Keizo Kanasaki,1 Sen Shi,1 Megumi Kanasaki,1 Jianhua He,1 Takako Nagai,1 Yuka Nakamura,2

Yasuhito Ishigaki,2 Munehiro Kitada,1 Swayam Prakash Srivastava,1 and Daisuke Koya1

Linagliptin-Mediated DPP-4Inhibition Ameliorates KidneyFibrosis in Streptozotocin-Induced Diabetic Mice byInhibiting Endothelial-to-Mesenchymal Transition ina Therapeutic RegimenDiabetes 2014;63:2120–2131 | DOI: 10.2337/db13-1029

Kidney fibrosis is the final common pathway of allprogressive chronic kidney diseases, of which diabe-tic nephropathy is the leading cause. Endothelial-to-mesenchymal transition (EndMT) has emerged as one ofthe most important origins of matrix-producing fibro-blasts. Dipeptidyl peptidase-4 (DPP-4) inhibitors havebeen introduced into the market as antidiabetes drugs.Here, we found that the DPP-4 inhibitor linagliptin amelio-rated kidney fibrosis in diabetic mice without alteringthe blood glucose levels associated with the inhibi-tion of EndMT and the restoration of microRNA 29s.Streptozotocin-induced diabetic CD-1 mice exhibitedkidney fibrosis and strong immunoreactivity for DPP-4by 24 weeks after the onset of diabetes. At 20 weeksafter the onset of diabetes, mice were treated withlinagliptin for 4 weeks. Linagliptin-treated diabetic miceexhibited a suppression of DPP-4 activity/protein expres-sion and an amelioration of kidney fibrosis associatedwith the inhibition of EndMT. The therapeutic effects oflinagliptin on diabetic kidneys were associated with thesuppression of profibrotic programs, as assessed bymRNA microarray analysis. We found that the inductionof DPP-4 observed in diabetic kidneys may be associ-ated with suppressed levels of microRNA 29s in diabetic

mice; linagliptin restored microRNA 29s and suppressedDPP-4 protein levels. Using cultured endothelial cells, wefound that linagliptin inhibited TGF-b2–induced EndMT,and such anti-EndMT effects of linagliptin were mediat-ed through microRNA 29 induction. These results indi-cate the possible novel pleiotropic action of linagliptinto restore normal kidney function in diabetic patients withrenal impairment.

Diabetic nephropathy is a leading cause of kidney disease,which progresses into end-stage renal disease, requiringkidney replacement therapy (1,2). Glycemic control couldbe essential for therapies combating diabetic nephropa-thy, although normalizing the blood glucose levels in suchpatients with appropriate monitoring is challenging (2).Therefore, to prevent/retard diabetic nephropathy, in ad-dition to achieving proper glycemic control, strategiesthat are not directly related with blood glucose normali-zation are required.

Fibrosis in the kidney is the final common pathway ofprogressive kidney diseases and results in the destructionof the kidney structure and the deterioration of thekidney filtration function (3–8). Kidney fibrosis is caused

1Department of Diabetology and Endocrinology, Kanazawa Medical University,Uchinada, Ishikawa, Japan2Medical Research Institute, Kanazawa Medical University, Uchinada, Ishikawa,Japan

Corresponding author: Keizo Kanasaki, kkanasak@kanazawa-med.ac.jp, orDaisuke Koya, koya0516@kanazawa-med.ac.jp.

Received 30 June 2013 and accepted 3 February 2014.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db13-1029/-/DC1.

© 2014 by the American Diabetes Association. See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.

See accompanying article, p. 1829.

2120 Diabetes Volume 63, June 2014

COMPLIC

ATIO

NS

by prolonged injury associated with the dysregulation ofthe normal wound healing process and an excess accumu-lation of extracellular matrix. Kidney fibroblasts play animportant role in this fibrotic process, but the origin of thefibroblasts remains unclear and has become the focus ofintense debate (2,9). A significant heterogeneity of matrix-producing fibroblasts is thought to exist (2), and diverseorigins for fibroblasts have been described, such as res-idential fibroblasts or pericytes, epithelial-to-mesenchymaltransition, and endothelial-to-mesenchymal transition(EndMT) (2). Among these diverse origins of matrix-producing fibroblasts, EndMT seems to be an importantorigin of myofibroblasts or activated fibroblasts (9).

Dipeptidyl peptidase-4 (DPP-4) inhibitors enhancethe activity of endogenous glucagon-like peptide-1 andglucose-dependent insulinotropic polypeptide (10), whichhave emerged as important prandial stimulators of insulinsecretion and have many physiological actions (10,11).Additionally, DPP-4 is distributed throughout the bodyand cleaves numerous substrates other than incretinhormones (10). Aside from their glucose-lowering action,DPP-4 inhibitors are also associated with potentially organprotective effects due to their diverse, widely distributed,and pleiotropic action (12). Indeed, the kidney is whereDPP-4 is expressed at the highest level per organ weight(13). Interestingly, DPP-4 has been associated with cell sur-vival signaling and extracellular matrix remodelings (14–16).

Linagliptin, a new DPP-4 inhibitor, is mainly excretedvia the bile; therefore, this drug can theoretically beprescribed for patients with renal dysfunction withoutadjusting the dosage (17,18). In this study, we testedwhether the DPP-4 inhibitor linagliptin could exert itstherapeutic benefits in mice with kidney fibrosis associ-ated with type 1 diabetes.

RESEARCH DESIGN AND METHODS

ReagentsA rat polyclonal anti-mouse CD31 antibody was pur-chased from emfret ANALYTICS (Eibelstadt, Germany).Mouse monoclonal anti-human CD31 and goat polyclonalanti-mouse DPP-4 antibodies (for tissue labeling) werepurchased from R&D Systems (Minneapolis, MN). A rab-bit polyclonal anti-aSMA antibody was purchased fromGeneTex (Irvine, CA). A rabbit polyclonal anti-SM22a an-tibody and a monoclonal antibody for VE-cadherin wereobtained from Novus Biologicals (Littleton, CO).A polyclonal antibody for FSP1 (S100A4) was purchasedfrom Covance (Princeton, NJ). A goat polyclonal anti–DPP-4 antibody (for Western blotting in human cells)and a rabbit polyclonal anti-GAPDH antibody were ob-tained from Sigma-Aldrich (St. Louis, MO). Fluorescence-,Alexa Fluor 647–, and Rhodamine-conjugated secondaryantibodies were obtained from Jackson ImmunoResearch(West Grove, PA). A horseradish peroxidase–conjugatedsecondary antibody was purchased from Cell SignalingTechnology (Danvers, MA). TGF-b2 was purchased fromPeproTech (Rocky Hill, NJ). Apoptosis was detected with

an Annexin V assay kit, which was obtained from ClontechLaboratories (Mountain View, CA). DPP-4 activity wasmonitored by DPP-4 assay kit (BioVision, Milpitas, CA).

