Adventitial delivery of nanoparticles encapsulated with 1α, 25-dihydroxyvitamin D3 attenuates restenosis in a murine angioplasty model

Percutaneous transluminal angioplasty (PTA) of stenotic arteriovenous fistulas (AVFs) is performed to maintain optimal function and patency. The one-year patency rate is 60% because of venous neointimal hyperplasia (VNH) and venous stenosis (VS) formation. Immediate early response gene X-1 (Iex-1) also known as Ier3 increases in response to wall shear stress (WSS), and can cause VNH/VS formation in murine AVF. In human stenotic samples from AVFs, we demonstrated increased gene expression of Ier3. We hypothesized that 1α, 25-dihydroxyvitamin D3, an inhibitor of IER3 delivered as 1α, 25-dihydroxyvitamin D3 encapsulated in poly lactic-co-glycolic acid (PLGA) nanoparticles loaded in Pluronic F127 hydrogel (1,25 NP) to the adventitia of the stenotic outflow vein after PTA would decrease VNH/VS formation by reducing Ier3 and chemokine (C–C motif) ligand 2 (Ccl2) expression. In our murine model of AVF stenosis treated with PTA, increased expression of Ier3 and Ccl2 was observed. Using this model, PTA was performed and 10-μL of 1,25 NP or control vehicle (PLGA in hydrogel) was administered by adventitial delivery. Animals were sacrificed at day 3 for unbiased whole genome transcriptomic analysis and at day 21 for immunohistochemical analysis. Doppler US was performed weekly after AVF creation. At day 3, significantly lower gene expression of Ier3 and Ccl2 was noted in 1,25 NP treated vessels. Twenty-one days after PTA, 1,25 NP treated vessels had increased lumen vessel area, with decreased neointima area/media area ratio and cell density compared to vehicle controls. There was a significant increase in apoptosis, with a reduction in CD68, F4/80, CD45, pro-inflammatory macrophages, fibroblasts, Picrosirius red, Masson’s trichrome, collagen IV, and proliferation accompanied with higher wall shear stress (WSS) and average peak velocity. IER3 staining was localized to CD68 and FSP-1 (+) cells. After 1,25 NP delivery, there was a decrease in the proliferation of α-SMA (+) and CD68 (+) cells with increase in the apoptosis of FSP-1 (+) and CD68 (+) cells compared to vehicle controls. RNA sequencing revealed a decrease in inflammatory and apoptosis pathways following 1,25 NP delivery. These data suggest that adventitial delivery of 1,25 NP reduces VNH and venous stenosis formation after PTA.


Results
Ier3 expression was significantly Increased in stenotic clinical AVFs by endovascular biopsy and murine PTA treated vessels. The gene expression of Ier3 in the endothelium of stenotic AVFs is unknown. We performed endovascular biopsy of human stenotic AVFs prior to PTA procedures (N = 5, 4 males, 1 female, average age = 66 ± 7 years old, Supplementary Table 1) 12 . Samples were obtained from nonstenotic veins of AVFs of each patient to serve as their own control. There was a significant increase in the mean gene expression of Ier3 in endothelial cells removed from stenotic AVFs compared to controls (stenotic AVFs: 19.2 ± 18.8, control: 1.00 ± 0.04, average increase: 1920%, p < 0.01, Fig. 1E).
The gene and protein expression of IER3 is unknown after PTA. This was assessed in the murine PTA model in which an AVF stenosis is treated with PTA in mice with CKD ( Supplementary Fig. 1). We determined the gene

Serum changes in kidney and liver function in 1,25 NP and vehicle control animals.
We assessed the changes in blood urea nitrogen (BUN), creatinine, ALT, AST, total bilirubin, and calcium after local delivery of 1,25 NP or vehicle control to the outflow veins by measuring. There was no significant difference in these assessments at day 3 and 21 after 1,25 NP delivery compared to vehicle controls (Supplementary Table 2).
