Combination Therapy with a Sodium-Glucose Cotransporter 2 Inhibitor and a Dipeptidyl Peptidase-4 Inhibitor Additively Suppresses Macrophage Foam Cell Formation and Atherosclerosis in Diabetic Mice

Dipeptidyl peptidase-4 inhibitors (DPP-4is), in addition to their antihyperglycemic roles, have antiatherosclerotic effects. We reported that sodium-glucose cotransporter 2 inhibitors (SGLT2is) suppress atherosclerosis in a glucose-dependent manner in diabetic mice. Here, we investigated the effects of combination therapy with SGLT2i and DPP-4i on atherosclerosis in diabetic mice. SGLT2i (ipragliflozin, 1.0 mg/kg/day) and DPP-4i (alogliptin, 8.0 mg/kg/day), either alone or in combination, were administered to db/db mice or streptozotocin-induced diabetic apolipoprotein E-null (Apoe−/−) mice. Ipragliflozin and alogliptin monotherapies improved glucose intolerance; however, combination therapy did not show further improvement. The foam cell formation of peritoneal macrophages was suppressed by both the ipragliflozin and alogliptin monotherapies and was further enhanced by combination therapy. Although foam cell formation was closely associated with HbA1c levels in all groups, DPP-4i alone or the combination group showed further suppression of foam cell formation compared with the control or SGLT2i group at corresponding HbA1c levels. Both ipragliflozin and alogliptin monotherapies decreased scavenger receptors and increased cholesterol efflux regulatory genes in peritoneal macrophages, and combination therapy showed additive changes. In diabetic Apoe−/− mice, combination therapy showed the greatest suppression of plaque volume in the aortic root. In conclusion, combination therapy with SGLT2i and DPP4i synergistically suppresses macrophage foam cell formation and atherosclerosis in diabetic mice.


Introduction
Dipeptidyl peptidase-4 inhibitors (DPP-4is) and sodiumglucose cotransporter 2 inhibitors (SGLT2is) have been increasingly used for the treatment of type 2 diabetes. Several preclinical studies [1][2][3][4] including ours [5,6] have demonstrated that DPP-4is confer antiatherosclerotic effects independent of antihyperglycemic actions. In addition, we recently reported that SGLT2i suppresses the acceleration of atherosclerosis in diabetic apolipoprotein E-null (Apoe −/− ) mice and macrophage foam cell formation in diabetic db/db mice in a glucose-dependent manner [7]. These findings raised the possibility that combination therapy with SGLT2i and DPP-4i could synergistically further suppress atherosclerosis and macrophage foam cell formation by both glucosedependent and glucose-independent mechanisms.
Subendothelial accumulation of lipid laden macrophagederived foam cells occurs during the early stage of atherosclerosis [8]. Accumulation of cholesterol esters in macrophages is a hallmark of foam cell formation that depends on the uptake of oxidized low-density lipoprotein (ox-LDL) via its receptors, lectin-like ox-LDL receptor-1 (Lox-1) and CD36, and on the efflux of free cholesterol by ATP-binding cassette transporter A1 (ABCA1) and ATP-binding cassette subfamily G member 1 (ABCG1) [8].
In the present study, we aimed to evaluate whether combination therapy with SGLT2i and DPP-4i was superior to monotherapy with either an inhibitor alone in suppressing macrophage foam cell formation, a critical process in atherosclerosis, or the expression of genes involved in foam cell formation in db/db mice, a model of type 2 diabetes. In addition, we evaluated atherosclerotic plaque lesion in streptozotocininjected Apoe −/− mice, a model of atherosclerosis exacerbated by hyperglycemia.

