The Effects of Dipeptidyl Peptidase-4 Inhibition On

2884 Diabetes Care Volume 37, October 2014

The Effects of Dipeptidyl Angelo Avogaro and Gian Paolo Fadini Peptidase-4 Inhibition on Microvascular Diabetes Complications

Diabetes Care 2014;37:2884–2894 | DOI: 10.2337/dc14-0865

We performed a review of the literature to determine whether the dipeptidyl peptidase-4 inhibitors (DPP4-I) may have the capability to directly and positively inﬂuence diabetic microvascular complications. The literature was scanned to identify experimental and clinical evidence that DPP4-I can ameliorate diabetic microangiopathy. We retrieved articles published between 1 January 1980 and 1 March 2014 in English-language peer-reviewed journals using the following terms: (“diabetes” OR “diabetic”)AND(“retinopathy” OR “retinal” OR “nephropathy” OR “renal” OR “albuminuria” OR “microalbuminuria” OR “neuropathy” OR “ulcer” OR “wound” OR “bone marrow”); (“dipeptidyl peptidase-4” OR “dipeptidyl peptidase-IV” OR “DPP-4” OR “DPP-IV”); and (“inhibition” OR “inhibitor”). Experimentally, DPP4-I appears to improve inﬂammation, endothelial function, blood pressure, lipid metab-

REVIEW olism, and bone marrow function. Several experimental studies report direct potential beneficial effects of DPP4-I on all microvascular diabetes-related complications. These drugs have the ability to act either directly or indirectly via improved glucose control, GLP-1 bioavailability, and modifying nonincretin substrates. Although preliminary clinical data support that DPP4-I therapy can protect from microangiopathy, insufficient evidence is available to conclude that this class of drugs directly prevents or decreases microangiopathy in humans independently from improved glucose control. Experimental findings and preliminary clinical data suggest that DPP4-I, in addition to improving metabolic control, have the potential to interfere with the onset and progression of diabetic microangiopathy. Further evidence is needed to confirm these effects in patients with diabetes.

Diabetes increases the incidence of cardiovascular disease (CVD) (1), but the role of hyperglycemia in the pathogenesis of CVD is still under debate. Recent large trials have shown that, at least for the duration of the trials, glucose lowering has modest or neutral effects on CVD in people with type 2 diabetes (T2D) (2–5). On the other hand, the relationship between hyperglycemia and microvascular outcomes is very strong, Department of Medicine, Division of Metabolic as high glucose promotes activation, dysfunction, and apoptosis of vascular and non- Diseases, University of Padova, Padova, Italy vascular cells (6). Importantly, diabetic microvascular changes affect the retina, kidney, Corresponding author: Angelo Avogaro, angelo. [email protected]. and nerves, but they can also be detected in other organs, such as the heart (7,8). For instance, we have shown that acute hyperglycemia in T2D significantly alters myocar- Received 7 April 2014 and accepted 26 June 2014. dial microvascular perfusion (9). Although the analogy between site-specificmicro- fi © 2014 by the American Diabetes Association. vascular complications is dif cult to assess, the association between retinal disease Readers may use this article as long as the work and nonretinal consequences of diabetes is supported by a fairly relevant amount is properly cited, the use is educational and not of clinical and experimental data (10). For instance, the presence of diabetic for profit, and the work is not altered. care.diabetesjournals.org Avogaro and Fadini 2885

retinopathy is associated with a two- to glucose lowering (27). Many efforts, and organ systems that may be affected threefold higher risk of incident fatal including preclinical studies, pooled by DPP-4 (32). Beside incretin hormones, and nonfatal coronary heart disease, analysis of phase III clinical trials, and these substrates include neuropeptides, even after adjustment for traditional randomized controlled trials, have been cytokines, growth factors, and chemo- risk factors, and up to 25-fold higher devoted to test the potential cardiovas- kines. Stromal cell-derived factor 1a prevalence of lower limb amputation cular benefits of DPP4-I, while their im- (SDF-1a) and 1b (CXCL12), macrophage- (11,12). A relationship between reti- pact on microangiopathy seems to be derived chemokine (CCL22), interferon- nopathy and the extent of coronary ar- relatively neglected. So far, randomized inducible T-cell a-chemoattractant tery calcium was observed (13), and controlled trials do not support cardio- (CXCL11), granulocyte chemotactic pro- microangiopathy is independently asso- vascular protective effects of DPP4-I tein 2 (CXCL6), and Grob (CXCL2) are ciated with presence, severity, and (23,28), but studies with longer follow- some relevant known substrates for pro- composition of carotid atherosclerosis up are likely needed to allow the effect teolytic modification by DPP-4. Neuro- (14). These data suggest the existence of good glycemic control achieved with peptide Y (NPY) and peptide YY belong of common, yet unknown, pathogenic DPP4-I to translate into improved cardio- to the pancreatic polypeptide family in- mechanisms and mutual relationships vascular outcomes (3). We hypothesize volved in neuroendocrine control of between microvascular disease and car- that the effects of DPP4-I on microangi- feeding-associated processes (33). Pep- diovascular risk. Diabetic nephropathy opathy may be one key element in this tide YY is secreted in response to neuro- is another strong CVD predictor. Micro- time-dependent effect. In this review, nalandhumoralfactors,aswellas albuminuria is itself recognized as an in- the role of DPP4-I on diabetes-related nutrients, and has vasoconstrictive prop- dependent determinant of mortality microvascular complications will be con- erties. The varieties of physiologic func- and CVD in both the general and dia- sidered and discussed from both exper- tions that can be theoretically modified betic populations (15,16), and chronic imental and clinical perspectives. by DPP-4 substrates indicate that DPP4-I kidney disease, starting from stage 1, can have important effects well beyond is well documented as an amplifier of DPP-4 STRUCTURE AND FUNCTION incretins and glucose control. the cardiovascular risk (17). More re- DPP-4 is a highly conserved peptidase Despite a common mechanism of ac- cently, diabetes has also been shown with high selectivity for peptides with a tion, there is a significant heterogeneity in to induce microangiopathy in the bone proline or alanine at the second NH2- the pharmacokinetic of different DPP4-I: marrow (BM) of mice and humans terminal position. The CD26 gene encodes they show differences in half-life, bio- (18,19). By providing regenerative vas- a type II transmembrane protein of 766 availability, metabolism, and excretion cular stem/progenitor cells, the BM amino acids, which is anchored to the route. Some DPP4-I act through compet- acts as a central housekeeper of cardio- lipid bilayer by a single hydrophobic seg- itive enzymatic inhibition (sitagliptin and vascular health, whereas BM microan- ment located at the N-terminus and has a alogliptin), while others are substrate- giopathy may impair cardiovascular short cytoplasmic tail of six amino acids enzyme blockers (saxagliptin and vilda- homeostasis (20). Large randomized (29). The extracellular part of CD26 contains gliptin) (34). DPP4-I might also differ in prospective studies have shown that, a glycosylation domain, a cysteine-rich do- their protective effect against microan- despite a neutral effect on macrovascular main, and a catalytic domain. Further- giopathy, especially in terms of nephro- disease, HbA1c control significantly re- more, CD26 presents interactions with protection, according to their route of duces microvascular end points (2,21– the extracellular matrix, namely, fibro- elimination: less than 5% of linagliptin is 24). However, either the short duration nectin and collagen. In 1993, CD26 was excreted through the kidney, while all of these studies or the adverse effects identified as the binding protein for aden- other DPP4-I are mostly excreted of glucose-lowering agents do not allow osine deaminase (30), the deficiency of through the renal route. Whether this us to (dis)prove the causal relationship which causes severe impairment of cellu- different route of excretion translates between the prevention of microvascular lar and humoral immunity. Formation of into a direct protective effect on kidney complications and the subsequent im- the complex between adenosine deami- function in people with diabetes is pres- provements in cardiovascular outcomes. nase and CD26 preserves the individual ently unknown. So-called incretinergic therapies have enzymatic activities of both molecules, boosted enthusiasm in the treatment of and a model for immunoregulation pro- Literature Search Strategy patients with T2D since both GLP-1 re- poses that CD26 modulates the concen- To review the current knowledge on ceptor agonists (GLP-1RA) and dipep- tration of local extracellular adenosine, DPP4-I in diabetic microangiopathy, po- tidyl peptidase-4 (DPP-4; also termed which provides negative signals to T cells tentially relevant articles were retrieved CD26) inhibitors (DPP4-I) have been (31). CD26 is strongly expressed on epi- from PubMed, ISI Web of Knowledge, and showntoexertsignificant glucose- thelial cells (kidney proximal tubulus, in- Scopus using the following combination lowering effect without inducing weight testine, bile duct), on endothelial cells, as of search terms: (“diabetes” OR “dia- gain and with risk of hypoglycemia com- well as on leukocytes. The soluble form of betic”)AND(“retinopathy” OR “retinal” parable to placebo (in monotherapy and CD26 (sCD26 or plasma DPP-4) lacks the OR “nephropathy” OR “renal” OR “albu- when associated with metformin) and intracellular tail and transmembrane re- minuria” OR “microalbuminuria” OR “neu- lower than that associated with sulfonyl- gions. A consistent number of proteins ropathy” OR “ulcer” OR “wound” OR ureas (25,26). Furthermore, the litera- have a penultimate alanine, proline, or “bone marrow”); (“dipeptidyl peptidase- ture offers lots of data showing positive serine at the N-terminus: these proteins 4” OR “dipeptidyl peptidase-IV” OR effects on vascular biology beyond act on many different cell types, tissues, “DPP-4” OR “DPP-IV”); and (“inhibition” 2886 DPP4-I and Microvascular Diabetes Complications Diabetes Care Volume 37, October 2014

