Effect of sex differences in treatment response to angioplasty in a murine arteriovenous fistula model

Abstract

Failure to mature and venous neointimal hyperplasia formation are the two major causes of hemodialysis arteriovenous fistula (AVF) vascular access failure. Percutaneous transluminal angioplasty (PTA) is the firstline treatment for both of these conditions, but, clinically, women have decreased patency rates compared with men. The hypothesis to be tested in the present study was that female mice after PTA of venous areas of higher intimal thickening have increased gene expression of transforming growth factor-β1 (TGF-β1) and TGF-β receptor 1 (TGFβ-R1) accompanied with histological changes of fibrosis compared with male mice. Seventeen male and eighteen female C57BL/6J mice were used in this study. Chronic kidney disease was induced by partial nephrectomy, and, 28 days later, an AVF was created to connect the left carotid artery to the right jugular vein. Two weeks later, the higher intimal thickening area was treated with PTA, and mice were euthanized 3 days later for gene expression analysis or 14 days later for histopathological analysis. Doppler ultrasound was performed weekly after AVF creation. At day 3, female AVF had significantly higher average gene expression of TGF-β1 and TGFβ-R1 compared with male AVF. At day 14, female outflow veins had a smaller venous diameter, lumen vessel area, decreased wall shear stress, lower average peak systolic velocity, and an increased neointima area-to-media area ratio. Moreover, female outflow veins showed a significant increase in α-smooth muscle actin and fibroblast-specific protein-1. There was a decrease in M1/M2 with an increase in CD68.

INTRODUCTION

In the United States, 124,111 new cases of end-stage renal disease were reported in 2015 and a total of nearly 500,000 patients received hemodialysis access maintenance treatment; ~45% were women (31). Women patients are less likely to use AVF for hemodialysis due to a slower maturation rate and more frequent salvage procedures (10, 30). Additionally, female sex is considered a risk factor in primary fistula failure (29). Percutaneous transluminal angioplasty (PTA) is the initial treatment for hemodialysis arteriovenous stenosis and failure to mature (35, 37). Overall, PTA has a poor patency rate, estimated to be ~42% at 1 yr after treatment (4). A higher percentage of women undergo single or repeat interventional salvage procedures compared with men (24). Moreover, hemodialysis access in women has a greater risk for loss of both AVF primary patency and postinterventional primary patency compared with men (36).

To understand the mechanisms responsible for venous stenosis formation after PTA, we recently created a novel murine model of venous restenosis in arteriovenous fistulas (AVFs) after CKD (7). There is little understanding of the molecular mechanisms responsible for restenosis formation after PTA procedures with respect to sex disparities. The purpose of the present study was to determine gene expression, hemodynamic, and histological changes after PTA in male and female animals using the murine PTA model of venous neointimal hyperplasia.

MATERIALS AND METHODS

Experimental animals.

Mayo Clinic Institutional Animal Care and Use committee approval was obtained before performance of any of the experiments. C57BL/6J male (n = 17) and female (n = 18) mice (age: 6–8 wk, Jackson Laboratories, Bar Harbor, ME) were housed with 12:12-h light-dark cycles and at 22°C and 41% relative humidity with access to food and water ad libitum. Mice were anesthetized using a combination of ketamine (120 mg/kg) and xylazine (10 mg/kg ip) before all procedures and maintained using ketamine (40 mg/kg) and xylazine (3 mg/kg). Before surgeries, one dose of buprenorphine (0.05–0.1 mg/kg body wt sq) was administered for pain relief.

Mouse surgeries.

To mimic clinical condition of PTA treatment in hemodialysis accesses, each animal underwent three surgeries, including partial nephrectomy for CKD, AVF creation, and a PTA procedure (Fig. 1A). A CKD model was created by surgical ligation of the arterial blood supply to the left renal upper pole accompanied by removal of the right kidney as previously described elsewhere (27). The AVF was created 28 days after partial nephrectomy by anastomosing end of the right external jugular vein to side of the left common carotid artery, as previously described (7). Two weeks later, venous neointimal hyperplasia was confirmed (see Fig. 7), and the murine PTA procedure was performed using a 1.25 × 6-mm coronary balloon catheter (Medtronic Sprinter Legend, Minneapolis, MN) to treat the outflow venous areas of higher intimal thickening as previously described elsewhere (7).

