A dual blocker of endothelin A/B receptors mitigates hypertension but not renal dysfunction in a rat model of chronic kidney disease and sleep apnea

Abstract

Obstructive sleep apnea is characterized by recurrent episodes of pharyngeal collapse during sleep, resulting in intermittent hypoxia (IH), and is associated with a high incidence of hypertension and accelerated renal failure. In rodents, endothelin (ET)-1 contributes to IH-induced hypertension, and ET-1 levels inversely correlate with glomerular filtration rate in patients with end-stage chronic kidney disease (CKD). Therefore, we hypothesized that a dual ET receptor antagonist, macitentan (Actelion Pharmaceuticals), will attenuate and reverse hypertension and renal dysfunction in a rat model of combined IH and CKD. Male Sprague-Dawley rats received one of three diets (control, 0.2% adenine, and 0.2% adenine + 30 mg·kg⁻¹·day⁻¹ macitentan) for 2 wk followed by 2 wk of recovery diet. Rats were then exposed for 4 wk to air or IH (20 short exposures/h to 5% O₂-5% CO₂ 7 h/day during sleep). Macitentan prevented the increases in mean arterial blood pressure caused by CKD, IH, and the combination of CKD + IH. However, macitentan did not improve kidney function, fibrosis, and inflammation. After CKD was established, rats were exposed to air or IH for 2 wk, and macitentan feeding continued for 2 more wk. Macitentan reversed the hypertension in IH, CKD, and CKD + IH groups without improving renal function. Our data suggest that macitentan could be an effective antihypertensive in patients with CKD and irreversible kidney damage as a way to protect the heart, brain, and eyes from elevated arterial pressure, but it does not reverse toxin-induced tubule atrophy.

INTRODUCTION

Obstructive sleep apnea (OSA) is associated with a high incidence of hypertension and peripheral vascular disease (20, 21, 23, 42). OSA, characterized by recurrent episodes of apnea and hypopnea during sleep attributable to obstruction of the pharyngeal airway, provokes intermittent hypoxia (IH) and daytime sleepiness (23). Moreover, the prevalence of OSA is high in patients with hypertension (21) and patients with chronic kidney disease (CKD) (1).

A potential mechanism for these associations is vascular damage because several studies have shown that IH damages the vascular endothelium (6, 7, 9, 32, 38, 41). However, the mechanisms linking OSA with cardiovascular diseases and CKD are still unclear. This study used an experimental animal model developed to mimic the cardiovascular-hemodynamic changes in OSA (5, 28), a model previously validated in our laboratory (8, 31).

IH causes oxidative stress, systemic inflammation, increased sympathetic activation, and elevated endothelin (ET)-1 levels (8). Our previous results demonstrated that acute administration of PD-145,065 (a nonselective ET-1 receptor antagonist) and BQ123 [a selective ET type A (ET_A) receptor antagonist] prevents IH-induced hypertension in rats (3, 24). Similarly, the dual ET receptor antagonist bosentan has the same effect in mice (17). ET-1, the most potent vasoconstrictor in the human cardiovascular system, has a particularly long-lasting action (29). Clinically, bosentan is one of the most used drugs to treat pulmonary arterial hypertension (PAH) and the first for pediatric PAH.

CKD is a highly prevalent condition (>20 million patients in the United States) and a heavy burden on health systems worldwide. Hypertension is a leading risk factor for the initiation and progression of CKD as well as a consequence of CKD. Using adenine oral administration to cause CKD in rats, we previously demonstrated that CKD and IH additively elevate blood pressure, but IH does not enhance adenine-induced renal fibrosis, inflammation, and decline of kidney function (31). In the same study, pre-pro-ET-1 mRNA levels were increased in the renal cortex and medulla of CKD rats, whereas IH did not further enhance ET-1 expression (31). In patients with CKD, COOH-terminal pro-ET-1 peptide plasma levels inversely correlate with estimated glomerular filtration rate (eGFR) (15). In the same randomized double-blind three-way crossover design study, individuals with varying degrees of CKD and minimal comorbidity were administered the selective ET_A receptor antagonist sitaxentan (100 mg QD) or the Ca²⁺ channel blocker nifedipine LA (30 mg QD). Sitaxentan significantly reduced mean arterial blood pressure (MAP) and improved impaired kidney function (15). These studies suggested that ET-1 plays an important role in the development and progression of CKD. However, the effect of a dual ET receptor blocker in CKD and/or CKD with IH has not been previously investigated. Furthermore, selective ET_A receptor antagonist therapy has not been approved by the Federal Drug Administration.

