Proprotein Convertase Subtilisin/Kexin Type 9 Inhibitor Improves the Vascular Function of Arteriovenous Fistula in Rats with Hyperglycemia

Abstract

Objectives

Few evidence-based medications to improve the primary patency of arteriovenous fistulas in patients with diabetes who require hemodialysis are available. We investigated whether proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i) could improve arteriovenous fistula function through pleiotropic effects in a rat model of hyperglycemia.

Methods

Ex vivo effects of PCSK9i on the aorta of Sprague-Dawley (SD) rats were investigated using an organ bath system. For in vivo experiments, an abdominal aortocaval (AC) fistula was generated in SD rats (200-250 g) after inducing hyperglycemia through streptozotocin administration (80 mg/kg, intraperitoneal). Alirocumab (50 mg/kg/week, subcutaneous) was administered on the day of fistula surgery and day 7. Echocardiography, blood flow through the aorta-limb, vasomotor reactivity, and serum biochemistry were examined on D14. Furthermore, enzyme-linked immunosorbent assay and immunoblotting were performed.

Results

PCSK9i induced aorta relaxation ex vivo through a potassium channel-associated mechanism. PCSK9i significantly improved blood flow and preserved endothelial function without changes in cardiac function and serum lipid levels in rats with hyperglycemia. The levels of lectin-like oxidized low-density lipoprotein receptor-1, superoxide dismutase, cyclooxygenase-2, caspase-1, and interleukin-1β were significantly reduced in the treatment group. PCSK9i decreased the ratio of phosphorylated to total p38 mitogen-activated protein kinase and extracellular signal-regulated kinase in the aorta of rats with hyperglycemia.

Conclusions

Short-term treatment with PCSK9i preserved endothelial function, induced vascular dilatation, and increased blood flow in the AC fistula of rats with hyperglycemia. The pleiotropic mechanisms were associated with the suppression of oxidative stress and tissue inflammation during hyperglycemia.

Abbreviations

AC, Aortocaval

ANOVA, One-way analysis of variance

AVF, Arteriovenous fistula

cGMP, Cyclic guanosine monophosphate

COX-2, Cyclooxygenase-2

D, Day

ECs, Endothelial cells

ELISA, Enzyme-linked immunosorbent assay

eNOS, Endothelial NO synthase

ERK, Extracellular signal-regulated kinase

HDL-c, High-density lipoprotein cholesterol

HG, Hyperglycemia

HO-1, Hem-oxygenase

IL, Interleukin

IVC, Inferior vena cava

LDL-c, Circulating LDL cholesterol

LDLR, Membrane-bound receptors for circulating LDL cholesterol

LOX-1, Lectin-like oxidized LDL receptor

LRR, Leucine-rich repeat

MAPK, p38 mitogen-activated protein kinases

MCP-1, Monocyte chemoattractant protein-1

MPO, Myeloperoxidase

NF-κB, Nuclear factor κ-light-chain-enhancer of activated B cells

NLRP3, Pyrin domain-containing protein-3

NO, Nitric oxide

NOD, Nucleotide oligomerization domain

oxLDL, Oxidation of low-density lipoprotein

PCSK9, Proprotein convertase subtilisin/kexin type 9

PCSK9i, PCSK9 inhibitors

PE, Phenylephrine

ROS, Reactive oxygen species

SD, Sprague-Dawley

SOD, Superoxide dismutase

VSMC, Vascular smooth muscle cells

INTRODUCTION

Arteriovenous fistula (AVF) is a common vascular access for hemodialysis in patients with end-stage renal disease. The generation and maintenance of AVF are closely associated with the quality of life in these patients. The non-maturation rate of AVF ranges from 20-60% within six months after surgery.^1-3 Over 10% of AVFs fail to maintain primary patency after one year of maturation.¹ As a leading cause of chronic kidney disease, diabetes is also a predictor of AVF maturation or re-intervention.^3-7 The underlying reasons for AVF malfunctioning include oxidative stress and the activation of proinflammatory cytokines under hyperglycemic conditions.^8,9 The overproduction of reactive oxygen species (ROS) leads to the uncoupling of endothelial nitric oxide (NO) synthase, thereby causing endothelial dysfunction. Therefore, the remodeling of AVF becomes incompetent, and blood flow through the fistula is reduced.¹⁰ Inflow of AVF from the arterial limb plays important roles in fistula maturation and maintaining adequate fistula function.⁶ Improving endothelial function and the preservation of adequate arterial blood flow are critical for a successful AVF and improving the quality of life of patients under dialysis, particularly those with diabetes. However, interventions to facilitate or improve the primary patency of AVF are lacking, as indicated in The National Kidney Foundation’s Kidney Disease Outcomes Quality Initiative Guidelines.³

The oxidation of low-density lipoprotein (oxLDL) generates oxidative stress under hyperglycemic conditions.^9,11 The cellular uptake of circulating LDL cholesterol (LDL-c) through membrane-bound receptors (LDLR) helps reduce circulating LDL-c and, thereby, oxLDL. However, the abundance of LDLR is limited by serine protease proprotein convertase subtilisin/kexin type 9 (PCSK9).¹² PCSK9 is associated with the induction of inflammation, mobilization of ROS, and apoptosis of endothelial cells (ECs)^12,13 through the activation of lectin-like oxidized LDL receptor (LOX)-1, an important scavenger receptor for the cellular uptake of oxLDL.^13,14 LOX-1 stimulates PCSK9 expression in ECs and vascular smooth muscle cells (VSMCs) by activating the nuclear factor κ-light-chain-enhancer of activated B cells (NF-κB) pathway.¹⁵ Taken together, PCSK9 may be an important mediator of fistula dysfunction in patients with diabetes.