Ier3 gene and protein expression is significantly decreased in 1,25 NP treated vessels compared to vehicle controls. In order to examine the therapeutic effect of IER3 inhibition in PTA treated vessels, we delivered 1,25 NP or vehicle to the adventitial layer of the outflow vein immediately after PTA (Supplementary Fig. 1 and Fig. 2). At day 3, the average gene expression of Ier3 and Ccl2 was significantly decreased in the 1, 25 Fig. 3D). The diameter of outflow veins was measured intraoperatively after AVF placement, at pre-and post-PTA, and at sacrifice (Fig. 3E). Twenty-one days after PTA, there was no difference in the average diameter of the outflow vein between the 1,25 NP group and vehicle control group (Fig. 3E).
There is an increase in the average peak velocity (PV) and wall shear stress (WSS) of 1,25 NP treated vessels compared to vehicle controls. Peak velocity of the blood through the outflow vein was determined using Doppler US. Before PTA, there was no significant difference in the average PV between the two groups ( Fig. 3F) There was an increase in the contractile SMCs with a decrease in the synthetic phenotype with a decrease in FSP-1 in 1,25 NP treated vessels compared to vehicle controls. Venous neointimal hyperplasia is characterized histologically by the accumulation of smooth muscle cells (SMCs) in the intima and media 5 . The presence of SMCs was determined using α-SMA. There were increased cells staining positive for α-SMA in the neointima of the outflow vein of the 1,25 NP treated vessels compared to vehicle group (Fig. 4A) with a significant increase in the average α-SMA index (1,25 NP: 6.98 ± 0.82, vehicle: 2.85 ± 0.50, average increase: 245%, p < 0.01, Fig. 4C). We wanted to assess if the α-SMA increase was due to contractile or synthetic MSCs. MHY11 staining was performed and this showed that there was decrease in the MYH11 index in 1,25 NP vessels compared to vehicle controls (1,25 NP: 3.39 ± 0.76, vehicle: 0.97 ± 0.29, average increase: 349%, p < 0.05, Fig. 4D). In order to determine the presence of contractile vascular SMCs, we performed double staining for α-SMA (+)/MHY11 (+). There was an increase in the co-staining for α-SMA (+)/MYH11 (+) cells in 1,25 NP treated vessels (Fig. 4G, left panel) compared to vehicle controls (Fig. 4F, left panel). These data suggest that there is more contractile VSMC in 1,25 NP treated vessels compared to vehicle treated vessels. The presence of synthetic SMC phenotype was determined using vimentin, MMP-2, and MMP-9 staining ( Supplementary Fig. 5). The synthetic phenotype is associated with increased fibrosis due to collagen secretion. In 1,25 NP treated vessels compared to control vehicles, there was a significant reduction in vimentin (1,25 NP: 17.5 ± 2.31, vehicle: 29 ± 2.93, average decrease: 40%, p < 0.05, Supplementary Fig. 5C), MMP-2 (1,25 NP: 2.81 ± 0.65, vehicle: 5.90 ± 0.44, average decrease: 52%, p < 0.05, Supplementary Fig. 5D), and MMP-9 staining (1,25 NP: 0.53 ± 0.12, vehicle: 1.16 ± 0.08, average decrease: 54%, p < 0.01, Supplementary Fig. 5E). These data suggest that there is less synthetic VSMC in 1,25 NP treated vessels compared to vehicle treated vessels.
Venous stenosis formation can occur because of differentiation of fibroblasts to SMCs 13 . We assessed the presence of fibroblasts by staining for fibroblasts specific protein-1 (FSP-1). There were more cells staining positive for FSP-1 in the vehicle control tissues compared to 1,25 NP treated vessels (Fig. 4A, B). There was a significant reduction in the average FSP-1 index of 1,25 NP treated vessels compared to vehicle controls (1,25 NP: 1.36 ± 0.38, vehicle: 4.23 ± 0.84, average decrease: 68%, p < 0.05, Fig. 4E). We performed staining for fibroblast activation protein-1 (FAP-1) to determine if the fibroblasts were activated and observed no difference in the staining between both groups ( Supplementary Fig. 4). In order to determine if FSP-1 cells also colocalized to α-SMA, we performed double staining for α-SMA (+)/FSP-1 (+) cells. There was no difference in the co-staining for FSP-1 (+)/α-SMA (+) cells in 1,25 NP treated vessels (Fig. 4G, right panel) compared to the vehicle controls (Fig. 4F, right panel).
