Hyperandrogenemia reduces endothelium-derived hyperpolarizing factor-mediated relaxation in mesenteric artery of female rats

Abstract

Women with polycystic ovary syndrome (PCOS) are often presented with hyperandrogenemia along with vascular dysfunction and elevated blood pressure. In animal models of PCOS, anti-androgen treatment decreased blood pressure, indicating a key role for androgens in the development of hypertension. However, the underlying androgen-mediated mechanism that contributes to increased blood pressure is not known. This study determined whether elevated androgens affect endothelium-derived hyperpolarizing factor (EDHF)-mediated vascular relaxation responses through alteration in function of gap junctional proteins. Female rats were implanted with placebo or dihydrotestosterone (DHT) pellets (7.5 mg, 90-day release). After 12 weeks of DHT exposure, blood pressure was assessed through carotid arterial catheter and endothelium-dependent mesenteric arterial EDHF relaxation using wire myograph. Connexin expression in mesenteric arteries was also examined. Elevated DHT significantly increased mean arterial pressure and decreased endothelium-dependent EDHF-mediated acetylcholine relaxation. Inhibition of Cx40 did not have any effect, while inhibition of Cx37 decreased EDHF relaxation to a similar magnitude in both controls and DHT females. On the other hand, inhibition of Cx43 significantly attenuated EDHF relaxation in mesenteric arteries of controls but not DHT females. Elevated DHT did not alter Cx37 or Cx40, but decreased Cx43 mRNA and protein levels in mesenteric arteries. In vitro exposure of DHT to cultured mesenteric artery smooth muscle cells dose-dependently downregulated Cx43 expression. In conclusion, increased blood pressure in hyperandrogenic females is due, at least in part, to decreased EDHF-mediated vascular relaxation responses. Decreased Cx43 expression and activity may play a role in contributing to androgen-induced decrease in EDHF function.

Summary Sentence

Hyperandrogenism in female rats reduced EDHF function via decrease in connexin 43 expression and activity in mesenteric arteries, providing a molecular mechanism linking elevated androgens and increased blood pressure.

Introduction

The endothelium plays a major role in control of vascular tone. This function is achieved by the release of nitric oxide (NO), prostacyclin (PGI2), and the endothelium-derived hyperpolarizing factor (EDHF) [1]. Of these autacoids, NO predominates in large conducting arteries, whereas the importance of EDHF increases as the size of the arteries decreases [2–4]. NO and PGI2 are well characterized with respect to chemistry and signaling cascades, but the nature of EDHF remains obscure [5,6]. The endothelial hyperpolarization mediated by Ca²⁺-activated K⁺ (K_Ca) channels, such as small and intermediate conductance K_Ca channels, has been suggested to play a critical role in initiating EDHF-type relaxation responses in the arteries of many species, including humans [7–10]. Gap junctional proteins, such as connexins (Cxs), play an important role in spreading this K⁺-channel-initiated hyperpolarization along the blood vessels, thus propagating vasodilatation or vasoconstriction [11]. In blood vessels, gap junctions allow direct coupling between adjacent endothelial cells and adjacent smooth muscle cells and can even connect endothelial to smooth muscle cells in structures corresponding to myoendothelial gap junctions. At least three Cxs are involved in the formation of vascular gap junctions: Cx37, Cx40, and Cx43 [11,12]. A close correlation between the distribution of myoendothelial gap junctions along the mesenteric arterial tree and EDHF-mediated responses has been suggested [13]. The functional role of gap junctions in EDHF-mediated responses has been confirmed in rat mesenteric artery [14–16] and in other arteries (e.g. omental arteries) [17].

Decreased generation of EDHF, which is particularly critical in resistance arteries, has been shown to contribute to impaired endothelium-dependent vasodilation in hypertension [18–25]. The expression levels of arterial Cxs are reduced in animal models of hypertension [26–29]. Studies using transgenic mice showed that Cxs exert a profound hyperpolarizing influence in resistance arteries and that suppression of Cx expression induced hypertension [30–32]. Overexpression of Cx is also shown to reverse vascular dysfunction and hypertension [33]. These observations indicate the significance of Cxs in regulating vasomotor tone and blood pressure. However, the underlying mechanism that contributes to reduction in expression and function of Cxs during hypertension remains largely unclear.

