SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Objective

Investigation was conducted to understand the mechanism of action of diacylglycerol acyltransferase 1 (DGAT1) using small molecules DGAT1 inhibitors, compounds K and L.

Design and Methods

Biochemical and stable-label tracer approaches were applied to interrogate the functional activities of compounds K and L on TG synthesis and changes of carbon flow. Energy homeostasis and gut peptide release upon DGAT1 inhibition was conducted in mouse and dog models.

Results

Compounds K and L, dose-dependently inhibits post-prandial TG excursion in mouse and dog models. Weight loss studies in WT and Dgat1-/- mice, confirmed that the effects of compound K on body weight loss is mechanism-based. Compounds K and L altered incretin peptide release following oral fat challenge. Immunohistochemical studies with intestinal tissues demonstrate lack of detectable DGAT1 immunoreactivity in enteroendocrine cells. Furthermore, 13C-fatty acid tracing studies indicate that compound K inhibition of DGAT1 increased the production of phosphatidyl choline (PC).

Conclusion

Treatment with DGAT1 inhibitors improves lipid metabolism and body weight. DGAT1 inhibition leads to enhanced PC production via alternative carbon channeling. Immunohistological studies suggest that DGAT1 inhibitor's effects on plasma gut peptide levels are likely via an indirect mechanism. Overall these data indicate a translational potential towards the clinic.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Obesity is one of the primary risk factors in diabetes and cardiovascular diseases. Of the several metabolic pathways identified in body weight (BW) homeostasis [1], the triacylglycerol synthesis machinery has emerged as a potential pharmacological target for modulating BW. Excess accumulation of triglyceride (TG) in blood and tissues is a critical mediator in metabolic diseases. Abnormal levels of TG in non-adipose tissues such as liver, skeletal muscle, and myocardium present significantly higher risk associated with insulin resistance and cardiovascular diseases [4]. The two major pathways that mediate TG biosynthesis are the glycerol phosphate pathway [5] and the monoacylglycerol pathway [6]. DGAT, which catalyzes the joining of diacylglycerol (DG) with a fatty acyl CoA, mediates the final step of TG synthesis in both pathways. Two Dgat genes have been identified, Dgat1 and Dgat2. The expression patterns of both genes are conserved across humans and mice [7, 8]. DGAT1, while broadly expressed, is markedly enriched in intestine, where the monoacylglycerol pathway is believed to be the major route for TG synthesis.

Initial evidence demonstrating that DGAT1 inhibition could have a beneficial metabolic effect came from studies of a mouse model with genetic deletion of Dgat1. Dgat1 knock out (Dgat1/−) mice are resistant to diet-induced obesity (DIO) and have improved insulin sensitivity [9, 10]. Dgat1/− mice also showed markedly reduced levels of intestinal TG synthesis and postabsorptive chylomicronemia following an oral lipid challenge, indicating that Dgat1 deficiency has a significant impact on intestinal TG metabolism in the presence of a high dietary fat load [11]. Interestingly, intestine-specific expression of Dgat1 in Dgat1/− mice rescues their sensitivity to DIO, indicating the key role of intestinal DGAT1 activity in controlling BW homeostasis [12]. Additionally, these results point to the importance of the postprandial triglyceridemic response in determining susceptibility to high fat diet (HFD)-induced obesity. More recently it has been demonstrated, from Dgat1−/− phenotyping studies, that DGAT1 mediates the prolongation of dietary fat-induced increase in plasma gut peptide levels, and deficiency of Dgat1 resulted in a delay of gastric emptying (GE) [13].

The positive metabolic phenotype of Dgat1−/− mice has led to significant efforts among the pharmaceutical industry toward the development of small molecule DGAT1 inhibitors for the treatment of metabolic diseases. Preclinical efficacy of small molecule DGAT1 inhibitors was demonstrated in rodent proof of concept studies for diminished postprandial hyperlipidemia and resistance to both DIO and liver steatosis [14]. More advanced DGAT1 inhibitors have progressed into clinical development including, Novartis (LCQ-908), Pfizer (PF 04620110), and AstraZeneca (AZ-7687) [15].