Previous animal and in vitro studies have shown that PCSK9 inhibition using monoclonal antibodies such as alirocumab and evolocumab protects vascular endothelium through anti-inflammatory and anti-oxidative stress mechanisms.^16,17 However, the direct action of PCSK9 inhibitors (PCSK9i) on vascular function is largely unexplored, and the pleiotropic effects of PCSK9i on AVF in patients with diabetes require further investigation. This study aimed to investigate whether PCSK9i treatment preserves vascular function in arterial sites of an aortocaval (AC) fistula rat model under short-term hyperglycemic conditions.

MATERIALS AND METHODS

Induction of hyperglycemia in rats

This study was approved by the Animal Care and Use Committee of National Cheng Kung University, Taiwan (IACUC Approval No.: 111027). Age-matched (6-8-week-old) male Sprague-Dawley (SD) rats (weighing 200-250 g) were maintained in an animal house with 13-h light and 11-h dark cycles. They were fed a standard chow diet and provided with water ad libitum. All procedures were performed in accordance with the IACUC. Hyperglycemia (HG, or type I diabetes) was induced in the rats using streptozotocin (80 mg/kg, intraperitoneal) as previously described (Supplementary Materials).^6,18

Vascular effects of PCSK9i ex vivo

We tested the ex vivo vascular effects of alirocumab (Sanofi Winthrop Industry, France), a monoclonal antibody against PCSK9, as a preliminary analysis. Normal or HG rats without AC fistula were euthanized to obtain segments of the aorta, which were then mounted in organ chambers as previously described.^6,19 Detailed methods are described in Supplementary Materials.

AC fistula rat model and treatment protocol

The detailed procedures for generating AC fistula have been previously described.^6,20 The rats were randomly allocated to three groups before generating AC fistula (control, HG, and HG + PCSK9i groups). Animals in the HG + PCSK9i group were treated with alirocumab (Sanofi Winthrop Industry, France, 50 mg/kg/week, subcutaneous)²¹ on the day of operation (D0) and day 7 (D7), followed by euthanasia on D14 (Supplementary Figure 1). Serum was then collected for biochemical analysis and protein quantification (n = 9 for control, n = 9 for HG, and n = 8 for HG + PCSK9i groups). The aorta tissues proximal to the fistula were harvested from rats with AC fistula to assess vasomotor function. The aorta specimens from other animals with AC fistula were used for protein or histology analysis.

Supplementary Figure 1.

Study protocol of in vivo treatment with alirocumab, a proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor in aortocaval fistula (ACF) rat model with hyperglycemia (HG). A total of 111 rats were used in the in vivo experiment, with 37 each for control, HG, and HG + PCSK9i. No surgical mortality was noticed, and the experimental mortality before the sacrifice was 5% owing to hyperglycemia-associated complications in the HG group (with and without PCSK9i treatment). Placebo indicates the solvent of PCSK9i, which contains histidine (8 mM), polysorbate 20 (0.1 mg), and sucrose (100 mg) per 1 mL of sterile water. ip, intraperitoneal; sc, subcutaneous; S-D, Sprague-Dawley.

Analysis of serum chemistry

The levels of serum glucose, blood urea nitrogen, creatinine, cholesterol, LDL-c, and high-density lipoprotein cholesterol (HDL-c) were analyzed. Enzyme-linked immunosorbent assay (ELISA) detection kits (Elabscience Biotechnology Inc., USA) were used to measure serum oxLDL concentrations according to the manufacturer’s instructions.

Hemodynamic measurements and assessment of vasomotor function

Echocardiography and blood flow measurements

Transthoracic echocardiography was performed in anesthetized rats as previously described.¹⁹ Blood flow in the arterial site of the AC fistula was determined using an ultrasonic flow probe (Transonic System, Ithaca, NY, USA) as previously described.^6,20 All measurements were conducted on D14, before euthanasia, by an investigator who was blinded to the treatment groups.⁶

Organ chamber experiments for the aortic limb of AC fistula

Segments of the aorta, proximal to the fistula, were harvested from rats with AC fistula on D14 for vasomotor functional analysis as previously described.⁶

Tissue morphology and protein analysis

Immunohistochemistry and connective tissue staining

Isolated arterial tissues from rats with AC fistula, above the fistula level, were immersed in 4% formaldehyde before tissue fixation in paraffin. The paraffin-embedded tissues were sectioned and stained with hematoxylin-eosin and Masson’s trichrome to evaluate the thickness of the aorta wall as previously described.^6,19

Determinations of oxidative stress, inflammatory proteins, and secondary messengers of NO

Aortas were frozen as serial sections (5 μm) in an optimal cutting temperature compound, and the generation of superoxide anions was measured using chemiluminescence and diethidium assays.⁶

Soluble proteins (50 μg) extracted from the homogenized aorta were prepared for immunoblotting as described in Supplementary Materials.^6,19,22 Tissue levels of superoxide dismutase (SOD) and cyclic guanosine monophosphate (cGMP) were quantified using ELISA assays as previously described.^6,22

Statistical analysis

Unless otherwise specified, data are presented as the mean ± standard deviation. Detailed statistical methods are described in Supplementary Materials. Statistical significance was set at p < .05.

