Cilostazol attenuates intimal hyperplasia in a mouse model of chronic kidney disease

Intimal hyperplasia (IH) is a common cause of vasculopathy due to direct endothelial damage (such as post-coronary revascularization) or indirect injury (such as chronic kidney disease, or CKD). Although the attenuation of coronary revascularization-induced IH (direct-vascular-injury-induced IH) by cilostazol, a phosphodiesterase III inhibitor, has been demonstrated, our understanding of the effect on CKD-induced IH (indirect-vascular-injury-induced IH) is limited. Herein, we tested if cilostazol attenuated CKD-induced IH in a mouse model of ischemic-reperfusion injury with unilateral nephrectomy (Chr I/R), a normotensive non-proteinuria CKD model. Cilostazol (50 mg/kg/day) or placebo was orally administered once daily from 1-week post-nephrectomy. At 20 weeks, cilostazol significantly attenuated aortic IH as demonstrated by a 34% reduction in the total intima area with 50% and 47% decreases in the ratios of tunica intima area/tunica media area and tunica intima area/(tunica intima + tunica media area), respectively. The diameters of aorta and renal function were unchanged by cilostazol. Interestingly, cilostazol decreased miR-221, but enhanced miR-143 and miR-145 in either in vitro or aortic tissue, as well as attenuated several pro-inflammatory mediators, including asymmetrical dimethylarginine, high-sensitivity C-reactive protein, vascular endothelial growth factor in aorta and serum pro-inflammatory cytokines (IL-6 and TNF-α). We demonstrated a proof of concept of the effectiveness of cilostazol in attenuating IH in a Chr I/R mouse model, a CKD model with predominantly indirect-vascular-injury-induced IH. These considerations warrant further investigation to develop a new primary prevention strategy for CKD-related IH.


Introduction
Intimal hyperplasia (IH) is a vasculopathy characterized by a differentiation of any cells that form a multilayer compartment at the tunica intima of blood vessels. IH is responsible for a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 vascular complications in several chronic disorders including coronary artery disease (CAD), peripheral arterial disease (PAD), and chronic kidney disease (CKD) [1][2][3]. Direct endothelial damage is the most important factor that induced IH in CAD and PAD [3]. On the other hand, both direct vascular damage (hypertension) and indirect vascular injury (uremia, anemia and chronic inflammatory state) are responsible for IH in CKD [1,4]. Because most animal studies on IH have been performed in experiments on a 5/6 nephrectomy model, or a model with chronic glomerulopathy with hypertension and heavy proteinuria, IH in these mice was influenced not only by CKD but also by the hypertension-associated direct endothelial injury. To see if the indirect-vascular-injury-induced IH of CKD exists without hypertension, another CKD model is needed. Interestingly, a CKD model with predominant tubulointerstitial damage, less albuminuria and hypertension is developed by ischemia-reperfusion injury with unilateral nephrectomy (Chr I/R) [5]. Hence, we examine indirect-vascularinjury-induced IH in this model.
Thus, a therapeutic strategy with multiple effects might be a more appropriate treatment of IH. Indeed, cilostazol, a phosphodiesterase (PDE) III inhibitor, attenuates the direct vascular injury-induced IH through multiple mechanisms, including vasodilation and antiplatelet action [15], anti-inflammation and the reduction of platelet-leukocyte interaction [16] and the inhibition of vascular proliferation [17,18] through the up-regulation of hepatocyte growth factors and enhancement of p53 oncogene [19][20][21]. Moreover, cilostazol might modulate vascularrelated growth factors and oxidative stress molecules such as VEGF [7], PDGF [8], and nitric oxides (NO) [22]. Nevertheless, the data on cilostazol's effects on indirect-vascular-injuryinduced IH is still limited. Therefore, we have conducted experiments to determine the therapeutic effects of cilostazol on indirect-vascular-injury-induced IH in a Chr I/R mouse model.

