Dipeptidyl Peptidase Inhibition Prevents Diastolic Dysfunction and Reduces Myocardial Fibrosis in a Mouse Model of Western Diet Induced Obesity

Abstract

Objective

Consumption of a high-fat/high-fructose Western diet (WD) is linked to rising obesity and heart disease, particularly diastolic dysfunction which characterizes early obesity/metabolic cardiomyopathy. Mounting evidence supports a role for inflammation, oxidative stress and fibrosis in the pathophysiology of metabolic cardiomyopathy. Dipeptidyl peptidase-4 (DPP-4) is a circulating exopeptidase recently reported to be elevated in the plasma of patients with insulin resistance (IR), obesity and heart failure. We hypothesized that a model of WD induced obesity/metabolic cardiomyopathy would exhibit increased DPP-4 activity and cardiac fibrosis with DPP-4 inhibition preventing cardiac fibrosis and the associated diastolic dysfunction.

Materials/Methods

Four-week-old C57BL6/J mice were fed a high-fat/high-fructose WD with the DPP-4 inhibitor MK0626 for 16 weeks. Cardiac function was examined by high-resolution cine-cardiac magnetic resonance imaging (MRI). Phenotypic analysis included measurements of body and heart weight, systemic IR and DPP-4 activity. Immunohistochemistry and transmission electron microscopy (TEM) were utilized to identify underlying pathologic mechanisms.

Results

We found that chronic WD consumption caused obesity, IR, elevated plasma DPP-4 activity, heart enlargement and diastolic dysfunction. DPP-4 inhibition with MK0626 in WD fed mice resulted in >75% reduction in plasma DPP-4 activity, improved IR and normalized diastolic relaxation. WD consumption induced myocardial oxidant stress and fibrosis with amelioration by MK0626. TEM of hearts from WD fed mice revealed abnormal mitochondrial and perivascular ultrastructure partially corrected by MK0626.

Conclusions

This study provides evidence of a role for increased DPP-4 activity in metabolic cardiomyopathy and a potential role for DPP-4 inhibition in prevention and/or correction of oxidant stress/fibrosis and associated diastolic dysfunction.

Introduction

The World Health Organization estimates that more than 1.4 billion people were overweight in 2008 and current trends predict a doubling of the overweight population by 2030. Worldwide, more deaths are now linked to being overweight or obese than underweight [1, 2]. The rise in obesity and insulin resistance appears to parallel increased consumption of a Western diet (WD) high in fat and fructose [3]. Accumulating experimental evidence supports a role for components of the WD in the development of heart disease that is characterized initially as delayed myocardial relaxation in diastole [4–6]. The driving forces in the pathogenesis of this diastolic dysfunction include; increased myocardial inflammation, oxidative stress and interstitial fibrosis [7–9]. Unfortunately, no evidenced based treatments exist to reduce the mortality of the diastolic dysfunction coupled to obesity and associated with these metabolic abnormalities [10–13]. Systemic inflammation is being increasingly implicated in the pathology of diastolic dysfunction particularly of over-nutrition induced metabolic cardiomyopathy [14–17]. The linkage of metabolic cardiomyopathy with inflammation has spurred research into new treatments for diastolic dysfunction of obesity and insulin resistance (IR) [18, 19]. Therapies aimed primarily at improving insulin responsiveness and substrate utilization in diastolic dysfunction have yielded mixed results [20–22]. However, recent research has uncovered an important association between increased dipeptidyl peptidase-4 (DPP-4) cellular expression/plasma activity and inflammation in states of obesity, insulin resistance and diabetes [23–25].