Therefore, we hypothesized that dual ET receptor antagonism with macitentan would attenuate the development of hypertension and kidney dysfunction in a combined model of CKD and sleep apnea. Macitentan, the most recent Federal Drug Administration-approved next-generation dual ET receptor antagonist, is more potent than bosentan, with longer receptor occupancy, and is converted to an active metabolite that contributes to improved pharmacodynamics and pharmacokinetics (29).

MATERIALS AND METHODS

Animals

Male Sprague-Dawley rats from Envigo Laboratories (200–225 g) were used in all experiments. The Institutional Animal Care and Use Committee of the School of Medicine at the University of New Mexico approved all animal protocols. All experimental procedures were conducted in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals.

Prevention Study

CKD model induction and prevention assessment protocol.

CKD was induced by feeding rats 0.2% adenine (Sigma-Aldrich) mixed in Nutra-gel (Bio Serv) for 2 wk (31), which is equivalent to a dose of 40 g·rat⁻¹·day⁻¹. Rats in CKD and control groups assigned to concurrent macitentan received 30 mg·kg⁻¹·day⁻¹ in the same food. Each week, plasma creatinine and blood urea nitrogen (BUN) were assessed with iSTAT Chem8 cartridges (Abbott) from blood collected through a saphenous vein puncture under isoflurane anesthesia. After 2 wk, all CKD rats reached 100 mg/dl BUN or more; thereupon, the Nutra-gel-based diet was replaced by normal rat chow (Teklad Global Soy Protein Free Diet 2920X, Envigo) or Teklad 2020X (nonirradiated) containing 0.24 mg macitentan/g of food (~30 mg·kg⁻¹·day⁻). All control rats were fed with an equivalent diet without adenine throughout the study.

IH exposure.

After 2 wk of recovery on control or macitentan-containing food, rats in the IH groups were exposed to 20 hypoxic episodes/h (90 s of air followed by 90 s of progressive hypoxia down to 5% inhaled O₂ to model moderate sleep apnea) during their sleep period for 6 wk as previously described (3, 4, 24, 31). CO₂ was added during hypoxia to prevent hypocapnia during the hypoxia-induced hyperventilation (34). Sham animals were exposed to air-air cycling.

Hemodynamic measurements.

HD-S10 telemeter devices (Data Sciences) were implanted under isoflurane anesthesia at 5% for induction and 2% for maintenance. Both buprenorphine (0.003 mg/100 g body wt sc) and 0.9% saline solution (1 ml/100 g body wt sc, USP, Nurse Assist) were administered for pain management and fluid loss compensation. Device catheters were placed into the left femoral artery and tunneled into the abdominal aorta. Rats recovered from surgery for at least 5 days before starting CKD induction.

MAP, heart rate (HR), and activity (not reported) were recorded every 30 min for 30 s over a 48-h period each week. MAP, HR, and activity values were averaged for the day and night periods to yield sleep and wake values (data not shown) and for 24 h to give weekly values.

GFR estimation.

Rats were placed in metabolic cages for a timed urine collection (16–20 h) the day before euthanasia. At the end of the urine collection period, rats were euthanized, and plasma samples were obtained to measure plasma creatinine (iSTAT Chem8 cartridges). Urine creatinine and protein were measured using Catalyst One Chemistry Analyzer (IDEXX). Creatinine clearance (C_Cre) was calculated per 100 g body wt as an estimate of GFR (eGFR) and milligrams per milliliter of protein were normalized for urine flow rate to calculate milligrams per hour excreted.

Osmolar clearance.

Osmolality was measured with an osmometer (Advanced Instruments) in plasma and urine samples using the freezing point depression method. Measurements were used to calculate osmolar clearance and free water clearance.

Kidney histology.

At the end of the study, rats were euthanized by cardiac exsanguination under isoflurane anesthesia. The kidneys, lung, and heart were collected and weighed. Half of the left or right kidney was fixed in 4% formaldehyde overnight at room temperature and then embedded in paraffin. Kidneys were cut into 5-µm sections and stained with Heidenhain’s AZAN stain to determine fibrosis. Slides were scanned electronically with Aperio CS2 (Leica Biosystems) at ×20. Images were analyzed using Halo Image Analysis software (Indica Laboratories). The software was tuned to identify fibrotic staining to perform unbiased calculation of percentage of renal tissue with fibrosis in three adjacent full kidney cross sections.

