Maternal obesity driven changes in collagen linearity of breast extracellular matrix induces invasive mammary epithelial cell phenotype

Abstract

Obesity has been linked with numerous health issues as well as an increased risk of breast cancer. Although effects of direct obesity in patient outcomes is widely studied, effects of exposure to obesity-related systemic influences in utero have been overlooked. In this study, we investigated the effect of multigenerational obesity on epithelial cell migration and invasion using decellularized breast tissues explanted from normal female mouse pups from a diet induced multigenerational obesity mouse model. We first studied the effect of multigenerational diet on the mechanical properties, adipocyte size, and collagen structure of these mouse breast tissues, and then, examined the migration and invasion behavior of normal (KTB-21) and cancerous (MDA-MB-231) human mammary epithelial cells on the decellularized matrices from each diet group. Breast tissues of mice whose dams had been fed with high-fat diet exhibited larger adipocytes and thicker and curvier collagen fibers, but only slightly elevated elastic modulus and inflammatory cytokine levels. MDA-MB-231 cancer cell motility and invasion were significantly greater on the decellularized matrices from mice whose dams were fed with high-fat diet. A similar trend was observed with normal KTB-21 cells. Our results showed that the collagen curvature was the dominating factor on this enhanced motility and stretching the matrices to equalize the collagen fiber linearity of the matrices ameliorated the observed increase in cell migration and invasion in the mice that were exposed to a high-fat diet in utero. Previous studies indicated an increase in serum leptin concentration for those children born to an obese mother. We generated extracellular matrices using primary fibroblasts exposed to various concentrations of leptin. This produced curvier ECM and increased breast cancer cell motility for cells seeded on the decellularized ECM generated with increasing leptin concentration. Our study shows that exposure to obesity in utero is influential in determining the extracellular matrix structure, and that the resultant change in collagen curvature is a critical factor in regulating the migration and invasion of breast cancer cells.

Keywords: Breast cancer; Collagen structure; Invasion; Leptin; Migration; Multigenerational obesity.

Figure 1:

Characterization of model groups (A) Depiction of the four different model groups: HH, HL, LH, LL. (B) Adipocyte diameter of the model group tissues. i) Representative images showing hematoxylin and eosin (top) and Masson’s trichrome (bottom) staining of the tissues. n=2. ii) Quantification of the average adipocyte diameter from the histology images. (C) Collagen fiber structure in the matrices. i) Representative second harmonic generation images showing the collagen fibers. n=3. ii) Quantification of the average fiber diameter from the SHG images. iii) Linearity of the collagen fibers. For statistical significance, one-way ANOVA was performed for (B) and (C) followed by Tukey’s HSD, and two-way ANOVA for (B). **P<0.01, and ****P<0.0001.

Figure 2:

Migration and invasion MDA-MB-231 cells are increased on matrices obtained from mice with maternal high-fat diet. (A) Schematic showing the experimental procedure for the transwell invasion assay. (B) Average cell motility of MDA-MB-231 cells based on the fluorescence images over six hours n=2. (C) Number of cells invaded through each of the model group matrices n=3. For statistical significance, one-way ANOVA was performed for (B) and (C) followed by Tukey’s HSD. * denotes statistical significance with P<0.05, ** P<0.01, and **** P<0.0001.

Figure 3:

Stretching ECM significantly decreases the curviness of collagen fibers. (A) Schematic depicting the set-up for the stretching experiments. (B) SHG images of unstretched and stretched model group matrices. (C) Quantification of the linearity for the stretched and unstretched matrices. (D) Average cell motility of the stretched matrices based on the fluorescence images over six hours n=2. (E) Number of MDA-MB-231 cells invaded through each of the model group matrices after stretching n=3. For statistical significance, one-way ANOVA was performed for (C), (D), and (E) followed by Tukey’s HSD. **** denotes statistical significance with P<0.0001.

Figure 4:

ECM generated by primary fibroblasts cultured with higher concentrations of leptin are curvier and increase cell motility. (A) SEM images of the ECM produced with the leptin concentration listed above the image. (B) Quantification of the linearity for the ECM generated with varying leptin concentrations, n = 2 (C) Average cell motility of MDA-MB-231 cells based on the fluorescence images over six hours, n = 3. For statistical significance, one-way ANOVA was performed for (B) and (C) followed by Tukey’s HSD. * denotes statistical significance with P<0.05, ***P<0.001, and **** P<0.0001.