Rat Mammary carcinoma susceptibility 3 (Mcs3) pleiotropy, socioenvironmental interaction, and comparative genomics with orthologous human 15q25.1-25.2

Abstract

Genome-wide association studies of breast cancer susceptibility have revealed risk-associated genetic variants and nominated candidate genes; however, the identification of causal variants and genes is often undetermined by genome-wide association studies. Comparative genomics, utilizing Rattus norvegicus strains differing in susceptibility to mammary tumor development, is a complimentary approach to identify breast cancer susceptibility genes. Mammary carcinoma susceptibility 3 (Mcs3) is a Copenhagen (COP/NHsd) allele that confers resistance to mammary carcinomas when introgressed into a mammary carcinoma susceptible Wistar Furth (WF/NHsd) genome. Here, Mcs3 was positionally mapped to a 7.2-Mb region of RNO1 spanning rs8149408 to rs107402736 (chr1:143700228-150929594, build 6.0/rn6) using WF.COP congenic strains and 7,12-dimethylbenz(a)anthracene-induced mammary carcinogenesis. Male and female WF.COP-Mcs3 rats had significantly lower body mass compared to the Wistar Furth strain. The effect on female body mass was observed only when females were raised in the absence of males indicating a socioenvironmental interaction. Furthermore, female WF.COP-Mcs3 rats, raised in the absence of males, did not develop enhanced lobuloalveolar morphologies compared to those observed in the Wistar Furth strain. Human 15q25.1-25.2 was determined to be orthologous to rat Mcs3 (chr15:80005820-82285404 and chr15:83134545-84130720, build GRCh38/hg38). A public database search of 15q25.1-25.2 revealed genome-wide significant and nominally significant associations for body mass traits and breast cancer risk. These results support the existence of a breast cancer risk-associated allele at human 15q25.1-25.2 and warrant ultrafine mapping of rat Mcs3 and human 15q25.1-25.2 to discover novel causal genes and variants.

Keywords: body mass; breast cancer susceptibility; genotype by environment interaction; human 15q25.1-25.2; mammary gland development; rat Mcs3.

Fig. 1.

Map of WF.COP congenic alleles delimiting Mcs3 to 7.2 Mb of R. norvegicus chromosome 1 (RNO1). Informative WF/COP rat genomic markers are listed on the vertical axis and placed in relative position along RNO1. Congenic strains (WF.COP) are represented by the respective COP allele (colored vertical bars), a strain-specific letter, and the WF.COP strain from which it was derived (superscripted letter). WF.COP strains with a mammary carcinoma resistance phenotype are denoted by red bars (strains D and J), while susceptible strains are shown with yellow bars (strains H, I, and K). Gray bars represent genomic regions of unknown genotype. WF.COP strain D was published previously by Le et al. as an Mcs3 containing strain.

Fig. 2.

Mammary carcinoma susceptibility phenotypes. Mean ± SD mammary tumor multiplicity for WF.COP and susceptible control WF/NHsd strains. Mammary tumors were counted at 15 weeks post-DMBA administration. Dunn’s nonparametric post hoc tests, comparing each line to the WF/NHsd strain phenotype, were performed following a significant Kruskal–Wallis test (P < 0.0001). Strain J (WF.COP-Mcs3^J) was the only strain that was significantly different from the WF/NHsd strain (P < 0.0001).

Fig. 3.

Histology of WF.COP congenic strain mammary tumors. All WF.COP strains tested developed expected DMBA-induced invasive cribriform and papillary carcinomas. Images are H&E-stained DMBA-induced representative mammary tumors forming in WF/NHsd and WF.COP strains. The top row is representative cribriform patterns, defined by neoplastic epithelial cell tumor nests (arrow A) within a surrounding desmoplastic stroma (arrow B). The bottom row contains representative papillary patterns, defined by fibrovascular cores that form a network throughout the tumor (arrow C), with neoplastic epithelial cells growing to expand the lesion via papillary projections (arrow D). Images were taken at 40× magnification on an Aperio ImageScope CS2.

Fig. 4.

