Milking It for All It's Worth: The Effects of Environmental Enrichment on Maternal Nurturance, Lactation Quality, and Offspring Social Behavior

Abstract

Breastfeeding confers robust benefits to offspring development in terms of growth, immunity, and neurophysiology. Similarly, improving environmental complexity, i.e., environmental enrichment (EE), contributes developmental advantages to both humans and laboratory animal models. However, the impact of environmental context on maternal care and milk quality has not been thoroughly evaluated, nor are the biological underpinnings of EE on offspring development understood. Here, Sprague Dawley rats were housed and bred in either EE or standard-housed (SD) conditions. EE dams gave birth to a larger number of pups, and litters were standardized and cross-fostered across groups on postnatal day (P)1. Maternal milk samples were then collected on P1 (transitional milk phase) and P10 (mature milk phase) for analysis. While EE dams spent less time nursing, postnatal enrichment exposure was associated with heavier offspring bodyweights. Milk from EE mothers had increased triglyceride levels, a greater microbiome diversity, and a significantly higher abundance of bacterial families related to bodyweight and energy metabolism. These differences reflected comparable transcriptomic changes at the genome-wide level. In addition to changes in lactational quality, we observed elevated levels of cannabinoid receptor 1 in the hypothalamus of EE dams, and sex-dependent and time-dependent effects of EE on offspring social behavior. Together, these results underscore the multidimensional impact of the combined neonatal and maternal environments on offspring development and maternal health. Moreover, they highlight potential deficiencies in the use of "gold standard" laboratory housing in the attempt to design translationally relevant animal models in biomedical research.

Keywords: RNA-sequencing; environmental enrichment; maternal brain; microbiome; milk quality; postnatal experience.

Figure 1.

Maternal care behaviors are different between EE and SD Sprague Dawley rat dams. A, Timeline of experimental procedures. B, Representative photographs of EE housing and litters. C, Representative photographs of SD housing and litters. D, Average number of pups born (male, female, and total pups) per SD and EE housing group (n = 8). E, Total time (seconds) SD and EE housed dams spent exploring P7 alien pups from different housing conditions (n = 9). Percent of time that dams spent on the nest across P1–P4 and P10 in the (F) light, (G) dark, and (H) light + dark periods combined. Total time (seconds) that dams spent on the nest collapsed across P1–P4 and P10 in the (I) light, (J) dark, and (K) light + dark periods combined. Stacked bars depict the frequency of pup directed nursing behaviors (arched back nursing, blanked/passive nursing, total nursing) collapsed across P1–P4 and P10 in the (L) light, (M) dark, and (N) light + dark periods. Stacked bars depict the frequency of other types of pup directed behaviors (licking/grooming, pup retrievals, nest building behaviors) collapsed across P1–P4 and P10 in the (O) light, (P) dark, and (Q) light + dark periods. Stacked bars depict the frequency of maternal self-directed behaviors (self-grooming, eating/drinking, tail/arm chases) collapsed across P1–P4 and P10 in the (R) light, (S) dark, and (T) light + dark periods (n = 8). Data are expressed as mean ± SEM; SD: open circles versus EE: closed circles. *p < 0.05, **p < 0.01, ***p < 0.001, SD versus EE; ^aap < 0.01, main effect of housing; ^bp < 0.05, main effect of postnatal day. See Extended Data Figure 1-1.

Figure 2.

Nutritional profile of milk and microbiome community distribution in EE and SD Sprague Dawley rat dams. A, Photograph depictions of maternal milk collection. Maternal milk concentrations (n = 8) of (B) % creamatocrit, (C) fat (g/l), (D) energy value (kcal/l), (E) protein (mg/ml), (F) lactose (ng/μl), (G) corticosterone (ng/ml), (H) IgA (mg/ml), and (I) triglycerides (mg/dl). Microbiome sequencing (n = 6) is demonstrated with (J) α diversity along the Shannon index and (K) β diversity using principle coordinate analysis (PCoA). This plot was created using the matrix of pair-wise distance between samples determined by Bray–Curtis dissimilarity using unique amplicon sequencing variants. Each dot represents an individual microbial profile. Samples that are closer together are more similar, while samples that are dissimilar are plotted further away from one another. L, Microbial composition of taxonomy in maternal milk at the family level for SD and EE-housed dams. M, Microbiome biomarkers plot. Taxa identified as significantly more abundant in the milk of the housing group where a bar appears; SD mothers (blue bars) and EE mothers (red bars). Significance was determined by LEfSe analysis, which identified taxa with distributions that were statistically significant (p < 0.05) and where the effect size (LDA score) was >2. Data are expressed as mean ± SEM; SD: open circles versus EE: closed circles. *p < 0.05; **p < 0.01, SD versus EE; ^bbbp < 0.001, main effect of postnatal day. See Extended Data Figure 2-1.

Figure 3.

Transcriptomic analyses of P10 milk samples from rat mothers living in EE or standard housing (SD). A, Volcano plot depicting the distribution of 756 genes based on log2 FC and -log10 p values. Each gray dot is a gene, and 110 dots highlighted in blue represent genes that displayed the highest magnitude of significance (padj < 0.05, FC > 1.3). Heatmaps of differentially expressed genes related specifically to milk (B) triglyceride concentration, (C), nutrient transport, (D), oxytocin signaling, (E) GR signaling, (F) GR binding, and (G) epigenetics. Gene expression is represented with the log2 transformation of counts recorded with a z score based on the average across experimental groups. Data are expressed as *p < 0.05 or **adjusted p < 0.05, with FC > 1.3. GR, glucocorticoid receptor. See Extended Data Figures 3-1 and 3-2.

Figure 4.

Juvenile offspring physiology and behavior following housing in EE or SD laboratory conditions. Data for (left side) male and (right side) female Sprague Dawley rats for (A, B) P21 body weights (grams). C, D, Total distance traveled (centimeters), and (E, F) percent of time spent in the center of an open field. G, H, Percent of time spent in social interaction, and (I, J) latency (seconds) to approach a novel rat in a social preference test (n = 7–8). Data are expressed as mean ± SEM; ^aap < 0.01, ^aaap < 0.001, main effect of prenatal experience (SD vs EE); ^bp < 0.05, ^bbp < 0.01, main effect of postnatal experience (SD vs EE). See Extended Data Figures 4-1 and 4-2.

Figure 5.

Summary of findings and proposed mechanisms.