Most studies have investigated the beneficial role of sex steroid hormones, especially estradiol, on cardiovascular function. Studies have demonstrated that estradiol has beneficial vascular effects because it increases EDHF function as well as expression and activity of Cxs [34–36]. However, despite both endothelial and vascular smooth muscle cells express androgen receptors [37,38], little is known about the role of androgens. The relationship between testosterone and female cardiovascular function deserves special consideration because plasma levels of testosterone are elevated approximately 2- to 3-fold in women with polycystic ovary syndrome (PCOS) [39–41], who are at increased risk for developing vascular dysfunction and hypertension [42–44]. In animal models of PCOS, anti-androgen treatment decreased blood pressure [45,46], indicating a key role for androgens in the development of hypertension in females. However, the underlying mechanism that contributes to androgen-mediated vascular dysfunction and increased blood pressure is not known. Since EDHF is more important for vasodilation in females than males [34,47,48], we hypothesized that elevated testosterone decreases EDHF function and induces hypertension in female rats. We tested this hypothesis by administering dihydrotestosterone (DHT, a nonaromatizable form of testosterone) into female rats, mimicking the 2-fold elevation in plasma DHT levels observed in PCOS, to investigate (1) whether the systemic arterial pressure is enhanced in DHT-treated compared with control rats; (2) whether endothelium-dependent EDHF-mediated vascular relaxation, particularly in the resistance mesenteric arteries, is inhibited in DHT–treated compared with control rats; and (3) whether the DHT-induced changes in EDHF relaxation involve alterations in Cx expression and function.

Materials and methods

All experimental procedures were performed in accordance with the National Institutes of Health guidelines (NIH Publication No. 85–23, revised 2011) with approval by the Animal Care and Use Committee at the University of Texas Medical Branch. Eight-week-old female Sprague-Dawley rats (Envigo, Indianapolis, IN) were divided into two groups, and one group was implanted subcutaneously with DHT pellets (7.5 mg, 90-day release, n = 8). The other group received placebo pellets (n = 8). At the end of a 12-week treatment period, this DHT treatment regimen resulted in rats with fewer estrus cycles and higher numbers of large cystic and atretic follicles in their ovaries, as previously reported [49,50]. In these rats, mean arterial pressure (MAP) was monitored using a carotid arterial catheter. Following blood pressure measurements, the animals were sacrificed, plasma was separated, and mesenteric arteries were isolated. A portion of the mesenteric arteries was used for vascular reactivity studies, and the remaining was quickly frozen for RNA/protein analysis. Plasma DHT levels in the samples were measured using radioimmunoassay (Biovendor, Asheville, NC) as per manufacturer's instructions. The intra- and interassay variations are 5.4% and 7.5%, respectively, and the detection limit is 6 pg/mL.

Experimental procedures

Mean arterial pressure

MAP in conscious free-moving control and DHT rats was determined using indwelling carotid arterial catheters as described in our previous publications [51]. Briefly, rats under anesthesia (Isoflurane, Henry Schein Animal Health, Dublin, OH) were surgically instrumented with flexible catheters (PE 50 tubing) in the right carotid artery. The catheters were tunneled to the nape of the neck and exteriorized. After a 24-h recovery period, when the animals were fully conscious and in a free-moving state, the catheter was connected to a pressure transducer, and arterial blood pressure was obtained using a data acquisition system (DBP001 direct blood pressure system and Workbench for Windows software; both from Kent Scientific, Litchfield, CT). Following a 30-min stabilization period, the arterial pressure was monitored continuously for 30 min and averaged to determine the baseline values.

