Deletion of the aryl hydrocarbon receptor in endothelial cells improves ischemic angiogenesis in chronic kidney disease

Abstract

Chronic kidney disease (CKD) is a strong risk factor for peripheral artery disease (PAD) that is associated with worsened clinical outcomes. CKD leads to the accumulation of tryptophan metabolites that are associated with adverse limb events in PAD and are ligands of the aryl hydrocarbon receptor (AHR), which may regulate ischemic angiogenesis. To test if endothelial cell-specific deletion of the AHR (AHR^ecKO) alters ischemic angiogenesis and limb function in mice with CKD subjected to femoral artery ligation. Male AHR^ecKO mice with CKD displayed better limb perfusion recovery and enhanced ischemic angiogenesis compared with wild-type mice with CKD. However, the improved limb perfusion did not result in better muscle performance. In contrast to male mice, deletion of the AHR in female mice with CKD had no impact on perfusion recovery or angiogenesis. With the use of primary endothelial cells from male and female mice, treatment with indoxyl sulfate uncovered sex-dependent differences in AHR activating potential and RNA sequencing revealed wide-ranging sex differences in angiogenic signaling pathways. Endothelium-specific deletion of the AHR improved ischemic angiogenesis in male, but not female, mice with CKD. There are sex-dependent differences in Ahr activating potential within endothelial cells that are independent of sex hormones.

NEW & NOTEWORTHY This study provides novel insights into the mechanisms by which chronic kidney disease worsens ischemic limb outcomes in an experimental model of peripheral artery disease. Deletion of the aryl hydrocarbon receptor (AHR) in the endothelium improved ischemic angiogenesis suggesting that AHR inhibition could be a viable therapeutic target; however, this effect was only observed in male mice. Subsequent analysis in primary endothelial cells reveals sex differences in Ahr activating potential independent of sex hormones.

INTRODUCTION

Peripheral arterial disease (PAD) is a chronic disease primarily affecting the lower extremities, which is caused by atherosclerotic narrowing or occlusion of blood vessels. An estimated 8–12 million adults in the United States and ∼200 million globally have PAD (1–3). The obstruction of blood flow in the lower extremities can manifest via an array of symptoms including claudication, ischemic rest pain, gangrene, and nonhealing ulcers. Conventional risk factors for the development of PAD include smoking, age, hypertension, dyslipidemia, diabetes, and physical activity (4). Several nonconventional risk factors including inflammation, HIV, and exposure to toxic metals and air pollution have also emerged. In addition, an underappreciated risk factor that increases the risk of developing PAD is chronic kidney disease (CKD) (5). Recently studies have estimated that ∼25–35% of patients with CKD develop PAD (1, 2, 6). A meta-analysis of 44,138 patients demonstrated that patients with PAD and CKD have an approximate twofold increase in major limb amputation and ∼2.55-fold increase in mortality risk (7). These observations agree with an earlier examination of a large cohort of male veterans, which reported that patients with PAD and renal insufficiency were more likely to present with ischemic ulceration/gangrene and had greater mortality rates compared with patients with PAD and normal renal function (8). Unfortunately, the success rates of endovascular and surgical interventions are significantly lower in patients with PAD and CKD compared with those with normal kidney function (1, 8–10). Despite the widespread prevalence of renal insufficiency in patients with PAD and worsened clinical outcomes, most clinical trials investigating novel therapeutics in PAD exclude patients with CKD. Thus, there is a significant gap in knowledge in understanding the coalescence of these diseases.

Although an association between CKD and PAD exists, the mechanisms linking the two pathologies are ill-defined. Independent of PAD, CKD has also been shown to negatively impact the vasculature, including a reduction in capillary density in the heart (11) and skeletal muscle (12, 13). In addition, rodents with CKD have impaired angiogenic responses to ischemia (14–16) but mechanistic data to show how these impairments arise remain limited. Other factors that contribute to PAD pathogenesis and worsen limb outcomes include microvascular disease (17, 18), decreased vasodilatory capacity (19, 20), and increased inflammatory response in the endothelial cells (21), all of which may be exacerbated in patients with renal insufficiency. The retention of uremic solutes in the blood and tissues is a hallmark of CKD (22) and a recent study in patients with PAD reported that plasma concentrations of tryptophan-derived uremic solutes were strongly associated with an increased risk of a major adverse limb event (16). Most notably, plasma concentrations of indoxyl sulfate, kynurenine, and kynurenic acid had hazard ratios of 2.2, 4.2, and 2.5, respectively. Reports in the literature have provided evidence that indoles and kynurenine metabolites, which accumulate in CKD, can impair nitric oxide-dependent vasodilation and angiogenesis (23–26), but the molecular mechanisms are incompletely understood. Several tryptophan-derived uremic metabolites (27–29) have been identified as ligands for the aryl hydrocarbon receptor (AHR), a ligand-activated transcription factor involved in xenobiotic metabolism. Elevated expression of the AHR is present in humans and rodents with CKD (30). Activation of the AHR using high-affinity ligand 2,3,7,8-tetrachlorodibenzo-p-dioxin has been shown to promote atherosclerosis (31, 32). AHR has also been reported to regulate angiogenesis, although there are discrepancies in the literature with some studies reporting antiangiogenic (16, 23, 33, 34) and others reporting proangiogenic (35, 36) effects, suggesting that these effects could be tissue- and ligand-specific. However, in a recent study where female mice with CKD were subjected to hindlimb ischemia, treatment with an AHR inhibitor (CH223191) improved perfusion recovery and capillary densities in the ischemic limb (16). In this study, we sought to determine the impact of AHR activation, specifically in endothelial cells, on ischemia-induced angiogenesis and myopathy in mice with CKD. It was hypothesized that an endothelial cell-specific conditional deletion of the AHR would improve ischemic limb outcomes in mice with CKD.

MATERIALS AND METHODS

Animals

An endothelial cell-specific AHR knockout mouse (AHR^ecKO) was generated by breeding floxed AHR mice (AHR^tm3.1Bra/J, Jackson Laboratories, Stock No. 006203) with Cdh5(PAC)-Cre^ERT2 mice (Taconic, Stock No. 13073), both strains maintained on a C57BL6J background. At 12 wk of age of mice, the conditional endothelial-specific AHR deletion was induced by delivering intraperitoneal injections of tamoxifen (80 mg/kg) for consecutive 5 days. Littermates without the Cre transgene (AHR^fl/fl) were injected with tamoxifen and used as controls. Both the researchers and surgeon were blinded to the genotype and grouping of the animals. All experiments involved male and female mice. Female mice were ovariectomized 2 wk before enrollment to mimic the sex hormone-deficient condition seen in postmenopausal patients with PAD. Buprenorphine (0.05 mg/kg) was given postoperatively for analgesia. All mice were housed in temperature- (22°C) and light-controlled (12-h:12-h light/dark) rooms and maintained on standard chow (Envigo Teklad Global 18% Protein Rodent Diet 2918 irradiated pellet) with free access to food and water before enrollment. All animal experiments adhered to the Guide for the Care and Use of Laboratory Animals from the Institute for Laboratory Animal Research, National Research Council (Washington, D.C., National Academy Press). To validate endothelium-specific knockout of the Ahr, genomic DNA was extracted from CD31⁺ cells isolated from skeletal muscle and liver from AHR^fl/fl and AHR^ecKO mice using Qiagen DNeasy kit (Qiagen, Cat. No. 69504) according to the manufacturer’s instructions. DNA was amplified using Terra PCR Direct Red Dye Premix (Takara Cat. No. 639286) and ran on a 1% agarose gel to confirm DNA recombination. Primer sequences are listed in Supplemental Table S1 (all Supplemental materials may be found at https://doi.org/10.6084/m9.figshare.24050244).