Immunohistochemistry Studies

Primary antibodies against CD68 (ab31630, Abcam) or CD3 (1:400, A052, lot 20024875, Dako) followed by horseradish peroxidase (HRP)-labeled secondary antibodies in conjunction with 3, 3-diaminobenzidine (DAB) (HRP-DAM or HRP-DAR from Jackson ImmunoResearch) were used to label macrophage and T cells, respectively. For all studies, antibody specificity was determined by omitting primary antibody and by replacing primary antibodies with mouse and rabbit IgG isotype controls (31). Antigen retrieval was performed by heating slide-mounted formalin-fixed, paraffin-embedded kidney sections to 95°C for 20 min in 10 mM Tris at pH 9.0 with 1 mM EDTA and 0.05% Tween 20 before antibody labeling.

The developed slides were scanned at ×20 with Aperio CS2 (Leica Biosystems). Images were analyzed using Halo Image Analysis software (Indica Laboratories). The software was tuned to identify brown positively marked cells (DAB staining) and used to perform unbiased calculation of percentage of label-positive cells versus total cells in midsagittal kidney cross sections.

Transcript levels.

Medulla and cortex regions of the unfixed kidney were collected and placed separately in RNALater solution (Ambion). RNA was extracted using RNeasy Fibrosis Tissue Mini Kit (Qiagen). cDNA was synthesized using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative real-time PCR was performed using β-actin as the endogenous control gene and the relative expression of pre-pro-ET-1, IL-6, and nitric oxide (NO) synthase isoform 2 (NOS2) calculated. The relative quantification for each sample was divided by the relative quantification of the control/sham pool sample. The log of the relative quantification ratio was plotted.

ET-1 plasma levels.

ET-1 plasma levels were measured in the blood samples collected at the time of euthanasia using an ET-1 ELISA kit (Quantikine, R&D Systems) as previously described (31).

Reversal Study

To initiate more rapid CKD development, a second group of rats received 100 mg adenine in a solution of 0.5% methylcellulose in water by oral gavage or vehicle for 3–4 days until BUN reached ~100 mg/dl. After CKD was established, rats were instrumented with telemetry devices to record hemodynamic parameters and activity and allowed 2 wk of recovery. After recovery, rats were exposed to air (sham) or IH for 2 wk, and baseline values were obtained (time 0). Macitentan or vehicle feeding started immediately after time 0 recordings and continued for 2 more wk. Vehicle-treated groups were fed with the same food without macitentan. MAP, HR, body weight, BUN, blood and urine creatinine, and eGFR were measured at each time point. At time of euthanasia, kidneys were collected and weighed.

Experimental Design and Statistical Analysis

An incomplete factorial design was applied with two groups (control and CKD) and two factors (treatment and exposure). The treatment factor (vehicle and macitentan) and exposure factor (sham or IH) both have two levels. The control + macitentan condition was omitted because macitentan has no effects in control rats (22). All data are expressed as means ± SE. A minimum of P < 0.05 was used as the statistical significance level. Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test for multiple comparisons among groups. Condition and factor interaction were assessed by two-way ANOVA analysis followed by Tukey’s post hoc test. Two-way ANOVA for repeated measures was used to compare the effects over time. The analysis used is indicated in each figure.

RESULTS

Macitentan Prevents Hypertension in the CKD-Sleep Apnea Rat Model

The effect of macitentan on hemodynamic variables was evaluated in rats from control and CKD groups under sham or IH exposure. Hemodynamic variables were assessed after 2 wk of adenine diet, after 2 wk of recovery diet, and every week after recovery (7 wk).

Mean arterial blood pressure.

After 2 wk of adenine diet, rats from both CKD and CKD/IH groups had a significant increase in MAP (~30 mmHg) above baseline (Fig. 1A). MAP of macitentan-treated rats remained unchanged (CKD + macitentan and CKD + macitentan/IH) compared with baseline and their respective control groups. After 2 wk of recovery diet, MAP remained elevated only in rats from CKD groups that were not receiving macitentan. As previously shown, IH exposure (control/IH) increased MAP from baseline (~10 mmHg) and exacerbated the MAP increase in rats in the CKD/IH group (31). Rats from the macitentan/IH group had a smaller increase in MAP than IH alone (~5 mmHg), and macitentan prevented the exacerbation in CKD-induced increases in MAP (Fig. 1A). Rats exposed to IH showed nondipping MAP, as previously reported (24, 31).

Fig. 1.