Ex vivo vascular reactivity studies

Rats were sacrificed by CO₂ inhalation, and the mesenteric arcade was removed. Resistance mesenteric arteries (2-mm segments of the third-order branch of the superior mesenteric artery, 150 to 200 μm diameter) were dissected free of fat and connective tissue and mounted in a Mulvany-style isometric wire myograph (Danish Myotechnology, Aarhus, Denmark) for vessel reactivity assessment. Vessels were maintained at 37°C in a physiologic Krebs buffer of the following composition (mM): NaCl, 120; NaHCO₃, 25; KCl, 4.8; NaH₂PO₄, 1.2; MgSO₄, 1.2; dextrose, 11.0; CaCl, 1.8_. They were then aerated with 95% O₂ and 5% CO₂ (pH = 7.4). The rings were bathed in 6 mL Krebs buffer and allowed to equilibrate for 60 min before normalizing to an internal diameter of 0.9 of L_13.3kPa by using a normalization software package (Danish Myotechnology). The rings were then assessed for vascular function. Data were captured using a Power Lab data acquisition system (AD Instruments, Colorado Springs, CO). The presence of intact endotheliums in the vascular preparations was confirmed by observing the relaxation response to acetylcholine (ACh, 10⁻⁶ M) in rings precontracted with phenylephrine (PE, 10⁻⁶ M) as described previously [52]. In our preliminary experiment, the concentration–response curves for the PE-induced contractions were not different between the two groups. After the PE-induced contraction (approximately the ED₈₀ concentration) stabilized, relaxation responses to cumulative concentrations of ACh (10⁻¹⁰ to 10⁻⁵ M) were elicited. The EDHF-mediated component of ACh vasorelaxation was assessed after inhibiting NO production with NG-nitro-L-arginine methyl ester (L-NAME, 10⁻⁴ M) and PGI2 synthesis with indomethacin (10⁻⁵ M) for 30 min. In some experiments, the EDHF component of ACh relaxation was generated in the presence of Cx mimetic peptides (300 μM each, 90-min preincubation) to inhibit Cx37 (^40,37GAP26, sequence VCYDQAFPISHIR), Cx40 (⁴⁰GAP27, sequence SRPTEKNVFIV), or Cx43 (^43,37GAP27, sequence SRPTEKTIFII) [53]. These peptides and their scrambled version were custom synthesized by Biopeptide Co. Inc. (San Diego, CA) or were purchased from AnaSpec (San Jose, CA). The purity of these peptides was >95% by HPLC. Endothelium-independent relaxation responses to levcromakalim (10⁻⁹ to 10⁻⁵ M) in PE precontracted rings were also determined.

Quantitative real-time reverse transcription polymerase chain reaction

Total RNA was extracted using RNeasy mini kit (QIAGEN, Valencia, CA) according to manufacturer's instructions. RNA concentration and integrity were determined using a DS-11 spectrophotometer (DeNovix, Wilmington, DE). One microgram of total RNA was reverse transcribed using an iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). After dilution, cDNA corresponding to 100 ng of RNA was amplified by quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) using FAM (Invitrogen; Thermo Scientific, Grand Island, NY) as the fluorophore in a CFX96 real-time thermal cycler (Bio-Rad). PCR conditions for TaqMan Gene Expression Assay were 2 min at 50°C and 10 min at 95°C for 1 cycle, then 15 s at 95°C and 1 min at 60°C for 50 cycles. Results were calculated using the 2^–ΔΔCT method and expressed in fold change of the gene of interest in treated versus control samples. All reactions were performed in duplicate, and β-actin was used as an internal control. TaqMan assays were carried out in 10 μL volumes for real-time PCR at a final concentration of 250 nM TaqMan probe and 900 nM of each primer. Cx37 (Rn00572193_s1), Cx40 (Rn00570632_m1), Cx43 (Rn01433957_m1), and β-actin (Rn00667869_m1) assays were obtained by Assay-on-Demand (Applied Biosystems; Thermo Scientific).