OR “inhibitor”). As of 3 June 2014, this shown that DPP4-I with sitagliptin or guanosine monophosphate production search retrieved 160 articles for review. saxagliptin in T2D is able to increase en- in cardiomyocytes (57). Therefore, The title, abstract, and key words were dothelial progenitor cell (EPC) levels (44) DPP4-I administration can potentially used to screen items unrelated to the topic and function (45). Some authors have decrease blood pressure by two distinct of interest. Cross-reference and citations shown an improvement of endothelial mechanisms: one at the renal level, by of other relevant items were checked in function (46–49), which has been, how- inhibiting sodium/hydrogen exchange, articles selected for further analysis. ever, confuted by others (50). and the other at cardiac level, by inhib- Clinical studies have demonstrated iting BNP degradation. Whether the ef- Effects of DPP4-I on Endothelial some improvement in lipid profile by fect of DPP4-I on BNP has any role on Function and Risk Factors for DPP4-I–based therapies. In a meta-analysis endothelial function in humans is pres- Endothelial Dysfunction of available trials, Monami et al. (51) ently unknown. Endothelial cells exposed to high glu- reported that treatment with DPP4-I is Clinical data demonstrate a modest cose exhibit enhanced DPP-4 activity associated with a significant reduction blood pressure reducing effect of (35); conversely, DPP4-I is associated in total cholesterol and triglycerides DPP4-I (52,58,59). This contrast with with an increased biological activity without significantly affecting HDL. In a the well-known blood pressure–lowering of nitric oxide (36,37). DPP4-I also more recent review, van Genugten et al. effects of incretinergic therapy with GLP- induced a significant reduction of CD40 (52) report a fairly positive effect of 1RA (60), which is mediated by atrial na- (38), intracellular adhesion molecule 1 DPP4-I on plasma lipids, particularly triuretic peptide secretion (61). Rather, (ICAM-1), and transendothelial migra- for sitagliptin on HDL cholesterol and owing to complex interactions between tion of circulating mononuclear cells for vildagliptin on total cholesterol. concomitant ACE and DPP4-I, blood pres- (39). In spontaneously hypertensive Eventually, in nondiabetic subjects, sure regulators that are sequential sub- rats, DPP4-I improves endothelium- Noda et al. (47) correlated alogliptin- strates of these two enzymes may dependent vasodilatation of renal mediated improvement in endothelial mediate unexpected effects in particular arteries, renormalizes renal blood flow, function to its ability to suppress the clinical settings. For instance, it has been and reduces systolic blood pressure postprandial elevation of triglycerides, shown that, in healthy volunteers, sub- (37). A pivotal contributor to diabetic apolipoprotein B48, and remnant lipo- stance P increases sympathetic activity vascular damage is determined by the protein cholesterol. when ACE and DPP-4 are both inhibited overproduction of advanced glycation The effects of DPP4-I on blood pres- (62). Such effect may limit blood pressure end products (AGEs) that bind to their sure appear more complex and less control obtained with maximal ACE inhi- specific receptors (receptors for AGEs clear. DPP-4 converts the NPY(1-36), re- bition (63). However, it should be empha- [RAGEs]), inducing oxidative stress, in- leased by sympathetic renal fibers and sized that very few clinical studies flammation, and thrombogenicity. In a agonist of Y1 receptor, to NPY(3-36), the included reduction of blood pressure model of OLETF rats, Matsui et al. (40) selective agonist of Y2 receptor (53). as a primary or prespecified end point. showed that vildagliptin treatment signif- Since Y1 receptors potentiate renovas- Inflammation in dysmetabolic condi- icantly reduced expression of RAGE, com- cular response to angiotensin II (AT-II), it tions is the result of expanded fat mass, ponents of NADPH oxidase (gp91phox has been postulated that DPP4-I might and it plays a causative role in inducing and p22phox), and markers of oxidative sustain NPY(1-36) capacity to increase insulin resistance, as well as atheroscle- stress in the thoracic aorta. Ishibashi et al. the hypertensive response to AT-II. rotic plaque instability. DPP-4 plays an (41) showed that sitagliptin in combina- Animal studies have shown the DPP4-I important role in the immune and in- tion with GLP-1 completely blocked the improved endothelium-dependent re- flammatory responses (64). Both obese AGE-induced increase in RAGE mRNA laxation in renal arteries, restored renal humans and rodents demonstrated in- and protein, thus preventing reactive blood flow, and reduced systolic blood creased levels of DPP-4 expression in oxygen species generation and endo- pressure in spontaneously hypertensive dendritic cell/macrophage populations thelial nitric oxide synthase (eNOS) down- rats by increasing cAMP level and eNOS from visceral adipose tissue (65). Lamers regulation. Two weeks of vildagliptin (49). Approximately 70% of excreted et al. (66) have suggested DPP-4 can be treatment in type 1 diabetic (T1D) rats Na+ is reabsorbed in the proximal tu- considered a new adipokine, released by reduced oxidative stress and sup- bule, via a Na+/H+ exchanger 3: DPP-4 human adipocytes: DPP-4 expression is pressed ICAM-1, transforming growth forms a complex with Na+/H+ exchanger particularly high in visceral fat of obese factor b, and plasminogen activator in- 3 at the level of the brush membrane subjects and correlates with all metabolic hibitor 1 gene expression (42). Recently (54). DPP4 inhibition with sitagliptin syndrome components. It is therefore in- Zeng et al. (43) showed that mice treated administration may interfere with Na+ tuitive that DPP4-I has the potential to with sitagliptin developed smaller athero- resorption mechanism, significantly in- suppress such a proinflammatory state. sclerotic plaques than controls, had re- creasing natriuresis, thereby reducing In diabetic Zucker rats, the chronic admin- duced collagen content in plaques, and blood pressure levels (55,56). Another istration of sitagliptin reduced C-reactive reduced the expression of MCP-1 and in- DPP-4 substrate that can play an impor- protein and IL-1b levels, along with im- terleukin (IL)-6 in the aorta. The authors tant role in blood pressure regulation is provement of oxidative stress (67). Sita- suggest that these effects of sitagliptin the brain-derived natriuretic peptide gliptin, in the diet-induced obesity mouse may be carried out via regulation of the (BNP). DPP-4 converts the active form model, reduced the proinflammatory mi- AMPK and mitogen-activated protein ki- of BNP (1–33) into a form inactive on lieu, macrophage infiltration, and gene nase pathways. In humans, we have natriuresis but still active on cyclic expression of MCP-1, IL-6, IL-12 (p40), care.diabetesjournals.org Avogaro and Fadini 2887