DPP-4 is a circulating exopeptidase that cleaves and inactivates proteins at X-proline dipeptides residues. Inhibitors of DPP-4 have a role in the treatment of diabetes by preventing DPP-4 cleavage of glucagon like peptide-1 (GLP-1) and glucose-dependent insulinotrophic peptide (GIP) which stimulate insulin secretion and suppresses glucagon release [26, 27]. Additionally, DPP-4 is a pleiotropic enzyme with abundant expression on multiple cell types including T cells and macrophages. DPP-4 expression on these immune cells is increased in states of obesity, IR and in poorly controlled diabetes [28, 29]. DPP-4 inhibition has also been shown to reduce inflammation in the setting of IR and cardiovascular disease (CVD) [25]. While emerging evidence supports a role for DPP-4 in both impaired insulin action and inflammation, few studies have looked at DPP-4 in diastolic dysfunction. Recent studies in diabetic patients found a relationship between increased DPP-4 activity and diastolic dysfunction as well as an improvement in cardiac function following DPP-4 inhibition [30, 31]. We have also recently reported improvement in diastolic dysfunction in the insulin resistant Zucker obese rat model after chronic DPP-4 inhibition [32]. However, the effects of DPP-4 inhibition on left ventricular (LV) function in the metabolic cardiomyopathy of over-nutrition with a WD remain unexplored. Moreover, oxidant stress and impaired insulin action play crucial roles in cardiac tissue remodeling with interstitial fibrosis and inflammation contributing to impaired diastolic relaxation [4]. However, the potential effects of DPP-4 inhibition on fibrosis and oxidant stress in over-nutrition induced cardiomyopathy are largely unknown.

Accordingly, we posited that chronic DPP-4 inhibition in the setting of a WD would attenuate the development of insulin resistance and cardiac diastolic dysfunction by reducing oxidant stress, inflammation and associated myocardial fibrosis. To test this hypothesis, mice were fed a WD, high in fat and fructose, for 16 weeks in the presence or absence of chronic treatment with a DPP-4 inhibitor in their diet. After 16 weeks of feeding (20 weeks of age), mice were examined for weight gain, heart weight increase, IR and DPP-4 activity. After baseline characterization, we performed cine MRI assessment of heart function followed by detailed assessment for fibrosis, inflammation and oxidant stress.

Methods

Animals

C57BL/6J mice were purchased from Jackson Laboratories and cared for in accordance with National Institutes of Health guidelines. All procedures were approved in advance by the Institutional for Animal Care and Use Committee of the University of Missouri. Male mice were used for this study and divided into 4 groups; control diet (CD, n=22), control diet with DPP-4 inhibitor (CD+MK, n=20), Western diet (WD, n=22) and Western diet with DPP-4 inhibitor (WD+MK, n=22). The WD consisted of a previously published diet of high fat (46%) and high carbohydrate (41.8%) with sucrose (17.5%) and high-fructose corn syrup (17.5%) (Test Diet modified 58Y1). The control diet consisted of fat (10.2%) and carbohydrate (67.4%) with sucrose (33%) and no high fructose corn syrup similar to standard rodent chow (Test Diet 58Y2). Differences in total sodium were negligible in the two diets (supplementary table 1) [8]. MK0626, a new DPP-4 inhibitor provided by Merck pharmaceuticals, was added to mouse chow so that the final concentration in chow was 33 mg MK0626•kg⁻¹ chow to achieve a dose and plasma level of approximately 10 mg•kg ⁻¹•day⁻¹ and 100 nM, respectively. This dose was based on previous developmental studies of this DPP-4 inhibitor and clinical studies on the related FDA approved DPP-4 inhibitor, sitagliptin [33].

In Vivo Cine Magnetic Resonance Imaging (MRI)

Noninvasive magnetic resonance imaging (MRI) scans were performed using a Bruker AVANCE III BioSpec 7 T horizontal bore MRI (Bruker Corp., Billerica, MA) equipped with a four-channel phased array mouse cardiac radiofrequency coil. Cine MRI of the left ventricle was acquired as previously described [9]. Animals were weighed and anesthetized using 1.8–2.7% isoflurane on a nose-cone non-rebreathing system supplying continuous oxygen.

Anatomical and biochemical studies

After 16 weeks of feeding, mice underwent body composition analysis for whole body fat mass, lean mass and total body water utilizing an EchoMRI-500 for quantitative magnetic resonance analysis (Echo Medical Systems, Houston, TX, USA). Briefly, mice were placed into a thin-walled plastic container while awake and under no anesthesia or distress. All measurements were performed during the same time of day. Mice were then weighed and euthanized via exsanguination under isoflurane anesthesia (above). Heart weights and visceral fat weights were obtained after harvesting along with tibial lengths measured to normalize weights and eliminate confounding effects of differences in size. A venous blood sample was collected from a subset of fasting mice in each treatment group and plasma was stored at −80°C for glucose and insulin assay and homeostatic model assessment of insulin resistance (HOMA-IR) as previously described [9]. DPP-4 activity assay in plasma and the myocardium was determined using fluorogenic substrate, H-Ala-Pro-AFC according to published protocol [32].