Animal ExperimentsEight-week-old male CD-1 mice (Sankyo Laboratory Service,Tokyo, Japan) were used in all of the diabetic experi-ments. The mice were injected intraperitoneally withstreptozotocin (STZ; 200 mg/kg). The induction of di-abetes was confirmed as a blood glucose level .16 mmol/L2 weeks after STZ injection. By 20 weeks after the inductionof diabetes, the diabetic mice were divided into two groups(linagliptin [5 mg/kg body weight (BW)/day in drinkingwater] and untreated). Dose of linagliptin was decidedbased on the dose-dependent effects of this drug between3 and 30 mg/kg BW/day (19). Linagliptin was diluted di-rectly in drinking water. All mice were killed 24 weeks afterthe induction of diabetes. Linagliptin was provided byBoehringer Ingelheim (Ingelheim, Germany) with a materialtransfer agreement. Blood pressure was monitored by thetail cuff method with BP-98A (Softron Co., Beijing, China).

EndMT Detection in VivoFrozen sections (5 mm) were used for the detection of invivo EndMT. Cells undergoing EndMT were identified bydouble-positive labeling with CD31-aSMA or CD31-FSP1.The immunolabeled sections were analyzed by fluorescencemicroscopy (Biozero; Keyence, Osaka, Japan). For eachmouse, 3300 magnification pictures were obtained fromsix different areas, and quantification was performed.

Morphological EvaluationThe glomerular surface area was calculated in 10 glomeruliper mouse using ImageJ software. To evaluate the mesangialmatrix area (%), we used a point counting method. Weanalyzed 10 Periodic acid-Schiff (PAS)-stained glomerulifrom each mouse on a digital microscope screen gridcontaining 540 (27 3 20) points in Adobe PhotoshopElement 6.0. The number of grid points in the mesangialarea (both matrix and cells) was divided by the total num-ber of points in the glomerulus to obtain the percentage ofrelative mesangial matrix area in a given glomerulus. Massontrichrome stain (MTS) were imaged and analyzed by ImageJsoftware, and fibrotic areas were quantified. In each mouse,six pictures (original magnification 3100) were evaluated.

ImmunohistochemistryDeparaffinized (2 min in xylene, four times; 1 min in 100%ethanol, twice; 30 s in 95% ethanol; 45 s in 70% ethanol;and 1 min in distilled water) mouse kidney sections wereused for DPP-4 labeling. Immunohistochemistry was per-formed with a Vectastain ABC Kit (Vector Laboratories,Burlingame, CA). The DPP-4 primary antibody was diluted1:100. In negative controls, the primary antibody wasomitted and replaced with blocking solution.

In Vitro EndMTHuman dermal microvascular endothelial cells (HMVECs;Lonza, Basel, Switzerland) cultured in endothelial basalmedium (EBM) supplemented with endothelial growth

diabetes.diabetesjournals.org Kanasaki and Associates 2121

medium (EGM)-2 kit medium were used in this experi-ment. When the HMVECs on the adhesion reagent(Kurabo Bio-Medical, Osaka, Japan) reached 70% conflu-ence, 5 ng/mL recombinant human TGF-b2 for 48 h wasplaced in the experimental medium (HuMedia-MVG inserum-free RPMI at a 1:3 ratio) with or without linagliptin(100 nmol/L) preincubation for 2 h. In the control well,vehicle (DMSO) was added (3 3 1025 dilution of DMSOin final concentration). The protein lysate was harvestedfor Western blot analysis.

Wound Healing AssayThe passaged HMVECs were placed in six-well plates andcultured with EBM supplemented with EGM-2 until reach-ing 70–80% confluence, and then the cells treated withTGF-b2 (5 ng/mL) in the presence or absence of linagliptin(100 nmol/L) were incubated with a medium containingHuMedia MVG and RPMI-1640 (1:3). At the same time,the control group was incubated in the same medium with-out TGF-b2 or linagliptin. In the control well, vehicle(DMSO) was added (3 3 1025 dilution of DMSO in finalconcentration). Using a pipette tip at an angle of;30°, eachwell received a straight scratch, simulating a wound. After24 and 48 h had passed, the number of cells that migratedinto the wounded area was counted under a light micro-scope. Six different areas were evaluated in each group, andthe experiment was repeated twice with similar results.

Cell Migration Boyden Chamber AssayThe bottom side of the migration chamber (Cell CultureInsert; BD Falcon, San Jose, CA) was coated with Matrigel(BD), and 1,000 HMVECs were passaged in the uppermigration chamber. Twenty-four hours after passage, themedium was changed to the experimental medium (1:3HuMedia-MVG:RPMI1640) in both the upper and thebottom wells. Subsequently, cells were exposed to TGF-b2in the presence or absence of linagliptin (100 nmol/L),while the control group was incubated with the samemedium, lacking TGF-b2 and linagliptin. In the controlwell, vehicle (DMSO) was added (3 3 1025 dilution ofDMSO in final concentration). After 48 h, the cells werewashed with PBS, followed by fixation with formaldehyde(3.7% in PBS) at room temperature for 2 min. After wash-ing twice with PBS, the cells were permeabilized with100% methanol for 20 min at room temperature. Then,the cells were washed twice with PBS and stained withhematoxylin-eosin. After scraping off the nonmigratorycells (upper well) with a cotton swab, the number of mi-grated cells was counted under a light microscope. Sixdifferent areas were evaluated in each group, and theexperiment was repeated twice with similar results.

