Dgat1 and Dgat2 regulate enterocyte triacylglycerol distribution and alter proteins associated with cytoplasmic lipid droplets in response to dietary fat

Abstract

Enterocytes, the absorptive cells of the small intestine, mediate efficient absorption of dietary fat (triacylglycerol, TAG). The digestive products of dietary fat are taken up by enterocytes, re-esterified into TAG, and packaged on chylomicrons (CMs) for secretion into blood or temporarily stored within cytoplasmic lipid droplets (CLDs). Altered enterocyte TAG distribution impacts susceptibility to high fat diet associated diseases, but molecular mechanisms directing TAG toward these fates are unclear. Two enzymes, acyl CoA: diacylglycerol acyltransferase 1 (Dgat1) and Dgat2, catalyze the final, committed step of TAG synthesis within enterocytes. Mice with intestine-specific overexpression of Dgat1 (Dgat1^Int) or Dgat2 (Dgat2^Int), or lack of Dgat1 (Dgat1^-/-), were previously found to have altered intestinal TAG secretion and storage. We hypothesized that varying intestinal Dgat1 and Dgat2 levels alters TAG distribution in subcellular pools for CM synthesis as well as the morphology and proteome of CLDs. To test this we used ultrastructural and proteomic methods to investigate intracellular TAG distribution and CLD-associated proteins in enterocytes from Dgat1^Int, Dgat2^Int, and Dgat1^-/- mice 2h after a 200μl oral olive oil gavage. We found that varying levels of intestinal Dgat1 and Dgat2 altered TAG pools involved in CM assembly and secretion, the number or size of CLDs present in enterocytes, and the enterocyte CLD proteome. Overall, these results support a model where Dgat1 and Dgat2 function coordinately to regulate the process of dietary fat absorption by preferentially synthesizing TAG for incorporation into distinct subcellular TAG pools in enterocytes.

Keywords: Chylomicron; Cytoplasmic lipid droplet; DGAT; Enterocyte; Triacylglycerol.

Figure 1. Identification of subcellular TAG pools within enterocytes by TEM

(A) An ultrastructural overview of enterocytes from the jejunum of WT mice 2 hours after a 200 μl oral olive oil gavage. At this time point during dietary fat absorption, TAG is found in several subcellular pools involved in either CM or CLD synthesis. CMs are assembled and further expanded in the ER lumen ➂, then transported to the Golgi ➃ and Golgi-derived SVs ➄, and finally secreted from enterocytes into the lymph ➅. CLDs serve as temporary TAG storage pools in enterocytes ➆. (B) TAG present in the lumen of the ER is surrounded by smooth ER membrane. (C) The smooth ER containing TAG (black arrows) is continuous with the rough ER membrane, which can be distinguished by the presence of ribosomes (white arrows). (D) TAG is present within stacks of the Golgi apparatus. (E) CMs are carried within SVs (black arrow) and secreted into the intercellular space (*). (F) CLDs are surrounded by a phospholipid monolayer.

Figure 2. Varying levels of intestinal Dgat1 and Dgat2 alters intracellular TAG distribution in enterocytes

Representative TEM images of enterocytes from the jejunum of Dgat1^Int, WT, Dgat2^Int and Dgat1^−/− mice 2 hours after a 200 μl oral olive oil gavage. Representative CLD (*), TAG in ER lumen (black arrows), and TAG in Golgi/SVs (white arrows) are highlighted. Scale bar = 2 μm.

Figure 3. Varying levels of intestinal Dgat1 and Dgat2 alters the distribution of TAG in subcellular pools for CM synthesis and secretion

(A) TAG secretion was assessed by measuring plasma TAG levels in Dgat1^Int, WT, Dgat2^Int and Dgat1^−/− mice given an IP injection of Tyloxapol at 0, 2 and 4 hours after a 200 μl oral olive oil gavage (n = 6–16 mice/group). (B) Representative TEM images of the area above the nucleus in enterocytes from Dgat1^Int and Dgat2^Int mice 2 hours after a 200 μl oral olive oil gavage. A greater amount of TAG was found in the ER lumen of Dgat1^Int enterocytes, whereas a greater amount of TAG was found in Golgi/SVs of Dgat2^Int enterocytes. Representative CLD (*), TAG in ER lumen (black arrows), and TAG in Golgi/SVs (†) are highlighted. (C) TEM images highlighting distinct, small CLDs (*) and TAG within the ER lumen (black arrow) in enterocytes of Dgat1^Int mice 2 hours after a 200 μl olive oil oral gavage. (D) Mean area of TAG in the ER lumen and (E) mean area of TAG in Golgi/ SVs per cell were quantified using ImageJ. The value of each bar is the average of three biological replicates (20 cells/mouse, n=3 mice/group). Data are represented as mean ± SEM. Different letters denote significant differences, p < 0.05 (one-way ANOVA, Tukey HSD test), N.S = not significant.

