A Possible Mechanism: Vildagliptin Prevents Aortic Dysfunction through Paraoxonase and Angiopoietin-Like 3

The collected data have revealed the beneficial effects of dipeptidyl peptidase-4 (DPP-4) inhibitors on the vascular endothelium, including vildagliptin. However, the involved mechanisms are not yet clear. In this study, Sprague-Dawley rats were randomly divided into the following four groups: control, diabetic, diabetic + low-dose vildagliptin (10 mg/kg/d), and diabetic + high-dose vildagliptin (20 mg/kg/d). The diabetic model was created by feeding a high-fat diet for four weeks and injection of streptozotocin. Then, vildagliptin groups were given oral vildagliptin for twelve weeks, and the control and diabetic groups were given the same volume of saline. The metabolic parameters, endothelial function, and whole genome expression in the aorta were examined. After 12 weeks of treatment, vildagliptin groups showed significantly reduced blood glucose, blood total cholesterol, and attenuated endothelial dysfunction. Notably, vildagliptin may inhibit angiopoietin-like 3 (Angptl3) and betaine-homocysteine S-methyltransferase (Bhmt) expression and activated paraoxonase-1 (Pon1) in the aorta of diabetic rats. These findings may demonstrate the vasoprotective pathway of vildagliptin in vivo.


Introduction
The incidence and prevalence of diabetes mellitus are dramatically increasing worldwide [1]. Epidemiological research shows that diabetic patients have a higher risk of cardiovascular diseases [2]. The main cause of morbidity and mortality in diabetic patients is cardiovascular diseases.
Glucagon-like peptide-1 (GLP-1) is produced in gut Lcells. It contains 30 amino acids. The main physiological function of GLP-1 is stimulation of insulin secretion from pancreatic cells when glucose is orally taken up in the human body. GLP-1 can also inhibit glucose production and appetite, activate adipose and muscle glucose uptake and storage, and thus moderate insulin sensitivity. However, GLP-1 is quickly hydrolyzed by dipeptidyl peptidase-4 (DPP-4).
DPP-4 inhibitors are a new class of GLP-1 based antidiabetic drugs. As one type of DPP-4 inhibitors, vildagliptin controls blood glucose by inhibiting the enzymatic activity of DPP-4. DPP-4 is also found on endothelial cells in the cardiovascular system, and increasing research has focused on the benefit of DPP-4 inhibitors on cardiovascular function. Sitagliptin (one type of DPP-4 inhibitor) has been proven to significantly attenuate heart failure-related left ventricular (LV) end-diastolic pressure, systolic performance, and chamber stiffness in an ablation-induced cardiac dysfunction rat model [3]. In a clinical trial, sitagliptin enhanced global and regional LV function in type 2 diabetes mellitus (T2DM) patients with coronary artery [4]. Vildagliptin exerts cardioprotective effects in obesity-based insulin resistance [5,6], myocardial infarction (MI) [7], and ischemia-reperfusion (I/R) injury rat models [8]. However, the exact mechanism of the beneficial effect of vildagliptin on the aorta in diabetic rats remains to be elucidated.
In this study, we hypothesized that vildagliptin improved aorta function through multiple pathways. We employed a whole genomic expression array and bioinformatics method to explore the pathway involved in aorta vascular function moderation in diabetic rats.