Western BlottingThe protein lysates were denatured by boiling in SDSsample buffer at 100°C for 5 min and then centrifuged(17,000g for 10 min at 4°C); subsequently, the supernatantwas separated on SDS-polyacrylamide gels. Separated pro-tein lysates were blotted onto polyvinylidene fluoride(PVDF) membranes (Pall Corporation, Pensacola, FL)

by the semidry method. After blocking with TBS-T (Tris-buffered saline containing 0.1% Tween 20) containing 5%nonfat dry milk or 5% BSA, the membranes were incubatedwith the primary antibodies of the target molecules(1:1,000 for all primary antibodies) in TBS-T containing5% BSA at 4°C overnight. The membranes were washedthree times and incubated with 1:2,000 diluted horseradishperoxidase–conjugated secondary antibodies (Cell SignalingTechnology) for 1 h at room temperature. The immunore-active bands were visualized with an enhanced chemilumi-nescence (ECL) detection system (Pierce Biotechnology,Rockford, IL) by ImageQuant LAS 4000 (GE HealthcareLife Sciences, Uppsala, Sweden).

mRNA Array AnalysisThe total RNA was isolated using a commercially availablekit (RNeasy Mini Kit; QIAGEN, Hilden, Germany). Theconcentration of RNA was quantified by photometry at260/280 nm, and the quality of the RNA was determinedby the ratio of the 18S/28S ribosomal band intensities inan ethidium bromide–containing 1% agarose gel afterelectrophoresis. The sense cDNA was prepared using anAmbion WT Expression Kit (Ambion, Austin, TX), andtarget hybridizations were performed using a MouseGene 1.0 ST Array (Affymetrix, Santa Clara, CA), accord-ing to the manufacturer’s instructions. Hybridization wasperformed for 17 h at 45°C in a GeneChip HybridizationOven 640 (Affymetrix). After washing and staining in aGeneChip Fluidics Station 450, hybridized cDNAs weredetected using the GeneChip Scanner 3000. The digitalizedimage data were processed using the GeneChip OperatingSoftware version 1.4. Because replicate assays were notperformed, the signal intensities of the selected genesthat were upregulated or downregulated by at least twofoldcompared with a control group were extracted using Gene-Spring GX software package version 12.5 (Agilent Technol-ogies, Santa Clara, CA). After hierarchical clustering, theresults were illustrated as a heat map. Ingenuity PathwayAnalysis (Ingenuity Systems, Inc., Redwood City, CA) wasused to select the specific function-related genes.

MicroRNA Array AnalysisTotal RNA was isolated using the miRNeasy Kit (QIAGEN)according to the manufacturer’s instructions. Quality-confirmed total RNA samples were assayed and qualified induplicate using the microRNA microarray. The input for theAgilent microRNA labeling system was 100 ng total RNA.Dephosphorylated and denatured total RNA was labeledwith cyanine 3-pCp and subsequently hybridized to the Agi-lent mouse microRNA microarray release version 15 usingthe microRNA Complete Labeling and Hyb Kit (Agilent).After hybridization for 20 h, the slides were washed withGene Expression Wash Buffer Kit (Agilent) and measuredusing an Agilent Scanner G2565BA. Agilent Feature Extrac-tion Software version 9.5.1 and GeneSpring GX softwareversion 12.5 (Agilent) were used for data processing, anal-ysis, and monitoring.

2122 Linagliptin Inhibits Diabetic Kidney Fibrosis Diabetes Volume 63, June 2014

MicroRNA Isolation and Quantitative PCRFrozen kidney tissues (one each from a control, diabetic,and linagliptin-treated diabetic mouse; samples were keptat 270°C) were first placed on the RNA later-ICE (LifeTechnologies) for 16 h at 220°C before the subsequenthomogenization process to avoid RNA degradation whileextracting high-quality microRNA. MicroRNA was extractedusing miRNeasy Mini Kit (QIAGEN) according to the man-ufacturer’s instructions for homogenized samples. The com-plementary DNA was generated by a miScript II RT Kit(QIAGEN) using the hiSpec buffer method. MicroRNAexpression was quantified using miScript SYBR GreenPCR Kit (QIAGEN) using 3 ng of complementary DNA.The primers to quantify Mm_miR-29a, Mm_miR-29b,and Mm_miR-29c were the miScript primer assays prede-signed by QIAGEN. The mature microRNA sequences were59-UAGCACCAUCUGAAAUCGGUUA for Mm_miR-29a,59-UAGCACCAUUUGAAAUCAGUGUU for Mm_miR-29b,and 59-UAGCACCAUUUGAAAUCGGUUA for Mm_miR-29c.All experiments were performed in triplicate, and Hs_RNU6-2_1 (QIAGEN) was used as an internal control.

TransfectionFor the transfection studies, HMVECs, which were main-tained in EBM supplemented with EGM-2, were passaged insix-well plates with nonproliferative medium (HuMedia-MVGand RPMI at a ratio of 1:3). The HMVECs were transfectedwith 100 nmol/L of antagomiR for microRNA 29a, microRNA29c, microRNA 29a+c (FASMAC, Kanagawa, Japan), microRNA29b inhibitor (QIAGEN), or mimetics for microRNA 29s(29a-3p, UAGCACCAUCUGAAAUCGGUUA; 29b-3p, UAGCACCAUUUGAAAUCAGUGUU; 29c-3p, UAGCACCAUUUGAAAUCGGUUA) using Lipofectamine 2000 transfectionreagent (Invitrogen, Carlsbad, CA) according to the man-ufacturer’s instructions.

The cells were incubated for 6 h with Lipofectamineand the anti-microRNA complex in antibiotic-free me-dium, after which the medium was replaced with freshmedium before the cells were incubated for another 48 h.Upon the termination of the incubation, the cells werescraped using RIPA buffer (with the addition of phenyl-methylsulfonyl fluoride, sodium vanadate, and proteaseinhibitor) to assess DPP-4 protein expression (Sigma-Aldrich) using the Western blot technique.

Luciferase AssayFor the luciferase assay to analyze the activity of 39 un-translated region (UTR) in human DPP-4, we cloned thefragment of human DPP-4 39 UTR sequence by PCR usingthe primer set (forward, ATAGAGCTCAATAGCTAGCAGCACAGCACACCAAC; reverse, ATATCTAGA GTGTCCATATGCCAGTGCGGTTTAGG) and BAC clone human RP11178A14 as template. Both purified PCR fragment andpmirGLO Dual-Luciferase microRNA Target ExpressionVector (Promega) were enzyme digested (SacI and XbaI),purified, and ligated (Thermo Fisher Scientific). The se-quence of DPP-4 39 UTR was confirmed, and amplifiedvector DNA (300 ng/well in 12-well plate) was transfected

into HMVECs. In the presence of control microRNA, mi-metic, antagomiR, or inhibitor of 29s (75 nmol/L in finalconcentration), transcriptional activity was evaluated byDual-Luciferase Reporter Assay System (Promega) intriplicate samples.

