Diacylglycerol acyltransferase-1 (DGAT1) inhibition perturbs postprandial gut hormone release

Abstract

Diacylglycerol acyltransferase-1 (DGAT1) is a potential therapeutic target for treatment of obesity and related metabolic diseases. However, the degree of DGAT1 inhibition required for metabolic benefits is unclear. Here we show that partial DGAT1 deficiency in mice suppressed postprandial triglyceridemia, led to elevations in glucagon-like peptide-1 (GLP-1) and peptide YY (PYY) only following meals with very high lipid content, and did not protect from diet-induced obesity. Maximal DGAT1 inhibition led to enhanced GLP-1 and PYY secretion following meals with physiologically relevant lipid content. Finally, combination of DGAT1 inhibition with dipeptidyl-peptidase-4 (DPP-4) inhibition led to further enhancements in active GLP-1 in mice and dogs. The current study suggests that targeting DGAT1 to enhance postprandial gut hormone secretion requires maximal inhibition, and suggests combination with DPP-4i as a potential strategy to develop DGAT1 inhibitors for treatment of metabolic diseases.

Figure 1. Biochemical characterization of DGAT1 heterozygous and homozygous knockout mice.

(A) DGAT1 mRNA expression determined by quantitative RT-PCR in jejunal mucosa, white adipose tissue, and liver from Dgat1+/−, Dgat1−/−, and age-matched C57BL/6N wild type mice. N = 3. DGAT1 mRNA and were normalized against β-actin mRNA. All data are mean ± SEM. *P<0.05, **P<0.01, ***P<0.001 vs. wild type. ^##P<0.01, ^###P<0.001 Dgat1−/− vs. Dgat1+/−. (B) DGAT enzymatic activity in intestinal mucosa of Dgat1−/− and wild type mice measured by TG synthesis in the presence of various concentrations of DGAT1i. (C) Binding of intestinal microsomes of wild type, Dgat1+/−, and Dgat1−/− mice to ³H-labeled DGAT1i. Microsome preps pooled from 2–5 mice were used in the assay. (D) Scatchard analysis of ³H-labeled DGAT1i binding in intestinal microsomes of wild type and Dgat1+/− mice.

Figure 2. Postprandial triglyceridemia, gut hormones, and gastric emptying in Dgat1+/− and Dgat1−/− mice.

4- to 5-month-old Dgat1+/−, Dgat1−/−, and wild type mice maintained on normal chow diet were fasted overnight, then p.o. dosed with (A–F) corn oil (100% lipid load, 8.14 kcal/ml) or (G–L) a mixed meal containing corn oil:Ensure-plus = 3∶17 v/v (19.4% lipid load, 2.48 kcal/ml) at 10 ml/kg body weight. At 2 h after meal challenge, plasma levels of (A) TG, (B, H) active GLP-1, (C, I) total GLP-1, (D, J) PYY, (E, K) GIP, and (F, L) stomach weight were measured. (G) In a separate cohort, plasma TG at 0, 1, 2, and 4h after 19.4% lipid load was determined. N = 8–12. *P<0.05, **P<0.01, ***P<0.001 vs. wild type. ^#P<0.05, ^##P<0.01, ^###P<0.001 Dgat1−/− vs. Dgat1+/−.

Figure 3. Effects of DGAT1 inhibition on postprandial GLP-1.

3-month-old lean C57BL/6N mice were p.o. dosed with DGAT1i at 3 mg/kg body weight and fasted overnight. The next morning, mice were p.o. dosed with mixed meals containing 2.6 (water:Ensure-plus = 1:1), 5.2, 28.9, 68.4, or 100% (corn oil:Ensure-plus = 0∶1, 1∶3, 2∶1, 1∶0 v/v) lipid load at 10 ml/kg body weight (0.74, 1.48, 3.15, 5.92, or 8.14 kcal/ml). At 1, 2, or 3h after meal challenge, plasma levels of (A–C) active GLP-1, (D–F) total GLP-1, and (H–J) blood glucose were measured. (G) 3-month-old lean C57BL/6N mice were p.o. dosed with 3 mg/kg DGAT1i or vehicle and fasted overnight. 18h after dosing, jejunal mucosa was harvested, and Gcg mRNA expression was measured by realtime PCR. Data are normalized against β-actin mRNA. N = 8. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001 between groups.