and IL-12 in the adipose tissue (p35) (68). first to underline this complexity, also in (91), lending support to the protective Similar anti-inflammatory effects of DPP- the light of the pleiotropic effect of DPP- effects of linagliptin against diabetic ne- 4 have also been demonstrated in the 4. They showed that inhibition of DPP-4 phropathy. Liu et al. (92) assessed atherosclerotic plaque. Alogliptin treat- prevents the catabolism of NPY(1-36) and whether vildagliptin had a renoprotective ment, in ApoE2/2 diabetic mice, thereby increases the effects of NPY(1-36) activity in STZ-induced diabetic rats. Dia- induced a significant reduction of athero- released from renal sympathetic nerves betic and nondiabetic rats were treated sclerotic lesions and a concomitant reduc- on Y1 receptors, thus leading to an en- with oral vildagliptin or placebo for 24 tion of IL-6 and IL-1b (69). Vittone et al. hancement of the renovascular effects weeks, and renal injury was observed by (70) showed that in ApoE2/2 mice, sita- of AT-II (85). Along this line, Tofovic light and electron microscopy. Diabetic gliptin reduced plaque inflammation and et al. (84) showed that sitagliptin enhan- rats exhibited marked polyuria, increased increased plaque stability, potentially by ces renovascular responses to AT-II in urinary albumin and protein excretion, high GLP-1–mediated inhibition of chemo- SHR rats; they also demonstrated that serum creatinine and blood urea nitrogen kine-induced monocyte migration and this effect persists in rats with diabetic levels, enhanced albumin/creatinine ratio macrophage matrix metalloproteinase 9 nephropathy and metabolic syndrome. (ACR), and decreased creatinine clearance release. Interestingly, it was shown that Kirino et al. (86) assessed renal function at weeks 12 and 24. Repeated treatments linagliptin administered in a rat model of in F344/DuCrlCrlj rats, a substrain of the with vildagliptin significantly reduced dia- sepsis ameliorated lipopolysaccharide- inbred Fischer 344 strain lacking DPP-4 betic albuminuria, proteinuria, and serum induced endothelial dysfunction in addi- enzyme activity. Interestingly, they creatinine in diabetic rats at week 12. tion to reduced aortic infiltration with showed that DPP-4–deficient rats were The fractional mesangial area and extent inflammatory cells (71). relatively resistant to developing strep- of segmental glomerulosclerosis were In humans, Rizzo et al. (72) observed tozotocin (STZ) diabetes, but once dia- significantly higher in the untreated that vildagliptin determined reductions betic, they were more susceptible to diabetic group compared with the in nitrotyrosine, IL-6, and IL-18 concen- reduction of glomerular filtration rate. nondiabetic groups. Treatment with trations that were correlated with less- Conversely, Mega et al. (87) assessed vildagliptin significantly lowered the ened glycemic variability. This might the effect of chronic low-dose sitagliptin fractional mesangial area and reduced have a potential benefit of preventing on renal lesions in a T2D rat model of the glomerular sclerosis indexes in dia- atherosclerosis progression in patients diabetic nephropathy. Sitagliptin amelio- betic rats. Furthermore, vildagliptin re- with T2D (73). In another study, sitagliptin rated renal lesions, including glomerular, duced interstitial expansion of diabetic induced a significant reduction of tubulointerstitial, and vascular lesions. rats by 33%. A more recent study by proinflammatory cytokines, TNF-a,en- Whether these effects were direct or de- Vavrinec et al. (93) showed that vildaglip- dotoxin receptor, Toll-like receptors 4 pendent from glucose reduction was not tin, without affecting plasma glucose and 2, nuclear factor-kB, and C-reactive ascertained. Consistent with these obser- levels or proteinuria, was able to decrease protein and IL-6 concentrations (74). vations, another study reported that sita- glomerulosclerosis and restore myogenic Satoh-Asahara et al. (75) showed that gliptin decreased IL-1b and TNF-a levels arteriolar constriction to normal levels, sitagliptin50mgq.i.d.for3months and prevented the increase of BAX/Bcl-2 possibly due to reduced oxidative stress. decreased serum levels of amyloid A– ratio, Bid protein levels, and TUNEL-- This series of experimental studies shows LDL, C-reactive protein, and TNF-a.In positive cells. Such data indicate protective that DPP4-I at the kidney level may pro- conclusion, both experimental and hu- effects against inflammation and apopto- mote both negative and positive effects, man studies consistently report that sis in the kidney (88). In a study addressing with most data pointing to protective DPP4-I provides endothelial protection prevention of renal damage, pretreatment effects of DPP4-I on kidney function. and blunts inflammation. While the of diabetic animals with sitagliptin was as- Whether these effects are direct or par- link between inflammation and athero- sociated with normal serum creatinine, tially mediated by changing glucose con- genesis is well known (76), it should blood urea nitrogen, and expression of tis- centration warrants further scrutiny, but be noted that inflammation also pro- sue injury markers following renal ische- results obtained in models of T1D seem to motes development and progression mia/reperfusion injury, while leading to support a direct effect. of diabetic microangiopathy, including normalization of blood glucose to control Clinical Studies nephropathy, retinopathy, and neuropathy levels (89). The beneficial effects of DPP4- Diabetic nephropathy in humans is the (77–79). I have been observed in STZ eNOS–/– consequence of glomerular, tubular, mice, a rodent model of human diabetic vascular, and interstitial structural ab- DPP4-I and the Kidney nephropathy: the coadministration of li- normalities and dysfunctions. The main Experimental Studies nagliptin and the angiotensin receptor stem of preservation of renal function is Since the discovery of DPP-4 as an aden- blocker telmisartan was associated glucose-lowering strategies. In several osine deaminase binding protein (80), with a marked reduction in albuminuria, trials, the reduction in HbA1c has been the expression of DPP-4 has been though telmisartan or linagliptin alone paralleled by an improvement in renal considered a marker of renal injury, in- did not significantly lower this parameter function metrics. In the UK Prospective cluding diabetic nephropathy (81–83). (90). In another study performed in STZ Diabetes Study (UKPDS) trial, each 1% The relationship between DPP-4 activity diabetic rats, linagliptin reduced AGE and reduction in updated mean HbA1c was and kidney function is complex, and avail- RAGE levels, quenched oxidative stress, associated with reductions in risk of 37% able studies frequently report conflicting improved albuminuria, and ameliorated for microvascular complications (94); results. Tofovic et al. (84) were among the histological features of glomerulopathy specifically, the risk reduction for 2888 DPP4-I and Microvascular Diabetes Complications Diabetes Care Volume 37, October 2014

albuminuria was 34%. An even more re- to either linagliptin 5 mg/day or placebo. determined the content and/or distribu- markable effect (54%) was observed in The primary end point was the percent- tion of tight junction proteins occludin the Diabetes Control and Complications age of change in geometric mean ACR and claudin 5, the nitrotyrosine resi- Trial (DCCT) (95). In the Veterans Affairs from baseline to week 24. ACR at week dues, and retinal cells apoptosis as well Diabetes Trial (VADT), any worsening of 24 was reduced by 32% with linagliptin as EPC adhesion to retinal vessels. They albumin excretion as well as the pro- compared with 6% with placebo. Either found that treatment with sitagliptin gression to macroalbuminuria was HbA1c or systolic blood pressure, at base- prevented the changes in the endothe- lower in the intensively treated group line or after treatment, did not influence lial subcellular distribution of the tight than in the standard-treated group (4). the albuminuria-lowering effect of lina- junction proteins induced by diabetes. In the Action to Control Cardiovascular gliptin. However, the mechanisms ac- Sitagliptin also decreased the nitrosa- Risk in Diabetes (ACCORD) trial, a reduc- counting for the direct beneficial effect tive stress, inflammatory state, and ap- tion in microalbuminuria (29%) and in of linagliptin on album excretion remain optosis in diabetic retinas. Diabetic macroalbuminuria (229%) was ob- unclear (100). animals showed decreased circulating served in the intensively treated group Several trials have been performed to EPC levels and EPC adhesiveness to the (5). Similar effects were reported in the determine safety of DPP4-I in diabetic retinal vessels. Sitagliptin allowed a re- Action in Diabetes and Vascular Disease: patients with various degrees of renal covery of the number of EPCs present in Preterax and Diamicron MR Controlled impairment (Table 1). This is important the bloodstream to levels similar to their Evaluation (ADVANCE) trial, with a 21% because episodes of renal failure have number in controls and increased their reduction in new or worsening nephrop- been reported in patients using this cat- adhesive capacity. In another study, athy in those patients randomized to in- egory of glucose-lowering agents. How- Maeda et al. (104) showed that vilda- tensive treatment (2). Therefore, a key ever, it should be emphasized that in gliptin significantly increased retinal question is whether DPP4-I are able to few studies, a direct effect of DPP4-I gene expression of vascular endothelial improve renal metrics beyond their anti- has been assessed on kidney function growth factor, intercellular adhesion hyperglycemic effect. metrics, such as estimated glomerular fil- molecule 1, plasminogen activator inhib- Few studies have been devoted to di- tration rate and microalbuminuria. Note- itor 1, and pigment epithelium-derived rectly assessing the effects of DPP4-I on worthy, in the Saxagliptin Assessment of factor. Gonçalves et al. (105) have also renal functional measures. Hattori (96) Vascular Outcomes Recorded in Patients shown that in the retina of STZ-induced investigated the effect of sitagliptin (50 With Diabetes Mellitus (SAVOR) trial, pa- T1D rats, sitagliptin prevented the in- mg/day) on albuminuria in patients with tients treated with saxagliptin were sig- crease in blood-retinal barrier perme- T2D. Sitagliptin significantly lowered nificantly more likely than patients ability and inhibited the changes in both systolic and diastolic blood pres- receiving placebo to have an improved immunoreactivity and endothelial sub- sures, fasting blood glucose and post- ACR and less likely to have a worsening cellular distribution of occludin, clau- prandial blood glucose, HbA1c,and ratio (23). At the end of the trial, 13.3% din 5, and zonula occludens 1 proteins glycated albumin at 3 and 6 months. of those receiving saxagliptin had a induced by diabetes. Furthermore, They showed that the urinary ACR did worse ratio (,3.4, $3.4 to #33.9, or sitagliptin decreased the retinal inflam- not change in the 6 months before sita- .33.9 mg/mmol ratio categories, re- matory state and neuronal apoptosis, gliptin treatment and decreased signifi- spectively) versus 15.9% of those on pla- thus indicating a direct protective effect cantly in the 6 months after sitagliptin cebo. Among those on saxagliptin, 10.7% on diabetic retinal cells. treatment. They hypothesized that sita- had improved ratio versus 8.7% among Regarding clinical data, in a recent gliptin reduces albuminuria without those on placebo. These data suggest a small double-blind, placebo-controlled, lowering the estimated glomerular fil- protection of the DPP4-I on albumin crossover trial in 50 T2D patients with- tration rate, most likely depending on excretion rate. Though the end-of-trial out retinopathy, Ott et al. (106) found blood glucose reduction and improved difference in HbA1c between saxagliptin- that 6 weeks of saxagliptin treatment blood pressure control. In a crossover and placebo-treated patients was small, significantly reduced retinal capillary study with two DPP4-I, sitagliptin and it remains unclear whether these effects blood flow and improved vasodilation. alogliptin, in 12 T2D patients with micro- are determined by glucose control itself However, to the best of our knowledge, albuminuria, taking angiotensin receptor or a direct effect of DPP4-I, as suggested no study has so far evaluated the effects blockers, it was found that alogliptin re- by experimental studies. of DPP4-I on retinopathy end points in duced urinary albumin levels (97). Tani diabetic patients. et al. (98) evaluated the effects of the DPP4-I and Diabetic Retinopathy DPP4-I vildagliptin on atherogenic LDL Intensive glucose control has been DPP4-I and Neuropathy heterogeneity and albuminuria in dia- shown to provide beneficial effects on Only three experimental studies are betic subjects. After 8 weeks of treat- retinopathy in both T1D (101) and T2D available on the effect of DPP4-I on di- ment, the ACR decreased significantly by (21,102). Few experimental studies abetic neuropathy. In the first report, ;45%. Recently, Groop et al. (99), in a have assessed the effect of DPP4-I on the authors investigated the GLP-1 pooled analysis of four studies, identified diabetic retinopathy. In Zucker diabetic pathway effect on peripheral nerves 217 subjects with T2D and prevalent al- fatty rats, Gonçalves et al. (103) using vildagliptin in STZ-induced dia- buminuria while receiving stable doses assessed the efficacy of sitagliptin in betic rats (107). They showed that daily of renin-angiotensin-aldosterone system preventing glucose-mediated dam- administration of vildagliptin protected inhibitors. Participants were randomized age on the blood-retinal barrier. They from nerve fiber loss compared with care.diabetesjournals.org Avogaro and Fadini 2889