Endothelial Cell Isolation

Primary CD31-positive cells were isolated from mouse liver and skeletal muscle as previously described (37). Briefly, transcardial perfusions with ice-cold phosphate-buffered saline (PBS) were performed via the left ventricle at a rate of 2 mL/min for 5 min immediately followed by perfusion with digestion buffer, consisting of 0.1% wt/vol collagenase type I, 0.1% wt/vol collagenase type II, 0.25 U/mL Dispase, and 7.5 µg/mL DNAse in DMEM. All hindlimb skeletal muscles and the liver were next dissected and minced separately on ice. The minced liver and skeletal muscle were then transferred separately into 15 mL of digestion buffer and incubated at 37°C for 30 min. The digestion was stopped by adding 8 mL of wash buffer (0.5% wt/vol bovine serum albumin and 2 mM EDTA in PBS). Cell suspensions were then filtered through a 100-µm cell strainer and centrifuged at 300 g for 7 min. The supernatant was discarded, and the remaining cell pellet was resuspended in 5 mL of wash buffer and centrifuged at 300 g for 5 min. The supernatant was discarded, and this step was repeated until the supernatant was clear. The final pellet was resuspended in 90 µL of wash buffer and 10 µL of CD31 Microbeads (Miltenyi Biotec, Cat. No. 130-097-418) were added to the suspension and incubated on ice for 20 min. Following microbead incubation, the suspension was washed, and the pellet was resuspended in 0.5 mL of wash buffer. Suspension was passed through LS columns (Miltenyi Biotec, Cat. No. 130-042-401) on a magnetic separator (Miltenyi Biotec, Cat. No. 130-091-051). The columns were removed from the magnet and a fraction containing CD31-positive cells was eluted using 5 mL of wash buffer. CD31-positive fraction was centrifuged at 300 g for 10 min, the supernatant was discarded, and the remaining cell pellet was used for subsequent experiments.

Induction of Chronic Kidney Disease

All mice were fed a casein control diet for a 7-day acclimation period before being randomly assigned to remain on the casein diet or to be fed a 0.2% adenine-supplemented diet to induce CKD (38–40). Mice remained on their respective diet for 4 wk before femoral artery ligation (FAL) and during the 2-wk post-FAL recovery period. All diets were provided ad libitum for the duration of the study.

Assessment of Kidney Function

Kidney function was evaluated via glomerular filtration rate (GFR) and blood urea nitrogen (BUN) levels as previously described (38–43). In brief, GFR was evaluated by the rate of inulin-FITC clearance as previously described (41, 42). Inulin-FITC (Millipore-Sigma, Cat. No. F3272) was dissolved in 0.9% NaCl (5% wt/vol), and the solution was dialyzed with a 1,000-kDa dialysis membrane at room temperature in the dark for 24 h, followed by sterile filtering through a 0.22 µm filter. Mice were briefly anesthetized under isoflurane and FITC-inulin (2 µL/g body wt) was injected into the retroorbital sinus. Blood was then collected in heparinized capillary tubes via a ∼1-mm tail snip at 3, 5, 7, 10, 15, 35, 56, and 75 min following FITC-inulin injection. During blood collection, the mouse was conscious and allowed free movement within their cage. Blood was centrifuged for 10 min at 4,000 rpm at 4°C, the resulting plasma was collected and diluted (1:20) in 0.5 mol/L of HEPES buffer (pH 7.4) and loaded into a 96-well plate. The fluorescence was detected using a BioTek Synergy II plate reader against a FITC-inulin standard curve. GFR was calculated using a two-phase exponential decay curve fit in GraphPad Prism. At euthanasia, blood was collected via cardiac puncture, allowed to clot, and centrifuged at 4,000 g for 10 min at 4°C. The resulting serum was diluted (1:25 in DiH₂O) and BUN was measured using a commercial assay kit (Arbor Assays, Cat. No. K024).

Animal Model of Peripheral Artery Disease

Femoral artery ligation (FAL) (38, 44) was performed by anesthetizing mice with intraperitoneal injection of ketamine (90 mg/kg)-xylazine (10 mg/kg) and surgically inducing unilateral hindlimb ischemia by placing silk ligatures on the femoral artery just distal the inguinal ligament and immediately proximal to the saphenous and popliteal branches. Buprenorphine (0.05 mg/kg) was given postoperatively for analgesia.

Limb Perfusion Assessment

Limb perfusion was assessed by laser-Doppler flowmetry (moorVMS-LDF, Moor Instruments) before surgery, postsurgery, 3 days postsurgery, 7 days postsurgery, and just before euthanasia as previously described (38, 45, 46). Both hindlimbs were shaved and the laser-Doppler probe was placed ∼1 to 2 mm away from the middle of the posterior side of the paw and the posterior side of the lateral head of the gastrocnemius muscle. Perfusion recovery was reported as a percentage of the nonischemic limb.

Assessment of Hindlimb Grip Strength

Unilateral hindlimb grip strength was measured using a Grip Strength Test Instrument (BIOSEB; Model No. BIO-GS3) before surgery, 3 days postsurgery, 7 days postsurgery, and just before euthanasia. The mice were allowed to firmly grip a metal T-shaped bar with a single hindlimb paw and then were pulled straight back with increasing force until the mouse released the bar. Three trials were performed on both the control and surgical limbs. The trial with the highest force was used for analysis, and grip strength was reported as a percentage of nonsurgical control limb.

Nerve-Mediated Muscle Contraction

The maximal twitch and tetanic force levels of the tibialis anterior (TA) muscle were measured in situ by stimulation of the sciatic nerve as previously described (47, 48). Briefly, mice were anesthetized with an intraperitoneal injection of xylazine (10 mg/kg)- ketamine (100 mg/kg). The distal tendon was tied using a 4-0 silk suture attached to the lever arm of the force transducer (Cambridge Technology; Model No. 2250). Muscle contractions were elicited by stimulating the sciatic nerve via bipolar electrodes using square wave pulses of 0.02 ms (Aurora Scientific, Model 701 A stimulator). Lab View-based DMC program (version 5.500) was used for data collection and servomotor control. After optimal length of the muscle was obtained using twitch contractions, three isometric tetanic forces were performed using 500-ms supramaximal electrical pulses at a stimulation frequency of 150 Hz with at least 1 min of rest between contractions. The highest force among the three measurements was reported as the peak tetanic force. Following tetanic contractions, three isometric twitch contractions (1 Hz) were performed. Peak twitch and tetanic force levels were reported as absolute and specific (normalized to muscle weight) force levels.