Despite relative progress in the field, additional structurally diverse DGAT1 small molecule inhibitors are required to better understand the mechanism whereby DGAT1 inhibition positively impacts metabolism. In this report, we utilized novel, selective DGAT1 inhibitors, compound K and L, to investigate the role of DGAT1 in inducing weight loss and compared it to a marketed weight reducing agent, orlistat. Furthermore, using an immunohistochemical based approach, we aimed to understand the mechanisms by which attenuation of postprandial TG synthesis, mediated by blockade of DGAT1 activity, alters gut peptide release. Finally, we examined the fate of DGAT1 substrates upon DGAT1 inhibition with tracer-based approaches in cellular and in vivo systems.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Materials

Immunoassay kits from Meso Scale Discovery (MSD, MD); total TG colorimetric assay (Thermo Scientific, NJ); anti-mouse DGAT1 (Santa Cruz, CA); Cy3-conjugated secondary antibody (Jackson ImmunoResearch, PA); anti-GLP-1 (Bachem, CA); Cy5-conjugated secondary antibody (Jackson ImmunoResearch, PA); 4′,6-diamidino-2-phenylindole (DAPI), diolein and oleoyl CoA (Sigma-Aldrich, MO); 7-diethylamino-3-(4′-maleimidylphenyl)-4-methylcoumarin (CPM) (Invitrogen, CA).

Mice

Experiments were performed in male lean and DIO C57BL/6 mice. All WT mice were obtained from Taconic Farms (Germantown, NY) and maintained on either standard rat diet or HFD in a 12-h light/12-h dark cycle. Animal protocols used in these studies were approved by the Merck Research Laboratories Institutional Animal Care and Use Committee (Rahway, NJ). Dgat1−/− mice were received from Tsukuba Research Institute, Japan [13]. Six month old mice were placed on HFD for 10 weeks prior to study.

Cells

HT-29 and C2C12 cells were obtained from American Type Culture Collection (ATCC, VA).

DGAT1 enzyme assay

The activity of DGAT1 from different species were measured using DG/oleoyl CoA as substrates at Km concentrations in the presence of CPM, which is weakly fluorescent until reacted with free thiols of CoA released from oleoyl CoA after it is incorporated into DG to form TG. IC50s were calculated using either Assay Data Analyzer or GraphPad Prism4 software.

Cell-based functional assays

HT-29 and C2C12 cells were seeded in the presence or absence of DGAT1 inhibitors followed by two hour incubation with 13C-labeled oleic acid. The lipids were extracted with isopropanol containing internal lipid standard followed by measurement using liquid chromatography-mass spectrometry (LC/MS).

Chronic dosing of compound K in mice

Mice were acclimated to nonspecific stress for 7 days before the onset of the chronic dosing studies. DIO mice were dosed orally with 0.1, 0.3, 1, 3, and 10 mg/kg of compound K for 14 or 28 days, with BW and food intake (FI) monitored daily. Orlistat (Roche) was dosed orally at 50 mg/kg.

Analysis of plasma samples

Concentrations of compounds in plasma were determined by LC-MS/MS. Acetonitrile solution containing internal standard was added for protein precipitation followed by centrifugation and separation of the supernatants.

Body composition

Whole body composition analysis of conscious live mice was conducted using quantitative magnetic resonance method (Echo Medical Systems, TX).

Mouse post prandial TG excursion test

Lean C57BL/6 mice (10 week old, n = 8/group) fasted overnight, received 10 ml/kg corn oil via oral administration (PO) 60 min post compound administration. Plasma was collected at various time points. For gut peptide measurements, plasma was collected at 120 min post lipid challenge.

Dog post prandial TG excursion test

Lean beagle dogs (2-6 years old, n = 6-8/group Marshall BioResources, NY) were fasted overnight and received 4 ml/kg heavy cream (PO) 1 h post compound administration. Blood sampling was taken at −1, 0, 1, 2, 4, 6 h relative to the lipid challenge.

Plasma gut peptide measurements

Total and active GLP-1 levels were analyzed using immunoassay kits from MSD. Analysis of PYY and gastric inhibitory polypeptide (GIP) were conducted using milliplex mouse metabolic magnetic bead panel (Millipore, MA). The data were analyzed using Bio-Plex Manager 6.0 software.

Plasma triglyceride measurements

TG was measured using colorimetric assay from Thermo Scientific.

Fecal fat analysis

Feces were collected the last two days of chronic treatment of compound K in DIO mice. The fecal free fatty acids (FFAs) were extracted with ethanol:distilled water:potassium hydroxide at ratio of 4:1:2. Following incubation at 70°C for 2 h, 6N HCl and petroleum ether were added at ratio of 1:2. The upper phase after spinning was dried and dissolved in isopropanol followed by measurement using FFA kit (Wako Pure Chemicals, Japan).

Immunohistochemistry

Paraffin-embedded mouse intestine sections were incubated overnight with DGAT1 antibody (Ab) followed by incubation with Cy3-conjugated secondary Ab. The tissue sections were then incubated with GLP-1 or GIP Ab followed by incubation with Cy5-conjugated secondary Ab. The nuclei were stained with DAPI.

Tracer-based plasma lipid profiling study

DIO mice (24-26 weeks old, n = 7-8 per group) were chronically dosed in feed (HFD) for 3 weeks with compound K. Following an overnight fast, mice received an intravenous bolus of 13C-oleic acid as a tracer. Plasma samples were collected at various time points. Lipids were extracted and measured by LC/MS as described [21].

