Liraglutide protects cardiac function in diabetic rats through the PPARα pathway

Diabetes is a worldwide public health problem that has prevalence greater than 5.71% in adults [1]. Chronic hyperglycemia leads to a high risk of cardiovascular events [2]. Cardiovascular disease is a leading cause of morbidity and mortality worldwide. Diabetic cardiomyopathy (DCM) is defined as cardiac hypertrophy that is independent of hypertension and coronary artery disease (CAD). The three main risk factors of DCM are insulin resistance, hyperinsulinemia, and hyperglycemia. To date, some cellular and molecular defects, including impaired insulin signaling, hyperglycemia, glucotoxicity, cardiac lipotoxity, mitochondrial dysfunction, oxidative stress, endoplasmic reticulum (ER) stress, and cardiomyocyte apoptosis, have been reported as primary causes of DCM pathogenesis [3,4]. Treatments providing glycemic control and cardiovascular protection are important to improve the health of people all over the world [5].

Glucagon-like peptide-1 (GLP-1) is secreted from L-cells in the gut. In addition, to controlling blood glucose levels, GLP-1 also reduces gastric emptying and inhibits appetite. However, GLP-1 can be digested quickly by dipeptidyl peptidase-4 (DPP-4). In clinical practice, the GLP-1 receptor (GLP-1R) agonist liraglutide was effective at controlling blood glucose levels. In addition to pancreatic α and β cells, GLP-1Rs are also found in the heart. More and more clinical trials and animal experiments have shown evidence of the protective cardiac effects of liraglutide, independent of its effects on blood glucose levels. Short-term liraglutide treatment mildly improves left ventricular ejection fraction (LVEF) in ST-segment elevation myocardial infarction patients [6]. In type 1 diabetic rats, liraglutide inhibits cardiac steatosis, oxidative stress, and apoptosis through activating the activated protein kinase (AMPK)-sirtuin 1 (Sirt1) pathway [7]. However, the exact mechanism of the beneficial effects of liraglutide on cardiac tissue in diabetic rats remains to be elucidated.

Therefore, the present study aimed to investigate whether liraglutide has protective cardiac effects and its underlying mechanism in type 2 diabetic rats. We employed a global microarray analysis combined with bioinformatics to explore key genes and pathways affected by liraglutide in cardiac tissue from diabetic rats.

Five-week-old male Sprague–Dawley (SD) rats, provided by the Institute of Laboratory Animal Science, Chinese Academy of Medical Sciences and Peking Union Medical College (Beijing, China, SCXK-2014-0013), were maintained in a pathogen-free environment with a 12-h light/dark cycle and free access to food and water. Animal experiments followed the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH publication number 85-23, revised 1996) and were approved by the Animal Care Committee of the Peking Union Medical Hospital Animal Ethics Committee (Project XHDW-2015-0051, 15 February 2015). After acclimatization, the rats were randomly divided into four groups (n=6 for each group): control, diabetic, low-dose liraglutide (Llirag), and high-dose liraglutide (Hlirag) groups. The control group was fed a standard rodent diet (kcal%: 10% fat, 20% protein, and 70% carbohydrate; 3.85 kcal/g); the other three groups were fed a high-fat diet (HFD) (kcal%: 45% fat, 20% protein, and 35% carbohydrate; 4.73 kcal/g, Research Diet, New Brunswick, NJ, U.S.A.). After 4 weeks of HFD feeding, diabetes was induced in the rats by streptozotocin (STZ, 30 mg/kg) injection. Fasting blood glucose levels higher than 11.1 mmol/l were considered standard for the diabetic model. Then, 0.2 or 0.4 mg/kg/day liraglutide (i.h.) was administered to the Llirag and Hlirag groups, respectively. Control and diabetic groups were injected with the same volume of normal saline. Body weights and fasting blood glucose levels (Bayer Contour TS glucometer, Hamburg, Germany) were recorded monthly. After 12 weeks of treatment, an oral glucose tolerance test, in which 20% glucose was gavaged at a dose of 2 g/kg, was performed after 10 h of fasting. Tail vein blood glucose levels at 0, 30, 60, and 120 min were measured. At the end of the study, all rats were fasted for 10 h and anesthetized with ketamine (100 mg/kg i.p., Pharmacia and Upjohn Ltd, Crawley, U.K.), then blood samples were collected from the abdominal aorta. Finally, the rats were killed by decapitation. Cardiac tissue was quickly collected, frozen in liquid nitrogen, and stored at −80°C for gene microarray analysis.