Isolation of Skeletal Muscle Mitochondria

Skeletal muscle mitochondria were isolated as previously described (39, 43). The gastrocnemius was dissected, cleaned of connective tissue, and minced into a fine paste on ice. The minced muscle was next incubated in ice-cold mitochondrial isolation media (MIM), consisting of 50 mM MOPS, 100 mM KCl, 1 mM EGTA, and 5 mM MgSO₄, containing 0.025% wt/vol trypsin for 5 min, followed by centrifugation at 500 g for 5 min at 4°C. The resulting supernatant containing trypsin was decanted and the pellet was resuspended with MIM containing 0.02% wt/vol bovine serum albumin (BSA 2 g/L). The sample was then homogenized on ice using a glass Teflon homogenizer and subsequently centrifuged at 800 g for 10 min at 4°C. The resulting supernatant was collected and centrifuged again at 10,000 g at 4°C resulting in a mitochondrial-rich pellet. The pellet was washed with MIM to remove damaged mitochondria and gently resuspended in MIM without BSA. Protein concentration of the resuspension was assessed using bicinchoninic acid protein assay (Thermo Fisher Scientific, Cat. No. A53225)

Assessment of Mitochondrial Oxygen Consumption

High-resolution respirometry was performed using an Oroboros Oxygraph-2k (O2K) to measure oxygen consumption (JO₂) at 37°C. Twenty micrograms of mitochondria were added to the O2K chamber in 2 mL of buffer D, consisting of 105 mM K-MES, 30 mM KCl, 1 mM EGTA, 10 mM K₂HPO₄, 5 mM MgCl₂-6H₂O, and 2.5 mg/mL BSA (pH 7.2), supplemented with 5 mM creatine monohydrate. Mitochondria were energized by the addition of 5 mM pyruvate, 2.5 mM malate, and 0.2 mM octanoylcarnitine. We next added a clamp system containing ATP (5 mM), phosphocreatine (PCr; 1 mM), and creatine kinase (CK; 20 U/mL), which couples the interconversion of ATP and ADP to that of phosphocreatine (PCr) and free creatinine, to titrate the extramitochondrial ATP:ADP ratio, thus free energy of ATP hydrolysis (ΔG_ATP), to measure mitochondrial oxygen consumption at physiologically relevant levels of energy demand as done previously (49). Exogenous cytochrome-c (10 mM) was used to assess the outermembrane integrity of isolated mitochondria, and samples with more than a 25% increase in oxygen consumption were excluded from this study. The ΔG_ATP was plotted against the corresponding JO₂, and the slope was used to represent conductance throughout mitochondrial oxidative phosphorylation (OXPHOS), where lower conductance indicates impaired mitochondrial energetics.

Skeletal Muscle Morphology and Ischemic Lesion Area

The tibialis anterior muscle (TA) from both limbs was removed, embedded in optimal cutting temperature (OCT) compound, and immediately frozen in liquid nitrogen-cooled isopentane for cryosectioning. Using a Leica 3050S cryotome, 10-µm-thick transverse sections of the TA were cut and mounted on microscope slides. Skeletal muscle morphology and ischemic lesion area were assessed using light microscopy and standard methods of hematoxylin and eosin staining (H&E). Slides were imaged at ×20 magnification with an Evos FL2 Auto microscope (ThermoFisher Scientific), and tiled images of the entire section were obtained for analysis in using ImageJ software. The ischemic lesion area was quantified by manually tracing areas of the muscle section with signs of ischemic injury (necrotic or regenerating myofibers). The nonmyofiber area was measured by thresholding the image to quantify the area tissues between myofibers, which was expressed as a percentage of the total section area. Regenerating myofibers were quantified by manually counting myofibers that contained centralized nuclei by a blinded investigator.

Perfused and Total Muscle Capillary Density

Mice received a retro-orbital injection of 50 µL of 1 mg/mL Griffonia simplicifolia lectin (GSL) isolectin B4, Dylight 649 (Vector Laboratories; Cat. No. DL-1208) to fluorescently label α-galactose residues on the surface of endothelial cells of perfused capillaries. Following the injection, animals were returned to their cage and allowed 1 to 2 h of free movement before euthanasia and muscle harvest. The tibialis anterior muscle was frozen in OCT compound and cryosectioned as described above. Transverse muscle sections were fixed with 4% paraformaldehyde (Thermo Fisher Scientific, Cat. No. J19943-K2) for 10 min, and permeabilized with 0.25% triton X-100 (Millipore-Sigma, Cat. No. 93443). Following three washes with PBS, slides were blocked in PBS + 5% goat serum + 1% BSA for 4–6 h. Total capillaries were labeled with a primary antibody raised against PECAM1 (anti-CD31, Abcam, Cat. No. ab28364, 1:100 dilution in blocking solution) overnight at 4°C. The following day, slides were washed with PBS and incubated for 1 h at room temperature with Alexa-Fluor555 anti-rabbit secondary antibody (Thermo Fisher Scientific, Cat. No. A11034, 1:250 dilution) and wheat germ agglutin (WGA) conjugated with Alexa-Flour488 1 mg/mL (ThermoFisher Scientific, Cat. No. W11261, 1:100 dilution) to label myofiber membranes. The slides were next washed with PBS and subsequently coverslipped with Vectashield Hardmount containing DAPI (Vector Laboratories, Cat. No. H-1500-10). Images were obtained at ×20 magnification using an Evos FL2 Auto microscope (Thermo Fisher Scientific), and tiled/merged images of the entire muscle section were used for analysis. The number of perfused and total capillary density measurements was quantified on thresholded images using a particle counter in Fiji/ImageJ, and the quantified results were expressed as percent perfused and total number per area of the muscle section. Skeletal myofiber cross-sectional area (CSA) was determined using MuscleJ (7), an automated analysis software developed in Fiji.

Primary Endothelial Cell Culture

Primary endothelial cells were isolated from the liver of 4-wk-old male and female C57BL6J mice (Jackson Laboratory, Stock No. 000664). All experiments were performed in three biologically independent samples. Primary cells were grown to ∼80% confluency on 0.25% gelatin (ScienCell, Cat. No. 0423)-coated flasks in endothelial cell growth medium (PromoCell, Cat. No. C-22022) supplemented with 1% penicillin-streptomycin (ScienCell, Cat. No. 0503) using standard culture conditions (37°C with 5% CO₂). Cells were stained with anti-CD31 antibody (Abcam, Cat. No. ab28364, 1:100 dilution in blocking solution) and DAPI to validate that the cells were CD31-positive. Hypoxia and nutrient deprivation were induced by replacing culture media with Hank’s balanced salt solution (HBSS; ThermoFisher Scientific; Cat. No. 24020) and placing cells within a cake pan hypoxia chamber flushed with nitrogen gas for 10 min before sealing as previously described (50). To activate the AHR, cells were treated with indoxyl sulfate (100 µM, Millipore-Sigma, Cat. No. 13875) or vehicle control DMSO (ThermoFisher Scientific; Cat. No. BP231).