We determined M1 staining by performing iNOS staining and observed that the average iNOS staining (1,25 NP: 4.77 ± 1.12, vehicle: 8.18 ± 0.71, average decrease: 42%, p < 0.05, Fig. 5C) were significantly decreased in the 1,25 NP treated vessels compared to vehicle controls at day 21. There was no significant difference for Arg-1 staining at day 21 ( Supplementary Fig. 4). We determined the average M1/M2 ratio in the 1,25 NP treated vessels compared to vehicle controls and there was no difference (Fig. 5D).
Because local delivery of 1,25 NP delivery caused a transient increase in serum calcium concentration, we assessed the presence of calcification by performing Alizarin Red staining. We observed no evidence of calcification in both groups ( Supplementary Fig. 4).
There was reduced collagen staining in 1,25 NP treated vessels. Increased fibrosis has been identified in failed AVF specimens and experimental rodent PTA model 11 . We performed Picrosirius red, collagen, and Masson's trichrome staining to determine the fibrotic changes 21 days after PTA in 1,25 NP treated vessels and in vehicle controls. Most of the Picrosirius Red and collagen IV staining was observed in the neointima and media area in both groups (Fig. 6A). Picrosirius red staining indicated that there is a significant reduction in collagen 1 (yellow) and There was a significant reduction in the average collagen 1 staining in the 1,25 NP treated vessels compared to vehicle controls (1,25 NP: 17.03 ± 2.39, vehicle: 25.6 ± 2.26, average decrease: 33%, p < 0.05, Fig. 6B) and average collagen 3 staining (green) in the 1,25 NP treated vessels compared to vehicle controls (1,25 NP: 34.9 ± 6.48, vehicle: 62.5 ± 7.39, average decrease: 44%, p < 0.05). There was a significant reduction in the average collagen IV staining in the 1,25 NP treated vessels compared to vehicle controls (1,25 NP: 29.34 ± 2.53, vehicle: 37.13 ± 1.55, average decrease: 21%, p < 0.05, Fig. 6C). Finally, we performed Masson's trichrome staining and observed a significant reduction in 1,25 NP treated vessels compared to vehicle controls (1,25 NP: 23.3 ± 2.8, vehicle: 50 ± 3.8, average decrease: 53%, p < 0.01, Supplementary Fig. 8 Supplementary Fig. 9). We wanted to identify the phenotype of the proliferative and apoptotic cells  There were 311 genes that were up regulated and 873 genes that were down regulated in 1,25 NP treated outflow vein compared to vehicle. Panther analysis identified that the most affected signaling pathways associated with the up regulated genes included Wnt signaling, cadherin, and inflammation (Fig. 8C). Similarly, in the downregulated genes, Panther analysis demonstrated that the top three signaling pathways that were involved were CCKR, inflammation, and apoptosis (Fig. 8D).