Statistical AnalysisThe data are expressed as the mean 6 SEM. The non-parametric Mann-Whitney U test was used to determinesignificance, which was defined as P , 0.05, if not specif-ically mentioned. GraphPad Prism software (version 5.0f)was used for the statistical analysis.

RESULTS

Linagliptin Restored the Normal Kidney Structure ofSTZ-Induced Diabetic Kidney Fibrosis in CD-1 MiceWe used a fibrotic diabetic kidney disease model, STZ-induced diabetes in male CD-1 mice. The immunohisto-chemistry analysis revealed robust DPP-4 immunolabelingin the glomerular basement membrane, tubules, andperitubular vascular cells in the kidney of STZ-induceddiabetic CD-1 mice when compared with control CD-1mice (Fig. 1A and B). Such peritubular vascular cells werelikely endothelial cells upon evaluation by immunofluores-cence microscopy (Fig. 1C–E). Western blot analysis usingwhole kidney revealed that in the kidneys of diabeticmice, the DPP-4 protein levels were upregulated com-pared with control kidneys (Fig. 1F and G). As Sugimotoet al. (20) elegantly showed, STZ-induced diabetic CD-1mice exhibit kidney fibrosis, with both glomerulosclerosisand tubulointerstitial fibrosis occurring ;16 weeks afterthe onset of diabetes (Supplementary Fig. 1), and at 24weeks after the initiation of diabetes, diabetic miceexhibited severe fibrosis when compared with controlmice (Fig. 2A, B, D, and E). Linagliptin-treated diabeticmice exhibited restored normal kidney structures (Fig.2C and F). Morphometric analysis of the kidneys revealedthat diabetic mice displayed significantly enlarged glomer-uli, mesangial expansion, and relative areas of Massontrichrome–positive interstitial fibrosis (Fig. 2G–I), whereaslinagliptin restored normal kidney histology and normalarchitecture (Fig. 2G–I). Urine albumin levels were elevatedin STZ mice, and linagliptin treatment suppressed the trendof the urine albumin levels (Fig. 2J) as observed in humanclinical trial (21). These antifibrotic effects of linagliptinin diabetic mice are associated with a suppressed trend ofTGF-b1 and DPP-4, and significant suppression of TGF-b2protein levels in the kidney (Fig. 2K–N). Linagliptininhibited DPP-4 mRNA (Fig. 2O) and diabetes-enhancedDPP-4 activity in kidney and plasma of diabetic mice (Fig.2P and Q).

When compared with control mice, diabetic CD-1 miceexhibited lower blood pressure, lighter BW, higher bloodglucose, and increased kidney and liver weight (Fig. 3A–E).Linagliptin treatment in STZ-induced diabetic CD-1 micecaused no alteration in blood pressure, body weight, bloodglucose level, or organ weight (for kidney, liver, and heart)when compared with untreated diabetic mice (Fig. 3A–E).

diabetes.diabetesjournals.org Kanasaki and Associates 2123

The heart weight was lighter in all diabetic mice but showedinsignificant changes in all groups analyzed (Fig. 3F).

Antifibrotic Effects of Linagliptin Were Associated Withthe Inhibition of EndMT in Diabetic KidneysKidney fibroblasts play essential roles in kidney fibrosisand originate from diverse sources (2,9,22). In our anal-ysis, we analyzed EndMT, a recently described importantsource of kidney fibroblasts (2,9,23). When EndMT wasanalyzed by quantifying cells that coexpressed endothe-lial marker CD31 and a mesenchymal marker, eithera-smooth muscle actin (aSMA) or FSP1, diabetic CD-1mice exhibited a significantly increased number of cells inthe process of the EndMT when compared with controlkidneys (Fig. 4A, B, D–F, and H). Linagliptin-treated miceexhibited significantly fewer cells undergoing EndMT inthe kidney when compared with untreated diabetic mice(Fig. 4C, D, G, and H).

Linagliptin Inhibited EndMT and Apoptosis in CulturedEndothelial CellsIn cultured endothelial cells, TGF-b2 induced the suppres-sion of CD31 with concomitant upregulation of SM22ain HMVECs, suggesting that TGF-b2 induces EndMT

(Fig. 5A–C); TGF-b2–induced EndMT was inhibited bylinagliptin pretreatment (Fig. 5A–C). Wound healing cellinvasion assays revealed that TGF-b2 also induced themigration of fibroblast-like EndMT cells (Fig. 5D, E, J,G, and H), while linagliptin inhibited the invasion of thosecells (Fig. 5F, I, and J). The Boyden chamber cell migrationassay also revealed that linagliptin inhibited the endothe-lial cells’ transmigration through Matrigel (Fig. 5K–N).When analyzing the molecular mechanisms of TGF-b2–induced EndMT effects, we found that linagliptin inhibitedTGF-b2–induced Smad3 phosphorylation in endothelialcells (Fig. 5O). Linagliptin inhibited TGF-b2–induced pro-tein expression, mRNA expression, and activity of DPP-4 inendothelial cells (Fig. 5P–R). Linagliptin and generic DPP-4inhibitor KR62436 inhibited TGF-b2–induced endothelialcell apoptosis (Supplementary Fig. 2).

Linagliptin Inhibited Profibrotic Programming in theDiabetic KidneyA heat map indicated the differences among controls,diabetic mice, and diabetic mice treated with linagliptin(Supplementary Fig. 3A). The diabetes-associated changes inthe expression of some genes were restored by linagliptin

Figure 1—DPP-4 induction in diabetic kidneys. Immunohistochemical analysis of DPP-4 in control (A) or STZ-induced diabetic (B) CD-1mice. Eight-week-old CD-1 mice received a single injection of STZ. The mice were killed 24 weeks after the induction of diabetes. n = 3 ineach group, and representative pictures are shown. C–E: Immunofluorescence microscopy analysis for DPP-4 in diabetic mice. Frozenkidney samples were labeled with DPP-4 and endothelial marker CD31. Immunofluorescence analysis was performed by fluorescencemicroscopy. C: FITC-labeled DPP-4. D: Rhodamine-labeled CD31. E: Merged images are shown. Original magnification 3300. Scale bar,50 mm in each panel. Representative pictures from three mice are shown. F: Western blot analysis for DPP-4 in the kidney. The proteinlysate (15 mg) was separated in polyacrylamide gels and transferred onto a PVDF membrane. Immunoreactive bands were analyzed by theECL method. Representative data from six mice in each group are shown. Arrow indicates immunoreactive bands of DPP-4. G: Densitometricanalysis of the DPP-4 Western blot analysis. n = 6 in each group were analyzed. Data are expressed as the mean6 SEM and are shown in thegraph. DM or D, diabetic mice.