Macitentan (MACI) prevents increases in mean arterial pressure (MAP) in a rat model of chronic kidney disease (CKD) and sleep apnea. Over 7 wk of study, weekly 2-day averages of MAP (A) and heart rate (HR; B) and weekly body weight (C) and hematocrit (Hct%; D) are shown. Data are expressed as means ± SE. Results were analyzed by two-way repeated-measures ANOVA followed by Tukey’s multiple-comparison test. A: *P < 0.0001, CKD/sham and CKD/intermittent hypoxia (IH) vs. control/sham groups; +P < 0.0001, CKD + MACI/sham vs. CKD/sham or CKD + MACI/IH vs. CKD/IH groups. #Times 4–7 vs. time 3 (P = 0.02 control/IH; P < 0.01 MACI/IH; P = 0.03 CKD/IH; P < 0.02 CKD + MACI/IH). B: *P < 0.05 all CKD groups vs. the control/sham group. C: *P < 0.0005, all CKD groups vs. the control/sham group. D: *P < 0.005, all CKD groups vs. the control/sham group. *Times 5–6 P < 0.02 CKD/sham and CKD + MACI/sham vs. control/sham groups. n = 7 in all groups except for n = 6 in the control/IH group. B, baseline; A, after 2-wk adenine diet; R, after 2 wk of recovery.

Heart rate.

Adenine-fed rats had lower HR compared with baseline (CKD/sham, CKD + macitentan/IH, and CKD/IH vs. control/sham, control/IH, and macitentan/IH), which normalized after 2 wk of recovery diet, and HR was not different between groups thereafter (Fig. 1B).

Body weight.

Body weight was significantly lower in adenine-fed rats (CKD/sham, CKD + macitentan/sham, CKD/IH, and CKD + macitentan/IH vs. control/sham, control/IH, and macitentan/IH). However, after 2 wk of recovery diet and in all subsequent weeks, all groups had similar weight levels and a normal growth rate (Fig. 1C).

Hematocrit.

The hematocrit percentage was reduced after 2 wk of recovery diet in all CKD groups compared with controls, but there was no significant difference between groups at week 7. As previously reported by our group (34), eucapnic-IH did not affect the hematocrit percentage (Fig. 1D). Furthermore, macitentan did not affect the hematocrit percentage (Fig. 1D).

Right and left ventricular hypertrophy.

There were no statistical differences in the right ventricle and left ventricle + septum over body weight ratio among groups (Table 1).

Table 1.

Heart and kidney weights at 7 wk

	Right Ventricle/Body Weight, mg/g	Left Ventricle/Body Weight, mg/g	Right Kidney/Body Weight, mg/g	Left Kidney/Body Weigh, mg/g
Control/sham	0.52 ± 0.03	2.07 ± 0.07	2.80 ± 0.09	3.05 ± 0.34
CKD/sham	0.56 ± 0.03	2.25 ± 0.04	3.19 ± 0.15	3.84 ± 0.17
CKD + MACI/sham	0.54 ± 0.03	2.12 ± 0.05	4.68 ± 0.39*	4.89 ± 0.22*
Control/IH	0.54 ± 0.02	2.09 ± 0.05	3.08 ± 0.25	3.06 ± 0.25
MACI/IH	0.54 ± 0.03	2.17 ± 0.05	3.17 ± 0.17	3.14 ± 0.13
CKD/IH	0.57 ± 0.01	2.16 ± 0.06	3.70 ± 0.48	4.14 ± 0.44
CKD + MACI/IH	0.54 ± 0.02	2.12 ± 0.06	4.39 ± 0.31*	4.97 ± 0.22*

Data are means ± SE; n = 7 in all groups except for the control/intermittent hypoxia (IH) group with n = 6. CKD, chronic kidney disease; MACI, macitentan.

^*P < 0.0001 vs. the control/sham group.

Macitentan Does Not Prevent Adenine-Induced Kidney Dysfunction

BUN, blood and urinary creatinine, urinary volume, clearance of creatinine, and proteinuria were determined to evaluate renal function.

Blood urea nitrogen.

As previously reported (31), all rats from CKD groups had elevated BUN after 2 wk of adenine diet, and BUN remained increased but lower after 2 wk of recovery diet until the end of the study (Fig. 2A). Macitentan did not affect BUN in any group.

Fig. 2.