Discussion
In the present study, we demonstrated that there is increased gene expression of IER3 in endothelial samples removed from patients with stenotic AVFs. There was increased gene and protein expression of IER3, CCL2, and CD68 staining after PTA of VS in mice with AVF and CKD compared to control veins. Adventitial delivery of 1,25 NP to the outflow veins after PTA resulted in a significant decrease in the average gene expression of Ier3 and Ccl2. This was accompanied with positive vascular remodeling with as increased lumen vessel area, reduced N/M ratio and reduced cell density in the neointima at day. We observed that IER3 staining was decreased in cells staining (+) for CD68 and FSP-1 in the 1,25 NP treated vessels compared to vehicle controls. This was accompanied with a significant decrease in proliferation, inflammatory cellular phenotypes (CD68, F4/80, and www.nature.com/scientificreports/ CD45 (+), pro-inflammatory iNOS staining), FSP-1, fibrosis as assessed using Masson's trichrome, collagen staining, and Picrosirius red), with an increase in contractile VSMCs with decrease in synthetic VSMCs. There was a decrease in proliferation and increase in apoptosis after 1,25 NP delivery. We assessed which cells had decreased proliferation and increased apoptosis and found that there was a decrease in the proliferation of α-SMA (+) and CD68 (+) cells with increase in the apoptosis of FSP-1 (+) and CD68 (+) cells. There was an increase in average peak velocity and WSS in the 1,25 NP treated vessels compared to vehicle controls. Finally, unbiased whole transcriptomic analysis with RNA sequencing and Panther analysis was performed to identify the mechanisms involved with 1,25 NP delivery and we observed that inflammatory and apoptotic pathways were involved. Previous studies have shown that there is increased expression of Ier3 and Ccl2 in stenotic experimental and human failed AVFs, and it was localized to the adventitia 7,14 . IER3 expression has been shown to localize to the endothelium 10 . A recent study described the use of endovascular biopsy to obtain venous samples from stenotic AVFs 9 . We took advantage of this approach and determined the gene expression of Ier3 prior to PTA treatment in patients with stenotic hemodialysis AVFs. We observed a significant increase in average gene expression of Ier3 in stenotic endothelial cells compared to patient matched endothelial cells from non-stenotic veins from the same AVF. Because the expression of Ier3 and Ccl2 after PTA of AVF is unknown, we investigated the expression of Ier3, Ccl2 in our mouse model of PTA of VS and found that gene expression of Ier3 and Ccl2 increases after PTA.
Genetic deletion of Ier3 in mice with AVF and CKD demonstrated a reduction in VS/VNH formation 7 . 1α, 25(OH) 2 D 3 is an inhibitor of IER3 8 . Previous work from our lab has shown that Ier3 expression is reduced with the adventitial delivery of 1,25 NP accompanied with a gene expression of Ccl2 7 . We postulated that adventitial delivery of 1,25 NP after PTA of outflow veins would have significant decrease in Ier3 and Ccl2 gene expression. We observed this and were able to localize the cells expressing IER3 by immunofluorescent technique by performing double staining for CD68 and FSP-1 cells which reside in the adventitia. There was reduced staining of IER3 in 1,25 NP treated vessels in particular cells staining positive for CD68 and FSP-1. Taken collectively, this implies that 1,25 NP treatment after PTA reduces the expression of IER3 in CD68 (+) inflammatory cells.
Restenosis after PTA can lead to a decrease in the lumen vessel area and VNH/VS. We assessed for this possibility by assessing vascular remodeling using histomorphometric analysis in 1,25 NP treated vessels compared to control controls. There was a significant increase in the average lumen vessel area with a decrease in N/M www.nature.com/scientificreports/ ratio and cell density in the neointima. These observations are consistent with our previous study in murine AVFs in which we administered 1,25 NP to adventitia of the outflow vein 7 . Because WSS and blood flow velocity have been hypothesized as factors leading to the development of VNH associated with AVF 15 , the changes in PV and WSS after PTA were determined. A previous study from our lab showed differences in WSS after PTA compared to control vessels and in this model, increased PV and WSS are seen in vein segments with positive vascular remodeling 16 . In the present study, we extended these results by performing Doppler US to assess the hemodynamic function of the AVF. We observed a significant increase in the average PV and WSS in 1,25 NP treated vessels compared to controls. The present model has increased WSS in the outflow vein in part due to the angle of the anastomosis, which is approximately 90 degrees and this angle has been shown to cause increased shear stress when compared to an angle of 45 degrees 17 . www.nature.com/scientificreports/ Angioplasty can lead to VS formation due to an increase in α-SMA (+) cells accompanied with inflammatory cells and fibrosis. We assessed for changes in SMCs by staining for α-SMA. There was an increase in α-SMA staining in 1,25 NP treated vessels compared to controls. There are different phenotypes of SMCs. We assessed for the presence of contractile SMC phenotype by staining for MHY11 and synthetic SMCs by staining for vimentin, MMP-2, and MMP-9 staining 18 . Synthetic smooth muscles cells are hypothesized to cause restenosis after vascular injury 18 . Several cytokines have been shown to be involved including transforming growth factor-β 19 . We observed an increase in MHY11 (+) cells, which colocalized with α-SMA in 1,25 NP treated vessels. We next assessed for the presence of synthetic SMCs, and observed a significant reduction in in vimentin, MMP-2, and MMP-9 staining in 1,25 NP treated vessels compared to controls. The mechanism of how 1,25 NP modulates the phenotype of SMC by increasing the contractile SMCs while reducing synthetic SMCs is not well understood.