Animal Treatments and Diets.
Five-week-old male Sprague-Dawley rats were obtained from the Institute of Laboratory Animal Science, Chinese Academy of Medical Sciences, and Peking Union Medical College (Beijing, China, SCXK-2014-0013). The animal protocol was approved by the Animal Care Committee of the Peking Union Medical Hospital Animal Ethics Committee (Project XHDW-2015-0051, 15 Feb 2015), and all efforts were made to minimize suffering. All the rats were fed in a light/dark cycle (12 hours : 12 hours) environment and were free to drink water. Three days after arrival, rats were randomly divided into four groups ( = 6 per group): normal control group, diabetic group, low-dose vildagliptin (vil-low), and highdose vildagliptin (vil-high). The normal control group was fed a standard rodent diet (kcal%: 10% fat, 20% protein, and 70% carbohydrate; 3.85 kcal/gm). Other groups were fed a high-fat diet (kcal%: 45% fat, 20% protein, and 35% carbohydrate; 4.73 kcal/gm, Research Diet, New Brunswick, NJ, USA). After 4 weeks, diabetic, low-dose vildagliptin, and high-dose vildagliptin groups were given a single injection of streptozotocin (STZ, 30 mg/kg body weight, i.p., Sigma-Aldrich, St. Louis, MO, USA). Fasting blood glucose > 11.1 mmol/L was the standard for the diabetic model. Then, vil-low and vil-high groups were treated with 10 mg or 20 mg vildagliptin (Novartis Pharma AG, Basel, Switzerland)/kg of body weight by daily gavage for 12 weeks. Normal control and diabetic groups were given normal saline. After 12 weeks of treatment, the rats were anesthetized using ketamine (100 mg/kg i.p., Pharmacia and Upjohn Ltd., Crawley, UK), followed by withdrawal of food overnight. Blood samples were obtained from the abdominal aorta. Then, the rats were sacrificed by decapitation. The thoracic aorta was quickly removed. Some aortas were placed in Krebs solution (120 mmol/L of NaCl, 4.7 mmol/L of KCl, 1.18 mmol/L of KH 2 PO 4 , 2.25 mmol/L of CaCl 2 , 24.5 mmol/L of NaHCO 3 , 1.2 mmol/L of MgSO 4 ⋅7H 2 O, 11.1 mmol/L of glucose, and 0.03 mmol/L of EDTA) and aerated with 95% O 2 and 5% CO 2 . Other aortas were frozen in liquid nitrogen and stored at −80 ∘ C for a gene microarray experiment.

Body Weight and Fasting Blood Glucose Measurements.
The rats were weighed every 4 weeks. Fasting blood glucose levels were measured by Bayer Contour TS glucometer (Hamburg, Germany).

Oral Glucose Tolerance Test (OGTT).
An OGTT was performed after 12 weeks of treatment. Blood glucose levels were measured at 30, 60, and 120 min after an oral administration of 20% glucose at a dose of 2 g/kg. The area under the curve (AUC) was calculated by the linear trapezoid method [9].
2.6. RNA Extraction, Amplification, Labeling, and Hybridization. To look for the differentially expressed genes under vildagliptin treatment in diabetic rats, we performed gene whole transcript-based array in vil-high group and diabetic group ( = 3 in each group). Total RNA was extracted from aortas using a mirVana6 RNA Isolation Kit (Ambion, San Paulo, SP, Brazil). The RNA was purified using an RNeasy Kit (Qiagen, Hilden, Germany), quantified by NanoDrop ND-2000 spectrophotometry (Nanodrop Tech, Rockland, Del, Wilmington, DE, USA), and qualified by agarose gel electrophoresis. Total RNA (100 ng) was used for cDNA synthesis. cRNA was synthesized followed by two-strand cDNA. Biotin-labeled cRNA was hybridized to an Affymetrix GeneChip Rat Gene 2.0 ST whole transcript-based array (Affymetrix Technologies, Santa Clara, CA, USA). After hybridization, the microarrays were washed, stained, and scanned with an Affymetrix Scanner 3000 7G (Santa Clara, CA, USA).
The microarray signals were analyzed using Expression Console software (version 1.4.1, Affymetrix, Santa Clara, CA, USA). The significance of the difference in genes was determined by one-way ANOVA. Differentially expressed genes were defined as genes with a fold change > 2.0 and < 0.05. The microarray raw data were submitted to the Gene Expression Omnibus (GEO) repository (GSE102196).

Bioinformatics Analysis for
Microarray. DAVID (Database for Annotation, Visualization and Integrated Discovery) software (https://david.abcc.ncifcrf.gov/) [11] was used to perform gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. The Search Tool for the Retrieval of Interacting Genes (STRING) database (https://string-db.org/) [12] was used to analyze the interaction network for differentially expressed genes.

Validation of Differentially Expressed
Genes by Quantitative PCR. The expression levels of three differentially expressed genes were measured by qPCR to validate the results of the microarray. All primers are listed in Table 1.
Actin was used as an internal control.