Statistical analysis

Statistical analysis of data was conducted by Student's t-test, one way ANOVA using GraphPad Prism4 (Graph-Pad Software, CA). Data are presented as means ± standard errors of the mean (SEM) and P values <0.05 were considered significant. GraphPad software was also used to determine areas under curve (AUC).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Identification of novel DGAT1 small molecule inhibitor

An enzymatic-based assay, using human DGAT1 enriched membranes from DGAT1 overexpressing Picchia, was utilized for screening Merck's small molecule libraries. Subsequent lead optimization effort led to the identification of compound K as a potent inhibitor of DGAT1 activity with a 50% inhibition (IC50) at 60 nM (Figure 1A) and a related analog compound L with IC50 value of 24 nM (Figure 1B). Counter screens against closely related acyl transferases indicated that compounds K and L are highly specific DGAT1 inhibitors. To further understand compounds K and L activity across species, similar enzymatic based-assays were performed using DGAT1 from different species and demonstrated that compounds K and L were potently active on DGAT1 from these preclinical species (Figure 1A and B).

image

Figure 1. Chemical structure and activity profile of compound K (A) and L (B). Compound concentration-dependent inhibition of triglyceride (TG) synthesis in human intestinal epithelial HT-29 (C) and mouse myoblastic C2C12 cells (D). Cells were preincubated with various concentrations of compound K for 30 min followed by [13C-18] oleic acid incubation. A representative labeled TG species, TG13C18:1-13C18:1-13C18:1, was used to represent TG levels in the presence of different concentrations of compound K. Results are expressed as mean ± SEM. IC50: 150 ± 26 nM for HT-29 and 35 ± 2 nM for C2C12. Compounds K and L were evaluated against other acyltransferases. IC50s: >7,5500 nM (human DGAT2); >100,000 nM (human MGAT2, MGAT3, GPAT1); 1,933 nM (human ACAT1). The above study was repeated three or more times. DGAT2: diacylglycerol-O-acyl transferase 2; MGAT2: monoacylglycerol-acyl transferase 2; MGAT3: monoacylglycerol-acyl transferase 3; GPAT1: glycerol-3-phosphate acyltransferase 1; ACAT1: acetyl-CoA-acyltransferase 1.

Download figure to PowerPoint

To determine the functional activity of compound K inhibition in cellular systems, two physiologically relevant cells, human intestinal epithelial cells (HT-29) and mouse skeletal muscle cells (C2C12) were evaluated. 13C-labeled oleic acid was used as the substrate at Km concentration (100 μM). The incorporation of 13C-labeled oleic acid into TG was measured through LC/MS analysis. Under these conditions, compound K inhibited TG synthesis in a concentration-dependent manner with IC50s at 150 nM and 35 nM for HT-29 and C2C12 cells, respectively (Figure 1C and D).

In vivo effects of compound K demonstrate pharmacological proof of concept

Pharmacokinetic (PK) study of compound K was conducted to guide in vivo efficacy studies. Compound K showed 16 ml/min/kg clearance, 2 h half-life and 54% oral bioavailability in mice, which are suitable for evaluation of in vivo efficacy. Compound K's pharmacokinetic properties precipitated a series of pharmacological studies to assess the effects of DGAT1 inhibition by compound K on several metabolic parameters.

To determine the degree of compound K inhibition of DGAT1 required for the efficacy of BW loss, post prandial TG (PPTG) excursion was used as a pharmacodynamic measure of gut DGAT1 inhibition. PPTG levels were measured in lean mice at 1, 2, and 3 h after an oral corn oil bolus challenge, given 1 h after compound dosing (Cmax). Under these conditions, a significant TG excursion was observed in the plasma of vehicle-treated animals. Compound K treatment led to dose-dependent attenuation in the PPTG relative to vehicle group, where 1 mg/kg was determined to be the minimal efficacious dose (MED) for maximal efficacy (Figure 2A).

image

Figure 2. Effects of compound K on postprandial TG excursion, BW loss, fat mass, and FI. Compound K reduced post-prandial TG excursion in a dose-dependent manner after acute dosing of compound K in lean mice (A). Change in percentage BW (B). Changes of fat mass after 28 day compound K treatment (C) and daily FI (D) in DIO mice treated with daily PO dosing of either compound K at 0.1, 0.3, 1, and 3 mg/kg or vehicle for 28 days. Results are expressed as mean ± SEM (n = 8 per group). *P < 0.05. The average BW is 46 g at the beginning of the study across all groups. The average BW is 50 g in vehicle-control group and 47.5 g in 3 mg/kg compound K treated group at the end of the study. The statistical analysis was performed using Student t test for TG excursion and one way ANOVA for BW and FI.