Fig. 1.

Murine outcomes after surgeries. A and B: schema of female and male subgroups that underwent partial nephrectomy at day −42, arteriovenous fistula (AVF) creation at day −14, and percutaneous transluminal angioplasty (PTA) surgery at day 0. Twelve mice were used in the PCR (day 3) group, including six female and six male mice. Ten mice were used in the pre-PTA (day 0) group for baseline analysis, including five female and five male mice. Thirteen mice were used in the immunohistochemical (IHC; day 14) group, including seven female and six male mice. C: better patency was observed in male vessels (P = 0.05). D: male mice were always heavier than female mice at each time point (#P < 0.0001). E: after partial nephrectomy, both male and female mice had increased average blood urea nitrogen (BUN) and creatinine (Cr) without sex diversity. A Mantel-Cox test was performed for the results shown in C. Two-way ANOVA with Bonferroni's correction was performed for the results shown in D and E.

Fig. 7.

Morphometry result after 2 wk of arteriovenous fistula (AVF) creation. A: representative slides for female and male outflow veins at 2 wk post-AVF surgery. B and C: there were no significant differences for average lumen area and neointimal area-to-media area ratio (N/M) between female and male AVFs, respectively. L, lumen. The black dashed line indicates the neointimal area. Scale bars = 50 μm. NS, not significant.

Murine blood urea nitrogen and creatinine assay.

At day −42 (before CKD surgery) and day −11 (3 days post-AVF surgery), 0.2-mL blood specimens were collected from each mouse for kidney function examination and another 0.5-mL volume was extracted at euthanization. With the use of an Abaxis vet scan VS2 machine (Abaxis, Union city, CA), serum blood urea nitrogen (BUN) and creatinine (Cr) were determined by the QuantiChrom Urea Assay Kit (BioAssay Systems, Hayward, CA) and the Mouse Creatinine Assay Kit (Crystal Chem, Elk Grove Village, IL), respectively.

Doppler ultrasound examination.

After animals had been weighed and anesthetized, Doppler ultrasound measurements were performed at the time of AVF creation, pre- and post-PTA, and at time of euthanization (Fig. 1A). A Doppler Flow Velocity System (INDUS Instruments, Houston, TX) equipped with a high-frequency 20-MHz transducer probe was used to evaluate AVF patency and blood flow velocity. Wall shear stress (WSS) at different time points was calculated using the velocity and intraoperative vessel diameter according to the following equation: WSS = 4ηV/r, where η is blood viscosity, V is flow velocity (in cm/s), and r is the radius (in cm). Blood is assumed to behave as a Newtonian fluid and blood viscosity is assumed to be constant at 0.035 Poise (17).

Murine tissue collection and processing.

At day 3, AVF fistula tissues were removed for RT-PCR gene expression as described later. At day 14, AVFs were dissected and fixed in 10% formalin reagent (Fisher Scientific, Pittsburgh, PA) for histomorphometric analysis. Each outflow vein was embedded lengthwise in paraffin and cut into 4-μm-thick sections as previously described (40).

Real-time PCR.

In the PCR group, female and male outflow veins and contralateral veins were collected for gene expression assays, determined by RT-PCR as previously described (14). The designed primer information is shown in Table 1. cDNA was synthesized using the iScript kit (Bio-Rad, Hercules, CA) according to the manufacturer’s instructions. RT-PCRs were performed using iTaq Universal SYBR Green Master Mix (Bio-Rad) in a C1000 thermal cycler equipped with a CFX96 Real Time System (Bio-Rad), and cycle quantitication (cq) values were measured by Bio-Rad CFX Manager software (Bio-Rad). ∆cq values of female and male outflow veins and contralateral veins were normalized with TATA-binding protein (TBP)-1, and the fold change in gene expression was calculated following the $2^{-(\Delta \Delta {\text{C}}_{\text{T}})}$ method (where C_T is threshold cycle).