The vessel wall remodels to PTA by changes in fibrosis and deposition of extracellular matrix and thus reducing fibrosis and collagen formation after PTA and therefore this is a potential therapeutic strategy for improving outcomes. We assessed changes in fibrosis and collagen deposition by performing picrosirius red, Masson's trichrome, and collagen staining after 1,25 NP treatment compared to controls. We observed a reduction in Picrosirius red (staining of collagen 1 and 3), Masson's trichrome, and collagen 4 staining. These results imply that less constrictive remodeling is occurring after 1,25 NP treatment and this is allowing for improved positive vascular remodeling.
Because fibroblast to smooth muscle differentiation can cause to VS formation in murine AVF, we assessed the presence of fibroblasts using FSP-1 staining 16 . We observed a reduction in FSP-1 staining in 1,25 NP treated vessels compared to controls. This is consistent with previous study in murine AVFs using 1, 25 NP 7 . Several studies have demonstrated that reduced FSP-1 staining is associated with improved vascular remodeling in experimental AVFs 20 . Moreover, there was no difference in FAP-1 staining indicating 1,25 NP treatment had no effect on fibroblast activation.
It is well known that angioplasty can lead to vascular injury, which causes inflammatory cells including macrophages and monocytes to accumulate in the vessel wall 12,21 . We assessed for the presence of inflammatory cells by performing staining for CD68 in PTA treated vessels compared to contralateral control veins. There was an increase in CD68 (+) cells in PTA treated vessels. Studies have demonstrated that reducing inflammation after vascular injury caused by PTA or stent placement is associated with improved patency 22,23 . We next assessed the presence of inflammatory cells by performing staining for CD68 (+), F4/80 (+), CD45 (+) cells. We observed that there was a significant reduction in CD68 (+), F4/80 (+), CD45 (+), and iNOS (+) cells in 1,25 NP treated vessels compared to controls. These findings are consistent with prior studies that showed that reducing CD68 (+) cells after PTA was associated with improved vascular remodeling and lumen vessel area 24,25 .
Several studies have shown that Ier3 gene expression is involved with apoptosis 26 . We assessed for changes in apoptosis by performing TUNEL and cleaved caspase 3 staining after 1,25 NP treatment. We observed a significant increase in cells staining positive for TUNEL and cleaved caspase 3 staining. We further performed double staining to identify which cells were affected. Meanwhile there was increased TUNEL staining in both FSP-1 and CD68 (+) cells. Since we observed changes in apoptosis, we assessed proliferation by performing Ki-67 staining and observed it was significantly reduced in 1,25 NP treated vessels compared to vehicle controls. Next, we determined the phenotype of the cells undergoing proliferation and observed that proliferation was decreased in CD68, FSP-1, and SMCs in 1,25 NP treated vessels. We have previously reported that 1,25 NP decreases proliferation and increases cell death as assessed by TUNEL staining 7 . Collectively, these data imply that reduced proliferation with increase in apoptosis results in a decrease in cell density with a reduction in FSP-1 and CD68.
