Precision-cut liver slices as an ex vivo model to assess impaired hepatic glucose production

Abstract

Fasting hypoglycemia is a severe and incompletely understood symptom of various inborn errors of metabolism (IEM). Precision-cut liver slices (PCLS) represent a promising model for studying glucose production ex vivo. This study quantified the net glucose production of human and murine PCLS in the presence of different gluconeogenic precursors. Dihydroxyacetone-supplemented slices from the fed mice yielded the highest rate, further stimulated by forskolin and dibutyryl-cAMP. Moreover, using ¹³C isotope tracing, we assessed the contribution of glycogenolysis and gluconeogenesis to net glucose production over time. Pharmacological inhibition of the glucose 6-phosphate transporter SLC37A4 markedly reduced net glucose production and increased lactate secretion and glycogen storage, while glucose production was completely abolished in PCLS from glycogen storage disease type Ia and Ib patients. In conclusion, this study identifies PCLS as an effective ex vivo model to study hepatic glucose production and opens opportunities for its future application in IEM research and beyond.

Fig. 1. Glucose production.

a Simplified scheme of the main glucose production pathways: 1. Glycogenolysis and 2. Gluconeogenesis; b Experimental design and conditions; NP no precursor, DHA dihydroxyacetone, L/P lactate/pyruvate; c–f Without 24 h pre-incubation: 5 h net cumulative glucose production of PCLS from fed (c) and fasted (d) mice; e Net glucose production rate of PCLS from fed and fasted mice; f 24 h net cumulative glucose of PCLS from fed and fasted mice; g–j With 24 h pre-incubation: 5 h net cumulative glucose production of PCLS from fed (g) and fasted (h) mice; i Net glucose production rate of PCLS from fed and fasted mice; j 24 h net cumulative glucose of PCLS from fed and fasted mice; *p < 0.05, vs NP in the same condition; ^#p < 0.05, vs DHA group in the same condition; data presented as mean ± SEM; the individual data points are depicted in Supplementary Fig. S1; n = 3–4 mice per group.

Fig. 2. Effects of hormone stimulation on glucose production.

a Simplified scheme of mechanism of action of insulin, glucagon, forskolin, and dibutyryl-cAMP; 5 h net cumulative glucose production of PCLS derived from fed (b) and fasted (c) mice incubated with WE medium without precursor treated with insulin, glucagon, forskolin or dibutyryl-cAMP; d 5 h net glucose production of PCLS derived from fed and fasted mice incubated with WE medium supplemented with DHA and treated with forskolin or dibutyryl-cAMP; HCl was used as a vehicle for insulin and glucagon, and DMSO was used as a vehicle for forskolin; *p < 0.05, vs the corresponding vehicle; data presented as mean ± SEM; n = 3 mice per group. GPCR G protein-coupled receptor, AC adenylate cyclase.

Fig. 3. Time course of effects of insulin, glucagon and forskolin on downstream signaling pathways and net glucose production.

Immunoblotting of PCLS derived from fed mice incubated with WE medium supplemented with DHA and treated with insulin (a) or glucagon (b) at different time points (0, 15, 30, 60,150, and 300 min). The complete dataset (three biological replicates) is provided in Supplementary Fig. S2. a For insulin, total Akt and pAkt-T308 were monitored by immunoblot. b For glucagon, phosphorylation of PKA substrates was monitored by immunoblot. c–e Quantification of data shown in (a); total Akt (c), pAkt-T308 (d) and ratio pAkt-T308/ total Akt (e) were compared between vehicle (HCl) and insulin-treated PCLS; data presented as mean ± SEM. f Quantification of data shown in (b); phosphorylation of PKA substrates was compared between vehicle (HCl) and glucagon-treated PCLS; data presented as mean ± SEM. g–h 5 h cumulative net glucose production of PCLS treated with insulin g and glucagon h; data presented as mean ± SEM. i Immunoblotting of PCLS derived from fed mice incubated with WE medium supplemented with DHA and treated with forskolin; phosphorylation of PKA substrates was monitored. The complete dataset (three biological replicates) is provided in Supplementary Fig. S3. j Quantification of data shown in (i); phosphorylation of PKA substrates was compared between vehicle (DMSO) and forskolin-treated PCLS; data presented as mean ± SEM. k 5 h cumulative net glucose production of PCLS treated with forskolin. For all the analyses, vehicle- and hormone-treated groups were compared using two-way ANOVA followed by Tukey multiple comparison test; *p < 0.05, vs the corresponding vehicle; data presented as mean ± SEM; n = 3 mice per group.

Fig. 4. Sources of glucose in the PCLS derived from fed mice.

Glucose label fraction from PCLS incubated with a ¹³C-glycerol and b ¹³C-DHA for 24 h; Calculated contribution of glycogenolysis and gluconeogenesis in the net glucose production of PCLS incubated with c ¹³C-glycerol and d ¹³C-DHA; e Glycogen stores in PCLS derived from fed and fasted mice incubated with WE medium supplemented with DHA, Lactate Pyruvate (L/P), Glycerol for 5 h; *p < 0.05, vs T0H; n = 3 mice per group; f ¹³C-TrioseP enrichment in PCLS incubated with either ¹³C-glycerol (blue) or ¹³C-DHA (magenta); data presented as mean ± SEM. For (c, d, f) *p < 0.05; n = 3–4 mice per group.

Fig. 5. PCLS as an ex vivo model of GSD type Ib.

a Schematic representation of the mechanism of action and expected effects of S4048, a SLC37A4 inhibitor; b–g PCLS derived from fed and fasted mice were incubated in WE medium supplemented with DHA plus DMSO (black) or S4048 (red); b 5 h net cumulative glucose production of PCLS derived from fed (●) and fasted (○) mice; c Net glucose production rate; *p < 0.05, vs DMSO group in the same condition; ^#p < 0.05, vs DMSO group from fed mice; 5 h net cumulative lactate in the medium of PCLS from fed d and fasted e mice; *p < 0.05, vs DMSO; glycogen stores from fed f and fasted g mice; *p < 0.05, vs T0H; ^#p < 0.05, vs DMSO 5H; data presented as mean ± SEM, n = 3 mice per group.

Fig. 6. Human PCLS produce glucose ex vivo.

a PCLS derived from two control human livers, one GSD Ia and one GSD Ib liver were incubated with WE medium without the addition of any precursor NP or supplemented with DHA, Lactate/Pyruvate (L/P), and glycerol for 5 h; b–c 5 h net cumulative glucose production of PCLS derived from human liver I (b) and II (c) incubated in WE medium without or with DHA, and plus DMSO (dark color) or S4048 (light color); data presented as mean ± SEM of the three technical replicates. ND means non detectable.

Fig. 7. ¹³C-labeled trioseP enrichment in glucose.

Calculation of ¹³C-labeled trioseP enrichment (p) in glucose. Equations 1–3.