Histologic staining and immunohistochemistry

Five micron sections of the LV were stained with picrosirus red according to manufacturing procedure and analyzed according to our published protocol for interstitial and periarterial fibrosis [34]. To detect the presence of reactive nitrogen species (RNS), LV tissue was fixed, embedded in paraffin and immunostained with an antibody to 3-nitrotyrosine (3-NT) (AB5411; 1:150 dilution; Millipore, Billerica, MA) as previously described [35]. Immunohistochemistry was performed according to previously published protocols using antibodies previously described for collagen 1 (Col 1, ab6308; 1:100; Abcam), collagen 3 (Col 3, ab23746; 1:50; Abcam), transforming growth factor- β (TGF-β, ab92486; 1:50; Abcam) and monocyte chemoattractant protein-1 (MCP-1, SC-28879; Santa Cruz).

Ultrastructure analysis with transmission electron microscopy

Details of LV tissue preparation, sectioning, staining and viewing are as previously described [8]. A JOEL 1400-EX transmission electron microscope (Joel Ltd. Tokyo, Japan) was utilized to review three fields randomly chosen per mouse to obtain three 2,000 X images/LV.

Statistical Analysis

Results are reported as the mean ± SEM. One-way ANOVA and post hoc t-tests (Fisher’s LSD) were performed to examine differences in outcomes between CD, CD+MK, WD and WD+MK groups (SPSS v22.0, IBM software). All differences were considered significant when p < 0.05.

Results

DPP-4 inhibition decreased plasma DPP-4 activity and normalized heart enlargement associated with a WD

DPP-4 activity has been shown to be elevated in the setting of obesity and insulin resistance [36, 37]. We found a significant increase in plasma DPP-4 activity in WD compared with CD (31.8 ± 2.0 for WD compared to 24.7 ± 3.9 for CD, p < 0.05). Treatment with the DPP-4 inhibitor MK0626 markedly lowered plasma DPP-4 activity to ~20% of control levels (4.9 ± 0.7 for CD+MK and 5.7 ± 0.9 for WD+MK, p < 0.001) in both control diet and western diet fed mice (Figure 1A). HOMA-IR revealed significant insulin resistance in WD fed mice when compared to CD feeding (6.47 ± 1.14 in WD and 3.00 ± 0.27 for CD, p < 0.01 for WD compared to all other groups). MK0626 treatment of WD fed mice corrected HOMA-IR back to control levels (4.10 ± 0.59 for WD+MK, p value < 0.05 for WD+MK vs. WD) (Figure 1B). WD feeding also caused a significant increase in normalized heart weight (72.1 mg/mm ± 2.9 mg/mm for CD vs. 83.0 mg/mm ± 4.1 mg/mm in WD, p < 0.05). DPP-4 inhibition with MK0626 normalized heart weight back to control level (73.2 mg/mm ± 2.6 mg/mm for WD+MK and 70.6 mg/mm ± 3.3 mg/mm for CD+MK, p < 0.05 for WD+MK vs. WD) (Figure 1C). Bodyweight was significantly increased after 16 weeks of WD feeding (15.9 g/mm ± 0.2 g/mm for CD compared to 21.9 mg/mm ± 0.7 mg/mm for WD, p < 0.05 for WD compared to all other groups). MK0626 treatment of WD fed mice significantly improved body weight but not back to control levels (20.2 g/mm ± 0.7 g/mm for WD+MK, p < 0.05 for WD+MK compared to all other groups) (Figure 1D). To more robustly characterize the effect of WD feeding on phenotype, we examined visceral fat weight, body composition, fasting glucose and insulin. We found significant increase in fat consistent with obesity and no changes in lean body weights. Fasting blood glucose levels were significantly lower in WD mice treated with MK0626 but no significant changes were noted in insulin levels. (Table 1).

Figure 1. Effect of DPP-4 inhibition in setting of WD feeding on DPP-4 plasma activity, insulin resistance, heart weight and bodyweight.