Animal Experiment Number 1.
Six-week-old male db/db mice (a mouse model of type 2 diabetes) were purchased from Sankyo Labo Service (Japan) and kept on standard rodent chow. All the mice showed fasting blood glucose (FBG) levels above 200 mg/dL at the age of 8 weeks. At the age of 9 weeks, we administered a normal diet comprising none, SGLT2i (ipragliflozin, 1.0 mg/kg/day), DPP-4i (alogliptin, 8.0 mg/kg/day), or this combination at these same respective doses to the mice. Ipragliflozin and alogliptin were gifted from Astellas Pharma Inc. (Tokyo, Japan) and Takeda Pharmaceutical Company Limited (Osaka, Japan), respectively. After the treatment period for 4 weeks, peritoneal macrophages were harvested and the mice were sacrificed under general anesthesia with isoflurane [5-7, 9, 10]. All procedures were approved by the Animal Care Committee of Showa University School of Medicine.

Oral Glucose Tolerance Tests (OGTTs).
In a subset of animals, OGTTs were conducted as described previously [6,7]. In brief, glucose (0.5 g/kg body weight) was administered via oral gavage after a 6-hour fast, and the blood glucose levels were measured at 0 (fasting glucose level), 15, 30, 60, and 120 minutes.

Animal Experiment Number 2. Nine-week-old male
Apoe −/− mice (BALB/c background) were purchased from Sankyo Labo Service and kept on a standard rodent chow until the age of 15 weeks. The animals received intraperitoneal injections of streptozotocin to induce diabetes (80 mg/ kg/day at 15 weeks old and 50 mg/kg/day at 18 weeks old on each of 5 consecutive days). The fasting blood glucose levels were measured 10 days after the last injection. All the mice showed fasting blood glucose levels higher than 200 mg/dL and were enrolled in this experiment as diabetic mice. At the age of 20 weeks, the diet was switched to an atherogenic diet containing 30% fat, 20% sucrose, 8% NaCl, and 0.15% cholesterol (Oriental Yeast, Tokyo, Japan), and the mice were randomly assigned to treatment with none, SGLT2i (ipragliflozin, 1.0 mg/kg/day), DPP-4i (alogliptin, 8.0 mg/kg/day), or this combination. After 4 weeks of these treatments, the mice were sacrificed under general anesthesia with isoflurane, and the aortas were collected after perfusion fixation. The cross sections taken from the aortic roots were stained with oil red O to evaluate plaque volume or immunostained using antimouse MOMA-2 antibody (1 : 60; Abcam Japan, Tokyo, Japan) to evaluate macrophage infiltration in plaques. The aorta from the ascending aorta to the bifarcation of the common iliac arteries was longitudinally dissected and stained with oil red O. The images were quantified using

Statistical
Analysis. Values are expressed as mean ± SEM. Statistical analysis was performed using analysis of variance (ANOVA) followed by Tukey's post hoc test using JMP software (version 11; SAS Institute Inc., Cary, NC, USA). The correlation between two values was determined using Pearson's correlation test. The significance level was defined as p < 0 05.

Characteristics and Laboratory
Data of db/db Mice. Table 1 shows the characteristics and laboratory data of diabetic db/db mice treated with none, ipragliflozin, alogliptin, and the combination. Mice treated with ipragliflozin or the combination showed higher water intake and urine volume than the others as expected. Ipragliflozin strongly decreased FBG and HbA1c levels, while alogliptin led to modest decreases. Combination therapy did not lead to further decreases in FBG and HbA1c levels compared to ipragliflozin monotherapy. Plasma levels of insulin, active GLP-1, and total GIP significantly increased in mice receiving alogliptin or combination therapy, suggesting effective DPP-4 inhibition. There were no differences in plasma lipid levels between the groups.

Glucose
Tolerance. Treatment with ipragliflozin significantly improved glucose loading after OGTT and the AUC, while alogliptin led to only modest amelioration of glucose intolerance. Glucose intolerance was not further suppressed in mice receiving combination therapy compared to ipragliflozin alone consistent with FBG and HbA1c levels (Figures 1(a) and 1(b)).

Foam Cell Formation in Peritoneal
Macrophages. The suppression of ox-LDL-induced cholesterol ester accumulation, indicative of the extent of foam cell formation, was comparable between the ipragliflozin and alogliptin monotherapy groups (31% and 35%, resp., Figure 1(c)); both significantly lower than the control group (p < 0 0005). Furthermore, the extent of foam cell formation suppression was enhanced by combination therapy (49%), compared to the ipragliflozin and alogliptin monotherapy groups (p < 0 05, Figure 1(c)).

