The adipokine FABP4 is a key regulator of neonatal glucose homeostasis

Abstract

During pregnancy, fetal glucose production is suppressed, with rapid activation immediately postpartum. Fatty acid-binding protein 4 (FABP4) was recently demonstrated as a regulator of hepatic glucose production and systemic metabolism in animal models. Here, we studied the role of FABP4 in regulating neonatal glucose hemostasis. Serum samples were collected from pregnant women with normoglycemia or gestational diabetes at term, from the umbilical circulation, and from the newborns within 6 hours of life. The level of FABP4 was higher in the fetal versus maternal circulation, with a further rise in neonates after birth of approximately 3-fold. Neonatal FABP4 inversely correlated with blood glucose, with an approximately 10-fold increase of FABP4 in hypoglycemic neonates. When studied in mice, blood glucose of 12-hour-old WT, Fabp4-/+, and Fabp4-/- littermate mice was 59 ± 13 mg/dL, 50 ± 11 mg/dL, and 43 ± 11 mg/dL, respectively. Similar to our observations in humans, FABP4 levels in WT mouse neonates were approximately 8-fold higher compared with those in adult mice. RNA sequencing of the neonatal liver suggested altered expression of multiple glucagon-regulated pathways in Fabp4-/- mice. Indeed, Fabp4-/- liver glycogen was inappropriately intact, despite a marked hypoglycemia, with rapid restoration of normoglycemia upon injection of recombinant FABP4. Our data suggest an important biological role for the adipokine FABP4 in the orchestrated regulation of postnatal glucose metabolism.

Keywords: Adipose tissue; Endocrinology; Glucose metabolism; Metabolism.

Figure 1. Maternal, fetal, and neonatal FABP4 circulating levels.

Serum samples were collected from 22 normoglycemic pregnant women and 18 women with GDM. All samples were collected at term, immediately before delivery. (A) FABP4 serum concentrations were determined using an ELISA assay and were correlated to BMI. (B) Comparison between FABP4 levels in normoglycemic pregnant woman (n = 22) and women with GDMA1 (n = 10) or GDMA2 (n = 8). (C) Serum samples, collected from umbilical artery (n = 22) and vein (n = 22) immediately after delivery of normoglycemic women were analyzed for FABP4 levels and compared with (normoglycemic) maternal (n = 22) concentrations. (D) Serum samples collected from neonates within the first few hours of life (n = 24) were analyzed for FABP4 levels and compared with fetal (n = 22) and maternal levels (n = 22). (E) FABP4 levels stratified to neonates who were small (SGA) (n = 7), appropriate (AGA) (n = 48), and large (LGA) (n = 5) for gestational age. (F) Birth weight of 40 neonates was correlated to FABP4 serum concentrations. Statistical analysis includes Spearman’s correlation test (A and F) and 1-way ANOVA (B–E). Data are shown as the mean ± SEM. *P < 0.05.

Figure 2. Neonatal FABP4 levels inversely correlated with blood glucose and are elevated among hypoglycemic neonates.

Serum samples collected from 21 normoglycemic and 29 hypoglycemic neonates (as detailed in Table 2) within the first few hours of life. (A) FABP4 circulating levels in neonates were correlated with blood glucose levels. Comparison of serum FABP4 between 31 normoglycemic and 28 hypoglycemic neonates is presented in (B). (C and D) C-peptide circulating levels in neonates were compared and analyzed between normoglycemic (n = 10) and hypoglycemic (n = 13) neonates (D) and insulin circulating levels in neonates were compared and analyzed between normoglycemic (n = 19) and hypoglycemic neonates (n = 21) (C). Statistical analysis was performed by nonparametric correlations using Spearman’s correlation test (A) and by Student’s t test (B–D). Data are shown as the mean ± SEM. *P < 0.05.

Figure 3. Levels of FABP4 and its effect on blood glucose, insulin, and hormonal network in mouse neonates.

Plasma samples were collected from 9 WT mouse neonates within 12 hours from birth and 7 adult WT female mice. (A) FABP4 concentrations were determined using an ELISA assay and were compared between neonates and adult mice. (B) Mice heterozygous for Fabp4-null mutation were crossbred, and the offspring’s fasting blood glucose levels were tested 12 hours following birth. Levels of Fabp4^WT (n = 19), heterozygous (n = 35), and Fabp4^–/– (n = 15) mice were compared. (C) Insulin (of 11 Fabp4^WT and 11 Fabp4^–/– mouse neonates), (D) glucagon (of 3 Fabp4^WT and 4 Fabp4^–/– mouse neonates), (E) catecholamines (of 10 Fabp4^WT and 10 Fabp4^–/– mouse neonates), and (F) corticosterone (of 9 Fabp4^WT and 8 Fabp4^–/– mouse neonates) plasma levels were compared between Fabp4^WT and Fabp4^–/– mouse neonates within 12 hours following birth. Statistical analysis was performed by Student’s t test (A and C–F) and by 1-way ANOVA test (B). Data are shown as the mean ± SEM. *P < 0.05, **P < 0.01. FABP4^WT; WT, HET; heterozygous.

Figure 4. Hepatic glucose production in Fabp4^WT and Fabp4^–/– mouse neonates.

Expression levels of key gluconeogenic genes in livers of mouse neonates: phosphoenolpyruvate carboxykinase 1 (Pck1) (n = 25 for Fabp4^WT and 23 Fabp4^–/–) (A) and glucose-6-phosphatase (G6pc) (n = 23 for Fabp4^WT and 22 Fabp4^–/–) (B). Statistical analysis was performed by Student’s t test. Data are presented as mean ± SEM. *P < 0.05.

Figure 5. Differential hepatic gene expression analysis of FABP4-deficient mice.

Differential hepatic gene expression analysis showing the effect of the FABP4-deficient state on various lipid- and carbohydrates-related pathways and the over/underexpression of genes in each related pathway. (A) Heatmap. (B) Bar chart. (C) The effect of FABP4 on genes that are known to be regulated in glucagon-treated hepatocytes. n = 3 in each study group.

Figure 6. Glycogen content of Fabp4^WT and Fabp4^–/– mouse neonates and recombinant FABP4 effect on blood glucose.

Fabp4^WT and Fabp4^–/– liver and muscle (left thigh) tissues were harvested and snap frozen at liquid nitrogen. (A and B) Glycogen content was determined in Fabp4^WT (n = 19) and Fabp4^–/– (n = 18) mouse neonatal liver (A) and Fabp4^WT (n = 5) and Fabp4^–/– (n = 5) mouse neonatal muscle (B) tissues using the Glycogen Assay Kit. (C) The ratio between blood glucose and liver glycogen content in each subject was compared between Fabp4^WT (n = 14) and Fabp4^–/– (n = 14) neonates. Recombinant FABP4 or saline was injected in Fabp4^–/– mice (n = 7 for FABP4 and 8 for saline injected) within 12 hours from delivery. (D) Blood glucose was measured from mouse tail vein at 0, 15, and 30 minutes following injections. Statistical analysis was performed by Student’s t test. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01.