A) DPP-4 plasma enzyme activity was significantly increased in mice fed a western diet (WD) compared with control diet (CD). CD group treated with MK0626 (CD+MK) and WD group treated with MK0626 (WD+MK) showed decreased DPP-4 activity by more than 75% in both groups. N = 3, 8, 10, & 7 respectively for CD, CD+MK, WD, and WD+MK. B) HOMA-IR revealed insulin resistance in WD compared to CD. Treatment with MK0626 improved IR in WD fed mice. N = 4 for all groups. C) Normalized heart weight was significantly increased by WD compared to CD. Treatment with MK0626 ameliorated this increase in heart weight. N = 22, 20, 22, and 22 respectively for CD, CD+MK, WD, and WD+MK. D) Normalized bodyweights showed significant increase with WD compared to CD. Treatment of WD fed mice with MK0626 improved body weight significantly but not to control diet level. N = 22, 20, 22, & 22 respectively for CD, CD+MK, WD, and WD+MK. (* indicates p < 0.05 for WD compared with all other groups, # denotes p < 0.001 for CD+MK and WD+MK vs. CD and WD, ^† indicates p < 0.05 for WD+MK compared to all other groups.)

Table 1.

Phenotypic Parameters

Phenotypic Variables	CD	CD+MK	p value1 (CD+MK vs CD)	WD	p value1 (WD vs CD)	WD+MK	p value1 (WD+MK vs WD)
Body Weight (g)2	27.8 ± 1.7	28.6 ± 3.1	0.459	38.6 ± 4.1	<0.001	35.8 ± 5.1	0.013
Heart Weight (mg)2	126.0 ± 21.2	123.0 ± 27.3	0.700	146.4 ± 29.9	0.010	130.2 ± 22.7	0.038
Visceral Fat Mass (g) 3	0.69 ± 0.16	1.11 ± 0.21	0.422	2.48 ± 0.26	0.001	2.40 ± 0.40	0.842
Total Body Fat Mass (g) 4	3.1 ± 0.2	4.3 ± 0.9	0.381	9.3 ± 1.1	<0.001	9.9 ± 1.0	0.618
Total Body Lean Mass (g) 4	23.0 ± 0.3	24.2 ± 0.5	ND	23.1 ± 0.7	ND	23.5 ± 0.4	ND
Fasting Plasma Glucose (mg/dL) 5	315 ± 31	277 ± 46	0.766	423 ± 13	0.056	329 ± 27	0.007
Fasting Plasma Insulin (ng/dL) 5	215 ± 3	232 ± 17	ND	343 ± 56	ND	278 ± 20	ND

¹p values results from post-hoc testing by LSD following ANOVA multi-group comparison with p <0.05.

²N = 22, 20, 22 and 22 for CD, CD+MK, WD and WD+MK respectively.

³N = 5, 9, 11 and 10 for CD, CD+MK, WD and WD+MK respectively.

⁴N = 7, 6, 8 and 7 for CD, CD+MK, WD and WD+MK respectively.

⁵N = 4 for all groups.

ND = post-hoc testing not done due to ANOVA p value >0.05

WD induced cardiac diastolic dysfunction that was prevented by MK0626 treatment

Utilizing high resolution cine-MRI we found a significant prolongation of cardiac diastolic relaxation time when mice were subjected to WD feeding (35.3 ms ± 1.0 ms for WD compared to 26.0 ms ± 1.8 ms for CD, p < 0.001). Treatment with DPP-4 inhibitor MK0626 completely normalized diastolic relaxation time altered by WD feeding (24.4ms ± 0.7 ms for WD+MK and 24.1 ms ± 1.9 ms for CD+MK, p < 0.001 for WD+ MK compared to WD) (Figure 2B). WD feeding also markedly reduced initial LV filling rate compared to CD fed mice (0.428 μl/μs ± 0.048 μl/μs for CD compared to 0.135 μl/μs ± 0.028 μl/μs in WD, p < 0.01). MK0626 treatment of WD fed mice corrected the initial LV filling rate to control levels (0.529 μl/μs ± 0.049 μl/μs for WD+MK and 0.753 μl/μs ± 0.149 μl/μs for CD+MK, p < 0.01 for WD+MK compared to WD) (Figure 2C). Peak filling rate was not significantly altered by WD feeding and no effect was seen with MK0626 treatment (1.094 μl/μs ± 0.130 μl/μs, 1.117 μl/μs ± 0.127 μl/μs, 1.074 μl/μs ± 0.099 μl/μs, and 0.944 μl/μs ± 0.037 μl/μs for CD, CD+MK, WD, and WD+MK respectively) (Figure 2D). Figure 2A shows representative early diastolic images from cardiac cycles of WD and WD+MK treated mice, highlighting the delayed filling of LV from impaired relaxation that was corrected by MK0626 treatment. Systolic function was unaffected by WD consumption (Supplemental Table 1).