Table 1.

Quantitative RT-PCR primers for genes

Gene	Forward	Reverse
TGF-β1	5′-CGAAGCGGACTACTATGCTAAA-3′	5′-TCCCGAATGTCTGACGTATTG-3′
TGFβ-R1	5′-TTGCTCCAAACCACAGAGTAG-3′	5′-ACACTAAGCCCATTGCATAGAT-3′
TBP-1	5′-AAGGGAGAATCATGGACCAG-3′	5′-CCGTAAGGCATCATTGGACT-3′

TGF-β1, transforming growth factor-β1; TGFβ-R1, TGF-β receptor 1; TBP-1, TATA-binding protein 1.

Immunohistochemical staining.

Staining was performed on paraffin-embedded sections from female and male PTA-treated vessels using the EnVision (Dako, Carpinteria, CA) method with a heat-induced antigen-retrieval step (39). The antibodies used and their sources are shown in Table 2. Peroxidase activity was visualized using 3,3′-diaminobenzidine as a chromogen. Slides were stained for normal IgG corresponding to the primary antibody species to serve as negative controls.

Table 2.

Antibodies used in the present study

Antibody	Host	Catalog Number	Provider	Dilution
IgG	Rabbit	sc-2027	Santa Cruz Biotechnology
CD31	Rabbit	ab29364-100	Abcam	1:200
CD68	Rabbit	ab125212	Abcam	1:2,000
CD68	Mouse	ab955	Abcam	1:2,000
α-SMA	Mouse	ab7817	Abcam	1:400
α-SMA	Rabbit	ab5694	Abcam	1:1,000
MYH11	Rabbit	ab53219	Abcam	1:1,000
FSP-1	Rabbit	07-2274	EMD Millipore	1:1,000
FSP-1	Mouse	188-11191	RayBiotech	1:500
Collagen type IV	Rabbit	600-401-106	Rockland	1:2,000
Arg-1	Rabbit	NBP1-32731	Novus Biologicals	1:1,500
iNOS	Rabbit	NB300-605	Novus Biologicals	1:2,000
pSMAD3	Rabbit	ab52903	Abcam	1:200
HIF-1α	Rabbit	ab2185	Abcam	1:800
TGF-β1	Rabbit	Sc-146	Santa Cruz Biotechnology	1:300
Alexa Fluor 488	Donkey	711-545-152	Jackson Immunoresearch	1:1,000
Alexa Fluor 594	Donkey	715-585-151	Jackson Immunoresearch	1:1,000
Alexa Fluor 647	Goat	A32728	Invitrogen	1:1,000

α-SMA, α-smooth muscle actin; MYH11, myosin heavy chain 11: FSP-1, fibroblast-specific protein-1; Arg-1, arginine-1; iNOS, inducible nitric oxide synthase; pSMAD3, phosphorylated SMAD3; HIF-1α, hypoxia-inducible factor-1α; TGF-β1, transforming growth factor-β1.

Immunofluorescence staining.

Immunofluorescence staining was performed in paraffin-embedded sections following the same antigen-retrieval step as described above. The primary and second antibodies and their sources are shown in Table 2. Prolong gold anti-fade reagent with DAPI (Invitrogen, Eugene, OR) was used for nuclear staining and mounting for immunofluorescence staining. All images were captured using an Axio Imager M2 microscope (Carl Zeiss, Jena, Germany).

Masson’s trichrome staining.

Masson’s trichrome (Richard-Allan Scientific, Kalamazoo, MI) was performed on PTA-treated female and male outflow veins according to the manufacturer’s directions to evaluate vascular fibrosis.

Morphometric and image analysis.

PTA-treated vessels were stained for hematoxylin and eosin (H&E) and viewed with a Zeiss Imager. Images were digitized to capture a minimum of 1,936 × 1,460 pixels covering one entire cross section from each female or male vessel using an M2 Microscope (Carl Zeiss) equipped with an Axiocam 503 color camera (Carl Zeiss) and analyzed by ZEN 2.3-blue edition (Carl Zeiss). H&E staining was highlighted by selecting the appropriate red-green-blue color intensity range and then counted as previously described elsewhere (5).