The potential changes in kidney and liver function after 1,25 NP delivery was evaluated by performing liver function test assessment measuring AST, ALT, total bilirubin and BUN and creatinine levels. Systemic administration of 1α,25(OH) 2 D 3 can cause hypercalcemia 27 . In the present study, there was no difference between all these serologic measures at day 3 and 21 after delivery in the 1,25 NP and vehicle control groups.
Adventitial delivery of 1,25 NP has several advantages. It allows for the delivery of drug to the vessel wall with increased concentration and without the wash out of the drug as would be seen with endothelial delivery using drug coated balloons or stents 28 . Adventitial delivery is superior to systemic delivery as it reduces non-target delivery of the drug during the first pass effect by the liver 28 . We have utilized this approach for delivering cells, lentiviral agents, and drugs to the vessel wall for reducing vascular injury and VS formation 7,13,16,29 .
Finally, an unbiased whole transcriptomic analysis using RNA sequencing with Panther analysis was performed to identify mechanisms of 1,25 NP treatment. There were 311 genes that were upregulated and 874 genes that were downregulated after 1,25 NP delivery compared to vehicle controls. Panther analysis demonstrated that inflammatory pathways were involved in the upregulated genes and apoptotic pathways were involved in the down-regulated genes.
There are several limitations of the present study. First, we used a single dose; and thus, the effects of higher dose or multiple doses should be evaluated in the future. Second, these results need to be corroborated by using a larger animal model, as the murine model may not be the same as the clinical scenario. Changes in WSS and PV can occur as result of blood flow and mechanical factors. Drug treatment using 1,25 NP likely did not have a direct effect on WSS and PV changes. Rather, the changes in WSS and PV reflect the effects of 1,25 NP on reducing VNH.
In conclusion, we demonstrate that adventitial delivery of 1,25 NP to the outflow vein after PTA in a murine model results in a decrease in IER3 gene and protein expression. This was localized to CD68 (+) and FSP-1 (+) cells which have decreased proliferation and increased apoptosis. This leads to a reduction in pro-inflammatory macrophages, venous neointimal hyperplasia, accompanied by a decrease in cell proliferation and venous fibrosis with positive vascular remodeling. The clinical importance of this study is that this that it provides rationale for local drug delivery using 1,25 NP for improving interventional outcomes in hemodialysis patients after PTA.

Materials and methods
Patient population and endovascular biopsy. Mayo Clinic Institutional Review Board (IRB18-009466) approval was obtained prior to conducting the present study. All experiments were performed in accordance with relevant guidelines and regulations. Informed written consent was obtained from all patients with dysfunctional AVFs requiring an interventional procedure prior to performing endovascular biopsy of the stenosis and non-stenotic veins. Only patients > 18 years old were enrolled. A total of 5 patients (1 Female) underwent an angiogram with PTA of an AVF stenosis. The patient demographics and location of stenosis are presented (Supplemental Table 1). An angiogram was performed to identify a stenosis (Fig. 1A). Once the stenosis was identified, a 5F glide catheter (Boston Scientific, Natick, MA) was advanced across the stenosis carefully using a 0.035″ glide wire (Terumo Medical Corporation, Tokyo, Japan). The glide wire was removed and a 0.035″ Rosen wire (Cook Inc., Bloomington, IN) was advanced and used to gently cross the stenosis multiple times (Fig. 1B). The tip of the wire was withdrawn into the Glide catheter and removed as a unit. The tip of the wire was cut and placed in an RNA free plastic tube with 1 ml of lactated Ringer's solution and placed on ice and saved for cell isolation and qRT-PCR analysis (see later). Figure 1C shows an angiogram prior to biopsy and Fig. 1D is after biopsy. A portion of non-stenotic vein from the fistula was sampled in an identical fashion to serve as control.