Figure 2. WD caused cardiac diastolic dysfunction that was prevented by DPP-4 inhibition.

A) Representative mid-ventricle short-axis cine-MRI images correspond to end-diastole, end-systole, and early diastole phases (frame 1 and 9–12 of a total of 16 frames) of cardiac cycle from WD fed mouse (WD, upper row) and WD fed mouse treated with MK0626 (WD+MK, lower row). B) Diastolic relaxation time was significant prolonged in mice feed a WD. MK treatment completely normalized diastolic relaxation time. C) Initial filling rate was significantly lower with WD feeding. Treatment with MK0626 normalized initial filling rate. D) There was no change in the peak filling rate with WD feeding. N = 5, 8, 5 and 8 for CD, CD+MK, WD, and WD+MK, respectively for all panels. (* denotes p < 0.05 for WD compared to all other groups.)

MK0626 abolished the myocardial fibrosis associated with WD

Myocardial fibrosis was markedly increased, by picrosirus red staining, after WD feeding (p < 0.001 for WD vs. CD). DPP-4 inhibition with MK0626 prevented the accumulation of interstitial fibrosis (p < 0.01 for WD+MK vs. WD). There was no significant change in interstitial fibrosis on CD fed animals treated with MK0626 (Figures 3A and 3B). Analysis of perivascular myocardium, to characterize the degree of fibrosis, revealed significant increases in picrosirus red staining around the vasculature of WD fed mice (p < 0.05). Treatment with MK0626 ameliorated perivascular fibrosis in mice consuming a WD (p < 0.01 for WD+MK vs. WD). Again, MK0626 treatment had no affect on perivascular fibrosis in CD fed mice. (Figures 3C and 3D).

Figure 3. WD feeding resulted in myocardial fibrosis that was prevented by DPP-4 inhibition.

A) Representative images of myocardial staining for interstitial fibrosis by picrosirus red staining. Pink color represents collagen deposition. B) Quantification of interstitial collagen deposition by average gray scale intensities revealed significant increases in interstitial fibrosis in mice fed a WD compared with all other groups. MK0626 treatment abrogated the increase in interstitial fibrosis. C) Representative images of picrosirus staining for perivascular fibrosis. D) Myocardial perivascular fibrosis was significantly increased in mice fed a WD by quantification of average gray scale intensities from picrosirus red stained sections. MK0626 treatment prevented the increase in perivascular fibrosis. N = 4, 5, 4 and 5 for all panels for CD, CD+MK, WD, and WD+MK, respectively. (* indicates p<0.05 for WD vs. every other group.)

MK0626 treatment attenuated myocardial oxidative stress in WD fed mice

Oxidative stress is associated with diastolic dysfunction in models of obesity/metabolic cardiomyopathy [38, 39]. Analysis of oxidant stress by 3-NT in the myocardium revealed a significant increase with WD consumption (p < 0.001 for CD vs. WD). DPP-4 inhibition with MK0626 significantly lowered 3-NT staining intensity (p < 0.001 for WD+MK compared to WD). However, oxidative stress was not completely abrogated by MK0626 treatment as WD fed mice given MK0626 had significantly higher 3-NT staining compared to CD fed mice (p <0.01 for WD+MK vs. CD). MK0626 treatment of CD fed mice had no affect on oxidative stress. (Figure 4A and 4B).

Figure 4. WD feeding caused myocardial oxidative stress that was ameliorated by DPP-4 inhibition.

A) Representative images of myocardium stained for 3-NT, a marker of oxidant stress from accumulation of oxidant peroxynitrite (ONOO⁻). B) Quantification of 3-NT staining by average gray scale intensities showed marked increase in mice fed a WD that was normalized by MK0626 treatment. N = 5 for all groups. (* indicates p < 0.005 for WD compared to all other groups.)