Table 1—Summary of the effects of DPP4-I on microvascular end points at the various sites of diabetes complications Experimental Clinical Complication Model Effects Drug and patients End point Nephropathy STZ diabetic DPP-4–deficient (F344/ ↓ GFR Sitagliptin in 36 T2D patients (96) ↓ hs-CRP, ICAM-1 DuCrlCrlj) rats (86) ↓ Albuminuria Sitagliptin in Zucker diabetic rats ↓ Glomerular, tubulointerstitial, Alogliptin vs. sitagliptin (crossover) ↓ Albuminuria (87) and vascular lesions in 12 T2D patients (97) Sitagliptin in Zucker diabetic fatty ↓ Tubulointerstitial and Vildagliptin in 47 T2D patients (98) ↓ Albuminuria rats (88) glomerular lesions ↓ Apoptosis Sitagliptin in ischemia reperfusion ↑ GFR Linagliptin vs. placebo in 217 T2D ↓ Albuminuria injury in nicotinamide/STZ diabetic ↓ Oxidative stress patients with micro-/ rats (89) ↓ Tissue damage macroalbuminuria (pooled analysis) (99) Linagliptin + telmisartan in STZ ↓ Albuminuria Saxagliptin vs. placebo in .16,000 ↓ Microalbuminuria eNOS–/– mice (90) ↓ Glomerulosclerosis T2D patients (23) Linagliptin in STZ diabetic rats (91) ↓ AGE and RAGE ↓ Oxidative stress ↓ Albuminuria ↓ Glomerulosclerosis Vildagliptin in STZ diabetic rats (92) ↑ GFR ↓ Albuminuria ↓ Glomerulosclerosis ↓ Interstitial fibrosis Vildagliptin in Zucker diabetic fatty ↓ Glomerulosclerosis rats (93) ↑ Arteriolar function ↓ Oxidative stress Retinopathy Sitagliptin in Zucker diabetic fatty ↓ Nitrosative stress, Saxagliptin vs. placebo in 50 T2D ↓ Blood flow rats (103) inflammation, apoptosis patients without retinopathy (106) ↑ Vasodilation ↓ Blood-retinal barrier changes ↑ EPCs Sitagliptin in STZ diabetic rats (105) ↓ Permeability ↓ Blood-retinal barrier changes ↓ Inflammation, apoptosis Vildagliptin in OLETF T2D rats (104) ↓ VEGF, ICAM-1, PAI-1, and PEGF Neuropathy Vildagliptin in STZ diabetic rats (106) ↓ Nerve fiber loss No data available so far PKF275-055 in STZ diabetic rats ↑ Na+/K+-ATPase activity (108) ↑ Nerve conduction velocity ↑ Mechanical and thermal sensitivity Sitagliptin in nicotinamide/STZ T2D ↑ Strength and paw function rats (109) ↓ Nerve cell loss Foot ulcers Linagliptin in ob/ob T2D mice (112) ↑ Epithelialization Vildagliptin vs. placebo in 106 ↑ Granulation tissue ↑ Myofibroblasts diabetic patients with ulcers (113) ↑ Capillary density ↓ Inflammation ↑ HIF-1a and VEGF ↓ Nitrotyrosines ↓ Proteasome activity Experimental and clinical data are reported separately. GFR, glomerular filtration rate; HIF, hypoxia inducible factor; PAI, plasminogen activator inhibitor; PEGF, placenta-derived growth factor; VEGF, vascular endothelial growth factor.

untreated rats, and they also observed a PKF275–055 treatment restored me- reverse the clinical features of diabetic significantly lower decrease of intraepi- chanical sensitivity thresholds by ;50% neuropathy, these experimental evi- dermal nerve fiber density. In a second and progressively improved the alter- dences on the protective effects of study, the beneficial effects of PKF275– ation in thermal responsiveness. Finally, DPP4-I are promising and deserve future 055, a selective DPP4-I, were tested in in rats with nicotinamide-/STZ-induced attention. STZ-induced diabetic peripheral neurop- diabetes, Sharma et al. (109) observed athy (108). It was shown that this drug that sitagliptin and sitagliptin combined DPP4-I and Diabetic Foot Ulcers partially counteracted the nerve con- with metformin or amitriptyline re- Delayed wound healing in diabetes is a duction velocity deficit observed in un- sulted in neural protection and reversed major source of morbidity and mortal- treated diabetic rats but did not improve the alteration of biochemical parame- ity. It results from the combination of mechanical and thermal sensitivity. ters in diabetic rats. Given the paucity vasculopathy and neuropathy, and often When used in a therapeutic setting, of treatment strategies available to leads to minor and major amputations 2890 DPP4-I and Microvascular Diabetes Complications Diabetes Care Volume 37, October 2014

(110). Many diabetic patients with ische- complication (114). Nowadays, shortage in the ischemic muscle (125). Therefore, mic foot ulcers are not amenable to sur- of EPCs is considered a contributor to the although it is still unknown whether DPP-4 gical revascularization of lower limb development of diabetes complications. inhibition is able to counteract structural arteries because of multiple distal steno- Therefore, investigations into the mecha- BM remodeling induced by diabetes, sis. In addition, microangiopathy is a ma- nisms that impair EPC in diabetes are of modulation of the DPP-4/SDF-1a axis jor contributor to the shortage of oxygen great interest to devise new endogenous can reverse BM dysfunction and improve and nutrient supply to the granulation regenerative therapeutic strategies (115). microvascular health in distant organs. As tissue, thus contributing to delayed heal- Experimental modeling suggests that the the BM is emerging as a central house- ing (111). In this setting, it is of paramount low circulating EPC level in diabetes is keeper, able to affect cellular turnover at importance to devise therapeutic strate- attributable to impaired mobilization distant sites, this hitherto overlooked site gies to restore the structure and function from the BM. Indeed, in the last years, it of DPP-4 action might uncover interesting of the epithelium and granulation tissue has been recognized that the BM is a and unexpected potentials for microvas- to aid wound healing in diabetic patients. novel and hitherto neglected site of dia- cular protection in diabetes (20). For the possible improvement in micro- betic microangiopathy (116). Features of vascular outcomes obtained with DPP4-I the diabetic BM in rodents and humans CONCLUSIONS as outlined above, these drugs are novel include autonomic neuropathy (117,118); The current standards of care significantly candidates for the medical treatment of rarefaction of capillaries, arterioles, and reduce but unfortunately do not elimi- diabetic foot ulcers. sinusoids; increased microvascular per- nate the risk of diabetic microangiopathy. Experimentally, it was shown that meability; excess oxidative stress; and re- This has important implications because, DPP-4 expression and activity are in- location of stem cells relative to altered although microangiopathy is rarely the creased in the wounded skin of diabetic oxygen gradients across an extensively re- cause of death in diabetic patients, it is obese mice (112). In turn, excess DPP-4 modeled BM niche (18,19,119). As a con- one of the most important risk factors for activity is supposed to degrade the che- sequence of these profound structural CVD (126). Furthermore, this implies that mokine and angiocrine factor SDF-1a and functional abnormalities, BM stem in the past years, the most commonly (CXCL12), thus impairing vascularization cells show reduced survival and altered used glucose-lowering drugs were unable andgrowthofthegranulationtissue. responsiveness to mobilizing agents, a to effectively decrease plasma glucose in Consistently, DPP4-I with linagliptin re- condition now deemed as “diabetic order to avoid the onset and the progres- sulted in accelerated wound healing in stem cell mobilopathy” (120). Indeed, in sion of microvascular disease. Another diabetic obese mice compared with un- humans and animals, BM stem cell mobi- relevant issue is that many glucose- treated mice. This was associated with lization is impaired by diabetes (121,122). lowering agents need to be dose ad- improved reepithelialization, reduced Importantly, DPP-4 seems to play a major justed or should not be used in the set- inflammation, and enhanced the forma- role in the regulation of stem cell traffick- ting of stage III–IV CKD or in those tion of myofibroblasts, all features of a ing and its role in the diabetic BM dys- receiving dialysis (127). Indeed, the healthier granulation tissue (112). function has been shown recently (123). safety of DPP4-I has been demonstrated As a clinical counterpart of these ex- Indeed, DPP-4 proteolytic activity deter- in several trials in patients with different perimental observations, it was found mines the systemic and local concentra- degrees of renal impairment (128–133). that treatment with a DPP4-I for just tions of the chemokine and stem cell Extensive experimental data and pre- 12 weeks improved healing features in trafficking regulator SDF-1a.Onone liminary clinical studies indicate that T2D patients with chronic nonhealing hand, DPP-4 activity is required for the DPP4-I may improve microvascular foot ulcers, improving nitrosative stress, mobilizing effect of granulocyte colony- structure and function (Table 1 and response to hypoxia, and capillary den- stimulating factor (G-CSF), the most com- Fig. 1). Whether the effects of DPP4-I sity (113). Although the mechanisms monly used agent to stimulate the egress aremediatedbyimprovedglucosecontrol that translate DPP4-I into such a strong of stem and progenitor cells toward the or by pleiotropic off-target actions of improvement in granulation tissue bloodstream (124). Therefore, although DPP4-I on nonincretin substrates remains structure and function in vivo remain DPP4-I is not expected to affect G-CSF– unclear. A few hints can help answer this to be completely understood, these pre- induced mobilization, data obtained in question with available data (Table 2). liminary data are promising and, if con- patients with diabetes indicate that a Preclinical findings obtained in vitro firmed, would enrich our therapeutic maladaptive DPP-4 response to G-CSF and using animal models of T1D (e.g., armamentarium to accelerate wound contributes impaired mobilization of STZ-induced diabetes) suggest that favor- healing in diabetes. stem and proangiogenic cells (122). In able effects of DPP4-I are conveyed inde- addition, a screening of factors associ- pendently of glycemic effects. Moreover, DPP4-I and the Diabetic BM ated with abnormal stem cell trafficking short-term studies in T2D patients, show- In addition to being the major source of in patients with diabetes and high car- ing raised EPCs after just 4 weeks of hematopoietic cells in the adult organism, diovascular risk identified excess DPP-4 sitagliptin treatment, are likely exploring the BM is a reservoir of stem/progenitor activity as a candidate negative modu- pleiotropic rather than glycemic effects. cells involved in endothelial repair and lator of mobilization (125). Using Finally, while lowering HbA1c significantly angiogenesis (20). EPCs, together with the F344/DuCrlCrlj rat model, it was prevents microangiopathy, the reduced other progenitor cell phenotypes, are re- demonstrated that DPP-4 deficiency im- development and progression of micro- duced in diabetes, especially in the pres- proves postischemic BM EPC mobiliza- albuminuria in the SAVOR trial is unlikely ence of any macro- or microvascular tion and improved microvascular density to be fully explained by the marginal care.diabetesjournals.org Avogaro and Fadini 2891