RESULTS

PCSK9i induced direct vascular relaxation in the aorta through a potassium channel-associated mechanism in normal rats and rats with hyperglycemia

PCSK9i induced the relaxation of the aorta ex vivo from normal rats in a concentration-dependent manner (p = .004, Figure 1A). The maximum relaxation responses were inferior to those induced by acetylcholine, which stimulated ECs to release NO (p = .011). No statistically significant changes were observed in relaxation responses to PCSK9i when the endothelial layer was removed (denuded) (p = .230, Figure 1B). Pre-treatment with 300 nM PCSK9i improved the endothelial-dependent relaxation responses of the aorta of rats with HG ex vivo (Figure 1C). The direct dilatation effects were attenuated by adding tetraethylammonium (a non-selective potassium channel blocker), KCl, and glibenclamide (a K_ATP channel blocker) (p < .001, Figure 1D). A total of 53 rats, including 8 with HG and 45 normal rats, were used for the ex vivo pharmacological experiments.

Figure 1.

Vasomotor functional analysis of the aorta of rats with or without hyperglycemia (HG) to assess the ex vivo effects of the proprotein convertase subtilisin/kexin type 9 inhibitor (PCSK9i). (A) Relaxation responses to cumulative addition of acetylcholine or PCSK9i, 10^-9-10^-5 M. The relaxation responses were obtained during contraction to EC₆₀ (the concentration required to achieve 60% of maximum contraction) of phenylephrine. * p < .001 vs. PCSK9i and acetylcholine; ** p = .011 vs. acetylcholine determined using the Holm-Sidak post hoc method; n = 11 control rats. The placebo corresponds to the PCSK9i solvent, which contains histidine (8 mM), polysorbate 20 (0.1 mg), and sucrose (100 mg) per 1 mL of sterile water. (B) Relaxation responses to cumulative addition of PCSK9i, 10^-9-10^-5 M in the intact or denuded aorta. The relaxation responses were obtained during contraction to EC₆₀ of phenylephrine. *** p < .001 vs. PCSK9i and PCSK9i denuded; p = .23 for PCSK9i vs. PCSK9i denuded determined using the Holm-Sidak post hoc method; N = 12 control rats. (C) Endothelial-dependent dilatation responses of the aorta from rats with HG after ex vivo treatment with 300 nM PCSK9i. ^# p < .001 vs. HG + PCSK9i and normal aorta; ^## p < .001 vs. normal aorta using the Holm-Sidak post hoc method. n = 8 each for HG and control rats. (D) Relaxation responses to cumulative addition of PCSK9i, 10^-9-10^-5 M. The relaxation responses were obtained during contraction to EC₆₀ of phenylephrine. Some of the aorta tissues were pretreated ex vivo with KCl (20 mM), tetraethylammonium (TEA, 3 mM), or glibenclamide (Gliben, 10 μM). ^### p < .001 vs. other groups; ^ p < .001 vs. other groups; ^^ p = .011 for PCSK9i only vs. Gliben + PCSK9i, determined using the Holm-Sidak post hoc method. Data are presented as the mean ± standard error.

PCSK9i improved blood flow through the fistula in rats with hyperglycemia

A total of 37 normal rats and 74 rats with HG were used for the in vivo experiments (Supplementary Figure 1). Both the peak and mean blood flow of the AC fistula decreased in the rats with HG. Treatment with PCSK9i improved blood flow through the fistula (p = .002, Figure 2). Echocardiography showed no significant differences in the ejection fraction of the left ventricle and biventricular morphology between the control and PCSK9i-treated rats (Supplementary Table 1).

Figure 2.

Blood flow through the aortocaval fistula in rats with and without hyperglycemia (HG) on day 14, before euthanasia. (A) Representative tracing diagram of peak blood flow. (B) A significantly decreased mean flow rate through the fistula in the HG group, which improved after treatment with proprotein convertase subtilisin/kexin type 9 inhibitor (PCSK9i). * p < .001, ** p = .01. Bars represent median value, n = 9/group.

Supplementary Table 1. Echocardiography on day 14 after aortocaval fistula.

	Control (n = 4)	HG (n = 3)	HG + PCSK9i (n = 3)	p value
LVEF (%)	78.37 ± 8.48	73.02 ± 10.99	78.86 ± 6.48	.672
LVIDs (mm)	3.58 ± .67	3.60 ± .66	4.40 ± .17	.185
LVIDd (mm)	6.25 ± 1.14	6.10 ± .79	7.00 ± .85	.506
FS	42.25 ± 8.56	40.96 ± 7.57	42.07 ± 4.55	.971
RV diameter (mm)	4.38 ± .39	3.90 ± .40	4.03 ± .23	.249

Data are presented as the mean ± standard deviation.