Effect of WD and MK0626 treatment on collagen subtype

Collagen subtype changes also contribute to extracellular matrix (ECM) alterations that affect diastolic function. Col 1 increases myocardial stiffness while Col 3 contributes to a more elastic ECM [40, 41]. Analysis of myocardial collagen subtypes by immunohistochemistry revealed a trend toward increases in Col 1 in mice fed a WD with evidence for a reduction in Col 1 with MK0626 treatment. However, these shifts did not achieve statistical significance (p = 0.08) (Figures 5A and 5B). Col 3 staining showed a less robust trend, with no significant differences seen in WD consumption with or without MK0626 treatment (p = 0.322). Col 1 and Col 3 showed no significant increase in CD fed mice treated with MK0626. (Figures 5C and 5D).

Figure 5. Effect of WD feeding and DPP-4 inhibition on perivascular collagen subtype deposition.

A) Representative images of Col 1 staining of myocardial sections from CD fed mice and WD fed mice with and without MK0626. B) Quantification of Col 1 staining intensity showed a trend towards increased Col 1 deposition with WD feeding and decrease with MK0626 treatment (p = 0.085). N = 4, 5, 5, and 5 for CD, CD+MK, WD, and WD+MK groups, respectively. C) Representative images of Col 3 staining of myocardial sections. D) Col 3 staining showed a similar trend towards increased perivascular deposition with WD feeding and signs of improvement with MK0626 treatment (p = 0.322). N = 4, 5, 4, and 4 respectively for CD, CD+MK, WD, and WD+MK.

Ultrastructural abnormalities of the myocardium in WD feeding are improved by MK0626

Transmission electron microscopy (TEM) examination of myocardial ultrastructure revealed sarcomeric disorganization in the left ventricle. There was an increase in intermyofibrillar mitochondria, mitochondrial swelling with loss of cristae and increased mitochondrial electron density after WD feeding. Treatment with MK0626 did not improve sarcomeric disorganization or the alterations in mitochondrial biogenesis. No changes in sarcomeric arrangement or mitochondrial biogenesis were seen in control diet mice on MK0626 treatment (Figure 6A). Peri-capillary ultrastructure in the LV revealed a marked reduction in endothelial transcytotic vesicles after WD consumption. The capillary endothelium also showed regions of cytoplasmic thinning (open arrows) and dysmorphic transcytotic vesicles (closed arrows) after WD. MK0626 treatment of mice on a WD normalized endothelial transcytotic vesicles. Treatment of CD mice with MK0626 did not adversely alter perivascular ultrastructure in the LV (Figure 6B).

Figure 6. Ultrastructural analysis of myocardial architecture.

A) Representative transmission electron microscopy (TEM) images of sarcomere organization in LV. Top left panel showing control diet fed mouse (CD) with orderly rows of sarcomeres (S) with intermyofibrillar (IMF) mitochondria (Mt). Insert depicts electron dense Mt matrix and cristae at higher magnification. N, highlights nuclei. Z lines also highlighted. Top right panel showing CD fed mouse treated with MK0626 (CD+MK) with similar normal sarcomere arrangement and IMF with normal Mt architecture. Bottom left panel highlighting WD fed mouse (WD) with expansion and disorganization of sarcomeres. Mt also appeared enlarged and disorganized with loss of cristae and electron lucency. Bottom right panel shows WD mouse treated with MK0626 (WD+MK). MK0626 treatment failed to attenuate Mt expansion or ultrastructure. Additionally, sarcomeric disorganization was unchanged. Magnification X2000; bar = 1 μm. B) Representative TEM images of perivascular ultrastructure in LV. Top left panel illustrates CD fed mouse with normal endothelial transcytotic vesicles (white arrows). Top right panel demonstrates CD fed mouse on MK0626 treatment with unchanged endothelial transcytotic vesicles. Bottom left panel depicts a marked decrease in endothelial transcytotic vesicles in WD fed mice with capillary thinning indicative of immature capillaries. Bottom right panel showing normalization of endothelial transcytotic vesicles and capillary angiogenesis in WD fed mice treated with MK0626. CL = capillary lumen; dashed white lines – capillary abluminal endothelium; RBC = red blood cell; N = endothelial nucleus. Magnification X10,000; bar = 0.2 μm.