Figure 1—A schematic representation summarizing the roles of DPP-4 inhibition on diabetic microangiopathy. Experimental evidence indicates that DPP4-I affects inﬂammation, vascular responses, lipids, blood pressure, and BM function. In combination with increased levels of GLP-1 and improved glucose control, these effects can mediate protection from microvascular diabetes complications. BNP, B-type natriuretic factor; GFR, glomerular ﬁltration rate; HIF, hypoxia inducible factor; NO, nitric oxide; TG, triglycerides; VEGF, vascular endothelial growth factor.

0.2–0.3%reductioninHbA1c obtained studies did not show any significant effect heterogeneous population of diabetic pa- with saxagliptin compared with placebo on both blood pressure and lipids. If these tients, as long as results from specifically throughout the trial (23). However, it purported protective effects of DPP4-I designed randomized controlled trials must be made clear that preliminary translate into better outcomes in people are not available. In addition to the afore- data demonstrate that GLP-1RA have with diabetes, they should be verified by mentioned aspects, the effect of DPP4-I stronger efficacy in terms of correction the several ongoing clinical trials. Indeed, on BM stem cells is also promising to of the major risk factors for CVD (including caution should be paid when trying to achieve microvascular protection at dis- blood pressure and lipids) (135); indeed translate findings obtained in animal tant sites. Ultimately, reducing the burden SAVOR, EXAMINE, and other smaller models and small clinical studies to the of microangiopathy may translate into

Table 2—Potential methodological evidence that the effects of DPP4-I on microvascular protection may be conveyed independently of glucose control Mechanisms References DPP-4 inhibition improves microvascular end points in vitro (41,134) DPP-4 inhibition improves microvascular end points in animal models of T1D (90,92,107–109) Short-term (4–6 weeks) DPP-4 inhibition in T2D has provided effects that occur before full development of the antihyperglycemic effect (44,106) Prevention of microvascular end point is not fully explained by improved glucose control (23) 2892 DPP4-I and Microvascular Diabetes Complications Diabetes Care Volume 37, October 2014