RNA Isolation and Quantitative PCR

Total RNA was isolated from primary murine endothelial cells and mouse extensor digitorum longus muscle using TRIzol (Invitrogen, Cat. No. 15–596-018) and a Direct-zol RNA MiniPrep kit (Zymo Research, Cat. No. R2052). RNA concentration and purity were assessed using spectrophotometry. cDNA was generated from 500 ng of RNA (tissue samples) or 100 ng RNA (cell culture experiments) using the LunaScript RT Supermix kit (New England Biolabs, Cat. No. E3010L) according to the manufacturer’s directions. Real-time PCR (RT-PCR) was performed on a Quantstudio 3 (ThermoFisher Scientific) using Luna Universal qPCR master mix for SYBR Green primers (New England Biolabs, Cat. No. M3003X). All primer sequences are listed in Supplemental Table S1. Relative mRNA expression was calculated using 2^−ΔΔCT from the relevant control group.

RNA Sequencing

Library preparation and RNA sequencing were performed by Quick Biology (Pasadena, CA). RNA integrity was checked by Agilent Bioanalyzer 2100; only samples with clean rRNA peaks were used. Library for RNA-Seq was prepared according to KAPA Stranded mRNA-Seq poly(A) selected kit with 201–300-bp insert size (KAPA Biosystems, Wilmington, MA) using 250 ng total RNA as input. Final library quality and quantity were analyzed by Agilent Bioanalyzer 2100 and Life Technologies Qubit3.0 Fluorometer. One-hundred fifty base-paired end reads were sequenced on Illumina HighSeq 4000 (Illumina, San Diego, CA). The reads were first mapped to the latest University of California Santa Cruz transcript set using Bowtie2 (51), and the gene expression level was estimated using RSEM (52). TMM (trimmed mean of M values) was used to normalize the gene expression. Differentially expressed genes were identified using the edgeR program (53). Genes showing altered expression with P < 0.05 and more than 1.5-fold changes were considered differentially expressed. Goseq was used to perform the gene ontology (GO) enrichment analysis.

Study Approval

All procedures were approved by the Institutional Animal Care and Use Committee of the University of Florida (Protocol No. 201810484).

Statistical Analysis

All data are presented as the means ± SD. Normality of data was tested with the Shapiro–Wilk test and/or inspection of QQ plots. Data involving comparisons of two groups were analyzed using a Student’s t test when normally distributed and Mann–Whitney test when normality could not be assessed. When making comparisons with more than two groups, data were analyzed using two-way ANOVA with Sidak’s post hoc testing for multiple comparisons when significant interactions were detected. Mixed-effects analysis and three-way ANOVA were used to determine differences when three or more main effects were present. In all cases, P < 0.05 was considered statistically significant. All statistical testing, except for RNA sequencing analysis, was conducted using GraphPad Prism software (version 9.0).

RESULTS

Endothelium-Specific AHR Deletion Promotes Ischemic Perfusion Recovery and Alters Angiogenic-Associated Gene Expression in Male Mice with CKD

To begin to explore whether chronic AHR activation in the endothelium contributes to worsening limb outcomes in PAD, we generated a conditional endothelial cell-specific AHR knockout mouse (AHR^ecKO, Fig. 1A). In this context, CKD induces the accumulation of tryptophan-derived uremic toxins (indoxyl sulfate, kynurenine, and kynurenic acid), as we have demonstrated in previous studies (39, 40, 54), which are known ligands of the AHR (23, 27, 55). Following administration of tamoxifen, deletion of the AHR was confirmed by traditional PCR designed to detect recombination of exon 2 in the AHR gene (Fig. 1B), as well as qPCR analysis of Ahr expression (Fig. 1C), both performed in primary endothelial cells. In addition, endothelium-specific Cre-induced DNA recombination was confirmed using a fluorescent reporter mouse (Supplemental Fig. S1; Ai14, Jackson Laboratory, Stock No. 007914). Mice were next randomized to either casein (control) or adenine-supplemented (CKD) diet before the induction of limb ischemia via femoral artery ligation (FAL) surgery (Fig. 1D). Consistent with previous studies (38–40), mice fed adenine diet had lower glomerular filtration rate (GFR, Fig. 1E) and elevated blood urea nitrogen (BUN, Fig. 1F), which was accompanied by significantly lower kidney (Fig. 1G) and body (Fig. 1H) weights compared with mice fed the casein diet.

Figure 1.

Validation of endothelium-specific knockout of the aryl hydrocarbon receptor (AHR). A: generation of inducible, endothelial cell-specific AHR knockout mice. B: confirmation of DNA recombination in endothelial cells isolated from liver and skeletal muscle. C: confirmation of AHR deletion by qPCR analysis of mRNA expression. D: graphical depiction of the overall experiment design. E: glomerular filtration rate (GFR) measured by FITC-inulin clearance and normalized to body weight (n = 5/group/sex). F: blood urea nitrogen (BUN) measured in serum after 8 wk on adenine diet (n = 8/group/sex). G and H: kidney weights (G) and body weights (H) measured at euthanasia. C was analyzed using unpaired, two-tailed Student’s t test. E–H were analyzed using a three-way ANOVA. Error bars represent the standard deviation. Graphics in A and D were generated using a licensed version of BioRender.com.

Because previous work has implicated that AHR activation can impair angiogenesis (16, 34, 56), we first examined whether deletion of the AHR in endothelial cells alone would alter perfusion recovery following FAL in mice with and without CKD. In the paw, there was a significant effect of diet in both male and female mice, indicating that CKD impaired perfusion recovery (Fig. 2A). Interestingly, in the gastrocnemius muscle, the effect of diet was present only in male mice (Fig. 2B), suggesting that CKD was more detrimental to males than females. A significant time × diet × genotype interaction was detected (P = 0.0146) for perfusion recovery in the paw of male mice. Post hoc testing revealed significantly higher levels of paw perfusion in male AHR^ecKO mice with CKD compared with male AHR^fl/fl mice with CKD at day 7 (P = 0.0195) and day 14 (P = 0.0002) after FAL. However, an interaction was not observed in female mice demonstrating that deletion of the AHR in endothelial cells had no impact on paw perfusion recovery in female mice. Similar to the paw, gastrocnemius muscle perfusion recovery was greater in male AHR^ecKO mice with CKD compared with male AHR^fl/fl mice with CKD at day 3 (P = 0.0001) and day 7 (P = 0.0227) after FAL; however, no difference was detected in female mice (Fig. 2B).

Figure 2.