Triglyceride (TG), total cholesterol (TC), high-density lipoprotein (HDL), and low-density lipoprotein (LDL) levels were measured by an enzyme end point method (Roche Diagnostics, GmbH, Mannheim, Germany). Fasting serum adiponectin and insulin levels were determined by enzyme-linked immunosorbent assay (Millipore, Billerica, MA, U.S.A.). Homeostasis model assessment of insulin resistance (HOMA-IR) scores was calculated using the following formula: FBG (mmol/l) × fasting insulin (μIU/ml)/22.5.

Serum nitric oxide (NO) and GSH/GSSG levels were determined using Fluorometric Assay Kit (Cayman Chemical, Ann Arbor, MI, U.S.A.) and Thiol Green Indicator fluorometric method (Abcam, Cambridge, MA, U.S.A.), respectively.

Systolic blood pressure (SBP) and heart rate (HR) were measured by tail-cuff plethysmography (BP98A, Softron, Tokyo, Japan). After prewarming at 25°C for at least 5 min, the first five cycles were used as acclimatization cycles. After that, the mean blood pressure was recorded for the next five consecutive cycles.

Rats were anesthetized by inhaling 1% isoflurane with 99% O₂. During anesthesia, left ventricle (LV) diameter was determined using a Vevo 2100 Ultrasound System (Visual Sonics, Toronto, Ontario, Canada). Fractional shortening (FS) was calculated as follow: FS% = (LVEDD – LVESD)/LVEDD × 100, where LVEDD and LVESD are the LV end-diastolic diameter and the LV end-systolic diameter, respectively [8]. Increased FS indicated better LV contractile function [9].

Total RNA was extracted from cardiac tissue by using a mirVana™ RNA Isolation Kit (Ambion, Sao Paulo, SP, Brazil). Total RNA was transcribed into double-stranded cDNA and then synthesized into double-stranded cRNA. The second cycle cRNA was then labeled with biotin. The biotinylated cRNA was purified, fragmented, and hybridized to an Affymetrix GeneChip Rat Gene 2.0 ST whole transcript-based array (Affymetrix Technologies, Santa Clara, CA, U.S.A.). After washing and staining, the microarrays were scanned using an Affymetrix Scanner 3000 7G (Santa Clara, CA, U.S.A.).

Expression Console Software (version 1.4.1, Affymetrix, Santa Clara, CA, U.S.A.) was used to analyze the microarray signals. Differentially expressed genes were defined as having a fold change >1.5 and P-value <0.05 (one-way ANOVA). The raw microarray data have been submitted to the Gene Expression Omnibus (GEO) repository (GSE102194). The enrichment analysis of differentially expressed genes was performed by gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis with Database for Annotation, Visualization, and Integrated Discovery (DAVID) software (http://david.abcc.ncifcrf.gov/) [10]. The gene interaction network was drawn using String software (http://string-db.org/) [11].

Total RNA was extracted from cardiac tissue. Reverse transcription products were tested by real-time PCR. The primers are listed in Table 1. Real-time PCR was performed on an ABI Prism 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, U.S.A.). The cycling conditions were 95°C for 10 min, followed by 40 cycles of 95°C for 15 s, and 60°C for 30 s. β-actin was used as an internal control. Samples were run in triplicate. The 2^−ΔΔC_t method was used to calculate the relative expression levels.

GraphPad Prism software (version 5.0, San Diego, CA, U.S.A.) was used for statistical analyses. All values are shown as the mean ± S.D. Group data were analyzed using one-way ANOVA followed by Student’s t test. P<0.05 was considered to be statistically significant.

Liraglutide significantly reduced the body weights of diabetic rats (P<0.05, Figure 1A). In addition, liraglutide dose dependently reduced fasting blood glucose levels and blood glucose and area under the curve (AUC) values for oral glucose tolerance tests ( OGTTs, P<0.01, Figure 1B–D). The rats that underwent 12-week liraglutide treatment had lower TC (P<0.01, Figure 1E) and LDL-c levels (P<0.05, Figure 1H). Only the high dose of liraglutide reduced serum TG levels and increased serum adiponectin level in diabetic rats (P<0.01, Figure 1F,I). Compared with diabetic rats, liraglutide-treated rats had lower serum fasting insulin levels and HOMA-IR scores (P<0.01, Figure 1J,K).