2124 Linagliptin Inhibits Diabetic Kidney Fibrosis Diabetes Volume 63, June 2014

Figure 2—Inhibition of DPP-4 by linagliptin in diabetic kidneys is associated with the amelioration of kidney fibrosis. A–C: PAS staining forglomeruli. D–F: MTS staining in the indicated group of mice. Scale bar, 50 mm. G–I: A morphometric analysis of the kidney histology wasperformed as described in the RESEARCH DESIGN ANDMETHODS. Control, n = 5; DM, n = 7; DM+linagliptin, n = 5 were analyzed. J: Urine albuminexcretion was analyzed by albumin-creatinine ratios. Control, n = 5; DM, n = 8; DM+linagliptin, n = 6 were analyzed. K: Western blot analysisfor TGF-b1, TGF-b2, and DPP-4. A representative blot from four independent experiments is shown. L–N: Densitometric analysis ofindicated protein expression relative to actin levels is shown. n = 4 in each group were analyzed. O: qPCR analysis for DPP-4 in thekidney from six mice in each group. DPP-4 activity measurements in kidney (control, n = 3; DM, n = 6; DM+linagliptin, n = 6) (P) and plasma(all n = 6) (Q) were analyzed. The graphs in the figure are expressed as mean 6 SEM. DM, diabetic mice.

diabetes.diabetesjournals.org Kanasaki and Associates 2125

treatment (Supplementary Fig. 3A). To focus on fibrosis,we selected fibrosis-associated genes based in the Ingenu-ity Pathways Analysis database and plotted them ina scattered format (Supplementary Fig. 3B). Comparedwith controls, 13 genes were upregulated and 2 geneswere downregulated in the diabetic mouse (Supplemen-tary Fig. 3C). However, the expression of these genestended to be reversed into normal levels after linagliptintreatment.

Role of MicroRNA 29 Family Suppression in the DPP-4Induction of Diabetic Kidney

Finally, to identify the underlying mechanisms of howDPP-4 is increased in the diabetic kidney, we analyzed themicroRNA profiles of the animals and found that themicroRNA 29 family tended to be suppressed in diabeticmouse kidneys when compared with control kidneys(although no significant difference by Welch test yet)(Fig. 6A). According to the prediction of microRNA targets

Figure 3—Characteristics of animals. Systolic and diastolic blood pressure (A), BW (B), blood glucose level (C), kidney weight (both rightand left) (D), liver weight (E ), and heart weight (F ) are shown. Organ weights are shown as the weight per gram BW of each mouse. Control,n = 5; DM, n = 5; DM+linagliptin, n = 3 were analyzed. Graphs are expressed as the mean 6 SEM. DM, diabetic mice.

Figure 4—Linagliptin suppressed the EndMT in diabetic kidney. FSP1-CD31 (A–C) and aSMA-CD31 (E–G), immunolabeling visualized byfluorescence microscopy. Arrows indicate cells undergoing EndMT. D and H: Quantification of FSP1-CD31 or aSMA-CD31 double-positivecells. The sections are labeled as described in the RESEARCH DESIGN AND METHODS, and the percentages of cells undergoing the EndMT werecalculated among all DAPI-positive cells. Control, n = 5; DM, n = 5; DM+linagliptin, n = 3 were analyzed. Original magnification3300. Scalebar, 50 mm. Data are expressed as the mean 6 SEM in the graph. DM, diabetic mice.

2126 Linagliptin Inhibits Diabetic Kidney Fibrosis Diabetes Volume 63, June 2014

Figure 5—In vitro EndMT was inhibited by linagliptin. A: Western blot analysis. HMVECs were incubated with 5 ng/mL TGF-b2 in thepresence or absence of linagliptin (100 nmol/L) for 48 h. Cell lysates were harvested with RIPA buffer and denatured; 15 mg of cell lysatewas separated on the gel and transferred onto the PVDF membrane and underwent chemiluminescence for the indicated antibody to detectimmunoreactive bands. Representative results from three independent analyses are shown. B and C: Densitometric analysis of eachprotein’s expression, normalized as indicated in the figure. Data are shown as the mean 6 SEM in the graph (n = 3). D–I: Wound healingcell migration assay. EndMT cells at 80% confluence were incubated, scratched linearly, and subsequently stimulated with TGF-b2 in thepresence or absence of linagliptin. Cells that migrated into the scratched wound were measured, and five different areas were evaluatedin 3300 (original magnification) images. J: Quantification of the wound healing assay. K–M: Boyden chamber cell migration assays wereperformed as described in the RESEARCH DESIGN AND METHODS. Migrated cells in the bottom layer of the Boyden chamber were counted in fivedifferent areas, and quantification was performed (N). O: Smad3 phosphorylation was analyzed by Western blot analysis. Densitometricanalysis was also performed (n = 3). The protein (P) and mRNA (Q) expression in HMVECs analyzed by Western blot analysis (n = 4) or qPCR(n = 5). R: DPP-4 activity in HMVECs (n = 3). The graph in the figure is expressed as mean 6 SEM. DM, diabetic mice.

diabetes.diabetesjournals.org Kanasaki and Associates 2127

by TargetScan (http://www.targetscan.org/vert_60/), DPP-4may be regulated by the microRNA 29 family. Quanti-tative analysis revealed that microRNAs 29a, -b, and -cwere suppressed in the diabetic kidney when comparedwith control kidneys, and linagliptin restored such diabetes-suppressed microRNA 29s (Fig. 6B–D). Similarly TGF-b2–suppressed microRNA 29s were restored by linagliptin in invitro analysis (Supplementary Fig. 4). When an antagomiRfor microRNA 29a was transfected into endothelial cells,DPP-4 protein levels were indeed increased (Fig. 6E andF), whereas microRNA 29c antagomiR transfection didnot change the DPP-4 level (Fig. 6E and F). The cotransfec-tion of antagomiRs for microRNAs 29a and -c resulted inpersistently higher DPP-4 protein levels in the endothelialcells compared with controls (Fig. 6E and F). In addition,the inhibitor for microRNA 29b induced DPP-4 proteinexpression in HMVECs (Fig. 6G and H), suggesting thatmicroRNAs, specifically the microRNA 29 family, regulateDPP-4 in the diabetic kidney. TGF-b2–enhanced DPP-4 pro-tein expression was significantly suppressed by microRNA29b mimetics, and some trend of suppression was foundin microRNA 29a and -c mimetic–transfected cells (Fig. 6Iand J). TGF-b2–induced EndMT and endothelial cell mi-gration was largely inhibited by microRNA 29 mimetictransfection (Fig. 6I and K–T). In contrast, antagomiRsand inhibitor for microRNA 29s induced EndMT phenotypeand migration (Supplementary Fig. 5). TGF-b2–stimulatedluciferase activity of pmirGLO Dual-Luciferase microRNATarget Expression Vector containing 39 UTR fragment ofDPP-4, where a microRNA 29 binding site was involved,was significantly suppressed by microRNA 29 mimetics(Fig. 6U). In contrast, antagomiR or inhibitor of microRNA29s significantly induced DPP-4 39 UTR luciferase activity(Fig. 6V).

