Social isolation induces autophagy in the mouse mammary gland: link to increased mammary cancer risk

  1. Leena Hilakivi-Clarke1
  1. 1Department of Oncology, Georgetown University, Washington, District of Columbia, USA
  2. 2Department of Surgery, Wake Forest University, Winston-Salem, North Carolina, USA
  3. 3Faculty of Pharmaceutical Sciences, Department of Food and Experimental Nutrition, University of São Paulo, São Paulo, Brazil
  1. Correspondence should be addressed to L Hilakivi-Clarke; Email: clarkel{at}georgetown.edu
  1. Figure 1

    Experimental design. Three separate experiments were performed to assess the effects of social isolation, with or without feeding mice obesity-inducing diet (OID), on mammary tumorigenesis and autophagy.

  2. Figure 2

    Effect of OID and social isolation on food consumption, body weight and visceral fat weight. (A) Food consumption in mice that were group-housed and fed control diet (G-C), group-housed and fed obesity-inducing diet (G-OID), socially isolated and fed the control diet (I-C), and socially isolated and fed OID (I-OID). Consumption was measured by weighing the amount of food (g) in the cages every 2–3 days over the first three weeks of dietary and social isolation exposures (n = 5 cages/group). OID reduced food intake: P < 0.001 and social isolation increased it: P < 0.001. (B) Caloric intake (kcal) was calculated based on the amount of food consumed by these mice and the caloric content of the food. During the 3 weeks when measurements were done, both OID: P < 0.001 and social isolation: P < 0.001 increased kcal intake. (C) Body weight after 4 weeks of dietary and social isolation exposures (n = 8–12 mice per group). Body weight was higher by OID: P < 0.001 and social isolation: P = 0.048. (D) Visceral fat weight (epididymal fat pads) in the same mice as shown in (C); it was increased by OID: P < 0.001. (E) Ratio between visceral fat and body weight; it was increased by OID: P < 0.001. (F) Body weight gain from the start of OID and social isolation on week 3 until week 16 (n = 15–21 mice per group), shown as weight in grams per week and as a percentile change. Both measures indicated that body weights were increased by OID: P < 0.001 and social isolation: P < 0.001. In addition, interaction between OID and social isolation was significant (P < 0.001) and indicated that initially social isolation was more potent than OID in increasing body weight, but when G-OID group started to be significantly heavier than the control group, I-C group no longer was heavier than the controls. For details, see ‘Results’ section. Bars marked with different letters are significantly different from each other. Means and standard error of means (s.e.m.) are shown.

  3. Figure 3

    Effect of OID and social isolation on mammary tumorigenesis. MPA + DMBA-induced mammary tumorigenesis in mice that were group-housed and fed the control diet (G-C), group-housed and fed obesity-inducing diet (G-OID), socially isolated and fed the control diet (I-C), and socially isolated and fed OID (I-OID) (n = 15–21/group). (A) Tumor incidence, assessed using log-rank test, was significantly higher in the I-OID mice compared with all other groups: P < 0.001, and in G-OID and I-C groups, compared with G-C control group: P < 0.05. (B) Tumor latency was significantly shortened by OID: P = 0.004 and by social isolation: P = 0.002. (C) Tumor multiplicity and (D) tumor burden were not affected. Bars marked with different letters are significantly different from each other. Means and standard error of means (s.e.m.) are shown.

  4. Figure 4

    Effect of OID and social isolation on insulin and glucose tolerance, and on neuropeptide Y, PPARγ and adipokine levels. Insulin and glucose tolerance were studied in 12- to 14-week-old mice exposed to MPA + DMBA that were group-housed and fed control diet (G-C), group-housed and fed obesity-inducing diet (G-OID), socially isolated and fed control diet (I-C) and socially isolated and fed OID (I-OID) (n = 4–10 mice per group). Blood glucose levels were determined at baseline (0 min, before injection) and 30, 60, 90 and 120 min after injection. (A) Insulin tolerance: at 60 min, glucose levels were significantly higher in I-OID mice than the G-C mice: P = 0.015 (difference marked with *). (B) Area under the curve (AUTC) in insulin tolerance test was significantly increased in socially isolated OID-fed mice: P = 0.021. (C) Glucose tolerance: At baseline, glucose levels were significantly elevated by social isolation but only in control diet-fed mice: P for interaction = 0.013. At 60 min, OID elevated glucose levels: P = 0.015. In addition, there was a significant interaction between OID and social interaction at 60 min: P = 0.021 (for interaction), 90 min: P = 0.016 (for interaction) and 120 min: P = 0.038 (for interaction), indicating that OID elevated glucose levels in group-housed mice and social isolation in control diet-fed mice (differences marked with *). (D) There was a significant interaction between OID and social isolation in affecting AUTC in glucose tolerance test (P = 0.033). At the end of tumor-monitoring period, serum and mammary tissues were obtained and EIA and qRT-PCR were performed to determine the circulating levels of neuropeptide Y (NPY), leptin and adiponectin and mRNA expression of Pparγ in the mammary glands (n = 4–7 mice per group). (E) NPY levels were significantly elevated in I-OID group; interaction: P = 0.013. (F) Leptin levels were significantly elevated in OID-fed mice: P < 0.001, and (G) adiponectin levels were significantly reduced in OID-fed mice: P = 0.011 and socially isolated mice: P = 0.042, but post hoc analysis indicated that the difference was significant only between G-OID and G-C groups. (H) Pparγ mRNA expression was significantly reduced by OID: P = 0.010 and social isolation: P = 0.036, but post hoc analysis indicated that the difference was significant only between I-OID and G-C groups. Bars marked with different letters are significantly different from each other. Means and standard error of means (s.e.m.) are shown.

