Fetal malnutrition-induced catch up failure is caused by elevated levels of miR-322 in rats

Abstract

If sufficient nutrition is not obtained during pregnancy, the fetus changes its endocrine system and metabolism to protect the brain, resulting in a loss of body size. The detailed mechanisms that determine the success or failure of growth catch-up are still unknown. Therefore, we investigated the mechanism by which catch-up growth failure occurs. The body weights of rat pups at birth from dams whose calorie intake during pregnancy was reduced by 40% were significantly lower than those of controls, and some offspring failed to catch up. Short-body-length and low-bodyweight rats showed blood IGF-1 levels and mRNA expression levels of IGF-1 and growth hormone receptor (GHR) in the liver that were lower than those in controls. The next generation offspring from low-bodyweight non-catch-up (LBW-NCG) rats had high expression of miR-322 and low expression of GHR and IGF-1. The expression of miR-322 showed a significant negative correlation with GHR expression and body length, and overexpression of miR-322 suppressed GHR expression. We found that insufficient intake of calories during pregnancy causes catch-up growth failure due to increased expression of miR-322 and decreased expression of GHR in the livers of offspring, and this effect is inherited by the next generation.

Figure 1

Changes in maternal rat weight due to calorie restriction during pregnancy and the weight of offspring. (A) Changes in body weights of mother rats fed a low-carbohydrate, calorie-restricted diet throughout pregnancy. (B) Comparison of bodyweights of rat offspring on the day of birth. (C,D), Body length and weight at the weaning day. LC indicates low carbohydrate, calorie-restricted mother rats and their offspring. Paired t test (A) and unpaired t tests (B–D) were used. Data shown are means ± SEM.

Figure 2

Comparison of hormones in blood and liver of female rats on the weaning day. Concentrations of blood GH (A), blood IGF-1 (B), tissue IGF-1 (C), mRNA expression levels of IGF-1 (D) and GH receptor (E), and protein expression of GHR (F) in livers of female rats at weaning day were quantified. To normalize each sample for RNA and protein content, GAPDH and beta actin were used, respectively. NBW, control rats (normal birth weight); LBW-CG, catch-up growth offspring from low carbohydrate, calorie-restricted dams (low birth weight); LBW-NCG, non-catch-up growth offspring from low carbohydrate, calorie-restricted dams. One-way ANOVA followed by Turkey’s post hoc test was used. Data shown are means ± SEM.

Figure 3

Quantification of miR-322 in liver and blood exosomes. (A) All five miRNAs (miR-15b, miR-16, miR-195, miR-322, and miR-497) were quantified from RNA extracted from the livers of rat offspring on the day of weaning (n = 5). *Indicates p < 0.05. (B) miR-322 was quantified from RNA extracted from the livers of female rat offspring on the day of weaning (n = 10). (C) exosomes were extracted from serum of weaning rats, and miR-322 in the exosomes was quantified (n = 10). (D) Other microRNAs (miR-192, miR-142-3p, and miR-2020) were quantified from RNA extracted from the livers of rat offspring on the day of weaning (n = 5). To normalize each sample for RNA content, U6 snRNA was used. NBW, a control rat (normal birth weight); LBW-CG, catch-up growth offspring from low carbohydrate, calorie-restricted dams (low birth weight); LBW-NCG, non-catch-up growth offspring from low carbohydrate, calorie-restricted dams. Unpaired t tests were used. Data shown are means ± SEM.

Figure 4

Birth weight and body length and body weight at the weaning day of next-generation offspring. Low-birthweight and non-catch-up growth rats from low carbohydrate, calorie-restricted dams were mated as fathers and/or mothers. Pregnant dams were fed a standard diet ad libitum. Body weights at the day of birth (A), and body lengths (B) and body weights (C) at the weaning day of their offspring were measured. One-way ANOVA followed by Turkey’s post hoc test was used. Data shown are means ± SEM.

Figure 5

Analysis of blood hormone levels and gene expression in the livers of next-generation offspring. Blood growth hormone (A) and IGF-1 (B) levels in female rat offspring on the day of weaning, and expression levels of IGF-1 mRNA (C), GH receptor mRNA (D), GH receptor protein (E) and miR-322 (F) in the liver were analyzed. To normalize each sample for RNA and miRNA content, GAPDH and U6 was used, respectively. To normalize each sample for protein content, beta actin was used. NBW: control rats (normal birth weight); LBW: low-birthweight rats due to calorie restriction; M: male rats (father); F: female rats (mother). One-way ANOVA followed by Turkey’s post hoc test was used. Data shown are means ± SEM.

Figure 6

Birth weights of rat offspring after the next generation. We obtained offspring of the third and fourth generations from fathers and mothers of short-body-length, low-bodyweight rats whose growth did not catch up to normal (A). As in the second generation, pregnant dams were fed a standard diet ad libitum. The concentration of serum IGF-1 (B) and the expression levels of IGF-1 mRNA (C), GH receptor mRNA (D), GH receptor protein (E), and miR-322 (F) in the liver were analyzed. To normalize each sample for RNA and miRNA content, GAPDH and U6 were used, respectively. To normalize each sample for protein content, beta actin was used. NBW, control rats (normal birth weight); LBW, low birth weight rats produced by caloric restriction during pregnancy of the F₀ generation; F₁-F₄ are the respective generations. One-way ANOVA followed by Turkey’s post hoc test (A) and unpaired t tests (B–E) were used. Data shown are means ± SEM.

Figure 7

Suppression of GH receptor expression by miR-322 in primary cultured hepatocytes. The expression level of miR-322 in rat livers was correlated with body length (A) and the expression level of GH receptor (B). Decreased GH receptor mRNA (D) and protein (E) expression level by overexpression (C) of miR-322 in primary cultured hepatocytes, while the mRNA expressions of IGF-1 (F) and IGF1R (G) were not affected by miR-322 overexpression. To normalize each sample for RNA content, GAPDH or U6 snRNA were used. To normalize each sample for protein content, beta actin was used. Pearson correlation coefficient tests (A,B) and unpaired t tests (C–G) were used. Data shown are means ± SEM.

Figure 8

Reduction of luciferase activity by co-expression with miR-322. (A) Change in luciferase activity by co-expression with a luciferase construct in which the miR-322 binding region is mutated in HEK 293 cells or a construct in which the binding region is scrambled. (B) DNA sequence of rat GH receptor 3′-UTR. The binding region of miR-322 is indicated in red. (C) Changes in luciferase activity due to co-expression of a construct in which the full-length luciferase gene and the miR-322 forced expression construct are shown in (B). Unpaired t tests was used. Data shown are means ± SEM.