Background
Increased arterial stiffness is a cardiovascular (CV) biomarker strongly associated with hypertension, diastolic heart failure, chronic kidney disease and stroke [
1,
2]. Importantly, obese, insulin resistant and diabetic women develop vascular stiffness more frequently than men and this circumstance might account for the disproportionate incidence of cardiovascular disease (CVD) in insulin resistant women [
3,
4]. The widespread consumption of diets high in refined carbohydrates and fat [western diet (WD)] is one of the driving forces behind the alarming growth in the incidence of obesity and insulin resistance [
5,
6]. The progression to CVD in the setting of obesity and type 2 diabetes is likely initiated by endothelial dysfunction and several associated vascular abnormalities, including altered vascular tone, extracellular matrix remodeling and adventitial dysfunction, all of which could promote arterial stiffness [
7,
8]. We recently tested this notion and reported that after 4 months of WD feeding, the aorta of normotensive female C57BL/6J mice exhibits altered nitric oxide (NO)-dependent vasodilation associated with aortic and endothelial stiffness, increased aortic wall thickening and fibrosis, oxidative stress and inflammation [
9]. Recent reports from our laboratory indicate that therapies targeting vascular stiffness could potentially improve CV outcomes in insulin resistance models [
7,
9]. In this regard, the use of dipeptidyl peptidase-4 (DPP-4) inhibitors may be of interest. DPP-4 is an exopeptidase that circulates in plasma and is also expressed on the surface of multiple cell types including endothelial and vascular smooth muscle cells, as well as immune cells [
10]. Although DPP-4 inhibitors were developed to control hyperglycemia, it is now established that these inhibitors have effects beyond glycemic regulation [
11]. The rationale for using DPP-4 inhibitors to target vascular stiffness relates to the potential of these compounds to improve endothelial function [
11,
12]. Even though the beneficial effects of DPP-4 inhibition on the cardiovascular system have been explored in both animal models and in type 2 diabetic patients [
11], to the best of our knowledge, the potential role of these compounds to target vascular stiffness in a female rodent model of insulin resistance due to over-nutrition without frank hyperglycemia or blood pressure elevation has not been studied. In the present investigation, we tested whether administering linagliptin (LGT), a long acting and specific DPP-4 inhibitor, to female C57BL/6 J mice fed a WD for 4 months, could ameliorate the development of WD-induced aortic and endothelial cell (EC) stiffness. Herein, we report that DPP-4 inhibition prevents the development of WD-induced aortic and EC stiffness in overweight female mice. The mechanism, in part, appears a consequence of preventing abnormalities in aortic endothelium-dependent vasorelaxation, oxidative stress, medial wall thickening and fibrosis, as well as expression of FGF-23 and Klotho.
Discussion
Collectively, the results of this investigation support the hypothesis that DPP-4 activation plays an important role in the development of aortic stiffness in female mice in the setting of WD-induced insulin resistance. Furthermore, we present evidence that DPP-4 inhibition prevented female C57BL/6 J mice from developing WD-induced aorta stiffening, remodeling and dysfunction. We previously reported that WD feeding for 4 months does not increase blood pressure in female C57BL/6J [
9]. This previous finding enabled us to design an experiment that factored out a potential blood pressure effect to explain the CV protective effects of DPP-4 inhibition.
Beneficial vascular effects of DPP-4 inhibition have been attributed to both direct effects of DPP-4 inhibition and to the collateral increase in glucagon like peptide-1 (GLP-1) availability [
10]. In the vessel, DPP-4 is located in the cytoplasmic membrane of both EC and vascular smooth muscle cells (VSMC). In EC, DPP-4 expression is increased in models of diabetes [
25] and in VSMC, DPP-4 expression is preferentially increased in conditions of vascular remodeling [
11]. Importantly, the dose of LGT used in our investigation significantly inhibits DPP-4 activity [
12,
14,
26].
Our group previously showed that 4 months of WD feeding in female mice resulted in increases in aortic stiffness, measured non-invasively, as well as EC stiffness, measured ex vivo [
9]. Results of this study using a separate cohort of mice were nearly identical with our previous report and further demonstrate aortic and EC stiffening after 4 months of WD feeding. Notably, DPP-4 inhibition prevented the development of aortic stiffness, both in vivo and ex vivo, observed in untreated WD-fed mice. Although previous reports demonstrate that DPP-4 inhibitors decrease cardiac stiffness [
12,
27,
28], to the best of our knowledge this is the first report demonstrating protection from vascular stiffening in a clinically relevant non-diabetic female model of over-nutrition.