Serum NO, GSH level, and GSH/GSSG ratio in diabetic rats were lower than control rats (P<0.01, Figure 1L,M,O). Liraglutide treatment moderated this decrease (P<0.01, Figure 1L,M,O). Serum GSSG level in diabetic group was higher than control rats (P<0.01, Figure 1N). Liraglutide reduced serum GSSG level dose independently (P<0.01, Figure 1N).

SBP, HR, LVEDD, and LVESD levels in the diabetic group were significantly increased (P<0.01, Figure 2A–D). However, diabetic rats had lower %FS values than normal control rats (P<0.01, Figure 2E). Liraglutide treatment decreased SBP, HR, LVEDD, and LVESD levels and increased % FS values (P<0.01, Figure 2A–E). These results suggest that liraglutide moderated LV dysfunction and decreased blood pressure and HRs.

A total of 269 differentially expressed genes were screened out from Hlirag group (fold change >1.5, P<0.05); these included 166 up-regulated genes and 105 down-regulated genes. The differentially expressed genes were enriched in 11 pathways (P<0.001, Table 2). The top five pathways were cardiac muscle contraction, non-alcoholic fatty liver disease (NAFLD), oxidative phosphorylation, metabolic pathways, metabolic pathways, and Alzheimer’s disease. The significant biological processes (BPs) in the GO categories are listed in Table 3 (P<0.01). The top ten BP terms were hydrogen ion transmembrane transport, cholesterol homeostasis, lipid homeostasis, positive regulation of TG biosynthetic process, fatty acid β-oxidation using acyl-CoA dehydrogenase, fatty acid β-oxidation, response to cAMP, very long-chain fatty acid catabolic process, circadian rhythm, and lipid metabolic process.

All the 271 differentially expressed genes were mapped using String online software. The results showed that there were 253 interactions with a total of 256 joint edges (Figure 3). Twenty-one nodes had more than ten joint edges. These nodes involved 137 joint edges. These 21 genes are listed in Table 4. The top ten genes were citrate synthase (Cs), ubiquinol-cytochrome c reductase core protein I (Uqcrc1), acyl-CoA dehydrogenase, very long chain (Acadv1), 3-oxoacid CoA transferase 1 (Oxct1), NADH dehydrogenase (ubiquinone) flavoprotein 1 (Ndufv1), succinate-CoA ligase, alpha subunit (Suclg1), enoyl CoA hydratase, short chain, 1, mitochondrial (Echs1), acyl-CoA dehydrogenase, C2–C3 short chain (Acads), and branched chain ketoacid dehydrogenase E1, α-polypeptide (Bckdha).

To validate the microarray results, we analyzed the mRNA expression levels of representative gene by using qPCR. As shown in Figure 4, the relative mRNA levels of LXRα, nuclear receptor subfamily 1, group H, member 3 (Nr1h3), sterol regulatory element binding transcription factor 1 (Srebf1) and peroxisome proliferator activated receptor α (PPARα) in the Hlirag group were significantly higher, whereas the mRNA levels of angiopoietin-like 3 (Angptl3), diacylglycerol O-acyltransferase 1 (Dgat1), diacylglycerol O-acyltransferase 2 (Dgat2), and epoxide hydrolase 2 (Ephx2) were lower than those in diabetic rat group (P<0.01). These outcomes were consistent with the microarray results.

In the present study, as expected, liraglutide reduced blood glucose levels and moderated insulin resistance in diabetic rats. Moreover, we found that liraglutide reduced the body weights in diabetic rats. Liraglutide-treated rats also had lower TC and LDL-c levels and higher adiponectin levels. Clinical trials proved that liraglutide was effective at reducing blood glucose levels and body weights [12–14]. Liraglutide at 3.0 mg/day has been approved by the U.S. Food and Drug Administration (FDA) for treating obesity since 2014. In human studies, liraglutide treatment at 3.0 mg for 56 weeks decreased insulin resistance [15]. In addition, liraglutide treatment dose dependently increased plasma adiponectin in Chinese type 2 diabetes [16].