Statistical
Analyses. Data are presented as the means ± standard deviation (SD). One-way ANOVA analysis followed by Student's -test was used to compare differences among groups. < 0.05 was considered significant.

Vildagliptin Moderated Blood Glucose and Insulin
Resistance. After 12 weeks of treatment, vildagliptin significantly reduced fasting blood glucose AUC in OGTT ( < 0.05, Figures 1(b) and 1(f)). Additionally, fasting serum insulin and HOMA-IR index in the vildagliptin-treated group were lower than in the diabetic group ( < 0.05, Figures 1(c) and 1(d)).

Vildagliptin Depressed Vasorelaxation
Responses of Aortic Rings to Ach. As shown in Figure 1(k), endotheliumdependent vasodilation was impaired in diabetic rats compared with normal rats. Treatment with vildagliptin ameliorated this impairment ( < 0.01).

Identification of Differentially Expressed Genes.
After filtering genes that had a cutoff of a 1.5-fold change or greater, we found that 150 genes were upregulated and 120 genes were downregulated in the vil-high group compared with the diabetic group ( < 0.05).

GO Term Enrichment Analysis of Differentially Expressed
Genes. In general, 91 GO terms were significantly enriched ( < 0.05, Table 2). The top 20 most significant biological process (BP) terms are shown in Figure 2 ( < 0.01). They were negative regulation of endopeptidase activity, epoxygenase P450 pathway, acute-phase response, blood coagulation, oxidation-reduction processes, cytoskeleton organization, negative regulation of fibrinolysis, microtubule-based processes, phosphatidylcholine metabolic processes, fibrinolysis, organ regeneration, complement activation classical pathway, inflammatory response, positive regulation of lipid catabolic process, triglyceride homeostasis, aging, cellular response to interferon gamma, cholesterol metabolic process, cholesterol homeostasis, and tissue regeneration.

Pathway Enrichment Analysis of Differentially Expressed
Genes. In the pathway enrichment analysis, 12 pathway terms were significantly enriched ( < 0.05, Table 3). The top 10 most significant pathway terms were complement and coagulation cascades, retinol metabolism, steroid hormone biosynthesis, chemical carcinogenesis, linoleic acid metabolism, inflammatory mediator regulation of TRP channels, Gap junction, prion diseases, metabolic pathways, and arachidonic acid metabolism.

Gene Interaction
Network. Based on 270 differentially expressed genes, a gene interaction network was analyzed by the String software. We found that 257 nodes and 338 edges were constructed ( Figure 3). Twenty-five genes had more than 5 edges. The top 10 genes by degree are listed in Table 4.

Confirmation with qPCR.
From the gene expression array results, we found that Pon1 and Bhmt were in "metabolic pathway" (KEGG ID: rno01100), and Angptl3 and Pon1 were in several GO terms such as "positive regulation of lipid catabolic process" (GO: 0050996), "triglyceride homeostasis" (GO: 0070328), "cholesterol metabolic process" (GO: 0008203), "cholesterol homeostasis" (GO: 0042632), "response to hypoxia" (GO: 0001666), and "acylglycerol homeostasis" (GO: 0055090). So, we focused on these three genes in further study. As shown in Figure 4, the relative mRNA levels of angiopoietin-like 3 (Angptl3) and betaine-homocysteine S-methyltransferase (Bhmt) in diabetic rats were significantly higher than those of normal rats ( < 0.01), whereas the expression of these two genes was significantly reduced in vildagliptin-treated groups compared with those in diabetic rats ( < 0.01). Conversely, paraoxonase 1 (Pon1) decreased in the diabetic group. However, Pon1 increased in vildagliptin-treated rats ( < 0.01). These results were consistent with the microarray results.