Endothelial cell sorting. The wire fragment was removed and centrifuged at 1000 RPM for 10 min at 4 °C.
The cell pellet was washed 3 times with 1 ml of PBS with 1% BSA. The pellet was resuspended in 100 µl of PBS with 1% BSA. CD31 (+) endothelial cells were sorted using magnetic beads. We added 10-ul of CD31 antibody (cat# 555444, BD Biosciences) to the cell suspension and it was incubated for 15 min at 4 °C. Next, anti-mouse IgG coupled magnetic beads (10ul, Cat# S1431S, New England Bio labs) were added to the suspension and CD31 (+) cells were sorted out. cDNA was synthesized directly using SuperScript III Cells Direct cDNA Synthesis Kit (cat# 18080200, Thermo Fischer).
Mice study. Approval was obtained from the Mayo Clinic Institutional Animal Care and Use Committee prior to performing any experiments. All experiments were performed in accordance with relevant guidelines and regulations. All the authors complied with the ARRIVE guidelines experiments. Eighteen C57BL/J6 male mice (Jackson Laboratories, Bar Harbor, ME) aged 6-8 weeks were housed at 12/12 h light/dark cycles, 22 °C, and 41% relative humidity with access to food and water ad libitum. Mice were randomly assigned to vehicle or 1,25 NP group (Supplementary Fig. 1). Mice were anesthetized using a combination of ketamine (100 mg/kg) and xylazine (10 mg/kg) administered by intraperitoneal injection prior to all procedures. Before surgery, one dose of buprenorphine-SR (0.05-0.1 mg/kg body sc) was administered for pain relief.

Encapsulation of 1α,25(OH) 2 D 3 in nanoparticles composed of PLGA and loaded in a Pluronic
F127 Hydrogel. 1α,25(OH) 2 D 3 was encapsulated into nanoparticles composed of PLGA using the interfacial method and loaded into Pluronic F127 hydrogel as described previously by our laboratory 7 . Briefly, 100 mg of PLGA (Sigma-Aldrich, St. Louis, MO) and 0.1 mg of 1α,25(OH) 2 D 3 (Tocris Bioscience, Bristol, UK) was dissolved in 10-mL of acetone. Then, the solution was added drop wise to 500 mL of deionized water at constant stirring. After particle formation, the organic solvent was evaporated and the PLGA nanoparticles were dried by lyophilization. The dried PLGA nanoparticles were dispersed in an aqueous solution to make a 200-µM solution. For vehicle controls, an equivalent amount of PLGA particles without 1α,25(OH) 2 D 3 was used. A 40% aqueous solution of Pluronic F-127 (Sigma-Aldrich, St. Louis, MO) was prepared under sterile conditions at 4 °C. An equal portion of PLGA particles in water and equal portion of 40% Pluronic F-127 gel was mixed to obtain a final concentration of 100 µM, 1α,25(OH) 2 D 3 -PLGA nanoparticles in 20% hydrogel. We have previously showed that by day 3 and 28 the cumulative releases of 1α,25(OH) 2 D 3 in our 1,25 NP drug was 40% and 80%, respectively 7 .
Surgical creation of murine model. Chronic kidney disease was created by surgical ligation of the arterial blood supply to the upper pole of the left kidney accompanied by removal of the right kidney. We have used and described this model previously 4,11,16,30 (Supplementary Figs. 1A,B). Four weeks later, an AVF was created by connecting the end of the right external jugular vein to the side of the left common carotid artery using an 11-0 nylon suture (Aros Surgical Instruments Corporation, Newport Beach, CA). Two weeks later, the diameter of the jugular vein was measured intraoperatively and PTA was performed using a 1.25 mm by 6-mm long balloon inflated to 14-atmospheres for 30-s (Medtronic Sprinter Legend, Minneapolis, MN) 4 . Next, 10-μL of either 1,25 NP in hydrogel or control vehicle (PLGA without 1,25 NP in hydrogel) was layered circumferentially onto the adventitia of the outflow vein for a length of 6-mm.