Endothelium-specific aryl hydrocarbon receptor (AHR) deletion promotes ischemic perfusion recovery and alters angiogenic-associated gene expression in male mice with chronic kidney disease (CKD). A and B: laser-Doppler flowmetry quantification of perfusion recovery in the paw (A) and gastrocnemius muscle expressed as percentage of control limb (n = 10/group) (B). Perfusion recovery was analyzed using mixed-model analysis. C and D: total number of capillaries (C) and total number of capillaries normalized to total muscle area (D). E: representative immunofluorescence images of total capillaries labeled with the anti-CD31 antibody. F: perfused capillaries quantified as a percentage of total capillaries. G: perfused capillaries quantified as the number per area of the muscle. H: representative immunofluorescence images of perfused capillaries labeled with isolectin. All capillary density measurements were performed n = 8–10/group/sex. I: expression of angiogenic and vasoreactive genes in ischemic skeletal muscle (n = 6/group/sex). Statistical analyses performed using two-way ANOVA with Sidak’s post hoc testing for multiple comparisons when significant interactions were detected. Error bars represent the standard deviation.

Because of the superficial penetration depth of the laser Doppler, we also performed a retro-orbital injection of Dylight649-labeled isolectin to live mice allowing for labeling and quantification of both perfused (isolectin⁺) and total (CD31⁺) capillaries across the tibialis anterior muscle. In male mice, CKD resulted in a significant decrease in the total capillary number in AHR^fl/fl mice only (Fig. 2C); however, once normalized to the total muscle area, this difference was abolished (Fig. 2C). In female mice, total capillary number and total capillaries per area of muscle were unaffected by CKD or deletion of the AHR in endothelial cells (Fig. 2D). In agreement with the laser-Doppler findings, the percentage of perfused capillaries, as well as the number of perfused capillaries per muscle area, was significantly higher in male AHR^ecKO mice with CKD compared with AHR^fl/fl mice with CKD (Fig. 2F). However, this relationship was not present in the female mice (Fig. 2G). No significant differences in total or perfused capillaries were observed in the nonischemic limb (Supplemental Fig. S2, A and B). We next employed qPCR to evaluate how the endothelium-specific AHR deletion affects expression of angiogenic and vasoreactive genes in ischemic skeletal muscle (Fig. 2I). In males, CKD resulted in decreased expression of proangiogenic genes Vegfa (P = 0.0004), Vegf121 (P = 0.0057), Vegf165 (P = 0.0059), Angpt1 (P = 0.0022), and Egf (P = 0.0130) in AHR^fl/fl mice. In contrast, only Vegf121 was significantly reduced by CKD in AHR^ecKO male mice. In fact, AHR^ecKO mice with CKD had significantly higher expression of Vegfa (P = 0.0031) and Vegf121 (P = 0.0009, along with a trending increase in Vegf165 (P = 0.0778), when compared with AHR^fl/fl male mice with CKD. In addition to angiogenic genes, the expression of endothelin 1 (End1), a vasoconstricting gene, was significantly reduced in male AHR^ecKO mice with CKD when compared with their AHR^fl/fl littermates (Fig. 2I). Taken together, these findings are congruent with the observed increases in limb perfusion recovery and perfused capillaries in male AHR^ecKO mice with CKD. Neither CKD nor the deletion of AHR was found to have a significant impact on angiogenic or vasoreactive gene expression in female mice with or without CKD (Fig. 2I). In the nonischemic limb, the expression of angiogenic and vasoreactive genes displayed greater impact in males compared with females, but minimal effects of AHR deletion were observed (Supplemental Fig. S2, C and D).

Endothelium-Specific AHR Deletion Does Not Improve Ischemic Muscle Function

Skeletal muscle function has emerged as a critically important characteristic of PAD and has been shown to predict/associate with mortality risk and mobility loss (57–60). In addition, independent of PAD, CKD is known to manifest with a progressive skeletal myopathy characterized by muscle atrophy, weakness, and exercise intolerance (61, 62). Thus, we tested whether deletion of the AHR in endothelial cells influenced ischemic muscle function. First, we used voluntary unilateral hindlimb grip strength testing to evaluate paw function and strength. There was a significant main effect of CKD (diet) in both male (P = 2.72E-07) and female mice (P = 0.0430) demonstrating that CKD decreased voluntary grip strength in the ischemic limb (Fig. 3A). A significant time × genotype × diet interaction was detected only in male mice (P = 0.0475). However, post hoc testing did not reveal any significant differences in grip strength between AHR^fl/fl and AHR^ecKO male mice (P = 0.16 at day 14 post-FAL). To provide a more rigorous assessment of the ischemic limb muscle function, we used nerve-mediated contraction of the tibialis anterior in situ, where supramaximal stimulation could be applied to ensure complete recruitment of the motor neurons. Consistent with previous findings in CKD mice subjected to FAL (38), male mice with CKD had lower twitch (Fig. 3B) and tetanic (Fig. 3C) forces, regardless of whether peak force was normalized to the muscle weight or not (Fig. 3, D–G). Despite the improved perfusion recovery and increased perfused capillary density in male AHR^ecKO mice with CKD, muscle contractile function was not different from AHR^fl/fl mice (Fig. 3, D–G). In female mice, CKD did not significantly impair muscle contractile function following FAL and consequently, the deletion of AHR in endothelial cells had no significant impact on muscle strength (Fig. 3, H–M). In the nonischemic limb of mice with CKD, deletion of the AHR in endothelial cells had no effect on muscle contractile function in either sex (Supplemental Fig. S3).

Figure 3.

Endothelium-specific aryl hydrocarbon receptor (AHR) deletion does not improve ischemic muscle function. A: hindlimb grip strength recovery measured in grams force and quantified as percentage of control limb. Grip strength recovery was analyzed using mixed model analysis (n = 10/group/sex). B and C: representative twitch (C) and tetanic contractions (D) in male mice with and without chronic kidney disease (CKD). D–G: maximal absolute twitch force (D), maximal absolute tetanic force (E), maximal specific (normalized to weight) twitch force (F), and maximal specific tetanic force in male mice (G). H and I: representative twitch (H) and tetanic contractions (I) in female mice with and without CKD. J–M: maximal absolute twitch force (J), maximal absolute tetanic force (K), maximal specific (normalized to weight) twitch force (L), and maximal specific tetanic force (M) in female mice. D–M were analyzed using two-way ANOVA with Sidak’s post hoc testing for multiple comparisons. Error bars represent the standard deviation. D–G and J–M contain n = 8–10/group.

Endothelium-Specific AHR Deletion Has No Impact on Muscle Mass or Histopathology but Increases Myofiber Area in Male CKD Mice

We next examined the impact of CKD and endothelium-specific AHR deletion on muscle size, myofiber area, and muscle histopathology in the ischemic limb. There was a significant effect of CKD on the weight of the tibialis anterior and gastrocnemius muscles in both male and female mice; however, there was no significant main effect for genotype or interaction detected (Fig. 4, A and B). Quantification of the total myofiber number within the tibialis anterior muscle revealed a significant impact of CKD in male mice only (P = 0.0268) but no significant effects of genotype (Fig. 4C). Representative images of ischemic muscle histopathology are shown in Fig. 4D. There was no significant difference in percentage of myofibers with centralized nuclei among any of the groups (Fig. 4E). Analysis of the mean myofiber cross-sectional area (CSA) revealed significantly smaller myofiber areas in male CKD mice (diet effect, P = 1.9E-05) but not female mice (diet effect, P = 0.27). Interestingly, male AHR^ecKO mice with CKD were found to have significantly higher mean myofiber CSA compared with AHR^fl/fl mice with CKD (Fig. 4F). In the nonischemic limb, CKD mice had lower muscle weights and myofiber CSA, but there were no significant effects of AHR deletion in mice with or without CKD regardless of biological sex (Supplemental Fig. S4).