Western blotting

Mesenteric arteries were homogenized in ice-cold radioimmunoprecipitation assay buffer (Cell Signaling Technology, Danvers, MA) containing a protease inhibitor tablet and phosphatase inhibitor cocktail-2 and -3 (Sigma-Aldrich, St. Louis, MO). Tissue lysates were centrifuged (14 000 × g for 10 min at 4°C), and the protein content was measured using the BCA protein assay kit (Pierce; Thermo Scientific). The supernatant was resuspended in NuPAGE lithium dodecyl sulfate sample buffer and reducing agent (Invitrogen; Thermo Scientific). Proteins (30 μg) alongside Precision Plus Standard (Kaleidoscope; Bio-Rad, Hercules, CA), and negative controls were resolved on 4%–12% gradient NuPAGE Bis-Tris gels (Invitrogen) at 100 V for 2 h at room temperature and then transferred onto Immobilon-P membranes (Millipore, Billerica, MA) at 100 V for 1 h. The membranes were blocked with 5% nonfat dry milk for 1 h and then incubated overnight at 4°C with primary antibodies of Cx37, Cx40, Cx43, and β-actin. Antibody details are in Supplementary Table S1. After being washed, the membranes were incubated with secondary antibodies conjugated with horseradish peroxidase at 1:5000 dilution and detected with the ECL detection kit (Pierce; Thermo Scientific). Densitometric measurement was done using ImageJ software. Briefly, the film was scanned using gray scale mode at 300 DPI and saved file in a TIFF format. Images were opened with ImageJ program and under gel analyzer option checked the boxes for Label with percentages and Invert peaks. Using rectangle selection tool selected each band and assigned numbers. After selecting all the bands, histograms for all the bands were generated and using line tool enclosed each peak from equal distance from the base line. Using magic wand tool, each enclosed peak was selected and through analyze tool the area under each peak was obtained which was used for further calculation in spreadsheet program and data were represented normalized with housekeeping loading control. Results were expressed as ratios of the intensity of a specific band to that of β-actin.

Immunofluorescence microscopy

Freshly isolated rat mesenteric arteries were cryopreserved in OCT compound (Tissue-Tek; Sakura Finetek, Torrance, CA) cooled by liquid nitrogen. Cryosections of transverse rings (8–10 μm thick) were prepared and mounted onto poly-l-lysine-coated slides, air-dried, and stored at −20°C. Immediately before immunostaining, sections were fixed in −20°C methanol for 10 min and then rehydrated in PBS (120 mM NaCl and 2.7 mM Na₂PO₄ · 2H₂O, pH 7.4) for 10 min. Sections were permeabilized with PBS containing 0.1% (vol/vol) Triton X-100 for 30 min followed by blocking with PBS containing 0.5% (wt/vol) BSA for 30 min at room temperature. After blocking, sections were stained with primary antibodies against Cx43 (details on Supplementary Table S1) overnight at 4°C and then washed for 30 min at room temperature in PBS. Sections were incubated with secondary antibody (Abcam, ab97050; 1:500 dilution) conjugated to fluorescein isothiocyanate for 1 h at 37°C and then washed to remove unbound antibody. All sections were counterstained with 4΄, 6-diamidino-2-phenylindole (DAPI, Thermo Fisher) per manufacturer's instruction to visualize cell nuclei. After washing, sections were air-dried and mounted under Dako fluorescence mounting medium (Agilent Technologies, Santa Clara, CA). After drying, immunofluorescent signals were viewed in an Olympus BX60 upright compound florescence microscope. The specificity of the immunostaining was evaluated by omission of the primary antibody and processed as described above.

Vascular smooth muscle cell isolation and culture

Mesenteric artery smooth muscle cells were isolated from adult female rats as described earlier [54]. Isolated cells were assessed for their purity by α-actin staining and were found to be >95% pure. Cells were routinely cultured in DMEM containing 4.5 g/L glucose supplemented with 25 mM HEPES, 2 mM L-glutamine, 10% FBS, and antibiotics without sodium pyruvate. Cells were changed to serum and phenol red-free media 48 h prior to treatments. Cells were treated with DHT (1–100 nM) for 5 days with fresh media changes every day. After treatments, cells were washed with PBS and lysed for RNA or protein preparation.