Macitentan (MACI) does not prevent adenine-induced kidney dysfunction. Over 7 wk of study, weekly blood urea nitrogen (BUN) (A) and blood creatinine (B) are shown. At time 7, urinary volume (C), free water clearance (C_H2O) (D), creatinine clearance (C_Cre; E), and urinary protein (F) are shown. Data are expressed as means ± SE. A and B show results analyzed by two-way repeated-measures ANOVA. C–F show results analyzed by one-way ANOVA followed by a Tukey’s multiple-comparison test. *P < 0.05 vs. the control/sham group. n = 7 in all groups except n = 6 in the control/intermittent hypoxia (IH) group. B, baseline; A, after 2-wk adenine-diet; R, after 2 wk of recovery; CKD, chronic kidney disease.

Blood creatinine.

After 2 wk of adenine diet, blood creatinine increased (1.4–1.5 mg/dl) in all CKD groups compared with baseline and compared with control groups. After 2 wk of recovery diet, creatinine decreased (0.7 mg/dl) but remained elevated above baseline and above control groups until the end of the study. Macitentan did not affect blood creatinine in rats in the CKD groups (Fig. 2B).

Urinary volume and free water clearance.

Urinary volume and free water clearance significantly increased in rats from CKD groups versus control groups. Macitentan treatment did not affect urinary volume and free water clearance (Fig. 2, C and D).

Urinary protein and eGFR.

eGFR significantly decreased and proteinuria increased in all CKD groups versus control/sham, control/IH, and macitentan/IH groups. Again, macitentan did not affect eGFR or proteinuria (Fig. 2G). A limitation of our study is that GFR was estimated using creatinine clearance instead of other more sensitive techniques (13).

Kidney weight/body weight.

Only the kidneys from CKD rats treated with macitentan showed an increase in the ratio of kidney weight to body weight (see Table 1).

CKD Increases Pre-Pro-ET-1 mRNA in the Renal Cortex

Pre-pro-ET-1 mRNA gives rise to pre-pro-ET-1, which undergoes a series of enzymatic cleavages to generate the mature ET-1 peptide (14). We assessed pre-pro-ET-1 mRNA levels in the kidneys (cortex and medulla) by RT-PCR. Pre-pro-ET-1 mRNA levels were significantly increased in the renal cortex of CKD/sham, CKD + macitentan/sham, and CKD + macitentan/IH versus control/sham groups (Fig. 3A). There was also a strong tendency (P = 0.07) for an increase in the CKD/IH group versus the control/sham group. However, pre-pro-ET-1 mRNA levels in the renal medulla were not different among groups (Fig. 3B).

Fig. 3.

Chronic kidney disease (CKD) increases pre-pro-endothelin (ET)-1 mRNA levels in the kidney cortex without affecting plasma ET-1 levels. At time 7, pre-pro-ET-1 mRNA levels in the kidney cortex (A), medulla (B), lungs (C), and ET-1 plasma levels (D) are shown. Data are expressed as means ± SE. Results were analyzed by one-way ANOVA followed by a Tukey’s multiple-comparison test. *P < 0.05 vs. the control/SHAM group. IH, intermittent hypoxia; MACI, macitentan.

Macitentan Increases Pre-Pro-ET-1 mRNA in the Lungs and ET-1 in the Plasma

The lungs are the main site of clearance and synthesis of ET-1. ET-1 clearance occurs via binding to and internalization of ET type B (ET_B) receptors (18). Previously, it has been shown that macitentan enhances synthesis and reduces clearance of ET-1 (22). Accordingly, pre-pro-ET-1 mRNA levels only increased in the lungs of rats from CKD + macitentan/sham versus control/sham groups. Surprisingly, this upregulation was not significant in the lungs of other macitentan-treated rats (Fig. 3C). However, plasma ET-1 levels were increased in all macitentan-treated rats (Fig. 3D).

Macitentan Does Not Prevent Adenine-Induced Kidney Fibrosis

Kidney fibrosis was assessed in Heidenhain’s AZAN-stained kidney sections in all groups. Figure 4 shows representative images of whole kidney sections at low magnification and zoomed in for the cortex and medulla regions with and without fibrosis mask generated with HALO software. The percentage of fibrotic area (percent blue stain) was greater in all adenine-fed rats (CKD groups), as previously reported (31), whereas macitentan treatment did not affect the area of fibrosis.

Fig. 4.

Macitentan (MACI) does not prevent adenine-induced kidney interstitial fibrosis. Representative images of AZAN-stained kidneys to identify fibrosis (blue) in each of the seven groups are shown. Images on the left are ×4, in the middle are ×20 with no mask, and on the right are ×20 and masked using the HALO system to indicate areas of fibrosis. A summary graph with HALO-calculated percentages of fibrosis for entire sagittal cross sections from 7−8 kidneys from each group is shown. Data are expressed as means ± SE. Results were analyzed by one-way ANOVA followed by a Tukey’s multiple-comparison test. *P < 0.05 vs. the control/sham group. CKD, chronic kidney disease; IH, intermittent hypoxia.