Discussion

Chronic overnutrition from a WD, high in fat and fructose, is increasingly linked to obesity, insulin resistance and increased systemic inflammation [3]. These changes carry an increased risk for CVD including heart failure (HF), initially manifesting as diastolic dysfunction [42–44]. While evidence based treatments exist for systolic dysfunction, strategies directed towards treating the diastolic dysfunction of metabolic cardiomyopathy are lacking [45]. DPP-4 inhibitors are increasingly being used for treatment of diabetes. Moreover, there is emerging evidence that these agents may have anti-inflammatory effects, which may provide protection from CVD [25, 27]. In this study, we investigated the potential of the novel DPP-4 inhibitor MK0626 in the treatment of chronic over-nutrition with a contemporary WD. In our investigation, chronic over-nutrition with a WD resulted in obesity, insulin resistance, elevated plasma DPP-4 activity as well as heart enlargement and dysfunction. Study of cardiac function with MRI revealed diastolic dysfunction with prolongation of LV diastolic relaxation and impaired initial filling rate. Systolic function was preserved after 16 weeks of WD feeding (Supplemental Table 1). Chronic treatment of WD fed mice with the DPP-4 inhibitor yielded marked inhibition of plasma DPP-4 activity and improved insulin sensitivity. IR was prevented by DPP-4 inhibition with improvement in body weight but, not back to control levels. Heart enlargement and cardiac diastolic dysfunction were also prevented by chronic DPP-4 inhibition. These findings recapitulate the effect of chronic WD consumption in humans, notably, the diastolic dysfunction seen with obesity and insulin resistance. Furthermore, these findings represent the first evidence for prevention of diastolic dysfunction with DPP-4 inhibition in a diet induced obesity model.

Multiple pathways have been implicated in the pathology of diastolic dysfunction including; LV fibrosis, increased oxidant stress, alterations in collagen content, inflammation and abnormal calcium handling. Oxidant stress and myocardial fibrosis have repeatedly been linked to impaired cardiac diastolic relaxation in rodent models of obesity and IR [9, 46–48]. In this investigation chronic WD consumption was associated with increased cardiac interstitial and perivascular fibrosis. Administration of the DPP-4 inhibitor along with the WD prevented this increased myocardial fibrosis. Multiple collagen isoforms exist with Col 1 and Col 3 isoforms representing the major components of the myocardial interstitial matrix. Changes in the levels of collagen isoforms are associated with alterations in passive mechanical properties of the heart with Col 1 a stiffer and stronger isoform than the more compliant Col 3 [41, 49]. We found a trend towards increased levels of Col 1 in WD but no significant differences were seen. Similarly, Col 3 showed no statistically significant change with WD feeding or DPP-4 inhibitor treatment. We have previously reported a role for increased oxidative stress in the underlying pathology of fibrosis and diastolic dysfunction with consumption of a WD [8]. Here, we showed marked elevation in 3-NT staining in myocardial tissue from mice fed a WD which was markedly attenuated with DPP-4 inhibition. Immunostaining for 3-NT was used as a surrogate for integrated accumulation of reactive oxygen species (ROS) and reactive nitrogen species (RNS) in cardiac tissue [47].

These findings, in the context of improved systemic IR with DPP-4 inhibition, are consistent with the notion that modulation of DPP-4 enzyme activity exerts cardiac tissue protective effects along with systemic effects on carbohydrate metabolism. Energy homeostasis and hormone regulation of carbohydrate metabolism, primarily mediated by leptin, play important roles in obesity [50]. Indeed, we recently reported evidence of increases in leptin levels with WD feeding [8]. Thus a recent study on the potential of the DPP-4 inhibitor sitagliptin to mediate direct beneficial effects on myocardial metabolism in the leptin receptor mutant db-/db-mouse provide some initial support on systemic effects on carbohydrate metabolism [51]. However, our study is the first report, to our knowledge, of improvement in myocardial derangements of oxidant stress and fibrosis with DPP-4 inhibition in a purely over-nutrition model of diastolic dysfunction.

To better understand the association of oxidant stress and fibrosis in WD induced diastolic dysfunction we examined ultrastructural changes in the myocardium with TEM. Indeed, mitochondrial dysfunction (increased uncoupling) may play a role in increased oxidant stress as well as diminished energy for myocardial functions [8]. We have previously shown altered sarcomeric organization and increases in abnormal/dysfunctional mitochondria with WD feeding [47]. In the current study, WD fed mice showed marked increases in mitochondria with altered architecture (electron lucent mitochondria with decreased matrix and cristae), which was not substantively corrected by DPP-4 inhibition. WD consumption did lead to a reduction in endothelial transcytotic vesicles with thinning of endothelial cytoplasm, which was improved with DPP-4 inhibition. Endothelial transcytosis has been shown in ischemia reperfusion models to be a mechanism of protection from oxidant stress [52]. The improvement in endothelial transcytotic vesicles with DPP-4 inhibitor treatment in combination with reduced oxidant stress suggest that DPP-4 inhibition may improve vascular angiogenesis and oxidant stress by promoting delivery of antioxidant molecules to the myocardium. A role for vascular angiogenesis and improved delivery of nitric oxide (NO) to limit the accrual of ROS has been shown recently in the db-/db- model of insulin resistance [53], highlighting the critical importance of oxidant stress and alterations in vascular architecture and function in the diastolic dysfunction of metabolic cardiomyopathy [54].