FS, fraction shortening; HG, hyperglycemia; LVEF, left ventricle ejection fraction; LVIDd, left ventricle internal diameter during diastolic phase; LVIDs, left ventricle internal diameter during systolic phase; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor; RV, right ventricle.

A significant reduction in body weight was observed in the rats with HG compared to that in the normal rats (p < .001, Table 1). However, body-weight changes in the PCSK9i-treated HG group were not statistically significant compared to those in the HG group (p = .167). Increased blood urea nitrogen, creatinine, and cholesterol levels were observed in the rats with HG (p < .001, p < .001, and p < .042, respectively). These levels were reduced after treatment with PCSK9i, although the differences were not statistically significant. No significant differences in HDL-c and LDL-c levels were observed in the rats with HG with or without PCSK9i treatment.

Table 1. Body weight and serum biochemistry profiles on the day of sacrifice.

	Control (n = 9)	HG (n = 9)	HG + PCSK9i (n = 8)	p value
Body weight (g)	384.78 ± 29.83#	267.44 ± 37.15	301.50 ± 44.34	< .001
Glucose AC (mg/dL)	267.22 ± 54.70*	639.33 ± 55.45	597.38 ± 151.87	< .001
BUN (mg/dL)	18.47 ± 1.67**	28.38 ± 3.68	25.98 ± 5.36	< .001
Creatinine (mg/dL)	0.42 ± 0.06†	0.68 ± 0.12	0.60 ± 0.13	< .001
Cholesterol (mg/dL)	57.44 ± 7.78‡	104.00 ± 57.40	72.13 ± 27.19	.042
HDL-c (mg/dL)	17.67 ± 3.94	22.89 ± 9.52	20.50 ± 5.66	.287
LDL-c (mg/dL)	8.22 ± 0.44	16.22 ± 9.78	11.63 ± 7.13	.071
LDL-c/HDL-c	0.39 ± 0.11	1.37 ± 1.98	0.63 ± 0.63	.231
t-Chol/HDL-c	3.33 ± 0.46	8.57 ± 12.00	3.98 ± 2.57	.269

Data are presented as the mean ± standard deviation; ^# p < .001 vs. HG; p < .001 vs. HG + PCSK9i; p = .167 for HG vs. HG + PCSK9i using Tukey’s post hoc test; * p < .001 vs. HG and HG + PCSK9i; p = .64 for HG vs. HG + PCSK9i using test; ** p < .001 vs. HG; p = .001 vs. HG + PCSK9i; p = .41 for HG vs. HG + PCSK9i using Tukey’s post hoc test; ^† p < .001 vs. HG; p = .005 vs. HG + PCSK9i; p = .28 for HG vs. HG + PCSK9i using Tukey’s post hoc test; ^‡ p = .037 vs. HG; p = .21 for HG vs. HG + PCSK9i using Tukey’s post hoc test.

AC, before meal; BUN, blood urea nitrogen; HDL-c, high density lipoprotein cholesterol; HG, hyperglycemia; LDL-c, low density lipoprotein cholesterol; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor; t-Chol, total cholesterol.

Vasomotor functional improvement in the aortic limb of the AC fistula in PCSK9i-treated rats with hyperglycemia

The maximum contractions achieved by KCl (40 mM) and phenylephrine (PE, 10^-9-10^-5 M) significantly increased in the HG rats (p < .001, and p = .006, respectively, Figure 3A). PCSK9i treatment significantly attenuated the maximum non-selective (to KCl) and selective (to PE, α1-agonist) contraction responses (Figure 3A, B). An increased concentration-dependent response of tension to PE was observed in the aorta of the HG group (p < .001, Figure 3B). However, the concentration of PE required to induce 50% contraction effects (EC₅₀) did not significantly change with PCSK9i treatment (p = .214, Figure 3C). Endothelium-dependent relaxation to acetylcholine was significantly impaired in the HG rats, and the relaxation response curve shifted to the left in the treatment group (p < .001, Figure 3D).

Figure 3.

Vasomotor function analysis of the aorta. (A) Contraction responses to KCl (40 mM) and phenylephrine (PE, 10^-5 M); bars represent the median value. (B) Contraction response of the aorta to the cumulative addition of PE (10^-9-10^-5 M); data are presented as the mean ± standard error (SE). (C) Effective concentration of PE (PE EC₅₀), which induced 50% of the maximum contraction; bars represent the median value. (D) Relaxation responses to the cumulative addition of acetylcholine (10^-9-10^-5 M). Relaxations were obtained during contraction to EC₆₀ (the concentration required to achieve 60% of the maximum contraction) of PE; data are presented as the mean ± SE. AVF, aortocaval fistula; Con, control without hyperglycemia (HG); PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor. * p = .038, ** p < .001, *** p = .026, ^# p = .008 using Tukey’s post hoc test; ^## p < .001 for HG vs. control and HG + PCSK9i, using the Holm-Sidak post hoc method; n = 8 rats/group.