Animal model
Thirty 6-week-old male CD-1 mice weighing 34±3 g were obtained from the National Laboratory Animal Center, Nakhon Pathom, Thailand. Animal care followed the National Institutes of Health criteria for the use and treatment of laboratory animals. The protocol for the experiment was approved by the Animal Care and Use Committee of the Faculty of Medicine, Chulalongkorn University (No.008/2556). The mice were separated to one animal per cage and maintained under conditions of 20-24˚C at 40-70% humidity with a 12-hour light-dark cycle. Food and water were given ad libitum.
was clamped through left flank incisions by atraumatic vascular clamp, and the incisions were closed layer by layer with nylon 2-0. The mice were then allowed to awaken in the box. Subsequently, the procedure was repeated to remove the clamps after a total ischemic time of 50 min. One week later (wk0), right nephrectomy was performed through right flank incisions ( Fig 1A). For the sham-operated group, mice underwent left renal sham operations at wk-1, followed by right nephrectomy. The body temperature during all operations, including anesthesia, incision/clamping, and suturing was maintained by setting the temperature of the heating plate (Kleintier-OP-Tisch, Medax GmbH, Germany) to 37˚C to avoid temperature effects on ischemic-reperfusion (I/R) injury [23]. At one week post-nephrectomy (wk1), Chr I/R mice were randomized to be administered either cilostazol (Pletaal 1 , PM129023, Otsuka, Tokushima, Japan) at 50 mg/kg/dose or a placebo control (reciprocal volume of normal saline) through daily oral gavage at 08:00 AM until euthanization at 20 weeks post-nephrectomy (wk20).
During the follow-up period, mice were monitored for clinical signs every day, including body condition, skin condition, estimated hydration, color of mucous membranes, heart rate and rhythm, respiratory rate, signs of diarrhea, and amount and appearance of urine and feces in the cage. Four mice died in wk15 and wk16 after randomization without preceding signs and symptoms. Therefore, 10 mice per group were available for analysis ( Fig 1B). Body weights were monitored twice per week. Blood was drawn via tail vein at 1, 4, 12, 16 and 20 weeks post-nephrectomy. Serum samples were immediately centrifuged at 8,000g for 10 min at 4˚C and kept at -80˚C until used. Spot urine samples were collected every week for 4 consecutive weeks, then every 4 weeks, and kept at -80˚C.
All mice were euthanized through cardiac puncture under isoflurane anesthesia, and internal organs (heart, lungs, and kidneys) were preserved in 10% neutral-buffered formalin for histological study. Despite the protective effect of isoflurane [24] and multiple isoflurane The effects of cilostazol on intimal hyperplasia of chronic kidney disease mouse exposures, renal fibrosis was demonstrated in the Chr I/R mice model in both placebo-and cilostazol-treated mice. Aorta from the arch of aorta to iliac bifurcation was removed. The proximal one-third of the aortic arch was preserved in 10% neutral buffered formalin, while the distal two-thirds were preserved in RNAlater (ThermoFisher Scientific, Waltham, MA, USA) and stored at -80˚C for miR study.

Systolic blood pressure measurement
A tail cuff plethysmography (IITC Life Scientific Instruments, Woodland Hills, CA, USA) was used for the measurement of systolic blood pressure (SBP) as previously described elsewhere [25]. SBP was measured before the clamping of left renal bundle at wk-1 and then again at 4, 12, and 20 weeks after right nephrectomy. The measurements of SBP in the perioperative period were omitted due to the confounding variable of perioperative stress. SBP measurements were carried out several times before the start of the experiment to precondition the mice to the procedure. SBP measurements were obtained after the mice were familiar with the process. The means of three SBP measurements at each time point, with a 10-minute rest interval between the readings, were used to represent the data.

Histology and assessment of glomerular and interstitial injuries
Kidney sections (3 μm thickness) were stained with hematoxylin and eosin (H&E), Periodic Acid-Schiff reagent (PAS), and Masson's trichrome color. Glomerular injury was scored blindly by 3 pathologists. For glomerular injury scoring, 50 randomly selected glomeruli in each slide of PAS-stained sections were explored, and the sclerosis area in each glomerulus was assessed by a semi-quantitative scoring method: sclerotic area 0% = normal glomeruli; sclerotic area up to 25% = minimal sclerosis; sclerotic area 25 to 50% = moderate sclerosis; sclerotic area 50 to 75% = moderately severe sclerosis; and sclerotic area 75 to 100% = severe sclerosis. Quantitative analysis of interstitial compartments occupied by fibrotic tissue was defined by the relative interstitial volume of the total interstitium in each slide of Masson's trichromestained sections. A standard point-counting method was used to quantitate the volume of the renal cortex [26]. The cortical region was analyzed in a stepwise fashion as a series of consecutive fields. For each field, a region of interest was traced that included cortical tubules and interstitial space. One point-count study evaluated interstitium only and was based on 1,200 points for each kidney. Glomeruli, large-or medium-sized blood vessels, and medullary tissue were excluded. The relative interstitial volume was scored as a percentage of affected renal parenchyma to total surface area of sampled cortical area. The average values of three consecutive measurements were used for all parameters.