Figure 4.

Endothelium-specific aryl hydrocarbon receptor (AHR) deletion has no impact on muscle mass or histopathology but increases myofiber area in male chronic kidney disease (CKD) mice. A and B: weight of tibialis anterior (A) and gastrocnemius muscles (B) of the ischemic limb (n = 10/group/sex). C: total number of myofibers within the ischemic tibialis anterior muscle (n = 8/group/sex). D: representative ×20 and tiled images of hematoxylin and eosin staining and laminin immunolabeling of ischemic tibialis anterior muscle. E: percentage of total myofibers with centralized nuclei (n = 8–10/group/sex). F: quantification of mean myofiber cross-sectional area (CSA) of the ischemic tibialis anterior muscle (n = 10/group/sex). Statistical analyses were performed using two-way ANOVA with Sidak’s post hoc testing for multiple comparisons when significant interactions were detected. Error bars represent the standard deviation.

Endothelium-Specific AHR Deletion Does Not Impact Ischemic Muscle Mitochondrial Function in CKD

Alterations in mitochondrial respiration and energy production have been identified as characteristics of skeletal muscle in both PAD and CKD independent of one another (39, 63–70). Thus, we sought to examine if deletion of the AHR in endothelial cells and the corresponding improvement in limb perfusion recovery in male AHR^ecKO mice altered skeletal muscle mitochondrial oxidative phosphorylation (OXPHOS). To accomplish this, we employed a creatine kinase clamp system that facilitates the assessment of mitochondrial oxygen consumption (JO₂) at various energy demands, akin to a stress test. Quantification of the slope of the relationship between JO₂ and energy demand (ΔG_ATP), termed OXPHOS conductance, represents the mitochondrion’s ability to respond to changes in energy demand by altering energy transduction. For these experiments, mitochondria were isolated from the ischemic gastrocnemius muscle and energized with saturating levels of pyruvate, malate, and the medium chain fatty acid octanoylcarnitine (Fig. 5A). Graphs of the JO₂ and ΔG_ATP relationship in for male and female mice are shown in Fig. 5, B and C. Quantification of the OXPHOS conductance revealed a significant genotype effect in male mice (P = 0.0021) only (Fig. 5C). Post hoc analysis revealed that CON male AHR^ecKO mice had higher OXPHOS conductance than CON AHR^fl/fl mice (P = 0.0316), but there was no difference between genotypes in the CKD group (P = 0.45, Fig. 5D). The respiration rate at the highest level of energy demand was also significantly higher in CON AHR^fl/fl mice (P = 0.0265), but there was no difference between genotypes in the CKD group (P = 0.25, Fig. 5E). Mitochondrial hydrogen peroxide (H₂O₂) production and an estimation of electron leak were not affected by CKD or genotype in male mice (Fig. 5, F and G). In female mice, there was a significant diet effect indicating that CKD mice had lower OXPHOS conductance (Fig. 5H) and respiration rates at the highest level of energy demand, but no genotype effect was observed (Fig. 5I). Mitochondrial hydrogen peroxide (H₂O₂) production and estimation of electron leak were not affected by CKD or genotype in female mice (Fig. 5, J and K). In the nonischemic limb, deletion of the AHR in endothelial cells did not impact muscle mitochondrial function in mice with or without CKD (Supplemental Fig. S5). In the ischemic muscle, the transcript levels of Cox7a1, Atp5k, Atp5d, Tfam, and Sod2 were reduced by CKD in AHR^fl/fl (Fig. 5L). Interestingly, male AHR^ecKO mice without CKD tended to have lower expression of these mitochondrial genes compared with AHR^fl/fl mice without CKD, but the impact of CKD in AHR^ecKO was absent except for Tfam expression, which was significantly decreased by CKD in male AHR^ecKO mice. These results suggest that CKD decreases mitochondrial gene expression in ischemic muscle from AHR^fl/fl male mice but not in AHR^ecKO male mice. Female mice did not have significant changes in mitochondrial gene expression in the ischemic limb due to either CKD or genotype (Fig. 5M). However, in the nonischemic limb, CKD was found to alter the expression of several of these mitochondrial genes in both sexes, but a genotype effect was only observed in a few mitochondrial genes with AHR^ecKO mice on the control diet having lower expression compared with AHR^fl/fl without CKD (Supplemental Fig. S5).

Figure 5.

Endothelium-specific aryl hydrocarbon receptor (AHR) deletion does not impact ischemic muscle mitochondrial function in chronic kidney disease (CKD). A: protocol for assessing mitochondrial function in the ischemic gastrocnemius muscle was assessed using a creatine kinase clamp system to measure oxygen consumption (JO₂) at physiologically relevant energy demands (ΔG_ATP). B and C: plot of JO₂ and ΔG_ATP in mitochondria isolated from male (B) and female (C) mice. D and E: quantification of the conductance (slope of JO₂ and ΔG_ATP relationship) (D) and JO₂ (E) at the highest energy demand in male mice. F and G: mitochondrial H₂O₂ production (F) and estimated percent electron leak (JH₂O₂/JO₂) (G) at each ΔG_ATP in males. H and I: quantification of the conductance (slope of JO₂ and ΔG_ATP relationship) (H) and (I) JO₂ (I) at the highest energy demand female mice. J and K: mitochondrial H₂O₂ production (J) and estimated percent electron leak (JH₂O₂/JO₂) (K) in females. L and M: expression of mitochondrial related genes in ischemic extensor digitorum longus muscle in male (L) and female (M) mice (n = 6/group/sex). Statistical analyses performed using two-way ANOVA with Sidak’s post hoc testing for multiple comparisons when significant interactions were detected. Error bars represent the standard deviation. B–K contain n = 10/group/sex.

Endothelial Cell AHR Displays Sex-Dependent Differences in Activating Potential in Normoxic and Hypoxic Conditions