A decreased ratio of phosphorylated to total endothelial NO synthase (eNOS) was observed in the aorta of rats with HG (Figure 4A, B). Tissue cGMP contents decreased under hyperglycemic conditions (p = .011, Figure 4C). Treatment with PCSK9i improved eNOS activity and cGMP abundance (p = .022 and p < .001, respectively).

Figure 4.

Protein levels for endothelial function in the aorta of hyperglycemia (HG) and non-HG rats after generation of an aortocaval fistula. (A) Representative immunoblots of phosphorylated (p-) and total (t-) endothelial nitric oxide synthase (eNOS). (B) Quantitative analysis of the ratio of p-eNOS/t-eNOS; n = 6/group. (C) Cyclic guanosine monophosphate (cGMP) level; n = 9/group; * p = .022, ** p = .011, *** p < .001 determined by Tukey’s post hoc test. Bars represent median values. Da, Dalton; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor.

PCSK9i reduced oxidative stress and some inflammatory responses in the aorta of rats with hyperglycemia

The abundance of LOX-1, hem-oxygenase (HO-1), inducible NO synthase, and the ratio of phosphorylated (p)-p47phox/total (t)-p47phox were significantly increased in the aorta of HG rats (Figure 5). These levels were significantly reduced in HG rats treated with PCSK9i (p < .001). Dihydroethidium fluorescence densities were notably enhanced in the aorta rings isolated from rats with HG, but were comparatively low in the controls and animals treated with PCSK9i (Figure 6A). Increased serum oxLDL levels and tissue SOD abundance were observed in the HG rats (p = .002, Figure 6B, C), and they were significantly reduced by PCSK9i treatment (p = .006 and p = .043, respectively).

Figure 5.

Proprotein convertase subtilisin/kexin type 9 inhibitor (PCSK9i) reduced protein levels associated with oxidative stress in the aorta tissues of rats with hyperglycemia (HG). (A) Representative immunoblots of lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1), heme-oxygenase (HO)-1, inducible nitric oxide synthase (iNOS), phosphorylated (p-), and total (t-) p47phox. Quantitative analysis for (B) LOX-1, (C) HO-1, (D) iNOS, and (E) Ratio of p-/t-p47phox. * p < .001, ** p = .002, *** p = .023, ^# p = .021 determined by Tukey’s post hoc test; n = 6/group. Bars represent the median value. AVF, aortocaval fistula; Da, Dalton; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Figure 6.

The extent of oxidation in the serum and aorta tissue of an aortocaval fistula (AVF) rat model with or without hyperglycemia (HG). (A) Dihydroethidium Fluorescence intensity of dihydroethidium in the aorta rings of AVF. Yellow scale bars, 500 μm for the upper row and 50 μm for the lower row. (B) Serum level of oxidized low-density lipoprotein; n = 6-8/group. (C) Superoxide dismutase contents in aortic tissue; n = 8/group. * p = .002, ** p = .006, *** p < .001, ^# p = .043 determined using Tukey’s post hoc test. Bars represent a median value. PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor.

Representative inflammatory proteins including cyclooxygenase-2 (COX-2) and myeloperoxidase (MPO) were increased in aorta tissue from the HG rats (p < .001 and p = .002, respectively), and they were reduced by PCSK9i treatment (Supplementary Figure 2A-C). No significant changes in monocyte chemoattractant protein-1 (MCP-1) levels were observed in the study animals (p = .318, Supplementary Figure 2D). Further analysis of nucleotide oligomerization domain (NOD)-, leucine-rich repeat (LRR)-, and pyrin domain-containing protein-3 (NLRP3) inflammasome-associated signal proteins revealed a significant reduction in caspase-1 and interleukin (IL)-1β levels (p < .001 and p < .001, respectively, Supplementary Figure 3). No significant changes in the levels of toll-like receptor 4, NF-κB, and NLRP3 were observed in the study animals (p = .442, .259, and .379, respectively, Supplementary Figure 3B-D).

Supplementary Figure 2.

Significant reduction of protein levels associated with inflammation, except monocyte chemoattractant protein (MCP)-1, p = .318. (A) Representative immunoblotting for cyclooxygenase (COX)-2, myeloperoxidase (MPO), and MCP-1. Quantitative analysis for (B) COX-2, (C) MPO, and (D) MCP-1. * p < .001, ** p = .002, *** p = .009 determined by Tukey’s post hoc test; n = 6/group. Bars indicate median values. Da, Dalton; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HG, hyperglycemia.

Supplementary Figure 3.

Nucleotide oligomerization domain-, leucine-rich repeat-, and pyrin domain-containing protein (NLRP3) inflammasome-associated signal proteins in aorta tissue from hyperglycemia (HG) rats. (A) Representative immunoblotting for toll-like receptor (TLR)4, nuclear factor κ-light-chain-enhancer of activated B cells (NF-κB), NLRP3, caspase-1, and interleukin (IL)-1β. Quantitative analysis for (B) TLR4, (C) NF-κB, (D) NLRP3, (E) Caspase-1, and (F) IL-1β. * p < .001, ** p = .004 determined by Tukey’s post hoc test. N = 8/group. Bars indicate median values. Da, Dalton; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor.