Histology and assessment of intimal hyperplasia
The vascular cross-sections of aorta (3 μm thickness) were stained with H&E and PAS, then quantitatively analyzed in ten sections per mouse, starting from a 5-μm 2 area of the left ventricle. To minimize bias in measurement, the average values of three consecutive measurements were used for all parameters. Aorta thickness was measured by two pathologists with blinded experimental data by using Image-Pro1 Premier 9.1 Software (Media Cybernetics Manufacturing, PA, USA). Lastly, the mean values of the scores of the two pathologists was taken. Tunica intima was an area inside internal elastic lamina and tunica media thickness was an area demarcated between internal and external elastic lamina. The tunica intima/tunica media ratio (I:M) and the ratio of tunica intima/(tunica intima + tunica media) (I:I+M) were used for the determination of IH.

RNA isolation and quantification
The protocol for this study followed a previous publication [28]. In brief, samples were placed into TRIzol (Invitrogen, Carlsbad, CA, USA), completely homogenized, and centrifuged at 10,000g for 15 min at 4˚C. Next, the supernatant was put in glycogen 1 μL in chloroform 250 μL. Then, the samples were centrifuged again at 10,000g for 20 minutes at 4˚C and supernatant was transferred into 1 mL of 70% ethanol and placed on an RNeasy column. Total RNA was evaluated by a miRNA easy kit (Qiagen, Valencia, CA, USA) for extraction. The samples were mixed again with denaturing buffer following the manufacturers' protocols. Both qualification and quantification of miRs was performed with the NanoDrop ND-1000 spectrometer (NanoDrop Technologies, Wilmington, DE, USA).

Reverse transcription, real-time PCR and graphical heat-map construction
Taqman probes for each gene and Taqman master mix were used to quantify synthetic miR spike-ins and cellular miR in real-time PCR assays according to the manufacturer's protocols.
Individual real-time PCR assays were performed in a 20 μL reaction volume on an ABI 7,500 real-time PCR system (Applied Biosystems, Waltham, MA, USA) based on the standard 7,500 run mode. The ΔΔCt method was used to calculate the relative expression (fold change) between sample groups. RNU6B expression was used for serum miR normalization [28]. For a better demonstration of miRs from aortic tissue in individual mice, a graphical heat-map construction of all RT-qPCR ΔCt values using Ward's method was performed.

Statistical analysis
All experimental data are expressed as mean±standard error (SEM). Analysis of variance (ANOVA) followed by post hoc Bonferroni's correction was used for the analysis of more than 2 groups. A two-way repeated measure ANOVA was used when two or more time points were assessed. Correlations between variables were analyzed using Pearson's correlation analysis, p-values of less than 0.05 were considered statistically significant. Statistical calculations were performed using SPSS for Windows 15.0 (SPSS for Windows; Chicago, IL) and GraphPad Prism 6.0 (GraphPad Software, La Jolla, CA, USA).