Download figure to PowerPoint

We next investigated compound K's effects on BW and FI in a DIO mouse model. DIO mice were treated daily for 28 days with 0.1, 0.3, 1, or 3 mg/kg of compound K. Compound K treatment at 1 mg/kg and 3 mg/kg led to 3.5% and 5% BW loss, respectively, relative to vehicle group (Figure 2B). The BW change was predominantly due to fat mass reduction (Figure 2C) with minimal effect on lean mass (data not shown). Furthermore, compound K treatment yielded modest but significant reductions in FI in the initial part of the study (Figure 2D), suggesting that compound K reduces BW partially by attenuating food consumption. Compound K-treated mice appeared normal with no overt behavioral changes or alteration in locomotor activities during the dosing period. PK analysis conducted at the conclusion of the study demonstrated that plasma concentration of compound K was 41 μM h as determined by the AUC at 3 mg/kg dose. Trough total plasma level of compound K was 1.2 μM. Since compound K is a highly-protein bound compound, the plasma-free fraction of the compound at trough is below Ki. However, full target engagement is observed at this drug dose (Figure 2A). Furthermore, 10 mg/kg dose (Figure 4) or 30 mg/kg (data not shown) did not lead to greater weight loss, suggesting that maximum DGAT1 engagement is achieved at 3 mg/kg dose. These data implicate that intestinal compound exposure levels may be a more accurate predictive of efficacy over plasma.

To examine the contribution of DGAT1 pharmacological inhibition to energy expenditure, the effect of acute DGAT1 inhibition was evaluated. No significant effects were observed on oxygen consumption (VO2) or respiratory quotient (RQ) between vehicle and compound L treated groups over a period of 24 h (Supporting Information Figure S1).

To confirm that the effect of compound K on BW is directly mediated by inhibition of DGAT1 activity, WT and Dgat1/− mice were placed on HFD and were monitored for 7 weeks. As previously reported, Dgat1/− mice were shown to be resistant to diet-induced obesity (Supporting Information Figure S2) [9]. At the 10th week on HFD, Dgat1/− and WT mice were treated with daily dosing of 3 mg/kg compound K or vehicle for 14 days. In WT mice, the 3 mg/kg dose of compound K led to approximately 3% reduction of BW gain relative to vehicle-treated controls. However, weight gain under vehicle or compound K treated Dgat1/− mice was virtually indistinguishable (Figure 3A). Consistent with BW results, compound K led to a significant 5% reduction in daily average FI in WT mice with minimal effect observed in compound K treated Dgat1/− mice (Figure 3B). BW loss in the compound-treated WT mice was at least partially due to the significant decrease in FI observed in the first two days of treatment (Figure 3C). To rule out potential confounding differences in compound K drug metabolism between WT and Dgat1/− mice, plasma compound levels were determined on day 14 of treatment. Cmax and trough levels of compound K were comparable between WT and Dgat1/ groups (data not shown). Taken together, these data support DGAT1 as the specific biological target of compound K-mediated pharmacological effects on weight loss.

image

Figure 3. Dgat1−/− mice demonstrate mechanism-based BW loss. Changes of BW (A), accumulative FI after daily PO dosing with compound K (3 mg/kg) in WT and Dgat1/− mice over the 14-day compound K treatment period (B), daily FI normalized as kcal/g/day (C). Compound K reduced FI and BW in WT, but not in Dgat1/− mice. Results are expressed as mean ± SEM (n = 8/group). *P<0.05. Both WT and Dgat1/− mice were on HFD for 10 weeks prior to and during 14 days of the study. The average starting BW at the time of the study for WT and Dgat1/− mice was 45 g and 40 g, respectively. The statistical analysis was performed using Student t test for accumulative FI and one way ANOVA for BW and daily FI.

Download figure to PowerPoint

Comparison of in vivo efficacy of compound K with orlistat

To understand the relative efficacy and tolerability of compound K to a marketed drug targeting a related mechanism, a chronic weight loss study was conducted comparing compound K to the pancreatic lipase inhibitor, orlistat. Pancreatic lipase is the key enzyme mediating breakdown of dietary TG in intestine, an essential step for intestinal fat absorption. By inhibiting intestinal TG breakdown, orlistat prevents absorption of dietary fat thereby decreasing absorption of dietary calories [22]. DIO mice were treated with either compound K or orlistat (50 mg/kg) for 14 days. Mice were monitored daily for BW and FI. Compound K at 3 or 10 mg/kg led to approximately 4% BW loss, similar to that observed in our earlier studies, whereas orlistat treated group lost roughly 6% compared to vehicle group (Figure 4A). In contrast to compound K, orlistat resulted in a modest but significant increase in cumulative FI (data no shown).

image

Figure 4. Comparison of compound K with orlistat suggests greater therapeutic margin. Changes of BW (A) and fecal FFA level (B) after 14-day chronic treatment with compound K (0.3, 1, 3, 10 mg/kg), orlistat (50 mg/kg) or vehicle in DIO mice. Fecal fatty acid contents trended higher with increasing doses of compound K, but significantly lower compared to the orlistat-treated group. Results are expressed as fold changes versus vehicle (mean ± SEM). ***P < 0.0001. The statistical analysis was performed using Student t-test for fecal fat output and one way ANOVA for BW measurements.