Figure 4. Gut-selective DGAT1 knockdown and its effects on TG and gut hormones.

(A) Design of the transgenic construct used to generate DGAT1 gKD mice. A transgenic vector was constructed that contains a 12.4 kb Villin1 promoter, a sequence encoding EGFP, followed by miR155-embedded Dgat1 shRNA and a poly-A tail. Sequence of the miR155-embedded Dgat1 shRNA is shown, with the sequences homologous to Dgat1 underlined. (B) DGAT1 mRNA expression in selected tissues from two DGAT1-shRNA transgenic lines and wild type control littermates. N = 3–5. Data are normalized against Rplp0 mRNA. ND: not determined. (C, D) 4-month-old female gKD mice and wild type littermates were fasted overnight, then p.o. dosed with corn oil at 10 ml/kg. Plasma TG was measured at 0, 1, 2, and 4 h after oil challenge, and net area-under-the-curve (AUC) compared to time 0 was calculated. Overnight fasted gKD mice and wild type littermates were p.o. dosed with a mixed meal containing (E–G) corn oil:Ensure-plus = 1∶1 (52.6% lipid load, 4.81 kcal/ml) or (H–J) corn oil:Ensure-plus = 3∶17 (19.4% lipid load) at 10 ml/kg. Plasma levels of active GLP-1, PYY, and GIP were measured at 2 h after meal challenge. N = 10–14. *P<0.05, **P<0.01 vs. wild type.

Figure 5. Body weight and food intake of gKD mice on HFD.

Male gKD mice and wild type littermate raised no normal chow diet were switched to HFD at 4 months of age, and (A) body weight and (B) food intake were monitored for 8 weeks. No statistical significant difference was observed. N = 17–18.

Figure 6. Combined effects of DGAT1 with DPP-4 inhibition on GLP-1.

(A, B) 3-month-old lean Dgat1−/− and wild type mice were fasted overnight, then p.o. dosed with vehicle or a DPP-4 inhibitor (10 mg/kg) at -1 h. At time 0, mice were p.o. dosed with a meal containing 28.9% lipid load. (A) Plasma active GLP-1 and (B) total GLP-1 were measured at 2 h after meal challenge. N = 8. **P<0.01, ***P<0.001 vs. vehicle. ^###P<0.001 between indicated groups. (C, D) 5-month-old DIO mice were p.o. dosed vehicle or DGATi (1 mg/kg) at -18 h and fasted overnight, then received vehicle or DPP-4 inhibitor (10 mg/kg) treatment at -1 h. At time 0, mice were p.o. dosed a meal with 28.9% lipid load, and plasma (C) active GLP-1 and (D) total GLP-1 were measured at 2h after meal challenge. N = 8. **P<0.01, ***P<0.001 vs. vehicle. ^##P<0.01 between indicated groups. (E) Lean male beagle dogs were fasted overnight, then p.o. dosed the next morning (time -2 h) with vehicle, DGAT1i (0.3 mg/kg), DPP-4i (3 mg/kg), or a combination of DGAT1i and DPP-4i. At time 0, animals were p.o. dosed heavy cream (4 ml/kg). Serial blood samples were collected at indicated time points, and plasma active GLP-1 was measured. N = 4 for vehicle treatment, N = 6 for all other groups. *P<0.05, **P<0.01 DGAT1i/DPP-4i combo vs. vehicle. ^#P<0.05 DPP-4i vs. vehicle. ^&P<0.05 DGAT1i vs. vehicle. ^$P<0.05 combo vs. DGAT1i alone. ^@P<0.05 combo vs. DPP-4i alone.