Regarding cardiac function, we found that liraglutide reduced SBP, HR, LVEDD, and LVESD levels, increased %FS values and serum NO levels in diabetic rats. These results indicate the beneficial effects of liraglutide on cardiac function. HFD-fed mice have cardiac ceramide accumulation. One week of liraglutide treatment improved cardiac ER homeostasis and cardiac function [17]. A clinical trial revealed that liraglutide reduced LVEF in patients with heart failure (HF) [18]. Interestingly, liraglutide reduced both systolic and diastolic blood pressure in hypertensive mice through the atrial natriuretic peptide (ANP) axis [19]. In a meta-analysis of 16 randomized controlled trials, GLP-1 receptor agonists (exenatide and liraglutide) reduced systolic pressure (SBP) and diastole pressure (DBP) by 1–5 mmHg compared with other antidiabetic drugs in diabetic patients [20]. Reductions in blood pressure were not related to weight loss or hemoglobin A1c (HbA1c) improvement [21].

We found that liraglutide increased PPARα expression in the cardiac tissue of diabetic rats. PPARs have three forms: α, γ, and δ. They can bind with retinoid X receptor (RXR) to regulate energy utilization and storage [22]. Recent results implicate PPARs in the regulation of inflammation and atherosclerosis [23]. In the heart, both PPARα and PPARδ can regulate lipid metabolism. In addition to lipid metabolism, PPAR-γ also modulates glucose metabolism [24–26]. Previous studies found that PPAR-α expression is down-regulated in diabetic rat hearts [27–30]. Many studies indicate that oxidative stress increases in diabetic status [31,32] and contributes to inhibition of PPAR-α in cardiomyocytes [33]. Our data also showed that diabetic rats had lower GSH/GSSG ratio, and liraglutide treatment increased serum GSH/GSSG ratio. GSH/GSSG ratio is an important antioxidant biomarker [34].

Interestingly, in our study, liraglutide reduced Dgat1 and Dgat2 expression in the hearts of diabetic rats. DGAT has two isoforms: DGAT1 and DGAT2. They are the enzymes that catalyze the final step in the biosynthesis of TG [35]. DGAT is the target gene of PPARα. DGAT2 appears to be a key enzyme that controls TG homeostasis in vivo and regulates fatty acid storage [36]. A previous study found that DGAT1 and DGAT2 expression was increased in diabetic rat hearts [37]. Increased DGAT expression generated reactive oxidative stress, and caused myocardial damage in DM cardiomyopathy [38]. Thus, our results indicate that liraglutide reverses oxidative stress generated by diabetic status to increase PPARα expression, leads to reduce the expression of DGAT, and also finally inhibits reactive oxidative stress.

Our research found that liraglutide increased Nr1h3 (LXRα) and Srebf1 expression in diabetic rat hearts. LXRs have important role in the regulation of cholesterol and fatty acid metabolism. It forms heterodimer with RXR [39]. Srebf is directly induced by LXRs through an RXR/LXR-binding site on the Srebf gene promoter [40,41]. SREBF is a transcription factor that regulates lipogenic enzymes by binding to sterol response elements [42]. Thus, our data supports that liraglutide treatment activates cardiac LXRα and Srebf1 expression in diabetic rats.

We also found that liraglutide reduced Ephx2 expression in diabetic rat hearts. Soluble epoxide hydrolase (sEH) is an Ephx2 gene product. Arachidonic acid (AA) can be transferred to epoxyeicosatrienoic acids (EETs) by cytochrome P-450 (CYP) epoxygenases. EETs are signaling molecules that regulate blood pressure [43–47], inflammation [44,45,47,48], and glucose homeostasis [49,50]. However, EETs have a short half-life and may be metabolized by sEH into dihydroxyeicosatrienoic acids (DHETs) with relatively weak activity. Hearts in sEH null mice had improved post-ischemic recovery of Lv developed pressure (LVDP), reduced infarct size after ischemia and reperfusion [51], and reduced survival after cardiac arrest and cardiopulmonary resuscitation [52]. Therefore, sEH is a promising target for treating cardiovascular disease. sEH inhibitors have been indicated as beneficial treatments in animal models of high blood pressure [43,44,46], inflammation [44,48,53], myocardial injury [44,54–56], ischemia–reperfusion [57,58], pathological cardiac hypertrophy [44], and insulin resistance [49,50] and have been shown to protect heart structure and function [59]. Our study provided evidence that as a sEH ingibitor, liraglutide can preserve cardiac function.