improved cardiovascular outcomes in amputations in a diabetic population. Diabetes 29. De Meester I, Korom S, Van Damme J, diabetes. Care 1999;22:951–959 Scharpe´ S. CD26, let it cut or cut it down. Im- 13. Wong TY, Cheung N, Islam FM, et al. Rela- munol Today 1999;20:367–375 tion of retinopathy to coronary artery calcifica- 30. Morrison ME, Vijayasaradhi S, Engelstein D, tion: the Multi-Ethnic Study of Atherosclerosis. Albino AP, Houghton AN. A marker for neoplas- Duality of Interest. A.A. received speaker Am J Epidemiol 2008;167:51–58 tic progression of human melanocytes is a cell honoraria from Sanofi, Boehringer Ingelheim, 14. de Kreutzenberg SV, Coracina A, Volpi A, surface ectopeptidase. J Exp Med 1993;177: Eli Lilly, Merck Sharp & Dohme, Novartis, Novo et al. Microangiopathy is independently associ- 1135–1143 Nordisk, Servier, Recordati, Bristol-Myers ated with presence, severity and composition of 31. Morimoto C, Schlossman SF. The structure Squibb, and AstraZeneca and is an advisory carotid atherosclerosis in type 2 diabetes. Nutr and function of CD26 in the T-cell immune re- member for Boehringer Ingelheim, Novartis, Metab Cardiovasc Dis 2011;21:286–293 sponse. Immunol Rev 1998;161:55–70 Bristol-Myers Squibb, AstraZeneca, and Sanofi. 15. Bigazzi R, Bianchi S, Baldari D, Campese VM. 32. O’Leary H, Ou X, Broxmeyer HE. The role of G.P.F. received speaker honoraria from Eli Lilly, Microalbuminuria predicts cardiovascular dipeptidyl peptidase 4 in hematopoiesis and Sanofi, Novo Nordisk, AstraZeneca, and Bristol- events and renal insufficiency in patients with transplantation. Curr Opin Hematol 2013;20: Myers Squibb. No other potential conflicts of essential hypertension. J Hypertens 1998;16: 314–319 interest relevant to this article were reported. 1325–1333 33. Medeiros MD, Turner AJ. Processing and Author Contributions. A.A. researched data 16. Karalliedde J, Viberti G. Hypertension and mi- metabolism of peptide-YY: pivotal roles of and wrote, reviewed, and edited the manuscript. croalbuminuria: risk factors for cardiovascular dis- dipeptidylpeptidase-IV, aminopeptidase-P, and G.P.F. researched data and wrote the manuscript. ease in diabetes. Curr Hypertens Rep 2005;7:1–2 endopeptidase-24.11. Endocrinology 1994;134: 17. Schiffrin EL, Lipman ML, Mann JF. Chronic 2088–2094 References kidney disease: effects on the cardiovascular 34. Baetta R, Corsini A. Pharmacology of dipep- system. Circulation 2007;116:85–97 1. Seshasai SR, Kaptoge S, Thompson A, et al.; tidyl peptidase-4 inhibitors: similarities and dif- 18. Oikawa A, Siragusa M, Quaini F, et al. Di- – Emerging Risk Factors Collaboration. Diabetes ferences. Drugs 2011;71:1441 1467 abetes mellitus induces bone marrow microan- mellitus, fasting glucose, and risk of cause-specific 35. Pala L, Rotella CM. The role of DPP4 activity giopathy. Arterioscler Thromb Vasc Biol 2010; death. N Engl J Med 2011;364:829–841 in cardiovascular districts: in vivo and in vitro – 2. Patel A, MacMahon S, Chalmers J, et al.; 30:498 508 evidence. J Diabetes Res 2013;2013:590456 ADVANCE Collaborative Group. Intensive blood 19. Spinetti G, Cordella D, Fortunato O, et al. 36. Matsubara J, Sugiyama S, Sugamura K, et al. fl glucose control and vascular outcomes in pa- Global remodeling of the vascular stem cell A dipeptidyl peptidase-4 inhibitor, des- uoro- tients with type 2 diabetes. N Engl J Med niche in bone marrow of diabetic patients: im- sitagliptin, improves endothelial function and 2008;358:2560–2572 plication of the microRNA-155/FOXO3a signal- reduces atherosclerotic lesion formation in apo- – fi 3. Holman RR, Paul SK, Bethel MA, Matthews ing pathway. Circ Res 2013;112:510 522 lipoprotein E-de cient mice. J Am Coll Cardiol – DR, Neil HA. 10-year follow-up of intensive glu- 20. Fadini GP, Avogaro A. It is all in the blood: 2012;59:265 276 cose control in type 2 diabetes. N Engl J Med the multifaceted contribution of circulating pro- 37. Mason RP, Jacob RF, Kubant R, Ciszewski A, 2008;359:1577–1589 genitor cells in diabetic complications. Exp Di- Corbalan JJ, Malinski T. Dipeptidyl peptidase-4 4. Duckworth W, Abraira C, Moritz T, et al.; abetes Res 2012;2012:742976 inhibition with saxagliptin enhanced nitric oxide VADT Investigators. Glucose control and vascu- 21. Ismail-Beigi F, Craven T, Banerji MA, et al.; release and reduced blood pressure and sICAM- lar complications in veterans with type 2 diabe- ACCORD Trial Group. Effect of intensive treat- 1 levels in hypertensive rats. J Cardiovasc Phar- tes. N Engl J Med 2009;360:129–139 ment of hyperglycaemia on microvascular out- macol 2012;60:467–473 5. Gerstein HC, Miller ME, Genuth S, et al.; comes in type 2 diabetes: an analysis of the 38. Mason RP, Jacob RF, Kubant R, et al. Effect ACCORD Study Group. Long-term effects of in- ACCORD randomised trial. Lancet 2010;376: of enhanced glycemic control with saxagliptin – tensive glucose lowering on cardiovascular out- 419 430 on endothelial nitric oxide release and CD40 comes. N Engl J Med 2011;364:818–828 22. Agrawal L, Azad N, Emanuele NV, et al.; Vet- levels in obese rats. J Atheroscler Thromb 6. Brownlee M. The pathobiology of diabetic erans Affairs Diabetes Trial (VADT) Study Group. 2011;18:774–783 complications: a unifying mechanism. Diabetes Observation on renal outcomes in the Veterans 39. Masuyama J, Yoshio T, Suzuki K, et al. Char- 2005;54:1615–1625 Affairs Diabetes Trial. Diabetes Care 2011;34: acterization of the 4C8 antigen involved in 7. Factor SM, Borczuk A, Charron MJ, Fein FS, 2090–2094 transendothelial migration of CD26(hi) T cells van Hoeven KH, Sonnenblick EH. Myocardial al- 23. Scirica BM, Bhatt DL, Braunwald E, et al.; after tight adhesion to human umbilical vein terations in diabetes and hypertension. Diabe- SAVOR-TIMI 53 Steering Committee and Inves- endothelial cell monolayers. J Exp Med 1999; tes Res Clin Pract 1996;31(Suppl.):S133–S142 tigators. Saxagliptin and cardiovascular out- 189:979–990 8. Gonzalez-Quesada C, Cavalera M, Biernacka comes in patients with type 2 diabetes 40. Matsui T, Nishino Y, Takeuchi M, Yamagishi A, et al. Thrombospondin-1 induction in the di- mellitus. N Engl J Med 2013;369:1317–1326 S. Vildagliptin blocks vascular injury in thoracic abetic myocardium stabilizes the cardiac matrix 24. Gerstein HC, Bosch J, Dagenais GR, et al.; aorta of diabetic rats by suppressing advanced in addition to promoting vascular rarefaction ORIGIN Trial Investigators. Basal insulin and car- glycation end product-receptor axis. Pharmacol through angiopoietin-2 upregulation. Circ Res diovascular and other outcomes in dysglycemia. Res 2011;63:383–388 2013;113:1331–1344 N Engl J Med 2012;367:319–328 41. Ishibashi Y, Matsui T, Takeuchi M, 9. Scognamiglio R, Negut C, De Kreutzenberg 25. Karagiannis T, Paschos P, Paletas K, Yamagishi S. Sitagliptin augments protective ef- SV, Tiengo A, Avogaro A. Postprandial myocar- Matthews DR, Tsapas A. Dipeptidyl peptidase- fects of GLP-1 against advanced glycation end dial perfusion in healthy subjects and in type 2 4 inhibitors for treatment of type 2 diabetes product receptor axis in endothelial cells. Horm diabetic patients. Circulation 2005;112:179–184 mellitus in the clinical setting: systematic review Metab Res 2011;43:731–734 10. Kawasaki R, Cheung N, Islam FM, et al.; and meta-analysis. BMJ 2012;344:e1369 42. Maeda S, Matsui T, Yamagishi S. Vildagliptin Multi-Ethnic Study of Atherosclerosis. Is dia- 26. Monami M, Iacomelli I, Marchionni N, inhibits oxidative stress and vascular damage in betic retinopathy related to subclinical cardio- Mannucci E. Dipeptydil peptidase-4 inhibitors streptozotocin-induced diabetic rats. Int J Car- vascular disease? Ophthalmology 2011;118: in type 2 diabetes: a meta-analysis of random- diol 2012;158:171–173 860–865 ized clinical trials. Nutr Metab Cardiovasc Dis 43. Zeng Y, Li C, Guan M, et al. The DPP-4 in- 11. Cheung N, Wang JJ, Klein R, Couper DJ, 2010;20:224–235 hibitor sitagliptin attenuates the progress of Sharrett AR, Wong TY. Diabetic retinopathy 27. Fadini GP, Avogaro A. Cardiovascular ef- atherosclerosis in apolipoprotein-E-knockout and the risk of coronary heart disease: the Ath- fects of DPP-4 inhibition: beyond GLP-1. Vascul mice via AMPK- and MAPK-dependent mecha- erosclerosis Risk in Communities Study. Diabe- Pharmacol 2011;55:10–16 nisms. Cardiovasc Diabetol 2014;13:32 tes Care 2007;30:1742–1746 28. White WB, Cannon CP, Heller SR, et al.; 44. Fadini GP, Boscaro E, Albiero M, et al. The 12. Moss SE, Klein R, Klein BE; The Wisconsin EXAMINE Investigators. Alogliptin after acute oral dipeptidyl peptidase-4 inhibitor sitagliptin Epidemiologic Study of Diabetic Retinopathy. coronary syndrome in patients with type 2 di- increases circulating endothelial progenitor The 14-year incidence of lower-extremity abetes. N Engl J Med 2013;369:1327–1335 cells in patients with type 2 diabetes: possible care.diabetesjournals.org Avogaro and Fadini 2893