DISCUSSION

Both the inhibition of kidney fibrosis and the restorationof normal kidney structure are fundamental processes toresearch for developing therapies to combat progressivechronic kidney disease, including diabetic nephropathy.Although kidney fibroblasts have been implicated in thepathogenesis of kidney fibrosis, it would be challenging toinoculate only kidney fibroblasts as therapeutic targets. Inour analysis, DPP-4 inhibition by linagliptin in the diabetickidney seems to be a powerful therapy that inhibits thefibroblast-activating process.

We found that DPP-4 protein expression was increasedin the whole kidney lysate, endothelial cells, tubules, andglomeruli of STZ-induced diabetic kidneys from CD-1mice. Both the protein expression and activity levels ofDPP-4 are higher in kidneys of high fat–fed, STZ-injectedmice (24). Liu et al. (25) reported that the administrationof the DPP-4 inhibitor vildagliptin for 24 weeks preventedkidney damage in STZ-induced diabetic male Sprague-Dawley rats. In our analysis, we used fibrotic STZ-induceddiabetic CD-1 mice and tested whether intervention withlinagliptin starting 20 weeks after the induction of diabetes

could rescue fibrotic kidneys. We found that introducingthis intervention to such fibrotic kidneys significantlyameliorated kidney fibrosis without altering physiologicalparameters, suggesting that increased DPP-4 in fibroticdiabetic kidneys is a therapeutically valuable target. Ourmicroarray analysis clearly demonstrated that diabetickidneys exhibited profibrotic signaling and that linagliptinrestored a normal profile in the diabetic kidney.

We have not yet realized an entire series of DPP-4–mediated, kidney-damaging signals in our models, but webelieve that signaling associated with endothelial cell sur-vival and EndMT is involved. Linagliptin inhibited boththe matrix- and fibroblast-generating pathways associatedwith EndMT and fibroblast-like EndMT cell invasion. Suchenhanced DPP-4 activity is associated with matrix metal-loproteinases, which are responsible for tissue remodeling(26). Furthermore, Takahashi et al. (27) recently showedthat DPP-4 inhibition by vildagliptin ameliorates the heartfailure and collagen deposition associated with inhibitingthe TGF-b–Smad signaling pathway. We found that lina-gliptin inhibited TGF-b2–induced Smad3 phosphorylation.

Our analysis reveals that the microRNA 29 family isone of the potential molecular regulators of DPP-4 in thekidney and endothelial cells as well. The microRNA 29family protects organs from fibrotic damage (28–31);therefore, DPP-4 inhibition and the subsequent inhibitionof the fibroblast activation pathway may be involved inthe antifibrotic effects of the microRNA 29 family. DPP-4has been implicated in fibrogenic pathologies (32–36).Regarding this, DPP-4 inhibition by linagliptin increasedmicroRNA 29s both in vivo and in vitro. MicroRNA 29 sup-pressions in either diabetic kidneys or TGF-b2–stimulatedendothelial cells are associated with EndMT and inductionof DPP-4. The inhibition of each microRNA 29 can inducethe EndMT feature in HMVECs and migration; the role ofeach microRNA 29 in DPP-4 level regulation is somewhatcomplicated. Since microRNA 29a inhibition results in thestrong induction of DPP-4 39 UTR–luciferase activity, in basalconditions microRNA 29a likely emerges as the main regula-tor of DPP-4 39 UTR; TGF-b2–induced DPP-4 protein leveland cell migration were most efficiently inhibited bymicroRNA 29b, suggesting that each microRNA 29 couldplay distinct roles and cooperate in the homeostasis ofcells in a context-dependent manner. Also, our data in-dicate that TGF-b2/DPP-4/miR29s display cross-talkmechanisms in the onset of kidney fibrosis, and linaglip-tin-mediated inhibition of DPP-4 would diminish profi-brotic signaling cross-talk in the kidney. Among humanfibrotic diseases, patients with hepatitis C viral infections,the condition associated with microRNA 29 suppression,also exhibit high DPP-4 activity; such DPP-4 activity couldbe a potential target for combating such liver diseases (37).Further study is required to determine whether DPP-4activity is greater in type 1 or type 2 diabetic kidney dis-eases, or in other fibrotic chronic and acute diseases, includ-ing kidney disease, and their association with microRNA29s levels in humans.

2128 Linagliptin Inhibits Diabetic Kidney Fibrosis Diabetes Volume 63, June 2014

Figure 6—MicroRNA (miR) 29 alterations may be involved in the induction of DPP-4 in kidneys of diabetic mice. A: MicroRNA array analysisof the microRNA 29 cluster. n = 2 were used. B–D: qPCR analysis for microRNA 29 a–c. For microRNA 29a, control n = 6, diabetes n = 7,others n = 4 or 5 in each group. DM, diabetic mice. E–H: MicroRNA inhibition experiments. AntagomiR for 29a or 29c or an inhibitor formicroRNA 29b was transfected into HMVECs, and their DPP-4 protein levels were analyzed by Western blots. Densitometric analyses(F and H) for DPP-4 protein levels were performed. In each experiment, relative values to the control cells were analyzed. n = 3 in 29a and29c experiment; n = 4 in 29b experiment. I: HMVECs were stimulated with TGF-b2 in the presence of indicated microRNA mimetictransfection for 48 h. Cells were harvested and analyzed with Western blot analysis. J–N: Densitometric analysis was also performedand relative value against actin levels (n = 3). O–S: Migration assay using Boyden chamber as in Fig. 5. T: Migrated cells in the bottom layerof the Boyden chamber were counted in six different areas. DPP-4 39 UTR transcriptional activity measurement with pmirGLO Dual-Luciferase microRNA Target Expression Vector, cotransfected with microRNA 29 mimetics in the presence of TGF-b2 (U) or microRNA29 antagomiR/inhibitor (V). Experiments were performed twice, showing similar results. One-way ANOVA followed by Tukey test was usedfor statistics analysis. The graph in the figure is expressed as mean 6 SEM. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; sc,scramble control microRNA.