Download figure to PowerPoint

To determine the effects of orlistat or compound K on intestinal fat absorption, fecal FFA was analyzed at the end of the study. Compound K treatment led to a small, but dose-dependent increase in fecal FFA (1.4- to 3-folds versus vehicle). In contrast, a profound increase (11-fold versus vehicle) in fecal FFA was demonstrated upon orlistat treatment (Figure 4B). These results indicate that weight loss in the orlistat-treated group is likely due to fat malabsorption predominantly, which was compensated by the increase of energy intake. On the other hand, DGAT1 inhibition led to decreased FI with only modest increase in fecal fat, suggesting that weight loss is primarily due to a mechanism independent of fat absorption.

Prolongation of lipid-induced gut hormone release in rodent and canine models

It has been previously demonstrated that Dgat1 inactivation in mouse, rat, and canine models lead to modulation of gut peptide secretion and delay in GE post-prandially [13, 19, 20]. To further investigate these observations, after pharmacological inhibition of DGAT1 activity. PPTG and gut hormone secretion profile were interrogated in rodent and canine models. In the rodent study, C57/BL6 mice were treated with compound K or L, which has similar potency on inhibiting DGAT1 activity and slightly weaker PK profile compared to compound K (Figure 1B and data not shown). The dosed mice were then challenged 1 h later with corn oil. Plasma TG and the levels of gut hormones were monitored 2 h post challenge. Compound K or L treatment completely suppressed plasma TG elevation compared to vehicle-treated group (Figure 5A, data not shown). However, compound K or L treatment led to increased levels of plasma active and total GLP-1, as well as PYY (Figure 5B, C, Supporting Information Figure S5 and data not shown). Interestingly, GIP level was reduced upon compound L treatment (Figure 5D). In addition to modulation of gut peptide release, stomach weight was significantly elevated in the compound-treated mice (Figure 5E), indicating a delay in GE. Overall, these data suggest that pharmacological inhibition of DGAT1 leads to similar effects observed in the Dgat1−/− mice [13].

image

Figure 5. Postprandial gut peptide secretion and GE after compound L or K treatment in mouse and dog models. Two hour-postprandial mouse plasma levels of TG (A), active GLP-1 (B), PYY (C), and GIP (D). Increased in stomach weight 2 h after lipid load (E), suggesting a delay in GE after compound treatment. Results are expressed as mean ± SEM (n = 8/group). Time course of postprandial dog plasma levels of TG (F), active GLP-1 (G), PYY (H), and GIP (I). Results are expressed as mean ± SEM (n = 6/group). *P < 0.05, **P < 0.005, ***P < 0.0001. The statistical analysis was performed using Student t-test.

Download figure to PowerPoint

To determine whether the effects of compound K in rodents were translatable to higher species, we investigated beagle dog model to determine PPTG and gut hormone secretion. Compound K (1 mg/kg) was administered orally to lean beagle dogs 1 h prior to lipid challenge. Plasma TG and gut hormone levels were measured at −1, 0, 1, 2, 4, and 6 h relative to the time of lipid challenge. Consistent with the results observed in mice, compound K treatment completely suppressed the elevation of plasma TG at all time points tested (Figure 5F). On the other hand, plasma GLP-1 (total and active forms) and PYY levels were markedly elevated and these elevations were prolonged for up to 6 h after treatment (Figure 5G and H and data not shown). In contrast to the elevation of GLP-1 and PYY, GIP levels were initially attenuated, however, showed a significant increase relative to vehicle treated group at 6 h time point (Figure 5I). These results demonstrate a tight homology in response to DGAT1 inhibition between rodent and canine models, implying that the mechanism of modulating gut hormone release via DGAT1 inhibition could be conserved across species.