role of stromal-derived factor-1alpha. Diabetes randomized clinical trials. Diabetes Obes Metab 75. Satoh-Asahara N, Sasaki Y, Wada H, et al. A Care 2010;33:1607–1609 2013;15:112–120 dipeptidyl peptidase-4 inhibitor, sitagliptin, ex- 45. Poncina N, Albiero M, Menegazzo L, Cappellari 60. Katout M, Zhu H, Rutsky J, et al. Effect of GLP- erts anti-inflammatory effects in type 2 diabetic R, Avogaro A, Fadini GP. The dipeptidyl peptidase-4 1 mimetics on blood pressure and relationship to patients. Metabolism 2013;62:347–351 inhibitor saxagliptin improves function of circu- weight loss and glycemia lowering: results of a 76. Libby P. Inflammation in atherosclerosis. lating pro-angiogenic cells from type 2 diabetic systematic meta-analysis and meta-regression. Nature 2002;420:868–874 patients. Cardiovasc Diabetol 2014;13:92 Am J Hypertens 2014;27:130–139 77. Kern TS. Contributions of inflammatory pro- 46. Matsubara J, Sugiyama S, Akiyama E, et al. 61. Kim M, Platt MJ, Shibasaki T, et al. GLP-1 cesses to the development of the early stages of Dipeptidyl peptidase-4 inhibitor, sitagliptin, im- receptor activation and Epac2 link atrial natri- diabetic retinopathy. Exp Diabetes Res 2007; proves endothelial dysfunction in association uretic peptide secretion to control of blood 2007:95103 with its anti-inflammatory effects in patients pressure. Nat Med 2013;19:567–575 78. Navarro-Gonzalez´ JF, Mora-Fernandez´ C. The with coronary artery disease and uncontrolled 62. Devin JK, Pretorius M, Nian H, Yu C, role of inflammatory cytokines in diabetic ne- diabetes. Circ J 2013;77:1337–1344 Billings FT 4th, Brown NJ. Substance P in- phropathy. J Am Soc Nephrol 2008;19:433–442 47. Noda Y, Miyoshi T, Oe H, et al. Alogliptin creases sympathetic activity during combined 79. Vincent AM, Callaghan BC, Smith AL, ameliorates postprandial lipemia and postpran- angiotensin-converting enzyme and dipeptidyl Feldman EL. Diabetic neuropathy: cellular dial endothelial dysfunction in non-diabetic peptidase-4 inhibition. Hypertension 2014;63: mechanisms as therapeutic targets. Nat Rev subjects: a preliminary report. Cardiovasc Dia- 951–957 Neurol 2011;7:573–583 betol 2013;12:8 63. Marney A, Kunchakarra S, Byrne L, Brown NJ. 80. Weisman MI, Caiolfa VR, Parola AH. Aden- 48. Kubota Y, Miyamoto M, Takagi G, et al. The Interactive hemodynamic effects of dipeptidyl osine deaminase-complexing protein from bo- dipeptidyl peptidase-4 inhibitor sitagliptin im- peptidase-IV inhibition and angiotensin-converting vine kidney. Isolation of two distinct subunits. J proves vascular endothelial function in type 2 enzyme inhibition in humans. Hypertension 2010; Biol Chem 1988;263:5266–5270 diabetes. J Korean Med Sci 2012;27:1364–1370 56:728–733 81. Scherberich JE, Wiemer J, Schoeppe W. Bio- 49. Liu L, Liu J, Wong WT, et al. Dipeptidyl pep- 64. Kubota T, Flentke GR, Bachovchin WW, chemical and immunological properties of urinary tidase 4 inhibitor sitagliptin protects endothelial Stollar BD. Involvement of dipeptidyl peptidase angiotensinase A and dipeptidylaminopeptidase function in hypertension through a glucagon- IV in an in vivo immune response. Clin Exp Im- IV. Their use as markers in patients with renal like peptide 1-dependent mechanism. Hyper- munol 1992;89:192–197 cell injury. Eur J Clin Chem Clin Biochem 1992;30: tension 2012;60:833–841 65. Zhong J, Rao X, Deiuliis J, et al. A potential 663–668 50. Ayaori M, Iwakami N, Uto-Kondo H, et al. role for dendritic cell/macrophage-expressing 82. Ishii N, Ogawa Z, Itoh H, Ikenaga H, Saruta T. Dipeptidyl peptidase-4 inhibitors attenuate endo- DPP4 in obesity-induced visceral inflammation. Diagnostic significance of urinary enzymes for thelial function as evaluated by flow-mediated Diabetes 2013;62:149–157 diabetes mellitus and hypertension. Enzyme vasodilatation in type 2 diabetic patients. J Am 66. Lamers D, Famulla S, Wronkowitz N, et al. Protein 1994-1995;48:174–182 Heart Assoc 2013;2:e003277 Dipeptidyl peptidase 4 is a novel adipokine po- 83. Sun AL, Deng JT, Guan GJ, et al. Dipeptidyl 51. Monami M, Vitale V, Ambrosio ML, et al. tentially linking obesity to the metabolic syn- peptidase-IV is a potential molecular biomarker Effects on lipid profile of dipeptidyl peptidase drome. Diabetes 2011;60:1917–1925 in diabetic kidney disease. Diab Vasc Dis Res 4 inhibitors, pioglitazone, acarbose, and sulfo- 67. Ferreira L, Teixeira-de-Lemos E, Pinto F, 2012;9:301–308 nylureas: meta-analysis of placebo-controlled et al. Effects of sitagliptin treatment on dysme- 84. Tofovic DS, Bilan VP, Jackson EK. Sitagliptin trials. Adv Ther 2012;29:736–746 tabolism, inflammation, and oxidative stress in augments angiotensin II-induced renal vasocon- 52. van Genugten RE, Moller-Goede¨ DL, van an animal model of type 2 diabetes (ZDF rat). striction in kidneys from rats with metabolic Raalte DH, Diamant M. Extra-pancreatic effects Mediators Inflamm 2010;2010:592760 syndrome. Clin Exp Pharmacol Physiol 2010; of incretin-based therapies: potential benefitfor 68. Dobrian AD, Ma Q, Lindsay JW, et al. Dipep- 37:689–691 cardiovascular-risk management in type 2 diabe- tidyl peptidase IV inhibitor sitagliptin reduces 85. Jackson EK, Dubinion JH, Mi Z. Effects of tes. Diabetes Obes Metab 2013;15:593–606 local inflammation in adipose tissue and in pan- dipeptidyl peptidase iv inhibition on arterial 53. Kuo LE, Abe K, Zukowska Z. Stress, NPY and creatic islets of obese mice. Am J Physiol Endo- blood pressure. Clin Exp Pharmacol Physiol vascular remodeling: Implications for stress- crinol Metab 2011;300:E410–E421 2008;35:29–34 related diseases. Peptides 2007;28:435–440 69. Shah Z, Kampfrath T, Deiuliis JA, et al. Long- 86. Kirino Y, Sato Y, Kamimoto T, Kawazoe K, 54. Girardi AC, Knauf F, Demuth HU, Aronson term dipeptidyl-peptidase 4 inhibition reduces Minakuchi K, Nakahori Y. Interrelationship of PS. Role of dipeptidyl peptidase IV in regulating atherosclerosis and inflammation via effects on dipeptidyl peptidase IV (DPP4) with the devel- activity of Na+/H+ exchanger isoform NHE3 in monocyte recruitment and chemotaxis. Circula- opment of diabetes, dyslipidaemia and ne- proximal tubule cells. Am J Physiol Cell Physiol tion 2011;124:2338–2349 phropathy: a streptozotocin-induced model 2004;287:C1238–C1245 70. Vittone F, Liberman A, Vasic D, et al. Sita- using wild-type and DPP4-deficient rats. J Endo- 55. Rieg T, Gerasimova M, Murray F, et al. Na- gliptin reduces plaque macrophage content and crinol 2009;200:53–61 triuretic effect by exendin-4, but not the DPP-4 stabilises arteriosclerotic lesions in Apoe (-/-) 87. Mega C, de Lemos ET, Vala H, et al. Diabetic inhibitor alogliptin, is mediated via the GLP-1 mice. Diabetologia 2012;55:2267–2275 nephropathy amelioration by a low-dose sita- receptor and preserved in obese type 2 diabetic 71. Kroller-Sch¨ on¨ S, Knorr M, Hausding M, et al. gliptin in an animal model of type 2 diabetes mice. Am J Physiol Renal Physiol 2012;303: Glucose-independent improvement of vascular (Zucker diabetic fatty rat). Exp Diabetes Res F963–F971 dysfunction in experimental sepsis by dipeptidyl- 2011;2011:162092 56. Pacheco BP, Crajoinas RO, Couto GK, et al. peptidase 4 inhibition. Cardiovasc Res 2012;96: 88. Marques C, Mega C, Gonçalves A, et al. Dipeptidyl peptidase IV inhibition attenuates 140–149 Sitagliptin prevents inflammation and apoptotic blood pressure rising in young spontaneously 72. Rizzo MR, Barbieri M, Marfella R, Paolisso cell death in the kidney of type 2 diabetic ani- hypertensive rats. J Hypertens 2011;29:520– G. Reduction of oxidative stress and inflamma- mals. Mediators Inflamm 2014;2014:538737 528 tion by blunting daily acute glucose fluctuations 89. Vaghasiya J, Sheth N, Bhalodia Y, Manek R. 57. Vanderheyden M, Bartunek J, Goethals M, in patients with type 2 diabetes: role of dipep- Sitagliptin protects renal ischemia reperfusion et al. Dipeptidyl-peptidase IV and B-type natri- tidyl peptidase-IV inhibition. Diabetes Care induced renal damage in diabetes. Regul Pept uretic peptide. From bench to bedside. Clin 2012;35:2076–2082 2011;166:48–54 Chem Lab Med 2009;47:248–252 73. Barbieri M, Rizzo MR, Marfella R, et al. De- 90. Alter ML, Ott IM, von Websky K, et al. DPP-4 58. Cobble ME, Frederich R. Saxagliptin for the creased carotid atherosclerotic process by con- inhibition on top of angiotensin receptor block- treatment of type 2 diabetes mellitus: assessing trol of daily acute glucose fluctuations in ade offers a new therapeutic approach for di- cardiovascular data. Cardiovasc Diabetol 2012; diabetic patients treated by DPP-IV inhibitors. abetic nephropathy. Kidney Blood Press Res 11:6 Atherosclerosis 2013;227:349–354 2012;36:119–130 59. Monami M, Ahren´ B, Dicembrini I, 74. Makdissi A, Ghanim H, Vora M, et al. Sita- 91. Nakashima S, Matsui T, Takeuchi M, Mannucci E. Dipeptidyl peptidase-4 inhibitors gliptin exerts an antinflammatory action. J Clin Yamagishi SI. Linagliptin blocks renal damage and cardiovascular risk: a meta-analysis of Endocrinol Metab 2012;97:3333–3341 in type 1 diabetic rats by suppressing advanced 2894 DPP4-I and Microvascular Diabetes Complications Diabetes Care Volume 37, October 2014