Vascular stiffness is determined by EC and VSMC properties, as well as by extracellular matrix and adventitial characteristics [
7]. In our experiments, DPP-4 inhibition in the WD-fed cohort prevented the development of impaired endothelial-dependent vasodilation, an effect in the aorta that is known to be largely NO-dependent [
29,
30]. In this regard, DPP-4 inhibitors exert direct effects on vascular tone independently of GLP-1 [
11]. The DPP-4 inhibitors, LGT and alogliptin, in particular, exert vasodilatory effects in the aorta in the setting of enhanced inflammation and oxidative stress and these effects were mediated by the NO/cGMP pathway [
14]. We previously reported a similar improvement in endothelium-dependent vasodilation in skeletal muscle arterioles with LGT administration in a model of obesity and insulin resistance [
12]. More recently we showed that impaired NO signaling accompanies increased vascular oxidative stress in our WD-fed mouse model [
9]. Thus, it is likely that a decrease in oxidative stress in our WD-fed female mice administered a DPP-4 inhibitor, helped to preserve normal aortic vaso-relaxation by preventing NO scavenging by free radicals, such as superoxide, as evidenced by a decrease in 3-nitrotyrosine staining in the aorta compared to untreated mice [
7,
31]. In a different model of obesity and insulin resistance, LGT also ameliorated oxidative stress associated with cerebral ischemia independently of glycemic control [
32], findings that further support the concept that the antioxidant properties of LGT may be vasculoprotective.
Likewise, the beneficial role of DPP-4 inhibition in restoring endothelial-dependent vasodilation in conditions of increased oxidative stress has been documented in other murine models [
33,
34]. It should be noted that even though the magnitude of the aortic vasodilatory responses reported in this investigation seem low, similar results in mice have been reported from other laboratories [
35‐
37]. Furthermore, our ultra-structural findings suggest that DPP-4 inhibition reverses the loss of luminal endothelial cytoplasmic elongations and nuclear contraction associated with WD feeding. Additionally, our findings indicate that the beneficial effects of the DPP-4 inhibition are not limited to the endothelium. Here, we observed that DPP-4 inhibition decreases medial thickening and fibrosis related to WD-feeding. The vascular anti-fibrotic role of DPP-4 inhibitors has been reported previously in models of atherosclerosis and neo-intima injury [
34,
38].
FGF-23 is elevated in conditions of insulin resistance and obesity [
39,
40]. Importantly, FGF-23 levels are associated with impaired vascular relaxation, vascular calcification and stiffness [
22,
40,
41]. Moreover, FGF-23 is expressed in the vasculature [
42] and has been shown to enhance oxidative stress and induce inhibition of NO-dependent vasodilation [
43]. We recently reported that female WD-fed mice exhibit increased FGF-23 levels and this was associated with impaired aortic endothelial function, increased aortic PWV and increased oxidative stress [
24]. FGF-23 effects may be modulated, in part, via Klotho [
44]. In this regard, Klotho deficiency was recently reported to accelerate aortic stiffening in high fat fed mice [
45]. Whether WD leads to a deficiency in Klotho expression has not been examined previously. Herein, we explored a possible contribution of FGF-23 and Klotho to the genesis of vascular stiffening, endothelial dysfunction and oxidative stress by testing whether DPP-4 inhibition prevents the increase in aortic FGF-23 expression and modulates Klotho expression in the aorta of WD-fed mouse model. In the present work, FGF-23 expression was increased in the aorta of WD-fed mice and this was prevented with DPP-4 inhibition. Furthermore, the expression of Klotho was decreased in the aorta of WD-fed mice relative to control mice and this deficiency was prevented with administration of LGT. In this regard and relevant to our findings, others have reported that Klotho deficiency results in increased production of reactive oxygen [
46], while its over-expression ameliorates oxidative stress [
47]. Since oxidative stress has been implicated in the impairment of vascular relaxation and bioavailable NO, the prevention of abnormal expression of FGF23 and Klotho by DPP-4 inhibition in the vasculature may represent a new mechanism to explain the efficacy of DPP-4 inhibition in preventing the appearance of vascular stiffness. Whether FGF-23/Klotho signaling in the vasculature, in concert with its vascular receptors and cofactors, directly modulates vascular function or influences vascular function via regulation of putative effects on mineral metabolism is unclear.