We also found that liraglutide reduced Angptl3 expression in diabetic rat hearts. Lipoprotein lipase (LPL) metabolized TG into free fatty acids (FFAs). LPL overexpression is correlated with reduced plasma TG levels and decreased cardiovascular risks [60]. However, LPL null models have severe hypertriglyceridemia [60]. The angiopoietin-like protein (ANGPTL) family is a key regulator of LPL [61,62]. ANGPTL3 is an endogenous inhibitor of LPL. Rare loss-of-function variants for ANGPTL3 have been associated with decreased TG levels as well as decreased low-density lipoprotein-cholesterol (LDL-c) and high-density lipoprotein-cholesterol (HDL-c) levels in family and general population studies in humans [62–70]. In addition, these subjects demonstrate an absence of coronary atherosclerotic plaques [71]. Another human population study showed that plasma ANGPTL3 levels were increased in myocardial infarction patients [71]. Heterozygous carries of ANGPTL3 loss-of-function mutations had a 34% reduction in CAD risk [71]. Angptl3 deletion was also reported to reduce the development of atherosclerosis in apolipoprotein E (apoE)-deficient mice [72]. Recently, a human monoclonal antibody against Angptl3 in dyslipidemic mice and against ANGPTL3 in healthy human subjects with elevated levels of TGs or LDL-c significantly reduced serum TG, HDL-c, and LDL-c levels and decreased the odds of atherosclerotic cardiovascular disease [73]. Another research group showed that treating mice and human subjects with antisense oligonucleotides targeting Angptl3 messenger RNA reduced atherogenic lipoproteins and retarded the progression of atherosclerosis [74].

In conclusion, liraglutide prevents cardiac dysfunction by activating cardiac PPARα to inhibit Dgat expression and oxidative stress in diabetic rats (Figure 5). The present study provides a potential mechanism for the protective cardiac effects of a GLP-1 analog in a model of diabetes.

We thank Beijing Compass Biotechnology Company for excellent technical assistance with the microarray experiments.

X.X. designed the experiments, contributed reagents and materials. Q.Z., J.Z., T.W., and X.W. conducted the experiments. M.Y, M.L., and F.P. analyzed the data. Q.Z. wrote the manuscript.

This work was supported by the National Key R&D Program of China [grant number 2017YFC1309603]; the National Key Research and Development Program of China [grant number 2016YFA0101002]; the National Natural Science Foundation of China [grant numbers 81170736, 81570715]; the National Natural Science Foundation for Young Scholars of China [grant number 81300649]; the China Scholarship Council Foundation [grant number 201308110443]; the PUMC Youth Fund [grant number 33320140022]; the Fundamental Research Funds for the Central Universities; and the Scientific Activities Foundation for Selected Returned Overseas Professionals of Human Resources and Social Security Ministry.

The authors declare that there are no competing interests associated with the manuscript.

ANGPTL3

angiopoietin-like 3

AMPK

activated protein kinase

apoE

apolipoprotein E

AUC

area uncer the curve

BP

biological process

CAD

coronary artery disease

DBP

diastole pressure

DCM

diabetic cardiomyopathy

DGAT

diacylglycerol O-acyltransferase

EET

epoxyeicosatrienoic acid

ER

endoplasmic reticulum

FS

fractional shortening

GLP-1

glucagon-like peptide-1

GLP-1R

GLP-1 receptor

GO

gene ontology

HbA1c

hemoglobin A1c

HDL-c

high-density lipoprotein-cholesterol

HFD

high-fat diet

Hlirag

high-dose liraglutide

HOMA-IR

homeostasis model assessment of insulin resistance

HR

heart rate

KEGG

Kyoto Encyclopedia of Genes and Genomes

LDL-c

low-density lipoprotein-cholesterol

Llirag

low-dose liraglutide

LV

left ventricular

LVEDD

LV end-diastolic diameter

LVEF

LV ejection fraction

LVESD

LV end-systolic diameter

OGTT

oral glucose tolerance test

PPAR

peroxisome proliferator activated receptor

qPCR

quantitative polymerase chain reaction

RXR

retinoid X receptor

SBP

systolic blood pressure

sEH

soluble epoxide hydrolase

SBP

systolic pressure

Sirt1

sirtuin 1

STZ

streptozotocin

TC

total cholesterol

TG

triglyceride