Statistical analysis

All values are given as mean ± SEM. Differences in tensions between PE contraction and basal tension were considered as maximal tension (100%), and relaxation to ACh was expressed as the percentage of relaxation from the maximal response induced by PE. Cumulative concentration–response curves were analyzed by computer fitting to a four-parameter sigmoid curve using the Prism 6 program (GraphPad, San Diego, CA) to evaluate pD2 value (negative logarithm of the molar concentration producing the half maximum response) and E_max (maximum asymptote of the curve). For statistical comparison of single parameters, an independent t test was used. Multiple comparisons in cultured vascular smooth muscle cells were made using ANOVA with post hoc Bonferroni test. Data from two vascular rings of the same rat were averaged and presented as the datum for one rat, with the n value representing the number of rats. Statistical significance was assumed if P < 0.05.

Results

Dihydrotestosterone exposure leads to hyperandrogenism and hypertension in females

Plasma DHT levels were significantly increased in DHT females (230.4 ± 16.2 pg/mL; n = 6; P < 0.05) compared with controls (110.8 ± 11.6 pg/mL; n = 6). However, estradiol levels were comparable between control (13.1 ± 0.5 pg/mL; n = 6) and DHT (13.7 ± 0.6 pg/mL; n = 6) groups. MAP was increased significantly in DHT rats (131 ± 9.2 mm Hg; n = 6, P < 0.05) compared with control rats (105 ± 2.5 mm Hg; n = 6) (Figure 1A). There was a proportional increase in systolic and diastolic pressures in DHT females compared to controls (Figure 1B and C). Heart rate was not significantly different between control (420 ± 18.5 beats per minute; n = 6) and DHT (408 ± 6.0 beats per minute; n = 6) females (Figure 1D).

Figure 1.

MAP and heart rate in control and DHT rats. MAP (A), systolic pressure (B), diastolic pressure (C), and heart rate (D) were continuously monitored via carotid arterial catheters in rats treated with placebo or DHT pellets (7.5 mg, 90-day release, subcutaneous). Values are presented as mean ± SEM of six animals in each group. *P < 0.05 vs control.

Endothelium-derived hyperpolarizing factor-mediated relaxation in mesenteric arteries is decreased in dihydrotestosterone females

To investigate the EDHF-mediated relaxation in the rat mesenteric artery, we examined ACh-induced relaxation in the presence of L-NAME and indomethacin. The ACh-induced EDHF-mediated relaxation was significantly lower in rings from DHT rats (n = 6, P < 0.05) than in those from controls (n = 6) (Figure 2 and Table 1).

Figure 2.

Endothelium-dependent EDHF-mediated vascular relaxation in mesenteric arteries of control and DHT rats. Arterial rings were isolated from control and DHT rats, contracted with PE, and examined for relaxation to ACh in the presence of NOS inhibitor L-NAME (10⁻⁴ M) and with PGI₂ blocker indomethacin (10⁻⁵ M). Values are means ± SEM (n = 6 rats, two vessel segments/rat).

Table 1.

Vascular function (pD₂ and E_max values) in control and DHT rats.

Relaxation	pD2		Emax
	Control	DHT	Control	DHT
EDHF mediated	6.96 ± 0.05	6.36 ± 0.09*	86.5 ± 3.42	66.2 ± 5.75*
Cx37 inhibition	6.42 ± 0.09⁁	5.71 ± 0.08*⁁	75.9 ± 4.01⁁	54.65 ± 4.83*⁁
Cx40 inhibition	6.84 ± 0.06	6.24 ± 0.07*	86.1 ± 4.17	61.5 ± 3.81*
Cx43 inhibition	5.98 ± 0.12⁁	6.22 ± 0.04	66.9 ± 3.80⁁	62.6 ± 2.43
Levcromakalim induced	6.18 ± 0.04	6.11 ± 0.03	98.8 ± 0.35	97.5 ± 0.38

Values are expressed as mean ± SEM of 12 mesenteric arterial rings from six rats in each group. pD2 is presented as negative logarithm of the molar concentration producing the half maximum response. E_max is presented as a percentage of PE relaxation.

*P < 0.05 compared to control.

^⁁ P < 0.05 compared to EDHF-mediated relaxation in their respective control and DHT groups.