Figure 4. Varying levels of intestinal Dgat1 and Dgat2 alters enterocyte CLD morphology

(A, B) Size distribution of CLDs within enterocytes of mice 2 hours after a 200 μl oral olive oil gavage. All of the CLDs identified from 60 enterocytes (20 enterocytes/mouse, n=3 mice/group) were examined by TEM and the diameter of CLDs were determined using ImageJ. Data is reported as % of all the CLDs measured per group. * denotes a statistically significant difference, p < 0.05 (Kolmogorov-Smirnov tests). (C) Mean number of CLDs and (D) mean size of CLDs per cell (diameter) were determined. (E) The amount of TAG stored in CLDs is estimated by the equation of (mean number of CLD) x π x (mean radius)². The value of each bar is the average of three biological replicates (20 cells/mouse, n=3 mice/group). (F) qPCR analysis of mRNA levels of genes involved in CLD metabolism. WT mice were the reference group, with their mRNA levels set as 1 (n=3–5 mice/group). Different letters denote significant differences, p < 0.05 (one-way ANOVA, Tukey HSD test), N.S = not significant.

Figure 5. Varying levels of intestinal Dgat1 and Dgat2 alters the enterocyte CLD proteome

CLDs were isolated from enterocytes of Dgat1^Int, WT, Dgat2^Int, and Dgat1^−/− mice 2 hours after a 200μl oral olive oil gavage. (A) Negative stain EM of the intestinal CLD-enriched fraction from a representative sample for each mouse model. Scale bar = 0.5 μm. (B) Overlap of CLD-associated proteins for each mouse model, identified by proteomics. (C) Biological functions of CLD-associated proteins common to all four mouse models, classified based on GO terms (for biological process & molecular function). (D) Differentially expressed CLD-associated proteins common among all mouse models. Different letters denote significantly different relative protein levels (p < 0.05). (E) String map of all proteins with a GO term/functional classification related to lipid/lipoprotein metabolism. Black circles indicate proteins identified in all four genotypes.

Figure 6. Varying levels of intestinal Dgat1 and Dgat2 impacts enterocyte mitochondrial biology

(A) qPCR analysis of mRNA levels of genes involved in FAO. (B) qPCR analysis of the ratio of mitochondrial DNA to nuclear DNA for the assessment of mitochondrial content. WT mice was the reference group, with its DNA level set as 1 (n=3–4 mice/group). Different letters denote significant differences, p < 0.05 (one-way ANOVA, Tukey HSD test), N.S = not significant. (C) Representative TEM images of mitochondria (black arrow) within enterocytes of WT and Dgat1^{− /−} mice 2 hours after a 200 μl oral olive oil gavage. (D) Representative TEM images of normal mitochondria (left) and round, swollen mitochondria (right) found in Dgat1^−/− mice 2 hours after a 200 μl oral olive oil gavage.

Figure 7. Summary of results and proposed model of cellular roles of Dgat1 and Dgat2 in synthesizing TAG for CM and CLDs within enterocytes

(A) Results summary: In Dgat1^Int mice, greater amounts of TAG in the ER lumen, similar amounts of TAG in Golgi/SVs, a trend toward smaller CLDs, and a similar intestinal TAG secretion rate were found compared to WT mice. In Dgat2^Int mice, similar amounts of TAG in the ER lumen, greater amounts of TAG in Golgi/SVs, an increased number of CLDs, and a higher intestinal TAG secretion rate were observed compared to WT mice. In Dgat1^−/− mice, a trend towards less TAG in the ER lumen and Golgi/SVs and an increase in CLD size were found in enterocytes compared to WT mice. Dgat1^−/− mice also secrete TAG from the intestine at a slower rate within smaller sized CMs compared to WT mice. Some of the CLD-associated proteins are present in all genotypes, whereas others, such as ApoAIV and Cideb, are unique to certain genotype(s) and may be contributing to the altered subcellular distribution of TAG observed in these models. (B) Proposed model: A proposed model of coordinated, non-redundant roles of Dgat1 and Dgat2 in CM synthesis was generated by combining the results of the current study with previous knowledge of the CM synthesis pathway. CM assembly in enterocytes requires two distinct pools of TAG within the lumen of the ER: ApoB-containing nascent CMs and ApoB-free LLDs. In this model, Dgat1 is proposed to preferentially synthesize TAG for incorporation into ApoB-free LLDs, which determines the availability of TAG for nascent CM expansion and thus the size of CMs. Dgat2 is proposed to preferentially synthesize TAG for incorporation into ApoB-containing nascent CMs, which likely regulates the rate of intestinal TAG secretion. In addition, TAG storage in CLDs and its subsequent catabolism provides substrates for re-synthesis of TAG at the ER membrane that may contribute to CM synthesis and secretion at later times. Based on the results from the current study, Dgat1 is proposed to restrict CLD growth in enterocytes by preferentially synthesizing TAG that is delivered into the ER lumen, thus limiting TAG storage in CLDs. Dgat2, on the other hand, is proposed to preferentially increase TAG storage in CLDs, which may provide fatty acids for re-synthesis of TAG at the ER membrane for incorporation into CMs at later time points. Differential associations of proteins with enterocyte CLDs in these models may serve to regulate the localization and activity of proteins involved in dietary fat absorption.