Results
Clinical manifestations of chronic kidney disease in ischemic-reperfusion injury with unilateral-nephrectomy-induced chronic kidney disease (Chr I/R) mice Clinical manifestations of CKD, including body weight (BW), systolic blood pressure (SBP), proteinuria, Hct, and urine volume, were explored in Chr I/R mice. CKD's failure to thrive was demonstrated with the unchanged BW during the first 8 weeks post-nephrectomy in Chr I/R mice. In contrast, a significant weight gain was demonstrated in sham mice. At 20 weeks, there was a 6% decrease from the baseline BW in Chr I/R mice but a 38% increase in BW from the baseline in the sham mice (Fig 2A). Furthermore, the present Chr I/R mice model demonstrated no significant changes in SBP measurements (Fig 2B).
Regarding kidney function, BUN and SCr increased as early as 2 weeks post-nephrectomy in both cilostazol-treated and placebo-treated mice. At baseline, 1 week before performing I/R injury in left kidney (wk-1), BUN and SCr were 14±2 and 0.3±0.1 mg/dL in cilostazol-treated mice, and 13±6 and 0.2±0.1 mg/dL in placebo-treated mice, respectively. At 20 weeks, postnephrectomy, BUN and SCr progressed to 83±5 and 1.2±0.5 mg/dL in cilostazol-treated mice and 88±6 and 1.2±0.3 mg/dL in placebo-treated mice, respectively (Fig 2C). This finding conforms to the idea that the Chr I/R mice had gradual anemia as determined by the Hct reduction from the baseline (from 50±7% to 45±6%) 12 weeks post-nephrectomy. Ultimately, Hct at 20 weeks post-nephrectomy was 37±2% (Fig 2D). Despite progressive deterioration of kidney function, proteinuria did not differ between the sham-operated, placebo-treated and cilostazol-treated mice at 20 weeks (Fig 2E), suggesting the CKD model with predominant tubulointerstitial injury was proven by kidney pathology (Fig 3A). At 20 weeks, however, cilostazoltreated mice demonstrated lower relative interstitial volume than placebo-treated mice (p< 0.05) as determined by histopathology analysis (Fig 3B).
In addition, diuresis was showed in Chr I/R mice. Urine volume was 9±3 mL/day at 20 weeks after nephrectomy. Despite the diuresis in Chr I/R mice, 35% (7/20) of these mice showed frank pulmonary edema by gross examination. The mean of lung wet weight/body weight (LW/BW) ratio in Chr I/R mice was 10.4±3.9 mg/g compared to 3.3±0.1 mg/g in sham mice (p< 0.01).

Cilostazol ameliorates vascular intimal hyperplasia at 20 weeks of ischemic-reperfusion injury with unilateral nephrectomy-induced chronic kidney disease (Chr I/R) mice
Chr I/R mice demonstrated a positive correlation between aortic IH severity and relative interstitial volume (r 2 = 0.74, p = 0.01) as well as BUN (r 2 = 0.67, p = 0.004), but a weak correlation The effects of cilostazol on intimal hyperplasia of chronic kidney disease mouse between IH severity and SCr (r 2 = 0.12, p< 0.001) as shown in S1 Fig. Interestingly, cilostazoltreated mice showed a lower severity of IH (Fig 4A). At 20 weeks post-nephrectomy, IH in cilostazol-treated and placebo-treated mice was 99,704±22,310 and 150,556±5,669 μm 2 (p< 0.01), respectively (S1 Table). These data are consistent with a 50% lower I:M ratio and a 46% lower I:I+M ratio. In contrast, the diameters of aorta of the cilostazol-treated group were not different from the placebo-treated group (Fig 4B-4E).