Statistical analysis.

Results are expressed as means ± SD. Statistical differences were tested by either one- or two-way ANOVA followed by a post hoc Bonferroni's correction or Student’s t test. The levels of significance were set at P < 0.05, P < 0.01, P < 0.001, or P < 0.0001. Graph Pad Prism (version 8, GraphPad Software, La Jolla, CA) was used for all statistical analyses.

RESULTS

Animal outcomes.

Thirty-five C57BL/6J mice (female: n = 18 and male: n = 17) comprised the terminal study (Fig. 1B). The body weight of animals during this study was determined weekly, and at each time point, there was a significant difference in the average weight of female and male animals over the study (P < 0.0001; Fig. 1D). Outflow venous patency was assessed using Doppler ultrasound, and better patency was observed in male mice post-PTA treatment (P = 0.05; Fig. 1C). BUN and Cr assay showed significant renal dysfunction after partial nephrectomy for female and male mice; nevertheless, there was no difference for sex divergence (Fig. 1E).

Increase in the average diameter and outward remodeling of male AVFs compared with female AVFs.

Sequential changes in the diameter of outflow veins were measured intraoperatively after AVF placement, at pre- and post-PTA, and at euthanization (Fig. 2E). The average diameter of outflow veins immediately after AVF creation was larger in male than female mice (average increase: 20%, P > 0.05), and, 14 days later, it increased in male mice by 46% to 1.05 ± 0.08 mm. In female mice, the average diameter increased by 60%, to 0.96 ± 0.14 mm. Three days and fourteen days after PTA, the male average diameter was larger than female average diameter (day 3: P < 0.05; day 14: P < 0.01). At day 14, the average vessel lumen area of PTA vessels was significantly higher in male mice compared with female mice (average increase: 252%, P < 0.05; Fig. 2D). The average ratio of the neointima area to media area significantly decreased in the male PTA group compared with the female PTA group (average decrease: 36%, P < 0.05; Fig. 2G).

Fig. 2.

Sexual diversity in hemodynamic and morphometric analysis at day 14. A: representative images showing that there was no significant sexual difference in outflow vein after arteriovenous fistula (AVF) creation. B: representative sections at ×40 magnification after hematoxylin and eosin staining at day 14. C: by day 14, there was a significant decrease in the average peak velocity of female vessels (*P < 0.05). D: there was a significant decrease in the average lumen vessel area of female percutaneous transluminal angioplasty (PTA)-treated vessels (*P < 0.05). E: post-PTA procedure, there was a significant decrease in female intraoperative average outflow vein diameter (day 3, *P < 0.05; day 14, **P < 0.01). F: by day 14, the average wall shear stress (WSS) of female outflow veins decreased significantly (*P < 0.05). G: there was a significant increase in the neointimal area-to-media area (N/M) ratio of female outflow veins compared with male outflow veins (*P < 0.05). Each bar represents the mean ± SD of 6–7 animals. L, lumen. The black dashed line indicates the neointimal area. Scale bars = 1 mm in A and 50 μm in B.

Increase in the average peak velocity of PTA-treated vessels from male compared with female mice.

The peak velocity of outflow veins was determined weekly using Doppler ultrasound after AVF creation, at pre-PTA and post-PTA, and at euthanization (Fig. 2C). The average peak velocity of the outflow veins at AVF creation in male mice was higher than in female mice (P > 0.05), and 7 days later (day −7), in male mice it increased by 12% to 158.84 ± 24.98 cm/s and in female mice it increased by 33% to 140.79 ± 33.20 cm/s compared with AVF placement. After PTA, the average peak velocity increased in male mice to 163.47 ± 49.80 cm/s and in female mice it was 152.01 ± 43.68 cm/s. By day 14 after PTA, male average peak velocity was significantly higher than female peak velocity (P < 0.05).