diabetes.diabetesjournals.org Kanasaki and Associates 2129

In our analysis, linagliptin inhibited kidney fibrosis andrestored normal kidney structure. Although convincing,there are several limitations. First, the concentration weused was very high (5 mg/kg BW) when compared withthat used in the clinic (;100 mg/kg BW). Second, accord-ing to our data, we believe that DPP-4 inhibition by anyDPP-4 inhibitor can ameliorate kidney fibrosis. However,it is not yet clear whether other DPP-4 inhibitors canefficiently suppress the profibrotic program in the kidneydue to differences in the mechanisms/metabolism of eachdrug. Third, we focused on EndMT as the origin of acti-vated fibroblasts. However, such origins of fibroblasts inkidney fibrosis are still controversial and the focus of in-tense debate (2,9), and it could be possible that diverseorigins of the kidney fibroblast–activating pathway (38)were affected by linagliptin. Finally, the analysis for dis-tinct roles of each microRNA 29 in the regulation of DPP-4 and kidney fibrosis would require further investigation.

Diabetic kidney disease represents a serious healthproblem worldwide and can develop into end-stage renaldisease, which requires kidney replacement therapy. Pro-gressive kidney fibrosis determines residual kidney func-tion, and the restoration of the normal architecture to thefibrotic kidney would constitute a fundamental therapy.We reported here that the antifibrotic effects of linagliptinare beneficial for diabetic kidney disease, regardless of theblood glucose levels, via the suppression of the activatedfibroblast-generating EndMT pathway at least in part.Linagliptin may be safe for use among kidney diseasepatients, given its ability to be exclusively eliminated viathe bile; this specificity suggests that this DPP-4 inhibitorhas potential utility for therapeutic use in combatingkidney fibrosis in diabetes.

Funding and Duality of Interest. This work was partially supportedby grants from the Japan Society for the Promotion of Science to K.K.(23790381), M.Ka. (24790329), T.N. (24659264), M.Ki. (24591218), and D.K.(25282028 and 25670414), research grants from the Japan Research Founda-tion for Clinical Pharmacology to K.K. (2011), and the Takeda visionally researchgrant to K.K. (2013). This work was partially supported by a Grant for PromotedResearch awarded to K.K. (S2011-1 and S2012-5) and a Grant for CollaborativeResearch awarded to D.K. (C2011-4 and C2012-1) from Kanazawa MedicalUniversity. K.K. was also supported by several foundational grants, includinggrants from the Daiichi Sankyo Foundation of Life Science, the Ono MedicalResearch Foundation, the Novartis Foundation (Japan) for the Promotion of Sci-ence, the Takeda Science Foundation, and the Banyu Foundation. S.P.S. issupported by the Japanese Government Ministry of Education, Culture, Sports,Science and Technology Fellowship Program. S.S. and J.H. are supported byforeign scholar grants from Kanazawa Medical University.The authors thank Boehringer Ingelheim for providing the linagliptin with a ma-terial transfer agreement. K.K. and D.K. received lecture fees from BoehringerIngelheim and Eli Lilly and Company. Both Boehringer Ingelheim and Eli Lillyand Company donated to Kanazawa Medical University and were not directlyassociated with this project. No other potential conflicts of interest relevant tothis article were reported.Author Contributions. K.K. proposed the original idea and design of theexperiments, supervised experiments, provided intellectual input, and wrote themanuscript. S.S. performed the quantification of the histological analysis and

most of the in vitro analyses and DPP-4 activity measurements. M.Ka. took careof the mice, performed the histological analysis, provided intellectual advice, andedited the manuscript. J.H. performed some of the animal experiments and someof the in vitro analysis. T.N. performed some of the in vitro analysis regardingEndMT. Y.N. and Y.I. performed the microRNA and mRNA array analysis. M.Ki.participated in the discussions. S.P.S. performed in vivo and in vitro quantitativePCR for mRNA and microRNA, Western blot, transfection, cloning of DPP-439 UTR vector, and quantification. D.K. provided intellectual input. S.S. andS.P.S. are the guarantors of this work and, as such, had full access to all thedata in the study and take responsibility for the integrity of the data and theaccuracy of the data analysis.Prior Presentation. Parts of this study were presented in abstract form atthe American Society of Nephrology Renal Week, Atlanta, GA, 9 November 2013.