Macitentan Does Not Prevent Adenine-Induced Renal Inflammation

Because we have previously demonstrated that IH does not enhance CKD-induced inflammation, inflammatory markers were not evaluated in the rats from groups exposed to IH (31).

Renal CD3⁺ T cell infiltration.

T cell infiltration was evaluated by immunohistochemically labeling CD3, a pan marker of T cells. The percentage of CD3⁺ cells was increased in the kidneys of rats from the CDK versus control/sham groups. Macitentan had no effect on CD3⁺ cell infiltration (Fig. 5). Figure 5, A, D, and G, shows CD3⁺ cells in yellow and total cells in blue as detected by HALO software, whereas Fig. 5, B, E, and H, shows the same images without the mask. Figure 5, C, F, and I, shows a zoomed-in region.

Fig. 5.

Macitentan (MACI) does not prevent chronic kidney disease (CKD)-induced increases in renal CD3⁺ T cell infiltration. Left: total cell count (blue) and CD3⁺ signal (yellow) as detected by HALO software. Right: zoomed-in region depicted by the square in the middle. The summary graph shows HALO-calculated percentages of CD3⁺ cells. Data are expressed as means ± SE. Results were analyzed by one-way ANOVA followed by a Tukey’s multiple-comparison test. *P < 0.05 vs. the control/sham group.

Renal macrophage (CD68⁺) cell infiltration.

The percentage of CD68⁺ cells (macrophage marker) demonstrated that macitentan treatment did not affect adenine-induced macrophage renal infiltration (Fig. 6).

Fig. 6.

Macitentan (MACI) does not prevent adenine-induced increases in renal macrophage (CD68⁺). Left: total cell count (blue) and CD68⁺ signal (yellow) as detected by HALO software. Right: zoomed-in region depicted by the square in the middle. The summary graph shows HALO-calculated percentages of CD68⁺ cells. Data are expressed as means ± SE. Results were analyzed by one-way ANOVA followed by a Tukey’s multiple-comparison test. *P < 0.02 vs. the control/sham group. CKD, chronic kidney disease.

IL-6 and NOS2 mRNA levels.

Figure 7 shows that IL-6 and NOS2 mRNA levels (inflammatory markers) were significantly increased in the renal cortex of rats from the CKD groups, whereas only NOS2 was increased in the medulla. Macitentan did not affect adenine-induced upregulation of IL-6 and NOS2.

Fig. 7.

Macitentan (MACI) does not prevent adenine-induced kidney inflammation. IL-6 and nitric oxide synthase isoform 2 (NOS2) mRNA levels were measured in the kidney cortex and medulla. Data are expressed as means ± SE. Results were analyzed by one-way ANOVA followed by a Tukey’s multiple-comparison test. *P < 0.01 vs. the control/sham group. CKD, chronic kidney disease.

Overall, these results show that the dual blockade of ET receptors with macitentan does not prevent adenine-induced renal inflammation.

Macitentan Reverses Hypertension But Not Renal Dysfunction in Rats From IH, CKD, and CKD/IH Groups

To determine whether macitentan reverses established hypertension and renal failure, macitentan was administered in the food to rats from CKD, IH, and CKD/IH groups. Figure 8A shows that macitentan lowered MAP in IH-exposed (macitentan/IH) and IH-exposed adenine-treated rats (CKD + macitentan/sham and CKD + macitentan/IH). HR was not different among groups (Fig. 8B).

Fig. 8.

Macitentan (MACI) reverses hypertension but not renal dysfunction in rats from chronic kidney disease (CKD) and CKD/intermittent hypoxia (IH) groups. A: mean arterial pressure (MAP). B: heart rate (HR). C: blood urea nitrogen (BUN). D: blood creatinine. E: creatinine clearance (C_Cre). F: urinary protein. n = 7 except for MACI/IH and CKD/IH with n = 6. Data are expressed as means ± SE. Results were analyzed by one-way ANOVA followed by a Tukey’s multiple-comparison test. MAP decreased in all rats that received MACI (CKD + MACI/sham, MACI/IH, and CKD + MACI/IH groups, times 1–2 vs. time 0, P < 0.05). +P < 0.05, CKD groups without MACI vs. CKD groups with MACI. *P < 0.05, CKD groups vs. the control/sham group.