Increase in average WSS of male compared with female PTA-treated vessels.

It is hypothesized that WSS is responsible for venous neointimal formation after AVF placement (25, 26). However, it has not been evaluated in male and female experimental animals longitudinally after PTA of venous neointimal hyperplasia. The changes in average WSS of outflow veins are shown in Fig. 2F. The average WSS of outflow veins at AVF creation in male mice was similar to female mice (P > 0.05), and, 14 days later, in male mice it decreased to 406.00 ± 124.41 dyn/cm² and in female mice to 359.01 ± 108.95 dyn/cm². Fourteen days after PTA, there was a significant difference between male and female mice (P < 0.05). Female AVFs had lower average WSS at day 14 compared with male AVFs.

Upregulated transforming growth factor-β1 and transforming growth factor-β receptor 1 after PTA treatment in female AVFs.

Gene expression of transforming growth factor-β1 (TGF-β1) and TGF-β receptor 1 (TGFβ-R1) were assessed using quantitative RT-PCR in female and male vessels at day 3 (Fig. 3). First, there was no significant difference in TGF-β1 and TGFβ-R1 gene expressions between female and male contralateral veins (data not shown). At day 3 post-PTA procedure, average gene expressions of TGF-β1 and TGFβ-R1 in male AVFs increased by 165.86 ± 59.36-fold and 66.55 ± 19.09-fold compared with male contralateral veins, respectively. Average gene expressions of TGF-β1 and TGFβ-R1 were significantly higher in female outflow veins than male outflow veins (TGF-β1: average increase: 96%, P < 0.001, Fig. 3A; TGFβ-R1 average increase: 130%, P < 0.001, Fig. 3B).

Fig. 3.

Diverse gene expressions of female and male arteriovenous fistulas after percutaneous transluminal angioplasty treatment at day 3. A and B: transforming growth factor (TGF)-β1 (A) and TGF-β receptor 1 (TGFβ-R1; B) were increased by 96% and 130% in female outflow veins compared with male outflow veins, respectively. Fold changes of arteriovenous fistula gene expression were normalized to contralateral veins (CVs). Each data point represents the mean ± SD of 6 animals. ***P < 0.001.

Reduction in staining of α-smooth muscle actin, fibroblast-specific protein-1, CD68, and CD31 but increased myosin heavy chain 11 and arginase 1 staining in PTA-treated male vessels compared with female vessels.

Increased expression of fibroblast-specific protein (FSP)-1, α-smooth muscle actin (α-SMA), and CD68 staining has been identified in failed AVF specimens and experimental male PTA animal models (6, 7). Histopathological comparison between female and male PTA-treated outflow veins has not been determined. α-SMA⁺ cells were mainly localized in the neointima of female and male outflow veins (Fig. 4, A and B). There was a significant reduction in the average α-SMA index of male PTA-treated vessels compared with female vessels (average decrease: 58%, P < 0.01; Fig. 4E). The majority of FSP-1⁺ cells were observed in the adventitia of male outflow veins, but in female PTA-treated vessels, they were located throughout the vessel wall (Fig. 4, A and B). There was a significant reduction in the average FSP-1 index of male outflow veins compared with female outflow veins (average decrease: 63%, P < 0.05; Fig. 4F). Myosin heavy chain 11 (MYH11)⁺ cells were localized in the neointima of female vessels, whereas in male vessels, they were located in the neointima and media (Fig. 4, A and B). There was a significant increase in the average MYH11 index of male compared with female PTA-treated vessels (average increase: 141%, P < 0.01; Fig. 4G). More α-SMA and FSP-1 but less MYH11 staining on female paraffin sections was confirmed by costaining of α-SMA-MYH11 and α-SMA-FSP-1 (Fig. 4, C and D).

Fig. 4.