  5. Figure 5

    Effect of OID and social isolation on p53, determined in the mammary glands of mice group-housed and fed control diet (G-C), group-housed and fed obesity-inducing high-fat diet (G-OID), socially isolated and fed control diet (I-C), and socially isolated and fed OID (I-OID) (n = 4–7 mice per group for RNA and n = 6–11 mice per group for immunohistochemistry, IHC). All mice were exposed to MPA + DMBA, and glands were collected at the end of tumor monitoring period. (A) Social isolation upregulated Trp53 mRNA levels; social isolation: P < 0.001. (B and C) Nuclear staining of p53 in mammary adipocytes, determined by IHC, was significantly increased by social isolation: P = 0.016. (D) Social isolation increased mRNA levels of Dram1; social isolation: P = 0.003 and (E) Jnk1 (n = 5–7/group); social isolation: P = 0.003. (F) Social isolation also increased mRNA levels of Mdm2; social isolation: P < 0.001.

  6. Figure 6

    Effect of OID and social isolation on unfolded protein response (UPR), determined by immunoblotting UPR markers (A) GRP78; increased by OID: P = 0.016, but post hoc analysis indicated that only isolated mice fed OID had significantly higher expression of GRP78 than controls, (B) IRE1; increased by OID: P < 0.001 and social isolation: P = 0.012. (C) PERK; no significant changes, (D) CHOP; no significant changes. Effect of OID and social isolation on autophagy, determined by immunoblotting autophagy markers (H) Beclin1; increased by OID: P = 0.008, (G) Atg7; increased by social isolation: P < 0.0001. Although OID also increased Atg7 expression in group-housed mice, socially isolated mice fed control diet had higher expression than G-OID or I-OID: P for interaction = 0.002, (E) LC3II; increased by social isolation: P = 0.002. Although OID also increased LC3II expression in group-housed mice, socially isolated mice fed control diet had higher expression than G-OID or I-OID: P for interaction = 0.009, and (F) p62; downregulated by social isolation: P < 0.001. Although OID also reduced p62 expression in group-housed mice, socially isolated mice fed control diet had higher expression than G-OID or I-OID: P for interaction = 0.003. (I) LC3A/B puncta staining of mammary glands; increased by social isolation: P < 0.001. Bars marked with different letters are significantly different from each other. Means and standard error of means (s.e.m.) of 4–11 mice per group are shown; these mice were not exposed to DMBA.

  7. Figure 7

    Effect of OID and social isolation on cell proliferation and apoptosis, determined in the mammary glands of mice group-housed and fed control diet (G-C), group-housed and fed obesity-inducing diet (G-OID), socially isolated and fed control diet (I-C), and socially isolated and fed OID (I-OID) (n = 4–10 mice per group). All mice were exposed to MPA + DMBA, and glands were collected at the end of tumor monitoring period. Indicators of cell proliferation were elevated in socially isolated mice: (A and B) assessed by Ki67 staining; social isolation: P = 0.002 and (C) Ccnd1 mRNA; social isolation: P < 0.001. (D and E) Apoptosis was not altered in any of the groups (assessed by TUNEL assay). Bars marked with different letters are significantly different from each other. Means and standard error of means (s.e.m.) are shown.

  8. Figure 8

    Effect of OID and social isolation on body weight, autophagy and mammary tumorigenesis in Atg7+/− mice. Atg7+/− and their wild-type (WT) littermates were group-housed and fed control diet (G-C) or socially isolated and fed obesity-inducing diet (I-OID). Body weight gain from weaning until week 20 (n = 15–21 mice/group), (A) shown as a fold change from the body weight at weaning and (B) body weight in grams. Fold change in body weights was significantly higher in socially isolated, OID-fed mice than in group-housed mice fed the control diet: P = 0.023. (C) MPA + DMBA-induced mammary tumor incidence was significantly higher in WT mice kept in I-OID than in group-housed G-C mice (P = 0.037), but not in I-OID Atg7+/− mice, compared with G-C Atg7+/− mice (P = 0.267) (n = 16–20 mice/group). (D) Among socially isolated mice fed OID, mammary tumor multiplicity (number of tumors per mice) was significantly higher (P = 0.0014) in WT group than in Atg7+/− mice. Significant differences are marked with *. (E) LC3II or (F) p62 levels were determined in mammary glands at the end of mammary tumor monitoring period, and they were not different in group-housed WT and Atg7+/− mice. Social isolation increased LC3II levels in WT mice and reduced them in Atg7+/− mice, compared with control WT mice (P for interaction = 0.002). Social isolation reduced p62 levels in WT mice but increased them in Atg7+/− mice (P for interaction = 0.006). Bars marked with different letters are significantly different from each other. Means and standard error of means (s.e.m.) are shown.

| Table of Contents
Editor's Choice