Immunohistochemistry studies to interrogate DGAT1 expression in enteroendocrine cells

To further understand the mechanisms underlying modulations of gut peptide release mediated by DGAT1 inhibition, we conducted studies to assess whether DGAT1 inhibition directly affects the secretion of GLP-1 and PYY by enteroendocrine L cells and GIP release from enteroendocrine K cells. An alternative mechanism for DGAT1 effects on gut hormone release could be mediated by an indirect signaling between the enterocytes and enteroendocrine cells. An immunohistochemistry-based approach was utilized to address this question. Dual-labeling immunohistochemistry studies were conducted using antibodies against DGAT1 and GLP-1 or GIP. This strategy permitted an assessment of the degree of DGAT1 immunoreactivity in enteroendocrine cells. We first tested the specificity of antibodies against DGAT1 by comparing staining of the duodenum, jejunum, and ileum of WT and Dgat1−/− mouse small intestine. Expression of DGAT1 protein was detected throughout small intestine. The highest DGAT1 expression was detected in the duodenum relative to the proximal portion of the intestine, which is consistent with mRNA levels (data not shown). More importantly, expression of DGAT1 was only detected in the WT intestine and was entirely absent along the intestine of Dgat1−/− mice (Figure 6A and B), indicating that the antibody and the conditions used were specific for the detection of DGAT1 expression along the gut.

image

Figure 6. DGAT1 expression in enterocyte with non-detectable level in enteroendocrine cells. Representative photomicrographs of DGAT1 immunoreactivity in duodenum sections from WT mice (A) and Dgat1/− mice (B). DGAT1 expression was only observed in duodenum from WT mice. Dual-labeling analysis of DGAT1 and GLP-1 expression using jejunum sections from Dgat1/− and WT mice at low (C) and high magnification (D). Multichannel image of the intestine demonstrates lack of colocalization of DGAT1 (red) and GLP-1 (green) in WT intestinal sections. Multiple sections were analyzed in these studies, and the representative sections are shown in this figure.

Download figure to PowerPoint

To interrogate whether DGAT1 expression is detectable in the enteroendocrine cells, GLP-1 or GIP was utilized as L or K cell marker, respectively. Antibodies against DGAT1 and GLP-1 or GIP were used for the immunohistochemical staining to evaluate the colocalization of DGAT1 with GLP-1 or GIP in target cells. It has been previously reported that GLP-1 positive cells are enriched at the distal portion of intestine and GIP expressing cells are predominantly located in the mid portion of small intestine [23]. Similar observations were made in this study. However, survey of all sections across entire intestine reveals lack of colocalization of DGAT1 with GLP-1 (Figure 6C and D; Supporting Information Figure S3) or with GIP (Supporting Information Figure S4). Taken together, these data support the hypothesis of an indirect mechanism of communication between the enterocytes and enteroendocrine cells likely driving the gut peptide release mediated by DGAT1 inhibition.

Alternative carbon channeling due to DGAT1 inhibition

To determine the carbon flow upon inactivation of DGAT1, we interrogated the fate of DGAT1 substrates in response to DGAT1 inhibition. In contradiction to the expectation that the DGAT1 substrate, DG, would accumulate following DGAT1 inhibition, levels of DG in tissues of Dgat1−/− mice remain unchanged [24, 26]. Similarly, acute treatment of pharmacological inhibitor of DGAT1 yielded unchanged levels of DG in the heart [24]. This observation led to the hypothesis that carbon flow is diverted to a different pathway following DGAT1 inhibition. To address this hypothesis, HT29 and C2C12 cells were treated with various concentrations of compound K followed by pulsing with 13C-labeled oleic acid. Newly synthesized 13C labeled lipid species were measured at 2 h after tracer loading. Compound K treatment led to minimal accumulation of DG in HT-29 cells at all concentrations (Figure 7A). Surprisingly, modest but significant reduction in DG levels was observed in C2C12 cells after compound K treatment (Figure 7B). While addition of compound K resulted in small to modest changes in DG concentration in both cells types, the levels of phosphatidylcholine (PC) were elevated in these cells in a concentration-dependent manner (Figure 7C and D), suggesting a divergent route for substrate carbon flow during DGAT1 inhibition.

image

Figure 7. Alternative carbon channeling using a stable isotope as a tracer. HT-29 (A) and C2C12 cells (B) were preincubated with various concentrations of compound K for 30 min followed by the loading of [13C18]-oleic acid. Changes of levels of DG in the cells after compound K treatment were measured. One of labeled DG species, DG13C18:1-13C18:1, was shown to represent DG levels. A significantly different profile of DG was observed in HT-29 compared with C2C12 cells. One of representative labeled PC species, PC13C18:1-13C18:1, was shown after compound K treatment for HT-29 (C) and C2C12 cells (D). PC levels were enhanced with increasing concentrations of compound K. Changes of other PC species including PC16:0-13C18:1 and PC13C18:1-C18:1 exhibited the similar trend (data not shown). Results are expressed as mean ± SEM. IC50 for DG in C2C12 cells: 15 ± 6 nM. EC 50 for DG in HT-29 cells: >5,000 nM. EC50 for PC in C2C12 cells: 14 ± 2 nM. EC50 for PC in HT-29 cells: 130 ± 7 nM. The above study was repeated three times. Diet-induced obese C57/BL6 mice were chronically dosed with vehicle or compound K in feed for three weeks. Mice were then fasted for 18 h followed by an intravenous bolus of intralipid containing 50 mg/kg [13C18]-oleic acid via tail vein injection. Plasma PC levels at 0, 15, 30, 60 min in vehicle and compound K treated groups (E). Plasma PC levels area under curve (F) (1–60 min). PC levels were significantly higher in compound treated group compared with vehicle group. n = 10 mice/group, **P < 0.005. The statistical analysis was performed using Student t-test.