Immunohistochemical and immunofluorescent staining for α-smooth muscle actin (α-SMA), fibroblast-specific protein (FSP)-1, and myosin heavy chain 11 (MYH11). A and B: representative sections of female (A) and male (B) negative control, α-SMA, FSP-1, and MYH11 staining. C: costaining of female α-SMA-MYH11 and α-SMA-FSP-1 illustrating that the MYH11⁺ area was included within α-SMA⁺ staining. D: male costaining demonstrated that MYH11 staining overlapped with α-SMA staining and FSP-1⁺ staining clustered in the male outflow vein. E−G: semiquantitative analysis showed significant increase in the female α-SMA (**P < 0.01) and FSP-1 (*P < 0.05) index but a decrease in MYH11 (**P < 0.01), respectively. Each bar represents the mean ± SD of 6–7 animals. ADV, adventitia; L, lumen. The black arrow indicates positive staining; the red dashed line indicates areas of MYH11 positive staining. Scale bars = 50 μm.

Next, CD31 staining was performed to evaluate reendothelialization and angiogenesis after PTA treatment. There were more CD31⁺ cells forming microvessels localized to the adventitia and media in female compared with male PTA-treated vessels (Fig. 5, A and B). Although the PTA procedure denuded the endothelium, intact endothelium with continuous CD31⁺ cells was observed in both male and female vessels. Semiquantitative analysis demonstrated that there was a significant increase in the average CD31 index of female PTA-treated vessels (average increase: 94%, P < 0.05; Fig. 5C).

Fig. 5.

Immunohistochemical staining of CD31, hypoxia-inducible factor (HIF)-1α, CD68, inducible nitric oxide synthase (iNOS), and arginase 1 (Arg-1) at day 14. A and B: representative female (A) and male (B) sections of CD31, HIF-1α, CD68, iNOS, and Arg-1 staining. More microvessels (red arrow) with CD31⁺ cells were localized to the female adventitia and media areas. C and E: there were significant increases in CD31 index (*P < 0.05) and CD68 index (*P < 0.05) of female outflow veins, respectively. D and G: significant reductions of female HIF-1α index (**P < 0.01) and Arg-1 index (**P < 0.01) were observed, respectively. Each bar represents the mean ± SD of 6–7 animals. ADV, adventitia; L, lumen. The black arrow indicates positive staining; the red arrow indicates microvessel. Scale bars = 50 μm.

Finally, CD68⁺ cells were observed in the adventitia and media of female vessels, whereas in male PTA-treated vessels, they were localized in the adventitia (Fig. 5, A and B). There was a significant reduction in the average CD68 index of male outflow veins compared with female outflow veins 14 days after the PTA procedures (average decrease: 53%, P < 0.05; Fig. 5E). Arginase 1 (Arg-1)⁺ cells exert anti-inflammation function, and inducible nitric oxide synthase-positive cells indicate proinflammation (6). Although there was no significant difference for inducible nitric oxide synthase staining, a significant increase in Arg-1 staining was observed in male PTA-treated vessels (average increase: 103%, P < 0.01; Fig. 5G).

Increase in the average staining linked to fibrosis in male AVFs compared with female AVFs.

Masson’s trichrome staining was performed to assess collagen density (Fig. 6, A and B). Semiquantitative analysis demonstrated a significant increase in the average intensity of Masson’s trichrome staining in PTA-treated male vessels compared with female vessels (P < 0.05; Fig. 6C). Additionally, staining for collagen type I and type IV was performed. Although there was no significant sex difference for collagen type I staining (data not shown), there was a significant decrease in average collagen type IV staining in female compared with male outflow veins post-PTA procedure (P < 0.05; Fig. 6D). TGF-β1/phosphorylated (p)SMAD3 staining was performed because of the increased gene expression of TGF-β1 and its link to fibrosis (23). Higher pSMAD3 staining was observed in male compared with female outflow veins (P < 0.01; Fig. 6E). At day 14 there was no significant difference in TGF-β1 staining between male and female outflow veins (data not shown). Histomorphometric analysis on paraffin-embedded sections from PTA-treated male and female animals was performed. At day 0 before the PTA procedure, there were no significant differences for average lumen area and neointima area/media area (Fig. 7).

Fig. 6.