Correlation between Macrophage Foam Cell Formation
and Glycemic Control. We examined the relationship between macrophage foam cell formation and glycemic control in each animal. The extent of foam cell formation was significantly correlated with HbA1c levels in all treatment groups (control, r = 0 76; ipragliflozin, r = 0 73; alogliptin, r = 0 78; and combination, r = 0 8; p < 0 05 for all) (Figure 1(d)). The comparison of correlation curves revealed that the extent of foam cell formation was similar between the control and ipragliflozin monotherapy groups at corresponding HbA1c levels. However, the extent of foam cell formation suppression was disproportionately higher in the alogliptin monotherapy or combination therapy group compared to the control and ipragliflozin monotherapy groups at corresponding HbA1c levels (Figure 1(d)). There were no significant correlations between the extent of foam cell formation and body weight, SBP, insulin, total cholesterol, or triglyceride.
3.6. Atherosclerosis in Diabetic Apoe −/− Mice. Following the confirmation of the additive antiatherosclerotic effects of combination therapy in db/db mice, we also evaluated its efficacy on atherosclerosis in diabetic Apoe −/− mice. The physiological and biochemical data are shown in Table 2. Daily food intake was significantly higher in the combination therapy group than the other groups, and pulse rate was lower in the alogliptin group than the control and ipragliflozin groups. Treatment with ipragliflozin alone and the combination significantly reduced HbA1c levels compared with the control. The other parameters had no significant differences between the groups. Although ipragliflozin and alogliptin monotherapies tended to reduce atheromatous plaque volume in the aortic root, these changes did not reach statistical significance (p = 0 1-0 2). However, combination therapy significantly reduced atheromatous plaque volume by 67% compared to the control (Figures 4(a) and 4(b)). In addition, combination therapy showed the greatest suppression of intraplaque macrophage infiltration assessed by MOMA-2positive area (Figures 4(c) and 4(d)). Atherosclerotic plaque formation on the surface of the whole aorta tended to be suppressed by ipragliflozin, alogliptin, or the combination; however, their suppressions had no significant differences (Figures 4(e) and 4(f)).