glycation end products-receptor axis. Horm 105. Gonçalves A, Marques C, Leal E, et al. Di- 121. Fadini GP, Sartore S, Schiavon M, et al. Di- Metab Res. 7 April 2014 [Epub ahead of print] peptidyl peptidase-IV inhibition prevents blood- abetes impairs progenitor cell mobilisation after 92. Liu WJ, Xie SH, Liu YN, et al. Dipeptidyl pep- retinal barrier breakdown, inflammation and hindlimb ischaemia-reperfusion injury in rats. tidase IV inhibitor attenuates kidney injury in neuronal cell death in the retina of type 1 di- Diabetologia 2006;49:3075–3084 streptozotocin-induced diabetic rats. J Pharma- abetic rats. Biochim Biophys Acta 2014;1842: 122. Fadini GP, Albiero M, Vigili de Kreutzenberg col Exp Ther 2012;340:248–255 1454–1463 S, et al. Diabetes impairs stem cell and proangio- 93. Vavrinec P, Henning RH, Landheer SW, et al. 106. Ott C, Raff U, Schmidt S, et al. Effects of genic cell mobilization in humans. Diabetes Care Vildagliptin restores renal myogenic function saxagliptin on early microvascular changes in 2013;36:943–949 and attenuates renal sclerosis independently patients with type 2 diabetes. Cardiovasc Dia- 123. Fadini GP, Avogaro A. Dipeptidyl peptidase-4 of effects on blood glucose or proteinuria in betol 2014;13:19 inhibition and vascular repair by mobilization of Zucker diabetic fatty rat. Curr Vasc Pharmacol 107. Jin HY, Liu WJ, Park JH, Baek HS, Park TS. endogenous stem cells in diabetes and beyond. 2013 Effect of dipeptidyl peptidase-IV (DPP-IV) in- Atherosclerosis 2013;229:23–29 94. Stratton IM, Adler AI, Neil HA, et al. Associ- hibitor (vildagliptin) on peripheral nerves in 124. Christopherson KW, Cooper S, Hangoc G, ation of glycaemia with macrovascular and mi- streptozotocin-induced diabetic rats. Arch Broxmeyer HE. CD26 is essential for normal crovascular complications of type 2 diabetes Med Res 2009;40:536–544 G-CSF-induced progenitor cell mobilization as (UKPDS 35): prospective observational study. 108. Bianchi R, Cervellini I, Porretta-Serapiglia determined by CD26-/- mice. Exp Hematol BMJ 2000;321:405–412 C, et al. Beneficial effects of PKF275-055, a 2003;31:1126–1134 95. de Boer IH, Rue TC, Cleary PA, et al.; Diabetes novel, selective, orally bioavailable, long-acting 125. Fadini GP, Albiero M, Seeger F, et al. Stem Control and Complications Trial/Epidemiology of dipeptidyl peptidase IV inhibitor in streptozotocin- cell compartmentalization in diabetes and high Diabetes Interventions and Complications Study induced diabetic peripheral neuropathy. J Pharma- cardiovascular risk reveals the role of DPP-4 in Research Group. Long-term renal outcomes of col Exp Ther 2012;340:64–72 diabetic stem cell mobilopathy. Basic Res Car- patients with type 1 diabetes mellitus and micro- 109. Sharma AK, Sharma A, Kumari R, et al. Sitaglip- diol 2013;108:313 albuminuria: an analysis of the Diabetes Control tin, sitagliptin and metformin, or sitagliptin and am- 126. Gerstein HC, Ambrosius WT, Danis R, and Complications Trial/Epidemiology of Diabe- itriptyline attenuate streptozotocin-nicotinamide et al.; ACCORD Study Group. Diabetic retinopa- tes Interventions and Complications cohort. induceddiabeticneuropathyinrats.JBiomedRes thy, its progression, and incident cardiovascular Arch Intern Med 2011;171:412–420 2012;26:200–210 events in the ACCORD trial. Diabetes Care 2013; 96. Hattori S. Sitagliptin reduces albuminuria in 110. Ramsey SD, Newton K, Blough D, et al. In- 36:1266–1271 patients with type 2 diabetes. Endocr J 2011;58: cidence, outcomes, and cost of foot ulcers in 127. Flynn C, Bakris GL. Noninsulin glucose- 69–73 patients with diabetes. Diabetes Care 1999;22: lowering agents for the treatment of patients 97. Fujita H, Taniai H, Murayama H, et al. DPP-4 382–387 on dialysis. Nat Rev Nephrol 2013;9:147–153 inhibition with alogliptin on top of angiotensin II 111. Yamaguchi Y, Yoshikawa K. Cutaneous 128. Chan JC, Scott R, Arjona Ferreira JC, et al. type 1 receptor blockade ameliorates albumin- wound healing: an update. J Dermatol 2001; Safety and efficacy of sitagliptin in patients with uria via up-regulation of SDF-1a in type 2 dia- 28:521–534 type 2 diabetes and chronic renal insufficiency. betic patients with incipient nephropathy. 112. Schurmann¨ C, Linke A, Engelmann-Pilger K, Diabetes Obes Metab 2008;10:545–555 Endocr J 2014;61:159–166 et al. The dipeptidyl peptidase-4 inhibitor linagliptin 129. Kothny W, Shao Q, Groop PH, Lukashevich 98. Tani S, Nagao K, Hirayama A. Association be- attenuates inflammation and accelerates epitheli- V. One-year safety, tolerability and efficacy of tween urinary albumin excretion and low-density alization in wounds of diabetic ob/ob mice. J Phar- vildagliptin in patients with type 2 diabetes and lipoprotein heterogeneity following treatment macol Exp Ther 2012;342:71–80 moderate or severe renal impairment. Diabetes of type 2 diabetes patients with the dipeptidyl 113. Marfella R, Sasso FC, Rizzo MR, et al. Di- Obes Metab 2012;14:1032–1039 peptidase-4 inhibitor, vildagliptin: a pilot study. peptidyl peptidase 4 inhibition may facilitate 130. Lukashevich V, Schweizer A, Shao Q, Am J Cardiovasc Drugs 2013;13:443–450 healing of chronic foot ulcers in patients with Groop PH, Kothny W. Safety and efficacy of vil- 99. Groop PH, Cooper ME, Perkovic V, Emser A, type 2 diabetes. Exp Diabetes Res 2012;2012: dagliptin versus placebo in patients with type 2 Woerle HJ, von Eynatten M. Linagliptin lowers 892706 diabetes and moderate or severe renal impair- albuminuria on top of recommended standard 114. Albiero M, Menegazzo L, Boscaro E, ment: a prospective 24-week randomized treatment in patients with type 2 diabetes and Agostini C, Avogaro A, Fadini GP. Defective re- placebo-controlled trial. Diabetes Obes Metab renal dysfunction. Diabetes Care 2013;36:3460– cruitment, survival and proliferation of bone 2011;13:947–954 3468 marrow-derived progenitor cells at sites of de- 131. McGill JB, Sloan L, Newman J, et al. Long- 100. Haluz´ıkM,Frol´ık J, Rychl´ık I. Renal effects layed diabetic wound healing in mice. Diabeto- term efficacy and safety of linagliptin in patients of DPP-4 inhibitors: a focus on microalbumin- logia 2011;54:945–953 with type 2 diabetes and severe renal impair- uria. Int J Endocrinol 2013;2013:895102 115. Fadini GP. A reappraisal of the role of cir- ment: a 1-year, randomized, double-blind, 101. The Diabetes Control and Complications culating (progenitor) cells in the pathobiology of placebo-controlled study. Diabetes Care Trial/Epidemiology of Diabetes Interventions diabetic complications. Diabetologia 2013 2013;36:237–244 and Complications Research Group. Retinopa- 116. Fadini GP. Is bone marrow another target 132. Nowicki M, Rychlik I, Haller H, et al. Long- thy and nephropathy in patients with type 1 di- of diabetic complications? Eur J Clin Invest 2011; term treatment with the dipeptidyl peptidase-4 abetes four years after a trial of intensive 41:457–463 inhibitor saxagliptin in patients with type 2 therapy. N Engl J Med 2000;342:381–389 117. Albiero M, Poncina N, Tjwa M, et al. Diabetes diabetes mellitus and renal impairment: a ran- 102. Chew EY, Ambrosius WT, Davis MD, et al.; causes bone marrow autonomic neuropathy and domised controlled 52-week efficacy and safety ACCORD Study Group; ACCORD Eye Study impairs stem cell mobilization via dysregulated study. Int J Clin Pract 2011;65:1230–1239 Group. Effects of medical therapies on retinop- p66Shc and Sirt1. Diabetes 2013 133. Nowicki M, Rychlik I, Haller H, Warren ML, athy progression in type 2 diabetes. N Engl J 118. Busik JV, Tikhonenko M, Bhatwadekar A, Suchower L, Gause-Nilsson I; D1680C00007 In- Med 2010;363:233–244 et al. Diabetic retinopathy is associated with vestigators. Saxagliptin improves glycaemic 103. Gonçalves A, Leal E, Paiva A, et al. Protec- bone marrow neuropathy and a depressed periph- control and is well tolerated in patients with tive effects of the dipeptidyl peptidase IV inhib- eral clock. J Exp Med 2009;206:2897–2906 type 2 diabetes mellitus and renal impairment. itor sitagliptin in the blood-retinal barrier in a 119. Mangialardi G, Katare R, Oikawa A, et al. Diabetes Obes Metab 2011;13:523–532 type 2 diabetes animal model. Diabetes Obes Diabetes causes bone marrow endothelial bar- 134. Shah Z, Pineda C, Kampfrath T, et al. Acute Metab 2012;14:454–463 rier dysfunction by activation of the RhoA- DPP-4 inhibition modulates vascular tone 104. Maeda S, Yamagishi S, Matsui T, et al. Ben- Rho-associated kinase signaling pathway. through GLP-1 independent pathways. Vascul eficial effects of vildagliptin on retinal injury in Arterioscler Thromb Vasc Biol 2013;33:555–564 Pharmacol 2011;55:2–9 obesetype2diabeticrats.OphthalmicRes 120. DiPersio JF. Diabetic stem-cell “mobi- 135. Scheen AJ. Cardiovascular effects of gliptins. 2013;50:221–226 lopathy.” N Engl J Med 2011;365:2536–2538 Nat Rev Cardiol 2013;10:73–84