References1. Xu L, Kanasaki K, Kitada M, Koya D. Diabetic angiopathy and angiogenicdefects. Fibrogenesis Tissue Repair 2012;5:132. Kanasaki K, Taduri G, Koya D. Diabetic nephropathy: the role of inflam-mation in fibroblast activation and kidney fibrosis. Front Endocrinol (Lausanne)2013;4:73. Mackensen-Haen S, Bader R, Grund KE, Bohle A. Correlations betweenrenal cortical interstitial fibrosis, atrophy of the proximal tubules and impairmentof the glomerular filtration rate. Clin Nephrol 1981;15:167–1714. Nath KA. Tubulointerstitial changes as a major determinant in the pro-gression of renal damage. Am J Kidney Dis 1992;20:1–175. Risdon RA, Sloper JC, De Wardener HE. Relationship between renal functionand histological changes found in renal-biopsy specimens from patients withpersistent glomerular nephritis. Lancet 1968;2:363–3666. Schainuck LI, Striker GE, Cutler RE, Benditt EP. Structural-functional cor-relations in renal disease. II. The correlations. Hum Pathol 1970;1:631–6417. Striker GE, Schainuck LI, Cutler RE, Benditt EP. Structural-functional cor-relations in renal disease. I. A method for assaying and classifying histopathologicchanges in renal disease. Hum Pathol 1970;1:615–6308. Kanasaki K, Kitada M, Koya D. Pathophysiology of the aging kidney andtherapeutic interventions. Hypertens Res 2012;35:1121–11289. He J, Xu Y, Koya D, Kanasaki K. Role of the endothelial-to-mesenchymaltransition in renal fibrosis of chronic kidney disease. Clin Exp Nephrol 2013;17:488–49710. Drucker DJ. The biology of incretin hormones. Cell Metab 2006;3:153–16511. Marathe CS, Rayner CK, Jones KL, Horowitz M. Effects of GLP-1 andincretin-based therapies on gastrointestinal motor function. Exp Diabetes Res2011;2011:27953012. Cernea S, Raz I. Therapy in the early stage: incretins. Diabetes Care 2011;34(Suppl. 2):S264–S27113. Mentlein R. Dipeptidyl-peptidase IV (CD26)—role in the inactivation ofregulatory peptides. Regul Pept 1999;85:9–2414. Ghersi G, Dong H, Goldstein LA, et al. Regulation of fibroblast migration oncollagenous matrix by a cell surface peptidase complex. J Biol Chem 2002;277:29231–2924115. Broxmeyer HE, Hoggatt J, O’Leary HA, et al. Dipeptidylpeptidase 4 nega-tively regulates colony-stimulating factor activity and stress hematopoiesis. NatMed 2012;18:1786–179616. Arscott WT, LaBauve AE, May V, Wesley UV. Suppression of neuroblastomagrowth by dipeptidyl peptidase IV: relevance of chemokine regulation and cas-pase activation. Oncogene 2009;28:479–49117. Scheen AJ. Pharmacokinetics of dipeptidylpeptidase-4 inhibitors. DiabetesObes Metab 2010;12:648–65818. Blech S, Ludwig-Schwellinger E, Gräfe-Mody EU, Withopf B, Wagner K.The metabolism and disposition of the oral dipeptidyl peptidase-4 inhibitor, li-nagliptin, in humans. Drug Metab Dispos 2010;38:667–67819. Kern M, Klöting N, Niessen HG, et al. Linagliptin improves insulin sensitivityand hepatic steatosis in diet-induced obesity. PLoS ONE 2012;7:e38744

2130 Linagliptin Inhibits Diabetic Kidney Fibrosis Diabetes Volume 63, June 2014

20. Sugimoto H, Grahovac G, Zeisberg M, Kalluri R. Renal fibrosis and glo-merulosclerosis in a new mouse model of diabetic nephropathy and its re-gression by bone morphogenic protein-7 and advanced glycation end productinhibitors. Diabetes 2007;56:1825–183321. Groop PH, Cooper ME, Perkovic V, Emser A, Woerle HJ, von Eynatten M.Linagliptin lowers albuminuria on top of recommended standard treatment inpatients with type 2 diabetes and renal dysfunction. Diabetes Care 2013;36:3460–346822. Zeisberg EM, Tarnavski O, Zeisberg M, et al. Endothelial-to-mesenchymaltransition contributes to cardiac fibrosis. Nat Med 2007;13:952–96123. Potenta S, Zeisberg E, Kalluri R. The role of endothelial-to-mesenchymaltransition in cancer progression. Br J Cancer 2008;99:1375–137924. Yang J, Campitelli J, Hu G, Lin Y, Luo J, Xue C. Increase in DPP-IV in theintestine, liver and kidney of the rat treated with high fat diet and streptozotocin.Life Sci 2007;81:272–27925. Liu WJ, Xie SH, Liu YN, et al. Dipeptidyl peptidase IV inhibitor attenuateskidney injury in streptozotocin-induced diabetic rats. J Pharmacol Exp Ther 2012;340:248–25526. Ta NN, Li Y, Schuyler CA, Lopes-Virella MF, Huang Y. DPP-4 (CD26) inhibitoralogliptin inhibits TLR4-mediated ERK activation and ERK-dependent MMP-1expression by U937 histiocytes. Atherosclerosis 2010;213:429–43527. Takahashi A, Asakura M, Ito S, et al. Dipeptidyl-peptidase IV inhibitionimproves pathophysiology of heart failure and increases survival rate in pressure-overloaded mice. Am J Physiol Heart Circ Physiol 2013;304:H1361–H136928. Cushing L, Kuang PP, Qian J, et al. miR-29 is a major regulator of genesassociated with pulmonary fibrosis. Am J Respir Cell Mol Biol 2011;45:287–294

29. van Rooij E, Sutherland LB, Thatcher JE, et al. Dysregulation of microRNAsafter myocardial infarction reveals a role of miR-29 in cardiac fibrosis. Proc NatlAcad Sci U S A 2008;105:13027–1303230. Liu Y, Taylor NE, Lu L, et al. Renal medullary microRNAs in Dahl salt-sensitive rats: miR-29b regulates several collagens and related genes. Hyper-tension 2010;55:974–98231. Du B, Ma LM, Huang MB, et al. High glucose down-regulates miR-29a toincrease collagen IV production in HK-2 cells. FEBS Lett 2010;584:811–81632. Chaykovska L, von Websky K, Rahnenführer J, et al. Effects of DPP-4 inhibitorson the heart in a rat model of uremic cardiomyopathy. PLoS ONE 2011;6:e2786133. Boerrigter G, Costello-Boerrigter LC, Harty GJ, Lapp H, Burnett JC Jr. Des-serine-proline brain natriuretic peptide 3-32 in cardiorenal regulation. Am JPhysiol Regul Integr Comp Physiol 2007;292:R897–R90134. Kaji K, Yoshiji H, Ikenaka Y, et al. Dipeptidyl peptidase-4 inhibitor attenuateshepatic fibrosis via suppression of activated hepatic stellate cell in rats. J Gas-troenterol 12 March 2013 [Epub ahead of print]35. Connelly K, Zhang Y, Advani A, et al. DPP-4 inhibition attenuates cardiacdysfunction and adverse remodeling following myocardial infarction in rats withexperimental diabetes. Cardiovasc Ther 2013;31:259–26736. Lenski M, Kazakov A, Marx N, Böhm M, Laufs U. Effects of DPP-4 inhibitionon cardiac metabolism and function in mice. J Mol Cell Cardiol 2011;51:906–91837. Casrouge A, Decalf J, Ahloulay M, et al. Evidence for an antagonist form ofthe chemokine CXCL10 in patients chronically infected with HCV. J Clin Invest2011;121:308–31738. LeBleu VS, Taduri G, O’Connell J, et al. Origin and function of myofibroblastsin kidney fibrosis. Nat Med 2013;19:1047–1053

diabetes.diabetesjournals.org Kanasaki and Associates 2131