In the present investigation we did not observe that LGT prevented or reduced WD-induced weight gain suggesting that the efficacy of LGT in preventing WD-induced aortic stiffening, impaired vasoreactivity and abnormal remodeling did not result from beneficial weight loss. Some rodent studies have documented a similar neutral effect of LGT administration on body weight in association with improvement in cardiovascular function and structure [
25,
48]. On the other hand, higher doses of LGT led to modest weight loss in rats fed a high fat diet [
49]. Importantly clinical trials have supported the notion of weight stability with LGT treatment [
50‐
52]. One of the limitations of the present investigation is that we did not directly evaluate glucose homeostasis or insulin resistance. Nevertheless, we have previously demonstrated that this diet paradigm in female C57BL/6J mice does not result in frank hyperglycemia while it manifests significant whole body insulin resistance evaluated by the hyperinsulinemic euglycemic clamp [
53]. Herein, we further explored the possibility that DPP-4 inhibition reduces the anticipated increase in expression of AGE products as AGE/RAGE signaling can result in a pro-inflammatory and pro-oxidant environment [
54]. Furthermore, AGE can increase expression of DPP-4 in EC [
55]. In the present investigation, we documented increased AGE presence in the vascular wall of WD-fed females, and this was not prevented with DPP-4 inhibition. This suggests that the efficacy of DPP-4 inhibition in the vasculature of WD-fed female mice occurs without significant changes in AGE. To the best of our knowledge this is the only study demonstrating that a DPP-4 inhibitor prevents development of diet-induced vascular stiffness. Further studies are needed to clarify if other DPP-4 inhibitors share these beneficial effects.
Despite abundant preclinical evidence of cardiovascular benefits of DPP-4 inhibitors, including results presented herein, meta-analyses of clinical data raise concerns regarding evidence of increased risk of heart failure with administration of certain DPP-4 inhibitors [
56]. On the other hand, other recent analyses have produced differing conclusions regarding heart failure risk assessment [
57]. In the SAVOR-TIMI 53 trial the use of saxagliptin was associated with increased risk of heart failure [
58]. Importantly, the risk was higher in the subjects with previous heart failure, elevated natriuretic peptides and impaired renal function [
59]. In the EXAMINE trial that explored the use of alogliptin in type 2 diabetics with post-acute coronary events there was no increased risk of major CV events [
60]. Nevertheless, a recent post hoc analysis of the EXAMINE trial did find a statistically significant increase in hospitalization rate from heart failure in the group of subjects without heart failure history [
61]. On the contrary, the sitagliptin TECOS trial did not find differences in the rate of hospitalization for heart failure between the treatment and the control group [
62]. Similarly, a recent pre-specified patient-level pooled analysis of available trials of LGT did not report an association between the DPP-4 inhibitor and increased CV risk (including heart failure) [
63]. Results from two ongoing trials using LGT, the CAROLINA (NCT01243424) and CARMELINA (NCT01897532), will further clarify the role and safety of this agent in regard to diabetic CVD.
To summarize, the current investigation demonstrates that DPP-4 inhibition prevents abnormal increases in vascular stiffness, aortic fibrosis, oxidative stress and FGF-23/Klotho expression induced by a WD in female mice and these benefits occur independent of AGE reduction. Our findings have clinical relevance as obese diabetic women are more frequently affected by increased vascular stiffness that likely promotes higher incident CVD compared to their male counterparts. Ultimately, only the results of well-designed clinical trials will clarify the role of LGT in the management of diabetic CVD.
Authors’ contributions
VGD, AA, TK and JRS made substantial contributions to conception and study design. CM, VGD and AA were involved in drafting and revising the manuscript, including statistical analysis and data interpretation, and graphics. JH, AA, GJ, MRH, MG, LAM, FIR, GAM, JRS and VGD contributed to the acquisition and interpretation of data and associated intellectual content. All authors read and approved the final manuscript.