Blood urea nitrogen (BUN), creatinine (Cr), aspartate aminotransferase (AST), alanine aminotransferase (ALT), total bilirubin, and calcium assay. Blood was removed to determine the kidney function, liver function tests, and calcium concentration at sacrifice at day 3 and 21 after PTA as described elsewhere 31 . A QuantiChrom Urea Assay Kit (BioAssay Systems, Hayward, CA) and Mouse Creatinine Assay Kit (Crystal Chem, Elk Grove Village, IL) were used to determine the serum BUN and Cr, respectively. The Preventive Care Profile Plus rotor (Abaxis, Union City, CA) was used in a Vetscan VS2 machine (Abaxis, Union City, CA) to determine the AST, ALT, total bilirubin, and calcium levels 31 .
Doppler ultrasound (US) examination. As  Murine tissue collection and processing. At 3 days after PTA, mice were sacrificed for gene and RNA sequencing expression analysis, and the outflow vein samples were stored in RNA later solution (Qiagen, Hilden, Germany). Twenty-one days after PTA, we dissected the outflow veins and fixed the tissue in 10% formalin reagent (Fisher Scientific, Pittsburgh, PA) using a non-pressurized method of fixation. The vessels were analyzed for histomorphometric and immunobiological analyses. Each vessel was embedded in paraffin lengthwise as described previously. Typically, an average of 60-80 consecutive 4-μm sections were obtained per outflow vein per animal for analysis.
RNA isolation and cDNA synthesis. RNA was isolated using miRNeasy kit (Qiagen) from all mice tissue samples and cDNA was synthesized using iScript kit (BioRad, Hercules, CA) according to the manufacturer's instructions.
Quantitative real time polymerase chain reaction (qRT-PCR). The endothelial cells were isolated from each patient's sample and analyzed for gene expression of Ier3 using quantitative real time polymerase chain reaction (qRT-PCR) as described previously 20  Morphometric and image analysis. We analyzed 3-5 outflow veins per animal per group for histomorphometric analysis. Sections were stained for Hematoxylin and Eosin (H and E) to assess venous remodeling. Images were digitized to capture a minimum of 1936 × 1460 pixels covering one entire cross-section utilizing an M2 Microscope (Carl Zeiss) with an Axiocam 503 color camera (Carl Zeiss). These slides were analyzed using ZEN 2 blue edition version 2.0 (Carl Zeiss) as described elsewhere 4,11,16,30 . We measured the lumen vessel area, neointima, media, and adventitia areas, along with cell density in each of the layers.

Semiquantification of immunobiological stains and immunofluorescent stains. Sections that
were stained with immunobiological stains and immunofluorescent stains were digitized to capture a minimum www.nature.com/scientificreports/ of 1936 × 1460 pixels covering one entire cross-section utilizing an M2 Microscope (Carl Zeiss) with an Axiocam 503 color camera (Carl Zeiss). These slides were analyzed using ZEN 2 blue edition version 2.0 (Carl Zeiss) as described elsewhere 4,11,16,30 . Staining index was obtained by number of positive cells/ total number of cells × 100.

Statistical analysis.
Results are presented as the mean ± SEM. Statistical differences were tested by using either a one-way or two-way analysis of variance (ANOVA) followed by an unpaired Student t test with post hoc Bonferroni's correction. The level of significance was set at NS p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001 or # p < 0.0001. GraphPad Prism version 8 (GraphPad Software Inc., La Jolla, CA) was used for all statistical analysis.

Data availability
The RNA seq dataset generated during and analyzed during the current study will be made available on the GEO website at https ://www.ncbi.nlm.nih.gov/geo once the manuscript has been published. The remaining datasets generated during and/or analyzed during the current study are not publicly available due to because of the multiple different immunohistologic measurements, gene expression, serum, and histomorphometric analyses but are available from the corresponding author on reasonable request.