The triglyceride synthesis enzymes DGAT1 and DGAT2 have distinct and overlapping functions in adipocytes

Abstract

Mammals store metabolic energy as triacylglycerols (TGs) in adipose tissue. TG synthesis is catalyzed by the evolutionarily unrelated acyl-CoA:diacylglycerol acyltransferase (DGAT) enzymes DGAT1 and DGAT2, which catalyze the same reaction and account for nearly all TG synthesis. The reasons for their convergent evolution to synthesize TGs remain unclear. Mice lacking DGAT1 are viable with reduced fat stores of TGs, whereas DGAT2 KO mice die postnatally just after birth with >90% reduction of TGs, suggesting that DGAT2 is the predominant enzyme for TG storage. To better understand the functional differences between the DGATs, we studied mice fed chow or high-fat diets lacking either enzyme in adipose tissue. Unexpectedly, mice lacking DGAT2 in adipocytes have normal TG storage and glucose metabolism on regular or high-fat diets, indicating DGAT2 is not essential for fat storage. In contrast, mice lacking DGAT1 in adipocytes have normal TG storage on a chow diet but moderately decreased body fat accompanied by glucose intolerance when challenged with a high-fat diet. The latter changes were associated with the activation of ER stress pathways. We conclude that DGAT1 and DGAT2 can largely compensate for each other for TG storage but that DGAT1 uniquely has an important role in protecting the ER from the lipotoxic effects of high-fat diets.

Keywords: adipose tissue; diacylglycerol acyltransferase; endoplasmic reticulum stress; glucose intolerance; high-fat diet; induced obesity.

Fig. 1.

DGAT2 and SkiDGAT2 KO mice die shortly after birth. A: Strategy for generating DGAT2 KO, DGAT2 flox mice, and tissue-specific DGAT2 KO mice. A “knock-out first/conditional-ready” gene-targeting vector was used to generate targeted ES cells. The gene-trap cassette is located between two FRT sites. B, C: DGAT2 KO mice die within a few hours after birth due to defective skin-barrier function. D, E: SkiDGAT2 KO mice were generated by crossing DGAT2 flox mice with the mice expressing Cre recombinase under the control of the human keratin-14 promoter. SkiDGAT2 KO mice die within a few hours after birth due to defective skin-barrier function. F: The DGAT2 protein is absent in the skin of SkiDGAT2 KO mice (n = 3). G: Adipose tissue-specific ADGAT2 KO were generated by crossing DGAT2 flox mice with the mice expressing Cre recombinase under the control of the mouse adiponectin promoter. Dgat1 and Dgat2 mRNA levels in gWAT of DGAT2 flox and ADGAT2 KO mice (n = 5). H: The DGAT2 protein is absent in gWAT of ADGAT2 KO mice (n = 4). Data are presented as mean ± SD. ***P < 0.001.

Fig. 2.

DGAT1 and DGAT2 largely compensate for TG storage in adipose tissue in the chow-fed state. A: Body weights of mice fed a chow diet (n = 10–12). B: Organ weights of mice fed a chow diet (n = 8 per genotype). C: Plasma parameters of mice fed a chow diet (n = 7 per genotype). D: H&E-stained sections of gWAT, iBAT, and livers of mice fed a chow diet (representative images of eight mice per genotype). Scale bars: 50 µm. Data are presented as mean ± SD. *P < 0.05 by t-test. iBAT, intrascapular brown adipose tissue; NEFA, nonesterified fatty acid.

Fig. 3.

DGAT1 contributes more to diet-induced obesity than DGAT2. A: Body weights of mice fed an HFD (n = 15). B: Lean- and fat-mass analysis of mice fed an HFD (n = 10 per genotype). C: Weights of gWAT and iBAT of mice fed an HFD (n = 8). D: TG content of livers and skeletal muscle (n = 6 mice). E: Plasma parameters of mice fed an HFD (n = 6). F: Western blot analysis of tissues from mice fed an HFD (n = 4). G: Glucose- and insulin-tolerance tests were performed on mice fed an HFD (n = 10–16 per genotype). Data are presented as mean ± SD (C–E) or mean ± SEM (A, B, G, H). *P < 0.05, **P < 0.01 by t-test, and ***P < 0.001 by two-way ANOVA. iBAT, intrascapular brown adipose tissue; NEFA, nonesterified fatty acid.

Fig. 4.

ER stress in ADGAT1 KO mice fed an HFD but not in ADGAT2 KO mice. A, B: mRNA levels of lipogenic genes, ER stress genes, and inflammatory genes determined by real-time qPCR (n = 6 mice per genotype). Mice were fed an HFD for 16 weeks. Tissues were collected in the ad libitum-fed state. C: Protein levels of BIP and CHOP in gWAT of mice fed an HFD for 16 weeks (n = 4). Data are presented as mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001 by t-test.