However, renal function did not improve in rats treated with macitentan (Fig. 8, C–F). Body weight did not change in any of the groups, and CKD decreased the hematocrit percentage, which was not affected by macitentan (Table 2).

Table 2.

Reversal study: Hct% and body weight over time

	Time 0		Time = 1 wk		Time = 2 wk
	Body weight, g	Hct%	Body weight, g	Hct%	Body weight, g	Hct%
Control/sham	349 ± 12	47 ± 1	361 ± 13	47 ± 1	370 ± 12	46 ± 0
CKD/sham	338 ± 9	39 ± 2*	358 ± 8	41 ± 1*	368 ± 7	40 ± 1*
CKD + MACI/sham	336 ± 10	36 ± 1*	359 ± 10	37 ± 1*	373 ± 10	36 ± 1*
Control/IH	340 ± 11	47 ± 1	357 ± 8	49 ± 1	366 ± 7	48 ± 1
MACI/IH	350 ± 14	47 ± 1	368 ± 16	47 ± 1	379 ± 13	48 ± 1
CKD/IH	316 ± 7	41 ± 1*	333 ± 8	43 ± 1	347 ± 10	44 ± 1
CKD + MACI/IH	320 ± 9	41 ± 0*	345 ± 9	39 ± 1*	355 ± 9	41 ± 1

Data are means ± SE; n = 7 in all groups except for the control/intermittent hypoxia (IH) group with n = 6. Macitentan (MACI) diet was started after time 0. CKD, chronic kidney disease; Hct%, hematocrit.

^*P < 0.005 vs. the control/sham group.

DISCUSSION

The present study tested the hypothesis that a dual ET receptor antagonist, macitentan, would attenuate the development of hypertension and kidney dysfunction in a model of combined CKD and sleep apnea. The results show that macitentan both prevented and reversed hypertension in both CKD and CKD/IH rats, but it did not improve or reverse adenine-induced tubular damage, kidney inflammation, fibrosis, or function.

Importantly, all rats from CKD groups not treated with macitentan developed hypertension, increased renal cortex pre-pro-ET-1 mRNA, proteinuria, decreased eGFR, elevated fluid loss, kidney fibrosis, and inflammation. In addition, IH exacerbated the CKD-induced hypertension but not the kidney dysfunction and damage, as recently shown (31).

ET_A receptor activation causes vasoconstriction, cell proliferation, and extracellular matrix accumulation, whereas ET_B receptor activation promotes antiproliferative and antifibrotic pathways and vasodilation (for a review, see Ref. 26). The beneficial actions of ET_B receptors are in part mediated by ET-1 clearance and in part through activation of NO synthesis (26). In the kidney, ET-1 constricts afferent arterioles through ET_A receptors but dilates efferent arterioles through ET_B receptors. The ET_A-mediated vasoconstriction decreases GFR, which could lead to elevated blood pressure (40). However, ET_B receptors in the collecting ducts enhance diuresis and natriuresis, lowering blood pressure (10, 25, 36). In a renal artery stenosis model of chronic renovascular disease in swine, selective ET_A receptor blockade slowed the progression of renal injury to preserve renal hemodynamics, function, and microvascular density without changes in blood pressure (25). In a followup study, the same group showed that, in established renovascular disease, ET_A receptor blockade decreased renal inflammation and fibrosis and decreased but did not normalize blood pressure (10). However, ET_B receptor blockade did not prevent or reverse renovascular disease (10). Furthermore, macitentan improved kidney function and decreased renal damage in the same model of chronic renovascular disease (36).

These prior reports demonstrate that ET-1 participates in the development of fibrosis, proteinuria, and CKD progression caused by renovascular disease (15, 26). However, the results of our study demonstrate that macitentan does not prevent or reverse the kidney dysfunction caused by adenine-induced tubule damage despite restoring blood pressure to normal levels. The increased arterial pressure in adenine-treated CKD rats thus appears to be dependent on renal-generated ET-1, which activates ET_A receptors outside of the kidney, independent of GFR regulation. However, elevated ET-1 levels were observed at the end of the study but not evaluated at the time of induction of the tubular damage. Our data suggest that tubular damage by adenine led to renal synthesis of ET-1, which, in turn, elevated arterial pressure. However, we cannot be certain that the increase in blood pressure did not precede ET-1 upregulation. It is also possible that the lack of renal protection by macitentan was due to blockade of ET_B receptors to mask any ET_A protection by diminishing ET_B activation of antifibrotic and anti-inflammatory pathways (for a review, see Ref. 26). If that is the case, then it is conceivable that a selective ET_A blocker or combination therapy of dual ET receptor blocker and anti-inflammatory drugs might better treat CKD of tubular origin.

