Deficiency of the metabolic enzyme SCHAD in pancreatic β-cells promotes amino acid-sensitive hypoglycemia

Abstract

Congenital hyperinsulinism of infancy (CHI) can be caused by a deficiency of the ubiquitously expressed enzyme short-chain 3-hydroxyacyl-CoA dehydrogenase (SCHAD). To test the hypothesis that SCHAD-CHI arises from a specific defect in pancreatic β-cells, we created genetically engineered β-cell-specific (β-SKO) or hepatocyte-specific (L-SKO) SCHAD knockout mice. While L-SKO mice were normoglycemic, plasma glucose in β-SKO animals was significantly reduced in the random-fed state, after overnight fasting, and following refeeding. The hypoglycemic phenotype was exacerbated when the mice were fed a diet enriched in leucine, glutamine, and alanine. Intraperitoneal injection of these three amino acids led to a rapid elevation in insulin levels in β-SKO mice compared to controls. Consistently, treating isolated β-SKO islets with the amino acid mixture potently enhanced insulin secretion compared to controls in a low-glucose environment. RNA sequencing of β-SKO islets revealed reduced transcription of β-cell identity genes and upregulation of genes involved in oxidative phosphorylation, protein metabolism, and Ca²⁺ handling. The β-SKO mouse offers a useful model to interrogate the intra-islet heterogeneity of amino acid sensing given the very variable expression levels of SCHAD within different hormonal cells, with high levels in β- and δ-cells and virtually absent α-cell expression. We conclude that the lack of SCHAD protein in β-cells results in a hypoglycemic phenotype characterized by increased sensitivity to amino acid-stimulated insulin secretion and loss of β-cell identity.

Keywords: HADH; SCHAD; amino acids; congenital hyperinsulinism of infancy; hypoglycemia; insulin secretion; islets; knockout mouse model; short-chain 3-hydroxyacyl-CoA dehydrogenase; transcriptomics; β-cell; β-cell dedifferentiation.

Figure 1

Validation of disrupted SCHAD expression in β-SKO mice.A, PCR assay and agarose gel electrophoresis for detection of the floxed and knockout Hadh alleles in islets, liver, skeletal muscle, and hypothalamus of littermate control and β-SKO mice. Recombination was assessed using the primers P1, P2, and P3 described in the Experimental procedures section. The P1-P2 combination yields a PCR product of 600 bp length only when exon 3, which includes the binding site for P2, is intact. The P1-P3 combination yields a 450 bp PCR product only when exon 3 is deleted. B, Western blots for detection of SCHAD protein expression in lysates of isolated islets and whole pancreas of control and β-SKO mice. Lysates from wild-type mice (WT) and the global Hadh knockout mouse (SCHADKO) were included as positive and negative controls, respectively. Detection of tubulin alpha-1A chain (αTub) served as loading control. C, immunohistochemistry performed on pancreas sections using antibodies against insulin (red) or SCHAD (green). Representative islets from control, SCHADKO, and β-SKO mice are shown. White arrows indicate SCHAD-positive cells in a β-SKO islet. Nuclei are stained blue with DAPI. D, same as (C) using antibodies against glucagon (red) and SCHAD (green). E, same as (C) using antibodies against somatostatin (red) and SCHAD (green). White arrows indicate cells that are positive both for SCHAD and somatostatin in a β-SKO islet.

Figure 2

Glucose homeostasis and insulin tolerance in male control and β-SKO mice.A, plasma glucose in fed, 16-h fasted, and refed 13-week-old control and β-SKO mice (n = 6–10). B, difference (Δ) in percentage change of fasted versus refed plasma glucose in control and β-SKO mice (n = 6–7). C, food intake per mouse during the 4-h refeeding period in control and β-SKO mice (n = 5–7). D–F, plasma insulin (n = 7–10), C-peptide (n = 8–10), and glucagon (n = 7) in fed, 16-h fasted and refed control, and β-SKO mice (13 weeks old). G, GTT and AUC (area under the curve) of 10-week-old control and β-SKO mice (n = 10–11). H, same as (G) for 30-week-old mice (n = 9). I, GSIS and AUC of 12-week-old control and β-SKO mice (n = 10). J, plasma C-peptide levels for the mice in (I) (n = 10). K, ITT and AUC of 10-week-old control and β-SKO mice (n = 9). All data are represented as mean ± SEM. p-values: ∗ < 0.05, ∗∗ < 0.005.

Figure 3

Body and organ weights and glucose homeostasis of male control and L-SKO mice.A, bodyweight of 10-week-old control and L-SKO mice (n = 9–10). B, tissue weights of pancreas, liver, visceral (VAT), subcutaneous (SAT), and brown (BAT) adipose tissue as a percentage of body weight of 14-week-old control and L-SKO mice (n = 9–10). C and D, plasma glucose (n = 9–10) and insulin (n = 7–10) in fed and 16-h fasted control and L-SKO mice (12 weeks old). E, GTT and AUC of 10-week-old control and L-SKO mice (n = 9–10). F, ITT and AUC of 10-week-old control and L-SKO mice (n = 9–10). All data are represented as mean ± SEM.

Figure 4

Effect of amino acids on glucose homeostasis in the β-SKO model.A, random-fed plasma glucose levels and AUC of male control and β-SKO mice over the course of 10 weeks on a diet enriched in the amino acids alanine, glutamine and leucine (ED) compared with mice fed a normal diet (ND) (n = 6–9). B, fasting (16 h) plasma glucose levels of male control and β-SKO mice fed ED or ND for 14 weeks (n = 5–9). C, plasma insulin levels for the mice in (B). All data are represented as mean ± SEM. p-values: ∗ < 0.05, ∗∗∗ < 0.0005.

Figure 5

Effect of amino acids on insulin secretion in β-SKO animals and isolated islets.A, plasma glucose from intraperitoneal amino acid administration on 16-h fasted β-SKO (n = 9) and control (n = 10) mice. B, plasma insulin levels before and after the amino acid injection. C, insulin secretion from isolated control and β-SKO islets stimulated with 3.3 mM glucose (G3.3), G3.3 in combination with amino acid mixture (AA), 16.7 mM glucose (G16.7), or G16.7 in combination with AA. Data are multiplied by a factor of 100,000 for simplification (n = 5–6). D, insulin content of isolated control and β-SKO islets (n = 8). All data are represented as mean ± SEM. p-values: ∗ < 0.05, ∗∗ < 0.005.

Figure 6

RNA sequencing of islets isolated from male control and β-SKO mice.A, heat map representation of mRNAs differentially expressed between islets isolated from control and β-SKO mice (10 weeks old). Data are shown as Z-scores to indicate the deviation from the group’s mean value. Significantly differentially expressed genes are presented in rows with high to low expression being represented by a change of color from red to blue. B, top upregulated pathways in β-SKO islets identified using over-representation analysis of the top 100 up-regulated genes. Y-axis denotes regulated pathways. X-axis represents their log10 p-value. C, top downregulated pathways in β-SKO islets identified using over-representation analysis of the top 100 down-regulated genes. Axes as in (B). D–H, differentially expressed genes related to β-cell identity, calcium binding and signaling, and cell adhesion. Gene names are indicated on the left. Increased and decreased gene expression are indicated in red and blue color, respectively. The X-axis represents their log10 p-value. FC, fold change.