Cilostazol effects on aortic tissue miRs, serum vascular smooth muscle cells (VSMCs)-related cytokines and inflammatory cytokines at 20 weeks of ischemic-reperfusion injury with unilateral nephrectomyinduced chronic kidney disease (Chr I/R) mice
Increased expression of miR-143 and miR-145 in aorta tissue was demonstrated in cilostazoltreated mice compared with the placebo control. On the contrary, miR-221 showed a significantly reduced expression in the cilostazol-treated group (Fig 5). To support the association between cilostazol and miR expression, an in vitro test on HUVECs was performed. As demonstrated in Fig 6, cilostazol significantly enhanced both miR-143 and miR-145 expression but reduced miR-221 expression in comparison with the control group.
Due to the influence of VSMC-related cytokines and pro-inflammatory processes in IH pathogenesis, we measured VEGF, PDGF, hs-CRP, IL-6, and TNF-α. In comparison with placebo-treated mice, there were significantly lower levels of serum VEGF, hs-CRP, IL-6, and TNF-α, but not serum PDGF, in cilostazol-treated mice. In addition, serum ADMA, a surrogate marker for vascular vasodilatory effect, was significantly lower in cilostazol-treated mice compared with the placebo group (Fig 7). Despite the effectiveness of cilostazol at 20 weeks on the attenuation of IH in Chr I/R mice, renal function as determined by BUN and SCr was not different between the cilostazol and placebo groups (data not shown).
Because CKD increases pro-inflammatory mediators [29,30] and the association between PDGF and VEGF with miRs in IH pathogenesis has been reported [12,31,32], we evaluated the concentration in sera over time to determine the cascade of these mediators in Chr I/R mice. We found that serum miR-221 was the earliest detectable biomarker among these molecules (detectable in 5 mice (50%) at 12 weeks after right nephrectomy). Subsequently, other molecules, including VEGF, miR-143, and miR-145, were detectable at 16 weeks (S2 Fig). Then, other serum biomarkers, including ADMA, PDGF, and inflammatory cytokines (IL-6, TNF-and hs-CRP), were demonstrated at 20 weeks after right nephrectomy (data not shown). Furthermore, the increase in expression of VEGF, hs-CRP, IL-6, and TNF-α in the Chr I/R mice was reversed in animals treated with cilostazol. Interestingly, a positive correlation was demonstrated between VEGF and miR-221, whereas a negative correlation was noted between ADMA and miR-221. There was no correlation observed between other parameters (Fig 8).

Discussion
We demonstrated for the first time that cilostazol attenuates CKD-induced aortic IH in CKD with a Chr I/R mouse model. Cilostazol modulated several miRs responses, including reducing VEGF and ADMA. In addition, cilostazol also induced anti-inflammatory effects as demonstrated by the attenuation of several pro-inflammatory mediators (IL-6, TNF-α, and hs-CRP).
These contributions support the usefulness of cilostazol as an adjunctive strategy for the prevention of CKD-induced IH.