Download figure to PowerPoint

To determine whether the carbon re-channeling observed in cellular system also occurs in vivo, 13C-labeled oleic acid was administrated via intravenous bolus to mice treated with compound K (3 mg/kg) for 21 days. Labeled lipid species from the plasma were analyzed via LC/MS analysis. Consistent with the observations in the cellular system, plasma DG levels were unaltered (data not shown), whereas plasma PC levels were significantly increased to over 50% in compound K treated group compared with vehicle control (Figure 7E and F). Thus, both in vivo and in vitro data sets confirmed the observations that DG is being shunted toward PC or potentially other lipid synthetic pathways upon inhibition of DGAT1.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

In this report, we describe potent selective DGAT1 inhibitors, compound K and L, structurally diverse from previously reported compounds. Compound K-treated mice exhibited mechanism based weight loss and reduction of FI. PPTG was employed as a surrogate target engagement marker at the end of the study, at which compound K had reached trough levels in plasma. Interestingly, at doses where statistically significant weight reduction occurred, TG excursion was nearly completely inhibited. At the lower dose (0.3 mg/kg), the level of TG excursion was decreased to approximately 50%, whereas weight reduction was not observed at this dose. These data suggest that nearly complete inhibition of DGAT1 is required to achieve the beneficial effects on BW. Modest reduction of FI was observed with compound K treatment, suggesting that FI, at least in part, contributes to the weight loss. Although increased energy expenditure was reported in Dgat1−/− mice [9, 25], in our study, acute treatment of compound L led to no significant effect on energy expenditure. It remains possible that chronic DGAT1 inhibition could lead to increased energy expenditure. However it does imply that the pharmacological inhibition of DGAT1 may differentiate from germline deletion of Dgat1. Further investigations are needed to elucidate this possibility.

Although the mechanism by which DGAT1 inhibition leads to weight loss is not fully understood, there have been several reports interrogating the site of action of DGAT1 inhibition as it relates to BW reduction. Lee et al [12] described a transgenic mouse model with intestine-specific Dgat1 expression in a Dgat1 null background. These mice were reported to attenuate the resistance of DIO compared to Dgat1 null mice, suggesting that inhibition of DGAT1 in the gut may be the primary driver for BW regulation. However, other studies suggest that adipose could be a key site of action, as demonstrated by the transplantation of Dgat1-deficient white adipose tissue into mice prone to obesity and insulin resistance. The Dgat1−/− adipose recipient WT mice acquired DIO resistance and increased insulin sensitivity [27]. Additionally, the reported prolonged release of satiety hormones including PYY and GLP-1 as well as delayed GE [13] could also be the factors contributing to the metabolically beneficial effects of Dgat1 deficiency. Our studies demonstrated that administration of compound K led to delay of GE, greater and prolonged release of GLP-1 and PYY as well as attenuation of GIP release in rodent and canine models. The mechanism by which DGAT1 inhibition leads to delay in GE remains elusive. It is possible that GE delay is mediated directly by GLP-1 action, or perhaps stimulated by higher levels of FFAs in the lumen. The delay in GE and/or the prolongation of GLP-1 and PYY release could potentially explain DGAT1 blockade-induced suppression of FI and loss of BW. Further studies in mice with a deletion of GLP-1 receptor are underway to establish the correlation of magnitude and the duration of delay in GE, elevation of GLP-1 with the efficacy of BW reduction occurred following the treatment with DGAT1 inhibitor. It is encouraging to see a similar profile of gut peptide release and GE delay in both mouse and canine models, indicating that the observation could potentially be translatable to the clinic.