Effects of Mcs3 and social environment on body mass. a) Mammary carcinoma-resistant WF.COP-Mcs3^J females had lower body mass than susceptible WF/NHsd strain females at 23 weeks of age, which was when mammary carcinoma susceptibility phenotypes were measured (tumor multiplicity). A 1-way ANOVA was performed (P < 0.0001), followed by Dunnetts’s post hoc tests comparing each strain to the control WF/NHsd strain. Strain J was significantly different from WF/NHsd (P < 0.0001). b) The effect of Mcs3 on female body mass was lost when females were raised in social environment containing males. Male and female rats from WF/NHsd and WF.COP-Mcs3^J strains were weighed at weaning age (4 weeks), puberty (8 weeks), and breeding age (12 weeks). WF.COP-Mcs3^J males had significantly reduced body mass compared to WF/NHsd males at 8 and 12 weeks of age (2-way ANOVA followed by Tukey’s multiple comparisons post hoc test, P < 0.0001), but no differences in body mass were found between females at 4, 8, and 12 weeks of age. c) The effect of Mcs3 on body mass was evident at 8 and 12 weeks of age when, at weaning, females were placed in a social environment that did not contain males (2-way ANOVA, P < 0.0001).

Fig. 5.

Mammary gland development of WF/NHsd and WF.COP-Mcs3^J females raised in the presence of males appears normal. a–c) Representative H&E staining of virgin WF/NHsd (left) and WF.COP-Mcs3^J (right) mammary glands at 4, 8, and 12 weeks of age. Images were captured at 40× magnification on an Aperio ImageScope CS2. a, b) Histology at 4 and 8 weeks is mixed population of TEBs and ductal structures embedded throughout an adipocyte matrix in both strains. c) Histology at 12 weeks is majority of mature ducts, with loss of TEBs, and minimal lobuloalveolar structures, which is representative of adult rat virgin mammary glands. d) Quantitative analysis of % lobuloalveolar morphology in mammary gland cross-sections at 12 weeks of age revealed no difference in lobuloalveolar area between WF/NHsd and WF.COP-Mcs3^J mammary glands [2-tailed unpaired t-test P > 0.05 (ns)].

Fig. 6.

Effect of Mcs3 on mammary gland development of females raised in the absence of males. Mammary carcinoma susceptible WF/NHsd, but not resistant WF.COP-Mcs3^J females had enhanced lobuloalveolar mammary gland histology at 12 weeks of age when raised in the absence of males. a–c) Representative H&E staining of WF/NHsd (left) and WF.COP-Mcs3^J (right) mammary glands at 4, 8, and 12 weeks of age. Images were saved at 40× magnification using an Aperio ImageScope CS2. a) Mammary gland histology at 4 weeks is mixed TEBs and ductal structures embedded throughout an adipocyte matrix in both strains. b) Histology at 8 weeks is mixed TEB structures and maturing ductal structures with more pronounced stroma than at 4 weeks of age in both strains. c) Histological analysis at 12 weeks of age reveals Mcs3^J females have mature ducts and minimal lobuloalveolar structures (black arrows), which are representative of an adult virgin mammary gland. On the other hand, age-matched WF females display greatly enhanced lobuloalveolar structures. d) Quantitative analysis of % lobuloalveolar morphology in mammary glands at 12 weeks of age revealed increased lobuloalveolar area in susceptible WF compared to Mcs3^J females (n = 3; 2-tailed unpaired t-test P < 0.0001).

Fig. 7.

Transcript map of rat Mcs3 and orthologous overlap with human 15q25.1-2, a potential breast cancer risk locus. Map of Mcs3 locus transcripts annotated by the UCSC rat genome browser (build version RGSC 6.0/rn6). The horizontal axis represents RNO1 from bp 143,700,228–150,929,594, to which Mcs3 was delimited. Mammary carcinoma-resistant WF.COP-Mcs3^J (red bar) and susceptible WF.COP-I (yellow arrow with a gray bar extension representing a segment of undetermined WF/COP genotype) are shown for reference. Vertical dashed lines demark the terminal ends of Mcs3 (dark blue bar labeled “Rat Mcs3 - 7.2 Mb”) and the respective rat genomic region that contains Mcs3 overlap with orthologous human 15q25.1-25.2 (light blue bar labeled “Human 15q25.1 - 25.2”). The lollipop within 15q25.1-25.2 marks the approximate location of associated breast cancer risk variant rs6495623.

Fig. 8.

Lollipop maps of Mcs3 orthologous human 15q25.1-25.2 and GWAS-nominated variants associated with body mass and other breast cancer-correlated traits. a) Relative location of human 15q25.1-25.2 syntenic regions orthologous to rat Mcs3 (red boxes). b) Orthologous syntenic region 1 contains portions of 15q25.1 and 15q25.2 (chr15:80005820–82285404). c) Orthologous syntenic region 2 contains part of 15q25.2 (chr15:83134545–84130720). Black horizontal bars represent the genomic track with intervals indicating genomic position beneath. Each lollipop color represents a different trait as indicated in the legend inset. Variants with multiple trait associations share a stem.