Arterial intimal hyperplasia and characteristics of Chr I/R model
Despite the diversity of IH from different diseases, the pathogenesis of IH can be explained through the direct and indirect endothelial injury processes [33,34], which, for CKD, are hypertension and the metabolic consequences, respectively. Because direct-vascular-injuryinduced IH is well-known, we aimed to investigate the predominantly indirect-vascular-injury effects of CKD on IH progression. Predominant tubulointerstitial damage with other CKD- The effects of cilostazol on intimal hyperplasia of chronic kidney disease mouse related consequences, but with mild proteinuria, was demonstrated in Chr I/R in our study, supporting previous research [5]. Interestingly, although the protective effects of isoflurane on I/R injury has been demonstrated [24], both placebo-and cilostazol-treated mice developed CKD characteristics under its influence. Due to the minimized high blood pressure in this model, IH was possibly mainly due to indirect endothelial injury. Indeed, aortic IH of Chr I/R was demonstrated with a 30% increase in tunica intima thickness of all mice at 20 weeks postnephrectomy. As such, a Chr I/R mouse model might be one of the most suitable CKD models for exploring indirect-endothelial-injury-induced IH. Interestingly, medial calcification in this CKD mice model was not prominent. We hypothesize that the calciphylaxis in this model is not as severe as in patients with CKD. This implies that the influence on calciphylaxis of calcium, vitamin D and/or diets in patients with CKD and in other in vivo models [35].
Cilostazol attenuated chronic kidney disease-related IH, at least in part, through anti-inflammation and miRs modulation The protective effects of cilostazol on CKD-related IH were demonstrated by the reduction of total tunica intima area, seen in the lower I:M ratio and I:I+M ratio in the cilostazol-treated mice compared with the placebo group. The lower I:M ratio implies that cilostazol was targeted toward the intima, while the lower I:I+M ratio supported the effectiveness of cilostazol The effects of cilostazol on intimal hyperplasia of chronic kidney disease mouse on the intima layer. Although the attenuation of direct-endothelial-injury-induced IH by cilostazol has been previously demonstrated [36][37][38], this study shows for the first time the benefits on a CKD model with predominantly indirect-vascular-injury-induced IH. Our findings suggest that cilostazol might promote vascular protection predominantly through anti- The effects of cilostazol on intimal hyperplasia of chronic kidney disease mouse inflammatory effects (the attenuation of hs-CRP, IL-6, and TNF-α) (Fig 7B, 7E and 7F). Perhaps the anti-inflammatory effects were important for the amelioration of the indirect-vascular-injury-induced IH, while the vasodilatory effects were important for the attenuation of the direct-vascular-injury-induced IH [39][40][41].
Although cilostazol theoretically induces VEGF, there was lower VEGF in cilostazol-treated Chr I/R [7,42]. Indeed, VEGF has a paradoxical effect on endothelium. VEGF was induced by pro-inflammatory responses but important for the endothelial healing process [43]. As such, anti-VEGF is beneficial as an anti-inflammatory drug in pro-inflammatory states but disrupts the wound healing processes in a chronic injury model [44,45]. The VEGF-induced endothelial proliferation might worsen IH. The lower VEGF in cilostazol-treated mice demonstrates the more prominent anti-inflammatory effects over the induction of endothelial proliferation, which is beneficial for IH attenuation.
On the other hand, circulating ADMA, a uremic toxin with endogenous NOS inhibitory effects and that induces pro-inflammatory responses [46], was also significantly decreased with cilostazol treatment. As observed in the rabbit model, ADMA inhibited NO, resulting in relative vasoconstriction, increased leucocyte and platelet adhesion, and enhanced IH [47]. As such, ADMA reduction after cilostazol treatment could also be beneficial. However, cilostazol did not attenuate PDGF, an important IH activator though VSMCs activation [48] in the Chr I/R model. Furthermore, the association between miRs, VSMC regulation, and IH formation was recently demonstrated [10,12]. Mice with the gene-deficiency of miR-143 and/or miR-145 showed prominent IH [49]; miR-145-blocking remarkably inhibited IH [12]; and a knockdown of miR-221 also inhibited IH [11]. We hypothesized that cilostazol might also mediate IH through miRs, and explored the expression of these miRs in aorta of Chr I/R mice. Indeed, we found that cilostazol inhibited tissue miR-221 but promoted tissue miR-143 and miR-145 in aorta and the role of miRs in IH pathogenesis [50]. In addition, the increased expression of miR-143 and miR-145 along with reduced miR-221 expression after cilostazol treatment were also demonstrated in HUVEC experiments.
Our data support previous studies on the importance of miR-221 [11], but demonstrate an opposite direction of miR-143 and miR-145 in association with IH severity [10,12,49]. These data imply diversity in IH pathogenesis in different models. Despite these data, the earliest time point of miR activation by cilostazol in relation to IH suppression was not defined in our study. Thus, although miR activation appeared to be a downstream event related to IH inhibition [10], the link between these 3 miR targets in CKD-related IH after Chr I/R requires further study.
Nevertheless, the correlation analysis demonstrated the association trends of miR-221 to VEGF and ADMA in a positive and a negative direction, respectively. Cilostazol decreases miR-221 together with VEGF. Because miR-221 induces IH through increased VEGF by Vegfc/Vegfr-3 signaling activation [51], it is possible that cilostazol may attenuate VEGF through decreased miR-221. We found that miR-221 had an early onset that decreased in concentration compared with VEGF, and miR-221 also had a higher percentage reduction from the baseline than VEGF at 16 weeks post-nephrectomy (93.5% and 71.6%, respectively, p< 0.05). On the other hand, cilostazol attenuated both miR-221 and ADMA, but there was an inverse trend between these parameters. This implies that both miR-221 and ADMA may be implicated in VSMC regulation; however, miR-221 may not be a differentially expressed miR for ADMA. Although an increasing body of evidence reveals the effects of miR221 and ADMA in early endothelial progenitor cells response [52,53], their role in VSMCs is not well-known and is a fertile topic for future studies.
In conclusion, we demonstrated a proof of the effectiveness of cilostazol in attenuating IH in a Chr I/R mouse model, a CKD model with predominantly indirect-vascular-injuryinduced IH. These considerations warrant further investigation to develop a new primary prevention strategy for CKD-related IH. The identification of tissue miR-221, ADMA, and inflammatory cytokines could also be interesting surrogate biomarkers to demonstrate the transition of biochemical changes into significant pathological defects. These data can open new avenues to manipulate vascular damage from IH.