Discussion
In our study, we found that vildagliptin reduced blood glucose, TC, and LDL-C. In a clinical trial, sitagliptin-combined metformin add-on therapy led to greater improvement in HbA1c than the metformin monotherapy group after 18 * * * * * * control diabetic vil-low vil-high          negative regulation of endopeptidase activity epoxygenase P450 pathway acute-phase response blood coagulation oxidation-reduction process cytoskeleton organization negative regulation of fibrinolysis microtubule-based process phosphatidylcholine metabolic process fibrinolysis organ regeneration complement activation, classical pathway inflammatory response positive regulation of lipid catabolic process triglyceride homeostasis aging cellular response to interferon-gamma cholesterol metabolic process cholesterol homeostasis tissue regeneration weeks of treatment. Moreover, sitagliptin combined with metformin significantly reduced TG, TC, and LDL-C and increased HDL-C compared with the metformin group [13]. We used aortic rings to test the Ach-induced endothelium-dependent vasodilation. In diabetic rats, Achinduced endothelium-dependent vasodilation was impaired. Vildagliptin augmented endothelial function. Other DDP-4 inhibitors also had similar vasoprotective effects. Saxagliptin treatment for 8 weeks increased aortic nitric oxide (NO) release by 22% and reduced mean arterial pressure in spontaneously hypertensive rats [14]. Moreover, sitagliptin treatment for 2 weeks protected endothelial function and reduced systolic blood pressure in spontaneously hypertensive rats through GLP-1 signaling [15].
In our research, the expression of Angptl3 was downregulated in vildagliptin-treated rats. Angptl3 is a key regulator, which can inhibit lipoprotein lipase (LPL) activity [16,17]. The enzyme LPL hydrolyzes TG to free fatty acids (FFA). ANGPTL3 overexpression mice had high plasma TG levels [18]. Patients with a loss-of-function mutation of Angptl3 are characterized by low plasma TC, TG, HDL-C, and LDL-C [19,20]. In 2008, Kathiresan et al. first identified an SNP site near ANGPTL3 that was associated with plasma TG levels in a genome-wide association study [21]. In obesity and T2D subjects, the ANGPTL3 level was higher than that in healthy subjects [22]. Moreover, the level of ANGPTL3 was increased in the livers of insulin-deficient and insulinresistant diabetic mice [23]. In mice and monkeys, using ANGPTL3-specific antibodies led to reduced plasma TG [24][25][26]. The therapeutic targets of the inhibition activity of ANGPTL require further research to treat dyslipidemia [16,17].
We also found that the expression of Bhmt was reduced in the vildagliptin-treated group. Methionine (Met) was produced from methylate homocysteine (Hcy) by BHMT enzymes with betaine. Thus, BHMT can reduce Hcy levels and increase Met levels. Met can then be converted to Sadenosylmethionine (SAM). In vivo, SAM is a main methyl donor to regulate methionine metabolism in many reactions. An increase in BHMT activity and SAM levels was observed in the livers of both type 1 diabetic rats [27] and the type 2 diabetic model [28,29]. Insulin treatment in the rat hepatoma cell line and STZ-induced diabetic rats can inhibit the excess expression of BHMT and SAM [28,30].
Moreover, our results showed that Pon1 was upregulated in the vildagliptin-treated group. Pon1 is an enzyme that has antioxidant functions. In serum, Pon1 is located in HDL. Pon1 protects LDL from oxidation and hydrolyzed oxidized LDL [31][32][33][34]. HDL levels are negatively associated with the risk of developing coronary artery disease (CAD), but high levels of oxidized LDL (oxLDL) in the aorta lead to cholesterol accumulation, foam cell formation, and atherosclerotic lesions [31,35]. Low Pon1 expression accelerates aortic lesion development in mice [36]. In humans, serum PON level is reduced in patients with a history of myocardial infarction [32] and diabetic patients [37]. PON1 transgenic mice had decreased oxidative stress and atherosclerotic lesions [38,39], whereas PON1 knockout mice had increased serum oxidative stress and were more susceptible to high-fat-diet-induced atherosclerosis [34,40].

Conclusion
In conclusion, this study has confirmed that vildagliptin significantly attenuates endothelial dysfunction in diabetic rats. Importantly, our study indicates that vildagliptin may activate Pon1 expression and inhibit the expression of Bhmt and Angptl3 in the aorta. This study may increase the understanding of the pathways that contribute to which DPP-4 inhibitor attenuates endothelial dysfunction. More studies are needed to investigate whether or not vildagliptin treatment directly alters Agptl3, Bhmt, and Pon1 expression in the aorta of diabetic rats, independent of blood glucose ( Figure 5).

Data Availability
The data used to support the findings of this study are included within the article.

Disclosure
The funding institutions had no role in study design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.