Altered gut microbiota in female mice with persistent low body weights following removal of post-weaning chronic dietary restriction

Abstract

Background: Nutritional interventions often fail to prevent growth failure in childhood and adolescent malnutrition and the mechanisms remain unclear. Recent studies revealed altered microbiota in malnourished children and anorexia nervosa. To facilitate mechanistic studies under physiologically relevant conditions, we established a mouse model of growth failure following chronic dietary restriction and examined microbiota in relation to age, diet, body weight, and anabolic treatment.

Methods: Four-week-old female BALB/c mice (n = 12/group) were fed ad libitum (AL) or offered limited food to abolish weight gain (LF). A subset of restricted mice was treated with an insulin-like growth factor 1 (IGF1) analog. Food access was restored in a subset of untreated LF (LF-RF) and IGF1-treated LF mice (TLF-RF) on day 97. Gut microbiota were determined on days 69, 96-99 and 120 by next generation sequencing of the V3-5 region of the 16S rRNA gene. Microbiota-host factor associations were analyzed by distance-based PERMANOVA and quantified by the coefficient of determination R² for age, diet, and normalized body weight change (Δbwt). Microbial taxa on day 120 were compared following fitting with an overdispersed Poisson regression model. The machine learning algorithm Random Forests was used to predict age based on the microbiota.

Results: On day 120, Δbwt in AL, LF, LF-RF, and TLF-RF mice was 52 ± 3, -6 ± 1*, 40 ± 3*, and 46 ± 2 % (*, P < 0.05 versus AL). Age and diet, but not Δbwt, were associated with gut microbiota composition. Age explained a larger proportion of the microbiota variability than diet or Δbwt. Random Forests predicted chronological age based on the microbiota and indicated microbiota immaturity in the LF mice before, but not after, refeeding. However, on day 120, the microbiota community structure of LF-RF mice was significantly different from that of both AL and LF mice. IGF1 mitigated the difference from the AL group. Refed groups had a higher abundance of Bacteroidetes and Proteobacteria and a lower abundance of Firmicutes than AL mice.

Conclusions: Persistent growth failure can be induced by 97-day dietary restriction in young female mice and is associated with microbiota changes seen in lean mice and individuals and anorexia nervosa. IGF1 facilitates recovery of body weights and microbiota.

Keywords: Animal model; Anorexia nervosa; Dietary restriction; Gut microbiota; Insulin-like growth factor 1 (IGF1); Machine learning; Protein-energy malnutrition.

Fig. 1

Low body weights persist following correction of chronic dietary restriction initiated at post-weaning in female mice. a Time course of normalized body weight changes from day 0 (Δbwt) representing 4 weeks of age (n = 12 mice/group). AL ad-libitum-fed cohort, LF limited-fed mice subjected to dietary restriction titrated to prevent weight gain, LF-RF limited-fed-refed mice representing a subset of LF mice given unrestricted access to food following 97 days of dietary restriction, TLF-RF a subset of LF-RF mice treated with twice daily subcutaneous injections of LONG R³ recombinant human insulin-like growth factor 1 (LONG R³ rhIGF1), a potent IGF1 analog with reduced affinity for IGF-binding proteins, from day 13 of the study. b Time period identified by gray shading in A. Vertical lines indicate feces collection. c One-week average body weight changes centered on the day of the last feces collection (days 117–123). *, P < 0.05 by Student-Newman-Keuls multiple comparison tests. n.s. not significant. d Average food intake of TLF-RF mice between days 0 and 96 expressed as the percentage of average food intake of LF mice over the same period. n.s. not significant. e Two-day average food intake determined on days 117 and 123. *, P < 0.05 by Dunn’s multiple comparison tests. n.s. not significant. LF mice weighed ~60 % less than AL controls after 167 days of dietary restriction. Body weights did not recover for at least 10 weeks after ad libitum refeeding despite comparable food intake. LONG R³ rhIGF1 facilitated body weight recovery

Fig. 2

Age explains more microbiota variability than diet and body weight. a The first two PCs from the PCA on the unweighted UniFrac distance matrix are plotted. Symbols represent data from individual diet regimens color-coded by sampling days. The main axes of the ellipses correspond to the PCs of the group with the heights and widths representing variances in the corresponding components. b The percentage of microbiota variability explained by age, diet type, Δbwt, and their combination (total) based on different UniFrac distances. UniFrac, GUniFrac, and WUniFrac represent unweighted, generalized (α = 0.5), and weighted UniFrac distance, respectively. Non-linear age effects are assumed

Fig. 3

Predicting mouse chronologic age based on gut microbiota using Random Forests. a Heat map of the mean relative abundance of age-discriminatory OTUs selected by the Boruta algorithm for the AL diet group. Rows represent the OTUs and columns represent the sampling day (Age). Hierarchical clustering on the left was based on complete linkage and Euclidean distance. Importance Z-scores from the Boruta alogrithm are plotted on the right. A large importance Z-score indicates stronger ability of corresponding OTU to discriminate chronological age. Green and yellow colors indicate the significance level (‘confirmed’ and ‘tentative’, respectively). b Predicting the age of the microbiota samples from the other diet groups using samples from the LF group as the training set. The y axis represents the predicted age (microbiota age) by Random Forests. Colors represent individual diet groups. Mice under dietary restriction (LF-RF and TLF-RF groups before the reintroduction of the ad libitum diet) exhibited lower microbiota ages than AL mice

Fig. 4

Altered gut microbiota community structure persists after correction of chronic dietary restriction. a–d PCA on days 69 (a), 96–97 (b), 98–99 (c), and 120 (d). The first two PCs from the PCA on unweighted UniFrac distance matrix are plotted. Symbols and colors represent data from individual diet regimens. The main axes of the ellipses correspond to the first two PCs with the height and width representing variances in the corresponding coordinates. Note that the LF-RF and TLF-RF data remained different from the AL data on day 120 despite a significant separation from the LF group; and that IGF1 treatment (TLF-RF group) mitigated the difference from the AL mice

Fig. 5

Specific bacterial taxa show hysteresis effect under chronic dietary restriction. a Cladogram generated with GraPhlAn (http://huttenhower.sph.harvard.edu/galaxy/) showing “hysteresis” bacterial taxa identified by comparing their abundance in the AL group to LF-RF and TLF-RF mice (refed groups) on day 120. Red represents abundance increase in the AL group and green represents abundance increase in the refed groups. b Log₂ fold change (refed groups/AL) of the abundance of taxa identified at an FDR of 10 %. The horizontal fuzzy line represents the 95 % confidence interval of the log fold change estimate