Selective ET_A and some dual ET receptor blockers have been reported to cause fluid retention (26). More recent studies have suggested that the new generation of dual blockers is less likely to cause fluid retention by preventing ET_B overactivation after ET_A blockade (39). Consistently, macitentan did not cause fluid retention in our study, evidenced by no effect on free water clearance. Thus, mixed blockers offer some advantages as antihypertensive agents.

Macitentan has previously been shown to enhance synthesis and reduce clearance of ET-1 (22). Consistently, pre-pro-ET-1 mRNA levels increased in the lungs of rats from CKD + macitentan/sham and macitentan/IH versus control/sham groups. Surprisingly, this upregulation was not observed in the lungs of rats from the CKD + macitentan/IH group (Fig. 3C). However, plasma ET-1 levels were increased in all macitentan-treated rats (Fig. 3D). The reason for the lack of ET-1 upregulation in the CKD + macitentan/IH group is unclear but might be due to the timing of tissue collection for mRNA measurements, weeks after the initiation of the therapy, so that increases immediately after start of therapy could have declined by the time of tissue collection.

HR and body weight were significantly decreased in all rats that received adenine diet, but this was restored during the following period with regular diet. Body temperature (data not shown) also decreased, suggesting that the decreased HR was a consequence of reduced body weight and basal metabolism (27). It is likely that low palatability of the adenine-containing diet lowered food consumption. To determine whether weight loss was a confounding factor, adenine was administered by oral gavage in the reversal study, which did not reduce body wieght and HR but did cause similar increases in blood pressure and kidney damage.

Anemia is an early symptom in patients with CKD (33). As previously reported by others (2, 30), adenine-induced CKD causes anemia. In our study, adenine caused a decrease in the hematocrit percentage that was evident after the recovery period (week 3 in prevention study) when BUN levels were 50% lower than their maximal increase. The hematocrit percentage showed a progressive recovery to normal values by the end of the study. IH accelerated the recovery from the anemia. However, IH alone had no effect on the hematocrit percentage, as previously reported (34). In that prior study, we demonstrated that maintenance of eucapnia by provision of supplemental CO₂ attenuates IH-induced polycythemia. This result is consistent with reports showing that polycythemia is rarely observed in patients with OSA (12, 16, 19).

Among the mechanisms that explain the anemia in CKD, it has been proposed that renal pericytes, which produce erythropoietin, transdifferentiate to myofibroblasts with decreased erythropoietin production and therefore erythropoiesis (33, 35). However, myofibroblasts are plastic and can recover their function over time (35), especially if the insult is ameliorated. Furthermore, hypoxia through hypoxia-inducible factor 1 is a potent stimulus for erythropoietin production (35), which could account for the accelerated recovery in the IH group of rats (CKD/IH), although rats from the IH group did not show polycythemia. Uremia has also been shown to reduce red blood cell lifespan, and the CKD/IH group of rats also showed a progressive recovery on BUN and eGFR in the reversal study, parallel to the increase in the hematocrit percentage. Macitentan did not affect the hematocrit percentage, providing additional evidence that the blockade of ET receptor does not improve CKD-induced kidney dysfunction.

In conclusion, macitentan could be an effective antihypertensive treatment in patients with CKD with irreversible kidney damage in whom lowering blood pressure could prevent secondary organ damage. Direct tubule damage, such as that occurring in drug-induced crystalline nephropathy, can occur in the absence of elevated blood pressure.

GRANTS

This work was supported by Actelion Pharmaceuticals (to L. V. Gonzalez Bosc and N. L. Kanagy).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

N.L.K. and L.V.G.B. conceived and designed research; H.M.-L., D.J., A.B., P.J.M., J.G., C.P., M.M., N.L.K., and L.V.G.B. performed experiments; H.M.-L., D.J., A.B., P.J.M., C.P., M.M., N.L.K., and L.V.G.B. analyzed data; H.M.-L., D.J., C.P., N.L.K., and L.V.G.B. interpreted results of experiments; H.M.-L., N.L.K., and L.V.G.B. prepared figures; H.M.-L., N.L.K., and L.V.G.B. drafted manuscript; H.M.-L., D.J., A.B., P.J.M., C.P., M.M., N.L.K., and L.V.G.B. edited and revised manuscript; H.M.-L., D.J., A.B., P.J.M., C.P., M.M., N.L.K., and L.V.G.B. approved final version of manuscript.