Increased DPP-4 activity has been linked to inflammation through activation of T cells and macrophages in multiple tissues. However, there has been little study of cardiovascular inflammation and DPP-4 [28]. We analyzed the pro-inflammatory TGF-β and MCP-1 which promote macrophage activation and collagen deposition. There were trends toward an increase in TGF- β and MCP-1 in the myocardium with WD consumption and improvement with DPP-4 inhibition that did not achieve statistical significant (Supplementary Figure 1). These findings suggest that DPP-4 inhibition may regulate inflammation in concert with a reduction in oxidant stress, prevent diastolic dysfunction in obesity and IR.

There are important limitations in our study for identifying the exact mechanisms underlying the prevention of diastolic dysfunction with DPP-4 inhibition. For instance, DPP-4 inhibition has an important role in lowering postprandial elevations in blood glucose which may be involved in the improvements seen in our study. WD fed mice treated with MK0626 did show a decrease in fasting blood glucose and insulin resistance. However, a comparative study with insulin administration or other control of glucose levels is required to elucidate the specific effect of glycemic control. Improvement in blood pressure has also been seen in diabetic patients treated with DPP-4 inhibitors [55]. However, experimental models of obesity treated with DPP-4 have shown no change in blood pressure [23]. We have previously published that this model develops increased blood pressure with WD feeding but found no effect of DPP-4 inhibition with MK0626 in this model of diet induced obesity [Nistala et al 2014 Endocrinology in press]. Nevertheless, we acknowledge limitations in the translation of data in the WD mouse model to that of overweight humans with metabolic cardiomyopathy.

Salt intake may also play a role in the pathology of diastolic dysfunction seen in WD induced obesity given the association with hypertension and activation of the renin-angiotensin-aldosterone system (RAAS) [56]. We limited our study to the effect of fat and carbohydrates thus, urther study is needed to better understand the contribution of salt on the RAAS and diastolic dysfunction. Mouse strain effects may also play a role in the changes seen in this study and additional work in other models of diet induced obesity and diastolic dysfunction are required to fully assess the potential of DPP-4 inhibition for prevention of diastolic dysfunction. Finally, additional study is needed to fully examine the fibrosis seen with WD consumption and how DPP-4 inhibition may modulate fibrosis. Examination of immune cell infiltration for changes in macrophage or T cell population is one example of other factors which may modulate fibrosis in WD induced obesity [19, 28].

In summary, this investigation found that preventative treatment with a DPP-4 inhibitor significantly lowered systemic DPP-4 activity, improved IR and completely prevented the diastolic dysfunction induced by chronic WD consumption. Improvements in oxidant stress and fibrosis in the heart appear to play a role in this protection but further studies are needed to fully elucidate the underlying mechanisms.

Abbreviations

WD

Western Diet

CVD

cardiovascular disease

IR

insulin resistance

MetS

metabolic syndrome

HF

heart failure

LV

left ventricle

ROS

reactive oxygen species

RAAS

renin-angiotensin-aldosterone system

GLP-1

glucagon-like peptide 1

GIP

glucose-like insulinotrophic peptide

DPP-4

dipeptidyl peptidase-4

CD

control diet

CD+MK

control diet and MK0626

WD+MK

western diet and MK0626

MRI

magnetic resonance imaging

HOMA-IR

homeostatic model of assessment of insulin resistance

RNS

reactive nitrogen species

Col 1

collagen type 1

Col 3

collagen type 3

TGF-β

transforming growth factor- β

MCP-1

monocyte chemoattractant protein-1

3-NT

3-nitrotyrosine

ECM

extracellular matrix

TEM

transmission electron microscopy

NO

nitric oxide