BACKGROUND
This study examined whether different degrees of maternal food restriction during pregnancy alter mammary gland development in offspring during the embryonic and immediate postnatal period. The authors argue that embryonic life, puberty, and pregnancy are the 3 major stages of mammary development, but fetal mammary development has received relatively little attention compared with placenta or brain development under maternal undernutrition. In mice, mammary development begins on day 10 of embryonic life, and before birth the gland forms a small ductal tree with a dominant duct and 10–15 branches embedded in the nascent fat pad. The authors hypothesized that mild and severe nutrient restriction may have different effects because the placenta can buffer limited maternal dietary fluctuation, while mammary stem-cell-associated pathways may respond differently across levels of restriction.
METHODS
The experiment used 60 female 8-week-old CD-1(ICR) mice. After mating, the day a vaginal plug was identified was defined as day 0 of gestation. To reduce effects on very early embryonic growth, nutritional restriction started on day 9 of pregnancy. Pregnant mice were assigned to 5 groups (n = 12): 100% ad libitum intake, or 90%, 80%, 70%, and 60% of the ad libitum food weight daily. Maternal body weight was recorded daily. On the day of delivery, the number of litters, maternal body weight, and offspring body weight were recorded. Maternal and female offspring body fat percentage was measured by dual-energy X-ray absorptiometry. Female offspring were euthanized and the #4 inguinal mammary glands were processed for whole mount analysis; other mammary glands were stored for qPCR. Whole-mount analysis was quantified with Sholl analysis using a radius step size of 0.02 mm. The investigators measured enclosing radius, mammary epithelial area (MEA), sum inters, Sholl regression coefficient (k), and branch density. They also performed principal component analysis (PCA) using enclosing radius, MEA, sum inters, and k, plus linear regression analyses using enclosing radius against MEA, sum inters, and k. Gene-expression testing by qPCR included development-related and stem-cell-related genes (Sox10, Axin2, Elf5, Lgr5, Wnt5a, Aldh1a1, Procr), basal and luminal markers (K5, K18), and hormone receptors (ERα, ERβ, PR). Statistical analysis used one-way ANOVA with Sidak correction.
KEY RESULTS
Maternal restriction affected maternal body weight rapidly. A significant difference in body weight was observed from day ten of pregnancy and persisted until the end of gestation (p < 0.05). After delivery, adult female mice had significantly lower body weight than controls in all restricted groups except the 90% group (p < 0.05 for restricted groups other than 90%; p > 0.05 for 90%). Maternal body fat percentage after parturition did not differ significantly among groups (p > 0.05).
For offspring, the clearest whole-animal effects appeared with the most severe restriction. Only the 60% group had a significant decrease in litter size versus control (p < 0.05). Individual offspring weight in the 60% group was significantly lower than in the other groups (p < 0.05). Offspring body fat percentage was significantly higher in the control group than in the 80%, 70%, and 60% groups (p < 0.05), indicating that body fat responded earlier than body weight to maternal restriction.
Mammary whole-mount and Sholl analysis showed a threshold-like pattern. Enclosing radius was significantly greater in the control group than in the 70% and 60% groups (p < 0.05), and significantly greater in the 90% group than in the 60% group (p < 0.05). MEA did not differ among the control, 90%, and 80% groups, but it was significantly lower in the 70% and 60% groups than in those first 3 groups (p < 0.05). Sum inters was significantly higher in the control, 90%, and 80% groups than in the 70% and 60% groups (p < 0.05). The k value was significantly higher in the 70% and 60% groups than in the other 3 groups (p < 0.05), consistent with less complex distal branching. Branch density did not differ significantly among groups (p > 0.05). Visually and quantitatively, the major decline occurred between the 80% and 70% groups. At a radius of 0.5 mm, the control, 90%, and 80% groups reached similar peak intersection numbers, all higher than the 70% and 60% groups.
PCA supported this threshold effect. The first 2 principal components explained 79.1% and 10.2% of the variance, respectively. Control, 90%, and 80% samples clustered nearer to each other, while the 70% and 60% groups were more separated. The variable with the highest weight in the first principal component was enclosing radius, suggesting that reduced ductal extension was the main driver of the developmental shift between 80% and 70%. The PCA biplot showed a positive correlation of enclosing radius with MEA and sum inters and a negative correlation with k.
Gene-expression results suggested that mild restriction may stimulate some developmental programs. In the 90% group, Sox10 expression was significantly higher than in the other 4 groups (p < 0.05), and Elf5 was significantly higher than in the control and 60% groups (p < 0.05). Sox10 was significantly lower in the control group than in the 90%, 70%, and 60% groups (p < 0.05). Axin2 was significantly higher in the control group than in the 60% group (p < 0.05). Aldh1a1 was significantly higher in the 80% group than in the 60% group (p < 0.05). K5 expression was significantly higher in the control group than in the 80%, 70%, and 60% groups (p < 0.05). In the 60% group, ER1 was significantly lower than in the 90% group (p < 0.05), and ER2 was significantly lower than in the control group (p < 0.05). Other tested genes did not differ significantly (p > 0.05).
CLINICAL IMPLICATIONS
This was a mechanistic animal study rather than a clinical trial, so it should not be used directly to set human pregnancy nutrition targets. Its main translational message is that fetal mammary development may not respond linearly to maternal undernutrition; instead, mild restriction such as 90% of ad libitum intake may be tolerated or even associated with favorable developmental gene expression, whereas more severe restriction approaching 70% appears to cross a developmental threshold. The work is potentially relevant to developmental programming, lactation biology, and animal agriculture, but human clinical significance remains indirect and requires confirmation in species and study designs that better model human pregnancy and later-life breast development.