Potential tissue remodeling in PCSK9i-treated rats with hyperglycemia

The ratio of phosphorylated to total p38 mitogen-activated protein kinases (MAPK) and extracellular signal-regulated kinase (ERK) was significantly increased in the aorta of rats with HG. However, it was significantly reduced following PCSK9i treatment (p < .001, Supplementary Figure 4).

Supplementary Figure 4.

Proprotein convertase subtilisin/kexin type 9 inhibitor (PCSK9i) reduced the levels of proliferative proteins responsive to oxidative stress in aorta tissues of hyperglycemia (HG) rats. (A) Representative immunoblotting for phosphorylated (p-) and total (t-) mitogen-activated protein kinase (MAPK) and extracellular signal-regulated kinase (ERK). Quantitative analysis for (B) Ratio of p-/t-MAPK and (C) Ratio of p-/t-ERK. * p < .001 using Tukey’s post hoc test. n = 6/group. Bars indicate median values. Da, Dalton; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Increased aortic wall thickness was observed in the HG groups (Figure 7). Spindle-shaped nuclei were also noted in the HG group, which then returned to an ovoid shape similar to those in the normal aorta after PCSK9i treatment. However, PCSK9i treatment did not significantly reduce the thickness of aorta walls on D14 (p = 1.0, Figure 7, Supplementary Figure 5A). Measurement of the cross-sectional area of the aorta showed no significant differences between the rats with HG with or without PCSK9i treatment (p = .468, Supplementary Figure 5B)

Figure 7.

Histological remodelling of the aortic limb in the aortocaval fistula rat model with hyperglycemia (HG) using hematoxylin and eosin and Masson’s trichrome staining for media layer of the aorta. PCSK9I, proprotein convertase subtilisin/kexin type 9 inhibitor. Scale bar, 100 μm.

Supplementary Figure 5.

Quantitative analysis for (A) Wall thickness, * p = .034 using post hoc Dunn’s method. (B) Cross-sectional area of the aortocaval fistula in the aorta, p = .468. Each rat aorta was examined in 3 different sections, with 3 counts for each section, 9 counts for each rat, with a total of 3 rats/group, HG, hyperglycemia; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor.

DISCUSSION

The results of the present study showed the direct vasorelaxation effects of ex vivo treatment with PCSK9i in the aorta of normal rats and rats with HG. The ex vivo dilatation effects of PCSK9i were mostly associated with an increased opening of potassium channels, which was independent of NO released from ECs. PCSK9i injection weekly for two weeks did not induce significant changes in serum cholesterol, HDL-c, and LDL-c levels in the rats with HG. However, in vivo treatment with PCSK9i improved blood flow in the arterial limb of the AC fistula in the rats with HG. In contrast to the ex vivo effects, subcutaneous PCSK9i injection contributed to preservation of arterial compliance in the rats with HG by restoring endothelial function. In vivo vascular protection in the rats with HG was associated with the attenuation of oxidative stress and, at least in part, inflammatory responses. Collectively, PCSK9i treatment had potential pleiotropic effects that restored vascular function and improved blood flow in this AC fistula model of rats with HG (Central Illustration).

Central Illustration.

Proprotein convertase subtilisin/kexin type 9 inhibitor (PCSK9i) induces ex vivo arterial dilatation which is associated with potassium channel opening. In vivo treatment with PCSK9i improves blood flow through inhibition of oxidative stress and anti-inflammation mechanisms.

The excessive generation of free radicals under hyperglycemic conditions induces the production of oxidants such as oxLDL, which leads to endothelial dysfunction and results in AVF dysfunction.^6,9,23 PCSK9i neutralizes PCSK9 to reduce the intracellular sequestration of LDL-c receptors and facilitate the cellular uptake of LDL-c. Consequently, the serum level of oxLDL formed by the oxidation of LDL decreases. This mechanism is consistent with the findings of our study. Intracellular ROS generates LOX-1, the receptor for oxLDL,¹⁵ and PCSK9 inhibition has been shown to suppresses LOX-1 expression in VSMCs in vitro. In line with these findings, PCSK9i treatment reduced oxidative stress and LOX-1 expression in the aorta limb of the AC fistula in the present study, which attenuated eNOS uncoupling and improved downstream cGMP levels in the arterial tissues of the rats with HG. The increase in laminar blood flow by AC fistula has been shown to mobilize the expression of eNOS in previous studies.^6,24,25 This mobilization was balanced by the hyperglycemic conditions in the HG group. Consequently, there was a trend of a reduction in p/t-eNOS ratio between the HG and control group, but the reduction was not statistically significant (Figure 4B, p = .068 post hoc Tukey test).⁶ We further increased the number of data points in the cGMP experiment to confirm our assumption that the NO-cGMP axis was inhibited in the HG group (Figure 4). Pre-treatment with evolocumab, a PCSK9i, has been shown to attenuate oxidative stress induced by hydrogen peroxide in ECs, indicating the direct pharmacological effects of PCSK9i in ECs.¹⁷ However, our ex vivo experiments showed no significant changes in vascular relaxation responses to PCSK9i when the endothelial layer was removed from the vessel. Approximately 80% of direct aorta relaxation responses to PCSK9i are associated with the activation of potassium channels in the vasculature rather than mobilization of NO signaling from ECs. Therefore, we propose that the vascular protective effects of PCSK9i in the AC fistula of the rats with HG were indirect. Nevertheless, oxLDL is an important mediator in the endothelial protection mechanism under hyperglycemic conditions.