The exact mechanism of GLP-1/PYY elevation and GIP attenuation through DGAT1 inhibition remains to be investigated. Immunohistochemistry evaluation indicates that DGAT1 does not co-express with enteroendocrine markers GLP-1 or GIP, suggesting that the prolonged secretion of GLP-1 and PYY or inhibition of GIP release is likely not mediated by a direct inhibition of DGAT1 in the enteroendocrine cells. Rather, we propose that inhibition of TG synthesis via DGAT1 blockade alters luminal distribution of lipids leading to modulation of gut peptide secretion by an indirect mechanism. We have observed delay of fat absorption with compound K treatment, which could lead to higher levels of FFAs in the L cell enriched, distal portion of the intestine. These FFAs could potentially serve as agonistic factors to interact with G-protein coupled receptors with GLP-1 and/or PYY secretagogue functions, including GPR119 and GPR120, on L cells [28, 29]. These interactions may potentiate the release of incretin peptides. On the other hand, attenuated levels of FFAs in the proximal portion of intestine could contribute to the reduced level of GIP. Additional studies are required to link the connection between the inhibition of TG synthesis by DGAT1 inhibitor and the alteration of incretin release.

The primary function of orlistat, marketed as a weight reducing agent, is to prevent fat absorption which is also the cause of an array of reported gastro-intestinal adverse effects [22]. To compare the effect of DGAT1 inhibitor to orlistat on lipid excretion, we conducted a study for the direct comparison of compound K and orlistat's effects on BW and lipid levels in feces. DIO mice were administered chronically with either compound K at doses that led to maximal weight loss efficacy or with orlistat at comparable clinically efficacious dose. In comparison to compound K treated mice, orlistat treated mice demonstrated slight increase in FI, which is consistent with accelerated GE and enhanced energy intake observed in previous human studies [22, 30]. Despite these observations, orlistat treatment led to significant weight loss that was comparable to that obtained with compound K. As observed in the clinic, a profound increase in fecal fat excretion was demonstrated in orlistat treated group. In contrast, DGAT1 inhibition by compound K yielded minor increases in fecal fat and this was observed only at higher doses. With only modest weight loss effect, orlistat exhibits minimal margin regarding risk versus benefit ratio clinically. Importantly, compound K's markedly smaller effect on fecal lipids relative to orlistat occurred at doses with comparable BW loss. Our studies suggest that a DGAT1 inhibitor's effect on BW would be commensurate with a clinically relevant dose of orlistat but with a superior tolerability profile.

In contrast to the expectation that inhibiting DGAT1 will result in accumulation of DG levels, we and others have demonstrated that DG levels remain unaltered upon DGAT1 inhibition. The fate of the carbons upstream of DGAT1 remains a critical question to address. The utilization of a stable labeled tracer approach allowed us to interrogate the outcome of carbon flow upon DGAT1 inhibition in vitro and in vivo. HT-29 and C2C12 cells were selected to represent gut epithelial cells from human origin and skeletal muscle cells from mouse, respectively. We demonstrated that carbon flow is diverted toward phospholipid upon DGAT1 inhibition. While DG levels remain largely unchanged, different trend was observed in these cells, implicating a possible differential capacity of metabolizing DG in intestinal and muscle derived cells. Consistently, the EC50 of PC production is approximately tenfold more potent in C2C12 cells than in HT-29 cells. In addition, mouse primary enterocytes were investigated in our studies and the trend was similar to HT-29 cells (data not shown).

Blockade of DGAT1 activity promoted cellular PC synthesis in a compound concentration-dependent manner. Consistent with these observations, elevated PC levels in plasma was also demonstrated in vivo after chronic treatment of compound K. These findings may have significant physiological implications. Since increased DG levels in skeletal muscle was shown to exacerbate muscle insulin resistance [31], reduction of DG level due to alternative carbon channeling, through DGAT1 inhibition, could lead to improved insulin sensitivity. Moreover, PC is a crucial component of high density lipoprotein (HDL), thus implying a potential additional benefit in cardiovascular health via DGAT1 inhibition. The mechanism leading to altered DG levels, increased production of PC, and potentially other lipid species remains to be further investigated. It is likely that the kinetics of DG production and/or consumption varies in different cell types. Since DG has been shown to be a signaling molecule in a variety of signaling pathways [32], it is possible that DG differentially mediates cellular functions in these cell types.

Much of the interest in DGAT1 as a target for development of small molecule inhibitors resulted from reports describing the metabolic benefit of Dgat1−/− mice. The identification of selective and potent small molecule inhibitors allows interrogation of the mechanism of inhibition of DGAT1 and the effects on its upstream and downstream pathways. More importantly, the development of these and other small molecule DGAT1 inhibitors are key steps toward greater understanding the underlying mechanism of DGAT1 inhibition in conjunction with potential translation to clinical efficacy.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

All work described here was performed by authors while employed by Merck & Co., Inc. The authors would like to thank Timothy He for his contribution to the studies related to this manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
aic13789-sup-0001-SuppInfo.gif10KSupporting Information Figure S1
aic13789-sup-0002-SuppInfo.tif880KSupporting Information Figure S2
aic13789-sup-0003-SuppInfo.JPG196KSupporting Information Figure S3-S4
aic13789-sup-0004-SuppInfo.tif824KSupporting Information Figure S5

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.