4, and 10 diet groups were analyzed using Growth Factors and Inflammatory Cytokines and Receptors PCR arrays (SABiosciences). The week-3 sample group identified six significant genes, the week-4 sample group identified five significant genes, and the week-10 diet group identified twenty-five significant genes.