Oxidative stress is an important factor in activating VSMCs towards a synthetic phenotype and remodeling of the vascular wall.^6,19 The MAPK signaling pathway is a common cellular proliferation pathway that is responsive to oxidative stress.²⁶ In line with the mobilization of downstream MAPK-ERK proliferation signaling proteins under hyperglycemia, our histological analysis showed a statistically significant increase in vascular wall thickness in the aorta limb of AC fistula among the rats with HG. PCSK9i treatment attenuated MAPK-ERK abundance without changing wall thickness. Considering the short study period, we did not find reverse vascular remodeling of aortic wall thickness in the treatment group. In brief, PCSK9i treatment induced the remodeling of the AC fistula in rats with hyperglycemia, and this may have contributed to increased blood flow through the fistula.

Activation of systemic and tissue inflammation is common in patients with diabetes, even under optimal glycemic control.²⁷ oxLDL is associated with increased levels of adhesion molecules such as integrin, which increase the recruitment of macrophages in the vascular wall.²⁸ The internalization of oxLDL by macrophages contributes to the increased secretion of inflammatory cytokines such as NF-κB, NLRP3, IL-1β, and tumor necrosis factor-alpha.^14,28,29 We observed reduced tissue inflammationby COX-2 and MPO, but not MCP-1, in the arterial limb of the AC fistula of the rats with hyperglycemia. To reconfirm the anti-inflammatory potential of PCSK9i, we investigated the IL-1β-associated NLRP3 inflammasome pathway. PCSK9i treatment significantly reduced caspase-1 and IL-1β levels without changing the NF-κB-NLRP3 signaling axis in the rats with HG. Collectively, our study supports that PCSK9 may be an important mediator of vascular inflammation in the AC fistula of rats with HG. Furthermore, our findings suggest that the anti-inflammatory effects of pharmacological PCSK9 inhibition are associated with the suppression of oxidative stress and NF-κB-NLRP3 signaling.

Clinical data on PCSK9i in patients requiring hemodialysis are limited. Subgroup analysis of chronic kidney disease (estimated glomerular filtration rate > 30 mL/ min/1.73 m²) in the ODYSSEY OUTCOMES randomized clinical trial showed a lower incidence of serious adverse events in alirocumab-treated patients compared to placebo-treated patients.³⁰ East et al. reported a 45% reduction in LDL-c after 12 weeks of treatment with alirocumab among 14 patients who received maintenance hemodialysis (n = 12) or peritoneal dialysis (n = 2).³⁰ In addition, although no alirocumab-associated serious adverse events were noted in the cohort, significantly reduced serum IL-6 and TNF-α levels were noted at the end of the experimental period. In contrast, the reduction in high-sensitivity C-reactive protein was not statistically significant. Three patients (21.4%) suffered dialysis access dysfunction during the study period.²⁸ Although PCSK9i-associated adverse events appear to be of low concern, further investigations are needed to verify the potential benefits of PCSK9i in the functional improvement or durability of dialysis access.

This translational study has several limitations. First, the animal model mimicked hyperglycemic conditions in a short-term period rather than chronic diabetic conditions in chronic kidney diseases. Second, the animal model of AC fistula mimicked a certain clinical morphology of AVF using an endovascular method.³¹ This fistula morphology is not the most common form of hemodialysis. Third, we tested the potential effect of PCSK9i as a treatment option. However, an optimal dose of PCSK9i to achi-eve a maximum vascular effect was not determined in this study. Moreover, our results are gender-biased because only male rats were used in the experiments; thus, experiments on female rats may yield different observations. Fourth, we did not investigate morphological or neointimal changes in the venous limb of the AC fistula, which requires a longer observation period. Extrapolation of our study findings might be limited due to the short-term observations. Lastly, the acute hyperglycemic effects identified in this study could be different from chronic vascular remodeling among patients with diabetes in a clinical setting. Our study presents the short-term benefits of PCSK9i in AC fistula within 14 days after fistula surgery; long-term or potentially persistent beneficial effects have yet to be determined.

New knowledge gained

Our results revealed the pleiotropic effects of PCSK9i in arterial limb function of arteriovenous fistula in a rat model of HG. The improvement was associated with the suppression of oxidative stress and inflammatory responses after treatment with PCSK9i.

CONCLUSIONS

This study showed the pleiotropic effects of PCSK9i on AC fistula under hyperglycemic conditions. PCSK9i treatment induced vascular dilatation and increased blood flow through the AC fistula in the rats with HG. PCSK9i induced vascular dilatation by suppressing oxidative stress and tissue inflammation under hyperglycemic conditions. PCSK9 inhibition also contributed to reverse vascular wall remodeling in the aortic limb of the AC fistula. This translational study provides supportive evidence for the clinical application of PCSK9i in patients with diabetes undergoing hemodialysis.

DECLARATION OF CONFLICT OF INTEREST

The authors declare no conflict of interest.