An intriguing observation in this study was the sex-dependent effect of AHR activation on ischemic angiogenesis. Although the underlying mechanisms behind this sexual dimorphism in endothelial cells are unknown, sex-dependent effects of AHR activation have been shown in other tissues (71–74). To determine if these differences were intrinsic to biological sex and not dependent on sex hormones, we isolated primary endothelial cells from male and female mice, cultured them ex vivo where sex hormones were not present, and exposed cells to the potent AHR ligand indoxyl sulfate (IS) under normoxic and hypoxic conditions (Fig. 6A). The purity of primary endothelial isolates was confirmed by immunolabeling with an anti-CD31 antibody (Fig. 6B). Following treatment with indoxyl sulfate or DMSO (vehicle control) for 6 h, primary endothelial cells were harvested for qPCR analysis of mRNA levels of the Ahr, Cyp1a1, and the Ahrr (AHR repressor). Ahr expression displayed a significant main effect for hypoxia and a hypoxia × indoxyl sulfate interaction (Fig. 6C). Post hoc testing revealed a ∼12-fold increase in Ahr expression in hypoxic cells treated with indoxyl sulfate in both male and female primary endothelial cells (Fig. 6C). To examine the activating potential of the AHR, we measured expression of Cyp1a1, a canonical xenobiotic response gene whose expression is controlled by Ahr. As expected, there was a main effect of indoxyl sulfate (P = 0.0001) fully demonstrating that indoxyl sulfate treatment upregulates Ahr signaling. Interestingly, a significant main effect of sex, as well as an indoxyl sulfate × sex interaction was detected. Post hoc analysis revealed higher Cyp1a1 expression levels in male primary endothelial cells treated with indoxyl sulfate when compared with female primary endothelial cells under both normoxic and hypoxic conditions (Fig. 6D). The determine if the lower AHR activating potential in female endothelial cells was due to elevated expression of the AHR repressor system, we measured Ahrr expression levels. Importantly, Ahrr expression was found to be unaffected by either sex or hypoxia (Fig. 6E). Taken together, these findings demonstrate that male endothelial cells display greater AHR activating potential when compared with female endothelial cells, and this difference is not explained by repression of the AHR or differences in sex hormones.

Figure 6.

Endothelial cell aryl hydrocarbon receptor (AHR) displays sex-dependent differences in activating potential in normoxic and hypoxic conditions. A: graphical depiction of experimental design made with a licensed version of BioRender.com. B: confirmation of endothelial isolates via immunolabeling with an anti-CD31 antibody. C: relative mRNA expression of Ahr Cyp1a1 (D) and Ahrr (E) measured by qPCR analysis in primary cells treated with indoxyl sulfate or DMSO control in normoxic and hypoxic conditions (n = 3 biological replicates/group/sex). Statistical analyses were performed using two-way ANOVA with Sidak’s post hoc testing for multiple comparisons when significant interactions were detected. Error bars represent the standard deviation.

RNA Sequencing in Primary Endothelial Cells Uncovers Sex-Dependent Alterations in Angiogenic Signaling with AHR Activation in Hypoxia

To assess the broader impact of sex-specific effects on Ahr activation on the endothelial cell transcriptome, we performed RNA sequencing on male and female primary endothelial cells treated with indoxyl sulfate under conditions of hypoxia and nutrient deprivation. There were 1,076 upregulated (Log₂FC < −1.5, adjusted P < 0.05) and 690 downregulated (Log₂FC > 1.5, adjusted P < 0.05) genes in male compared with female endothelial cells (Fig. 7A). A complete list of differentially expressed genes can be found in Supplemental Data Set S1. Gene set enrichment analysis of genes upregulated in male endothelial cells revealed terms related to “morphogenesis of a branching structure,” “regulation of angiogenesis,” “wound healing,” “epithelial tube morphogenesis,” and “ameboidal-type cell migration” (Fig. 7B). A heatmap of the top 100 differentially expressed genes is shown in Fig. 7C. Despite being exposed to hypoxia and nutrient deprivation within the same culture plate, Hif1a expression was ∼4.6-fold higher in male endothelial cells compared with female endothelial cells (adjusted P = 0.0057). Correspondingly, several genes involved in angiogenesis were also expressed differently in the sexes. For example, males had higher expression of Vegfa, Vegfb, and Hgf, whereas females had higher expression of Fgf2 (Fig. 7D). These data also exposed sex-dependent differences in the Notch signaling pathway, a master regulator of sprouting angiogenesis (75), under conditions of Ahr activation. Male endothelial cells had greater expression of Notch1/4 and notch-related genes Dll4 and Cxcr4, as well as decreased expression of Usp10, an antiangiogenic factor that inhibits Notch (76). Male endothelial cells also had higher expression of matrix metalloproteinases (Mmp2/9/19), which are involved in remodeling of the extracellular matrix during vessel sprouting (77). Several genes that regulate redox balance and have been linked to angiogenesis (78) also displayed sex differences with males expressing less Trx1, Trxnd1 but more Txnip and Sod3 when compared with female endothelial cells (Fig. 7D). Male endothelial cells also had greater expression of nitric oxide synthase 3 (Nos3) compared with female endothelial cells (adjusted P = 2.8E-04) treated with indoxyl sulfate, suggesting that vasoreactivity may also be different between males and females with AHR activation. Future work is needed to mechanistically explore how these genes and pathways are regulated by the AHR.

Figure 7.

RNA sequencing in primary endothelial cells uncovers sex-dependent alterations in angiogenic signaling with aryl hydrocarbon receptor (AHR) activation in hypoxia. A–D: RNA extracted from primary endothelial cells isolated from male and female mice treated with indoxyl sulfate under hypoxic conditions. A: volcano plot of all detectable genes, differential expression significant if (Log₂FC < −1.5, adjusted P < 0.05) or (Log₂FC > 1.5, adjusted P < 0.05). B: gene set enrichment analysis of genes upregulated in male endothelial cells. C: heat map of top 100 differentially expressed genes. D: normalized expression of select differentially expressed genes from top gene ontology (GO) terms. Adjusted P values from a Wilcoxon test are shown in D.

DISCUSSION

The presence of CKD is known to exacerbate the pathobiology of PAD and increase the risk of major adverse limb events and mortality. Despite the robust clinical evidence, the cellular mechanisms that contribute to worsened outcomes are ill-defined. In this study, we used adenine diet and femoral artery ligation to explore the role of AHR activation, caused by uremic solute accumulation, in regulating ischemic angiogenesis in CKD. In male mice with CKD and PAD, endothelial cell-specific deletion of the AHR improved limb perfusion recovery, enhanced angiogenesis evidenced by more perfused capillaries, and attenuated myofiber atrophy. Gene expression analysis revealed greater mRNA levels of angiogenic and mitochondrial genes within the ischemic limb of AHR^ecKO mice with CKD compared with AHR^fl/fl mice with CKD. However, these effects did not translate to improvements in muscle strength or mitochondrial function in male AHR^ecKO mice. Conversely, female AHR^ecKO mice with CKD had no significant improvement in limb perfusion recovery, angiogenesis, or muscle health. Taken together, these findings demonstrate that endothelial cell AHR deletion improves ischemic angiogenesis in male but not female mice with CKD.

The results herein are in agreement with previous studies demonstrating that AHR activation impairs angiogenic processes in cultured endothelial cells (23, 79–81) and mice (14, 16). A recent study in mice with CKD provided evidence implicating activation of the AHR in the degradation of β-catenin leading to disruption of the Wnt signaling pathway that impaired ischemic angiogenesis (16). The authors also found that global AHR inhibition improved perfusion recovery and normalized ischemic angiogenesis in female mice with CKD. In contrast, the results herein demonstrated that genetic deletion of the AHR specifically in endothelial cells improved ischemic angiogenesis in male, but not female, mice with CKD. The discrepancy between our study and the study by Arinze et al. (16) may be explained by differences in the experimental approach. First, while both studies used adenine-supplemented diet to induce CKD, the duration of CKD was significantly longer in the current study (6 vs. 3 wk). Second, our study employed a genetic ablation of the AHR specifically in endothelial cells, whereas Arinze et al. (16) treated mice with CH223191, a pharmacological AHR inhibitor. Recent evidence has established that AHR activation has pathological effects on skeletal muscle (82), which can disrupt paracrine angiogenic signaling (54). Thus, it is likely that some of the observed benefits with CH223191 treatment were derived from the drug’s effects on limb muscle rather than being solely attributable to its effect on the endothelium.