Discussion
The present study is the first to demonstrate that combination therapy with SGLT2i and DPP-4i prevents macrophage foam cell formation in diabetic db/db mice. We also demonstrated that combination therapy with SGLT2i and DPP-4i suppressed atherosclerosis in diabetic Apoe −/− mice.
Suppression of macrophage foam cell formation was comparable with ipragliflozin-and alogliptin-treated mice despite the significant difference in amelioration of hyperglycemia between the two groups. Indeed, under similar glycemic conditions, combination therapy led to increased suppression of foam cell formation compared to ipragliflozin monotherapy. Further, the extent of foam cell formation suppression was disproportionally higher in mice treated with alogliptin alone, or in combination, at corresponding HbA1c levels. Taken together, these results suggest that DPP-4i exerts a pleiotropic suppressive effect on macrophage foam cell formation beyond glucose regulation.
Our results indicate that SGLT2i therapy suppressed macrophage foam cell formation in a glucose-dependent manner. Although treatment with ipragliflozin alone, or in combination, tended to reduce body weight and SBP, these changes did not correlate with foam cell formation. There were no significant correlations between plasma lipid levels and foam cell formation. These results provide additional evidence that the glucose-lowering effect of SGLT2i therapy was solely attributable to its suppressive effect on macrophage foam cell formation. Therefore, it is highly likely that the suppressive effect of combination therapy on macrophage foam cell formation is due to both the marked amelioration of hyperglycemia by SGLT2i and the pleiotropic effect of DPP-4i. Several studies demonstrated that high glucose upregulated the expression of scavenger receptors such as Lox-1, CD36, and scavenger receptor-A (SR-A) in cultured macrophages or dendritic cells [11][12][13][14] and that the expression of scavenger receptors in macrophages was suppressed by glucose control [11][12][13][14]. In our preliminary study, the expression of Lox-1 and CD36 were 2-3 times higher in db/db mice than in nondiabetic mice, but the expression of SR-A was not significantly increased in db/db mice compared with control mice (data not shown). Therefore, we did not examine SR-A expression in this study. The exact mechanism of hyperglycemia-induced upregulation of scavenger receptors remains unknown; one hypothesis is that hyperglycemia-mediated accumulation of reactive oxygen species induces proinflammatory signaling, including p38 mitogen-activated protein kinase (MAPK), which induces the nuclear transcription factor-kappa (NF-κ) B pathway and subsequent upregulation of scavenger receptor expression [14]. We previously reported that CD36 in mice macrophages treated with GLP-1, GIP, or DPP-4i was downregulated [5,9]. Dai et al. demonstrated that DPP-4i reduced the expressions of Lox-1 and CD36 in macrophages [15], suggesting DPP-4i directly suppressed the macrophage foam cell formation, independent of incretins. Thus, this is the first study to demonstrate that combination therapy can remarkably suppress expression of both Lox-1 and CD36 associated with the reduction of foam cell formation. Further, our results provide evidence that both DPP-4 inhibition and glucose-lowering effect might synergistically suppress the expression of scavenger receptors. The cholesterol efflux system plays a pivotal role in suppressing macrophage foam cell formation. Cholesterol efflux to apolipoprotein A-1 or HDL requires ABCA1 and ABCG1, respectively [8]. Accumulating experimental evidence suggest that ABCA1 or ABCG1 expression in macrophages is suppressed by high-glucose concentrations in vitro and by hyperglycemia ex vivo [16][17][18]. Thus, our data may provide support for the association of glycemia with ABCA1 and ABCG1 expressions. We previously reported that ABCA1 expression in human macrophages was increased by a GLP-1 receptor agonist [19]. The 3 ′ ,5 ′cyclic adenosine monophosphate (cyclic-AMP)/cyclic-AMP protein kinase (PKA) pathway is a major signaling cascade of incretin receptors, and ABCA1 and ABCG1 expressions are stimulated by the activation of cyclic-AMP/PKA pathway [20,21]. Therefore, it is possible that alogliptin treatment may stimulate the cyclic-AMP/PKA pathway in macrophages via increased endogenous incretins. Similar to results observed for the scavenger receptors, this is the first study to show that combination therapy leads to a significant increase in ABCA1 and ABCG1 expressions. Glucose-lowering action and incretin-mediated action may be both involved in enhancing the expression of key mediators of cellular cholesterol efflux.
We also found that combination therapy with SGLT2i and DPP-4i showed greatest suppression on atheromatous plaque in the aortic root. In addition, the intraplaque macrophage infiltration in the aortic root showed a similar tendency. These results indicate additive antiatherosclerotic effects of combination therapy consistent with the suppression of macrophage form cell formation observed in db/db mice. In contrast, combination therapy did not show any additive effect on atherosclerotic plaque formation on the surface of the aorta in Apoe −/− mice (BALB/ c background). In the present study, we used Apoe −/− mice with BALB/c background that has been shown to be resistant to atherosclerosis compared with those with C56BL6 background, which are more frequently used [22]. For this reason, it is possible that atherosclerotic plaque formation was too weak to detect suppression by these agents.

Conclusions
The present study is the first to demonstrate that combination therapy with SGLT2i and DPP-4i synergistically suppresses macrophage foam cell formation in type 2 diabetic mice and atheromatous plaque formation in diabetic, dyslipidemic mice. Our findings indicate that scavenger receptors and cholesterol efflux, two major players in foam cell formation, are regulated by both glycemic control and DPP-4 inhibition.

Disclosure
This study was presented in part at the 52nd Annual Meeting of the European Association for the Study of Diabetes, Munich, Germany, 12-16 September 2016. The funders had no role in this study.