SUPPLEMENTARY MATERIALS

METHODS

Induction of hyperglycemia in rats

HG, defined as a random blood glucose level over 250 mg/dL as previously described, was confirmed after 72 h.^S1

Animal model of AC fistula

Rats were anesthetized with isoflurane (2%-3% v/v in O₂).^S2,S3 Following a midline abdominal incision, the inferior vena cava (IVC) and aorta were exposed. The aorta was punctured with a 20G disposable needle below the renal vessels. The needle was introduced into the IVC lumen through the wall between the aorta and IVC. The needle was withdrawn, and the puncturing point was closed by a purse-string suture. Vascular clamps were removed, and the abdominal wall was closed in layers.^S3

Preliminary organ chamber experiments of PCSK9i in the aorta

Aorta rings (approximately 5 mm long) from rats with HG without AC fistula were mounted in organ chambers containing 25 mL Krebs solution (118.6 mmol/L NaCl, 4.7 mmol/L KCl, 2.5 mmol/L CaCl₂, 1.2 mmol/L MgSO₄, 1.2 mmol/L KH₂PO₄, 25.1 mmol/L NaHCO₃, 10.1 mmol/L glucose, and 0.026 mmol/L ethylenediaminetetraacetic acid) at 37 °C (94% O₂ and 6% CO₂). After a 45-min equilibration period, the rings were contracted by cumulative addition of PE (10^-9-10^-5 M). Changes in force were continuously recorded using an isometric force-displacement transducer (Grass FT03; Grass Instrument, D-79232 March-Hugstetten Germany). Concentration-response curves were then obtained by the cumulative addition of acetylcholine (10^-9-10^-5 M) during contraction to the concentration required to achieve 60% of maximum contraction (EC₆₀) of PE to test for endothelial-dependent relaxations.

The aorta rings from normal rats without AC fistula were mounted in organ chambers as described above. Concentration-response curves were then obtained by the cumulative addition of placebo or alirocumab (10^-9-10^-5 M) during contraction to the EC₆₀ of PE. Concentration-response curves were compared between the groups in the presence and absence of KCl (20 mM), tetraethylammonium (3 mM, a nonselective K⁺ channel blocker), or glibenclamide (10 μM, a K_ATP channel blocker) as previously described.^S4,S5 In some experiments, the endothelium was mechanically removed using a very gentle grip of the artery rings with stainless forceps and then gently rolling the inner surface of the artery against the organ bath metal holders.^S3,S6,S7

Measurement of cyclic adenosine monophosphate and cyclic guanosine monophosphate levels in tissues

Aorta tissue segments were incubated in Earle’s salts solution with 3-isobutyl-1-methylxanthine (10^-3 M) for 30 min to inhibit the phosphodiesterase-mediated degradation of cyclic nucleotides as previously described.^S7 Artery tissues were then snap-frozen and stored at -80 °C until assayed.

Immunoblotting

Soluble proteins (50 μg) extracted from the rat aorta were resolved on polyacrylamide gels (9%-12%) and transferred onto polyvinylidene fluoride membranes following previously described methods.^S3,S8 Mouse monoclonal antibodies against inducible nitric oxide synthase (1:1000, BD Biosciences, NJ, USA), MPO (1:1000, Abcam, Cambridge, UK), MCP-1 (1:1000, Proteintech Group, Inc, IL, USA), COX-2 (1:1000 GeneTex International Corporation, CA, USA), oxidized low density lipoprotein receptor (LOX-1, 1:500, ABclonal Technology, MA, USA), HO-1 (1:1000, Enzo Life Sciences, NY, USA), p47-phox (t-p47-phox, 1:1000, BD Biosciences), phosphorylated p47-phox (p-p47-phox, 1:1000, MilliporeSigma, Germany), p38 mitogen-activated protein kinases (t-p38 MAPK, 1:1000, Cell Signaling Technology, Inc, USA), phosphorylated p38 MAPK (p-p38 MAPK, 1:500, Cell Signaling Technology, Inc), extracellular signal-regulated kinase (t-ERK, 1:1000, Cell Signaling Technology, Inc), phosphorylated ERK (p-ERK, 1:1000, Cell Signaling Technology, Inc), the nucleotide-binding oligomerization domain, leucine-rich repeat-containing proteins (NLRP3, 1:1000, Novus Biologicals, Canada), toll-like receptor 4 (TLR-4, 1:1000, Abcam, Cambridge, UK), nuclear factor (NF-κB, 1:1000, BioLegend, USA), caspase-1 (1:1000, EMD Millipore, USA), IL-1β (1:500, GeneTex International Corporation), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 1:5000, GeneTex International Corporation) were used. Bands were visualized using enhanced chemiluminescence and quantified by performing scanning densitometry (ImageJ; v.1.48, National Institutes of Health, Bethesda, MD, USA).^S3,S7

Statistical analysis

All data sets were tested for normality assumption using the Shapiro-Wilk test prior to statistical analysis. Intergroup differences were analyzed using one-way analysis of variance (ANOVA) followed by appropriate post hoc tests for multiple comparisons. The concentration curves of vasomotor function were analyzed using two-way repeat measurement ANOVA with post hoc tests as appropriate. All statistical analyses were performed using SigmaPlot v.14.0 (Systat Software Inc., San Jose, CA, USA).