Unexpectedly, we found sex-specific differences when comparing the effect of AHR deletion in endothelial cells during CKD on ischemic limb pathology. Whereas male AHR^ecKO mice with CKD had improved limb perfusion recovery and angiogenesis, their female counterparts were not affected by the deletion of the AHR in endothelial cells. The notion that AHR signaling may be sex-dependent is supported by previous work in nonvascular tissue. For instance, treatment with the high-affinity AHR ligand, 2,3,7,8-tetrachlorodibenzo-p-dioxin, resulted in sex differences in the liver transcriptome of mice (72). In addition, in rodents with CKD, AHR abundance in the kidney has sexually dimorphic effects (71). Considering that all female mice underwent ovariectomies before enrollment in our study, we theorized that these stark differences may not be explained by sex hormones alone. To test this hypothesis, we isolated and cultured primary endothelial cells from male and female mice ex vivo where sex hormones were absent. Cells were treated with a known AHR ligand and uremic solute, indoxyl sulfate, in normoxic and hypoxic conditions. These experiments confirmed that male endothelial cells had greater AHR activating potential in both normoxic and hypoxic conditions compared with female endothelial cells. However, these differences were not related to the abundance of Ahr or its repressor (Ahrr), which were not different between males and females under these conditions. These data unequivocally establish that murine endothelial cells have sex-dependent differences in AHR activating potential, which are not explained by sex hormones. To further explore this, we performed RNA sequencing on male and female primary hypoxic endothelial cells treated with IS. This experiment revealed sex differences in several genes related to angiogenesis that agree with the observed sex differences in ischemic angiogenesis in our mouse studies. In totality, these findings implicate sex chromosomes as potential mediators of the sex-dependent differences in AHR biology; however, future mechanistic studies are needed to confirm this assertion.

The current study along with previous work from our group (45) demonstrates that vascular-targeted interventions can improve blood perfusion without impacting muscle function. Limb ischemia resulting from PAD induces angiogenesis, a process by which preexisting endothelial cells proliferate and migrate to form new blood vessels (83, 84). Yet, patients with PAD have diminished microvascular perfusion (18, 21, 85–87), suggesting that the angiogenic response is insufficient to meet the demands of the tissue. Numerous studies have used gene, protein, and cell-based therapies aimed to promote angiogenesis in patients with PAD; however, these have been unsuccessful in clinical trials (88–94). Although the lack of clinical success with angiogenic therapies is complex, it may indicate that targeting angiogenesis alone to improve PAD symptomology is suboptimal. It is important to recognize that it is challenging to adequately model the interaction between comorbidities, genetics, and the environment of patients with PAD in a preclinical setting. The poor modeling of PAD is likely to be a main contributor to the failure of translating preclinical PAD therapies to humans. It is indisputable that reduced blood flow contributes to walking performance in a preclinical setting; therefore, improving oxygen delivery to the limb should remain a primary clinical focus. However, there are several nonvascular cell types within the lower extremity (skeletal muscle, adipose tissue, connective tissue, nerves, etc.) that contribute to walking performance, and how these cell types interact with vascular cells in the ischemic microenvironment is incompletely understood. Skeletal muscle can regulate angiogenesis via paracrine signaling (95), and coincidentally muscle quality/function in patients with PAD is a strong predictor of morbidity and mortality (57–60). To make matters worse, common comorbidities in patients with PAD like CKD have deleterious effects on skeletal muscle independent of PAD (61, 62, 96). In fact, several preclinical studies have demonstrated that therapeutic interventions targeting ischemic skeletal muscle can improve limb outcomes in mice subjected to femoral artery ligation (44, 45, 48, 97, 98). The stagnant therapeutic development in PAD and the lack of success of angiogenic therapies indicate that further research is necessary to understand the complex pathobiology of PAD and the intercellular communication contributing to limb functionality.

We acknowledge that there are limitations present in the current study. First, we enrolled young mice despite PAD prevalence being strongly associated with advancing age. Unfortunately, due to the extensive breeding necessary to generate conditional, endothelium-specific AHR deletion, experiments in aged mice were not possible. Second, limb ischemia in patients with PAD occurs gradually over time, whereas the FAL surgery used herein causes abrupt and severe ischemia. In addition, mice were only studied at a single time point post-FAL surgery such that temporal differences in angiogenic processes could not be evaluated. Third, patients with PAD and CKD often present with additional comorbidities (hypertension, obesity, hyperlipidemia, and diabetes) that contribute to disease pathology but were not present in the mouse models employed in our study. Although our study examined muscle histopathology and strength, our contraction protocol only measured muscle force levels during brief contractions. Thus, changes in muscle function may not translate the sustained contractions, such as the commonly used 6-min walk test in patients with PAD. Future work is needed to develop better preclinical assessments of limb functionality as walking performance following a single-limb surgery (like FAL) is difficult to assess in quadrupedal species. Finally, although our primary endothelial cell experiments confirmed a sex difference in AHR activation following indoxyl sulfate exposure, these cells were isolated from the liver where dissociation and cell yield are high. It has been shown that endothelial cells from different organs have different phenotypes (99) and it is possible that the observed results could be different in primary endothelial cells from skeletal muscle.

In summary, deletion of the AHR in endothelial cells improves limb perfusion recovery and ischemic angiogenesis in male, but not female, mice with PAD and CKD. However, the improved limb blood flow in males did not result in better muscle strength or mitochondrial function in the PAD limb. Primary endothelial cell culture experiments demonstrate sex differences in AHR activation and angiogenic signaling that are not dependent on sex hormones.

DATA AVAILABILITY

All source data are available from https://doi.org/10.6084/m9.figshare.24050244. RNA sequencing data have been deposited to the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/) under Accession No. GSE234508.

SUPPLEMENTAL DATA

Supplemental Table S1, Supplemental Figs: S1–S5, and Supplemental Data Set S1: https://doi.org/10.6084/m9.figshare.24050244.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grant R01-HL149704 (to T.E.R.).

DISCLAIMERS

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

V.R.P. and T.E.R. conceived and designed research; V.R.P., J.T., Q.Y., J.M., O.L., and T.E.R. performed experiments; V.R.P., J.T., Q.Y., J.M., and T.E.R. analyzed data; V.R.P., O.L., and T.E.R. interpreted results of experiments; V.R.P. and T.E.R. prepared figures; V.R.P. and T.E.R. drafted manuscript; V.R.P., J.T., Q.Y., J.M., O.L., and T.E.R. edited and revised manuscript; V.R.P., J.T., Q.Y., J.M., O.L., and T.E.R. approved final version of manuscript.