Higher fibrosis after percutaneous transluminal angioplasty (PTA) procedure in male outflow veins from Masson’s trichrome, phosphorylated (p)SMAD3, and collagen type IV staining. A and B: representative sections of Masson’s trichrome, collagen type IV, and pSMAD3 for female (A) and male (B) outflow veins. C−E: there were significant increases in Masson’s trichrome index (*P < 0.05), collagen type IV index (*P < 0.05), and pSMAD3 index (**P < 0.01) from male outflow veins compared with female outflow veins, respectively. Each bar represents the mean ± SD of 6–7 animals. L, lumen. The black arrow indicates positive staining; the red dashed line indicates positive stained area. Scale bars = 50 μm.

Increased hypoxia-inducible factor-1α staining of male AVFs compared with female AVFs.

Hypoxia-inducible factor (HIF)-1α staining was performed to determine whether there was a difference in hypoxic injury in male and female AVFs after PTA treatment. In female AVFs, HIF-1α⁺ staining was present throughout the vessel wall (Fig. 5A), and in male AVFs, positive staining was observed primarily in the media (Fig. 5B). Semiquantitative analysis showed a significant increase in the average HIF-1α index of male AVFs compared with female AVFs (P < 0.01; Fig. 5D).

DISCUSSION

Angioplasty reduces vascular stenosis but causes vascular injury. It is known that angioplasty results with respect to primary patency are different among men and women patients with stenotic hemodialysis AVF (36). In the present study, we determined the gene expression, hemodynamics, WSS, and histomorphometric analysis in male and female mice with CKD with areas of higher intimal thickening AVF treated with PTA. We observed that female vessels after angioplasty had increased TGF-β1 and TGFβ-R1 gene expression accompanied with a smaller venous diameter, decreased WSS, and lower average peak systolic velocity with decreased lumen vessel area and an increased neointima-to-media ratio than male vessels. Moreover, female AVFs showed a significant increase in positive staining for α-SMA, FSP-1, CD31, and CD68 accompanied by reductions in MYH11, Arg-1, pSMAD3, and venous fibrotic staining.

TGF-β signaling has been demonstrated to have a core role in vascular remodeling of AVF (12, 34). Rare findings regarding fistula angioplasty treatment linked to the female sex have been reported. In our previous work (7), stronger picrosirius red staining was observed in male PTA-treated AVFs compared with male sham AVFs, and this implied that fibrosis was linked to vascular remodeling after the angioplasty procedure. Our hypothesis is that, after angioplasty procedures, excess collagen deposition may form an extra scaffold to circumvent constrictive shrinking in target vessels. In one scenario, active TGF-β signals through a canonical pathway mediated by SMAD1 to SMAD8 (21). Additionally, epithelial-to-mesenchymal transition (EMT) is a fundamental pathophysiological mechanism for the formation of neointimal hyperplasia associated with AVF and PTA restenosis (3, 38). TGF-β has been shown to induce the EMT process (33) accompanying increased higher expressions of α-SMA and FSP-1. TGF-β signaling promotes a rapid increase in phenotypic changes to biosynthesis from contractile protein in vascular smooth muscle cells confirmed in the rat balloon-injured artery model (9, 15). Moreover, TGF-β signaling stimulates angiogenesis (19, 20). Our murine PTA model had similar results showing increased α-SMA⁺, CD31⁺, and FSP-1⁺ but decreased MYH11⁺ and pSMAD3 cellular indexes at day 14 compared with male outflow veins. In the vascular system, TGF-β1 can be produced by endothelial cells, vascular smooth muscle cells, fibroblasts, and macrophages (11, 34). This is consistent with our costaining results. As shown in Fig. 8, TGF-β1⁺/FSP-1⁺, TGF-β1⁺/α-SMA⁺, and TGF-β1⁺/CD68⁺ cells were observed in PTA-treated outflow veins at day 3.

Fig. 8.

Transforming growth factor (TGF)-β1 costaining at day 3. A−C: TGF-β1⁺ staining was observed throughout the whole vessel wall. A: at day 3 after angioplasty, α-smooth muscle actin (α-SMA)⁺ staining was partially observed in the neointima with positive staining for TGF-β1. B and C: positive TGF-β1 staining was colocalized with fibroblast-specific protein (FSP)-1 and CD68, respectively. L, lumen. The white arrow indicates positive costaining. Scale bars = 50 μm.

