Since the early 1990s GLP-1 peptides and analogs have been investigated for treating T2DM because of their ability to enhance glucose-dependent insulin secretion. In addition, GLP-1 agonists are hypothesized to have effects on the cardiovascular system beyond glycaemic control, which may be exploited for therapeutic benefit [19, 20].
Cardiovascular disease is the leading cause of death in patients with type 2 diabetes. In a recent prospective cohort study of one million U.S. adults, diabetes was associated with a twofold increase in the risk of death from ischemic heart disease . Concerns about the cardiovascular safety of anti-diabetic drugs resulted in the FDA recommendation that the CV risk should be more thoroughly addressed during drug development . As a consequence of this new guideline cardiovascular outcome studies are currently ongoing in more than 26,000 type 2 diabetic patients directly treated with GLP-1 receptor agonists and more than 40,000 treated with a DPP-IV inhibitors which also may act indirectly via GLP-1. The overall goal of our study was to demonstrate in a pre-clinical set of models cardioprotective effects of the GLP-1 analog lixisenatide, building a further fundament for clinical testing in patients with reduced cardiac function.
In an acute model of global cardiac ischemia-reperfusion injury, lixisenatide treatment during the last 10 minutes of ischemia and during the whole reperfusion period significantly reduced infarct size. The observed cardioprotective effect was not associated with a significant change in cardiac hemodynamics, as assessed by rate-pressure product RPP (LVDP x HR), and particularly coronary flow. The results described here with lixisenatide are consistent to similar results in previous studies with another GLP1R agonist, exenatide . Ischemia reperfusion injury is a multifaceted damage, comprising effects on cardiomyocytes function and death and endothelial function. Due to the lack of effect on coronary flow by lixisenatide, we assume a direct effect on cardiomyocyte function and survival. Sonne and co-workers found that GLP-1 (9–36) amide, which is the first breakdown fragment of GLP-1 (7–36) amide but not a GLP-1 receptor agonist, did not reduce infarct size but however increased functional recovery of ischemia hearts . Based on these finding the authors proposed that beside the known GLP-1 receptor another hitherto unidentified receptor could be responsible for effects of GLP-1 like peptides in rat hearts.
Only a limited numbers of pre-clinical studies have investigated so far effects of GLP-1 peptides on long-term consequences after myocardial ischemia and reperfusion. Using liraglutide, Noyan-Ashraf and coworkers demonstrated cardioprotection after myocardial infarction in diabetic and non-diabetic mice . Liu and coworkers treated rats two weeks after myocardial infarction with either GLP-1 (7–36) or the exenatide analog AC3174 and noticed improved cardiac function and morphology . In contrast to those previous studies, we performed only a transient and not permanent ligation of the left coronary artery in our rat model in order to obtain results closer to the clinical situation in which almost all patients with myocardial infarction are reperfused. The transient ischemia reperfusion protocol resulted in only moderate changes of systolic function after a long-term recovery period. LVP and dp/dtmax were slightly reduced in the ischemia reperfusion group versus sham treated animals. Ramipril but not lixisenatide normalized LVP but not dp/dtmax. Major differences were observed for the active treatments regarding diastolic function, in particular on left ventricular end diastolic pressure (LVedP) and the relaxation time tau Weiss. These functional improvements occur despite lack of effect on increased cardiac fibrosis assessed by specific morphological staining and gene expression analysis. In addition, heart weight did not significantly differ between the various groups ruling out strong effects on hypertrophy of cardiomyocytes as potential mode of action. Moreover, we noticed no effects on plasma glucose and insulin by lixisenatide treatment in this non-diabetic model. Hence, indirect effects on heart metabolism via glucose uptake may not sufficiently explain the observed improvement in cardiac function.
A broader gene expression pattern of genes involved in rat cardiac remodeling delivered two major findings. First, the myocardial infarction is a major driver of gene expression changes and drug effects in the non-infarct regions of infarcted hearts are rather moderate. Second, lixisenatide and the ACE inhibitor ramipril are similar in their reaction pattern, indicating activation of more common protective signaling pathways for both treatments. A principal component analysis supported these conclusions. An ANOVA analysis provided a few number of genes differentially regulated by lixisenatide versus placebo in the infarct area (see Additional file 2).
Interestingly, lixisenatide was equipotent to GLP-1 and other GLP-1 receptor analogues on reducing infarct size in perfused isolated rat hearts. In contrast, lixisenatide at higher concentrations, but not GLP-1, improved fractional shortening in isolated cardiomyocytes. This might suggest that lixisenatide displays acute cardioprotection via the GLP-1 receptor, while effects of lixisenatide on contractility are mediated by a different signalling pathway. In line with our findings, Vila Petroff et al. showed the lack of contractility-promoting effect of GLP-1 in isolated cardiomyocytes . We supplemented their findings by demonstrating that even very high concentrations of the GLP-1 peptide, not amenable to rapid degradation by dipeptidyl-peptidase-IV (DPP4), were ineffective. Finally, when performing a similar set of experiments on cardiomyocytes from GLP-1 receptor knockout mice, a robust response to lixisenatide remained suggesting a GLP-1 receptor independent effect lixisenatide on the contractility response.
Lixisenatide displayed beneficial effects in our rodent models, but GLP-1 receptor mRNA was not detectable without ambiguity. Two different PCR Taqman probes, located to different parts of the rat gene, did not reveal abundant expression of the GLP-1 receptor in rat hearts and cardiomyocytes, but indicated a strong expression in other organs like rat pancreas. Overall, controversial data on the expression of the GLP-1 receptor in mammalian cardiac tissues exists. GLP-1 receptor mRNA expression in human heart samples was detected by an RNA protection assay . In contrast, using autoradiography with a radio-labeled GLP-1, no staining was detected in human heart despite several other organs being positive for this ligand . An antibody against the GLP-1 receptor stained in mouse heart several cell types . Later it was shown that available antibodies against the GLP-1 receptor display strong cross-reactivity in cells not expressing the GLP-1 receptor . Hence, immunostainings should be carefully considered as long as specificity is not clearly demonstrated. Simultaneously to immunostainings, Ban et al. detected mRNA for the GLP1-receptor, using an endpoint PCR with a high number of amplification cycles followed by specific hybridization . This methodology does not rule out a very low expression level in the samples. Bullock and co-workers used the RNAse protection assay, a more quantitative but less sensitive methodology, to assess mRNA distribution in rat tissues for the GLP-1 receptor. They could not confirm expression in the rat heart in contrast to other positive tissues like pancreatic tissues .
Currently, we cannot exclude that a GLP1R variant is expressed in the heart which is not detectable using the two different PCR primer assays. In addition, a normal GLP1R may be expressed in rodent hearts in other cell types than cardiomyocytes with overall low abundance, e.g. in resident or invading immune cells. Alternatively, another receptor, not related in its primary structure to the GLPR1 may exist in the rat heart that is responsive to lixisenatide and other GLP-1 analogs, mediating the cardioprotection seen in our studies. Clearly, further work needs to be invested here, e.g. testing of lixisenatide and related GLP-1 like analogs on ligand efficacy of a broad panel of receptors.
The pre-clinical effects described here provide a rationale for further clinical testing of lixisenatide in patients at cardiovascular risk. In a first randomized, double-blind, placebo-controlled, multicenter study patients are currently being recruited (ELIXA, ClinicalTrials.gov Identifier: NCT01147250). The primary objective of this study with approximately 6,000 patients is to demonstrate that lixisenatide can reduce cardiovascular morbidity and mortality compared to placebo in type 2 diabetic patients who recently experienced an acute coronary syndrome event.