Our PCR data showed that after the PTA procedure, female AVFs had higher TGF-β gene expression. Our hypothesis is that, in female PTA-treated AVFs, increased TGF-β expression induces a higher degree of the EMT process and increased CD68⁺ cells but decreased Arg-1⁺ cells, leading to inward vascular remodeling compared with male AVFs. In clinical research, increased serum TGF-β1 was correlated with symptomatic stenotic coronary artery disease (8). Moreover, increased TGF-β1 expression in human tissues has been linked to restenosis lesions (28). One interesting human wound healing research study demonstrated that a lower estrogen level was associated with improved quality of scarring healing (1). These clues imply that estrogen may be responsible in TGF-β1 signaling and sex differences.

There are some limitations in this study. First, female outflow vessels were smaller than male outflow vessels after AVF creation. According to the instructions for use of the Medtronic angioplasty catheter, the lesion entry profile is 0.41 mm, and at 14 atmospheres the balloon diameter is 1.35 mm. Even though angioplasty was performed following the same standard techniques using 14 atmospheres for 30 s, perhaps more vascular injury was induced for female outflow veins. Second, in female AVFs, there was increased TGF-β1 and TGFβ-R1 gene expression at day 3 and decreased fibrosis at day 14. SMAD3-null mice presented with constrictive remodeling with decreased extracellular matrix deposition (16). Furthermore, estrogen is capable of suppressing TGF-β signaling by inhibiting SMAD3 (13, 22). We hypothesize that there is cross talk between TGF-β and estrogen signaling. Upregulation of pSMAD3 after PTA has been shown to result in adaptive remodeling (18). Further mechanisms of estrogen on TGF-β1/pSMAD3 signaling and their implications on fistula fibrosis after PTA treatment need to be defined. Third, lower patency of female AVF was determined by Doppler ultrasound detection. At day 14 we included all patent and failed AVFs, and this may induce bias in this experiment. Finally, studies with animals with ovariectomy and estrogen replacement would need to be further performed to illustrate the hormone function in this interventional diversity.

In summary, the present study demonstrated that male and female mice have different hemodynamic, gene expression, and histological changes in response to PTA treatment. Sex disparities in murine interventional outcomes are not only due to vascular size but also due to gene changes related to inflammation, fibrosis, and cellular differentiation after a PTA procedure, as shown in Fig. 9. In the future, the sex variable should be considered in experimental and clinical trials for PTA treatment in hemodialysis AVFs.

Fig. 9.

Conclusion of sexual differences for female and male percutaneous transluminal angioplasty (PTA)-treated vessels. TGF-β1, transforming growth factor-β1; TGFβ-R1, TGF-β receptor 1; α-SMA, α-smooth muscle actin; FSP-1, fibroblast-specific protein-1; EMT, epithelial-to-mesenchymal transition.

GRANTS

This work was supported by National Institutes of Health Grants HL-098967 and DK-107870 (to S. Misra).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.C., C.Z., S.K., A.S., A.K.S., M.L.S., A.M., Y.L., and S.M. conceived and designed research; C.C., C.Z., and S.K. performed experiments; C.C., C.Z., S.K., A.S., A.K.S., M.L.S., A.M., Y.L., and S.M. analyzed data; C.C., C.Z., S.K., A.S., A.K.S., M.L.S., A.M., Y.L., and S.M. interpreted results of experiments; C.C., C.Z., and S.M. prepared figures; C.C., C.Z., S.K., A.S., A.K.S., M.L.S., A.M., Y.L., and S.M. drafted manuscript; C.C., C.Z., S.K., A.S., A.K.S., M.L.S., A.M., Y.L., and S.M. edited and revised manuscript; C.C., C.Z., S.K., A.S., A.K.S., M.L.S., A.M., Y.L., and S.M. approved final version of manuscript.