Cx43-mediated vasodilator response was reduced in mesenteric arteries of dihydrotestosterone females

To address the involvement of Cx activities in DHT-impaired mesenteric EDHF vasodilation, we examined ACh-induced EDHF relaxation in the absence or presence of specific inhibitors. To examine the contribution of Cx37 in the EDHF-mediated relaxation, rings were incubated with ^37,43Gap27, a Cx37 inhibitor. Preincubation with ^37,43Gap27 significantly attenuated the ACh-induced EDHF relaxation to a similar magnitude in mesenteric arteries from control and DHT rats (n = 6 in each group, Figure 3A and Table 1).

Figure 3.

Role of Cxs in EDHF-mediated vascular relaxation in mesenteric arteries from control and DHT rats. Arterial rings were contracted with PE and examined for EDHF-mediated ACh relaxation in the presence of Cx mimetic peptides to inhibit (A) Cx37 (^40,37GAP26, 300 μM), (B) Cx40 (⁴⁰GAP27, 300 μM), or (C) Cx43 (^43,37GAP27, 300 μM). Left panel is control and right panel is DHT group. Values are means ± SEM (n = 6 rats, two vessel segments/rat).

To examine the part played by Cx40 in EDHF-mediated relaxation, rings were incubated with ⁴⁰Gap27, a Cx40 inhibitor. Presence of ⁴⁰Gap27 did not alter EDHF relaxation responses in the mesenteric arteries of control and DHT rats (n = 6 in each group, Figure 3B and Table 1).

To examine the contribution of Cx43 in the EDHF-mediated relaxation, rings were incubated with ⁴³Gap27, a Cx43 inhibitor. Preincubation with ⁴³Gap27 did not affect EDHF relaxation in DHT rats, but it significantly attenuated the EDHF relaxation in control rats (n = 6 in each group, Figure 3C and Table 1, P < 0.05). Overall, these data imply that DHT treatment does not affect Cx37 or Cx40 components of EDHF relaxation, but only inhibit the Cx43 component of relaxation response to ACh.

Endothelium-independent levcromakalim-induced relaxation response was not altered in dihydrotestosterone females

Vascular relaxation to levcromakalim, an ATP-sensitive K⁺ channel opener used to determine the smooth muscle vasodilating capacity [55], was not significantly different between the control and DHT rats (n = 5 in each group, Table 1).

Cx43 expression is decreased in the mesenteric arteries of dihydrotestosterone females

We next determined the expression profile of Cxs in the mesenteric arteries. As shown in Figure 4, quantitative real-time PCR shows the expression of all three Cxs in rat mesenteric arteries. Compared to control rats, vessels in DHT rats showed significantly decreased expression levels of Cx43 mRNA only (↓48%, Figure 4, P < 0.05, n = 5 in each). The expression of the Cx37 and Cx40 mRNA was not different between control and DHT rats (n = 5, Figure 4).

Figure 4.

Cxs mRNA expression in mesenteric arteries of control and DHT rats. Cxs mRNA expressions were assessed using real-time reverse transcriptase PCR. Quantitation of vascular Cxs was normalized relative to β-actin levels. Values are given as means ± SEM of five rats in each group. *P < 0.05 vs control.

Western blotting showed that Cx37 and Cx40 protein levels were unaltered, while Cx43 protein levels were significantly decreased (↓46%) in the mesenteric arteries from DHT rats (full length western blot in supplementary Figure S1) than in those from the control (n = 5 in each) (Figure 5, P < 0.05). The immunofluorescent pattern of Cx43 was identified in endothelial and smooth muscle cells of mesenteric arteries of both control and DHT groups (Figure 6, n = 4). However, a local marked decrease in the number and intensity of fluorescent spots was observed significantly in the media than endothelium of DHT compared to control rats (Figure 6, n = 4).

Figure 5.

Cxs protein expression in mesenteric arteries of control and DHT rats. Cxs protein expressions were assessed using western blotting. Representative western blots for respective Cxs are shown at top; blot density obtained from densitometric scanning normalized to β-actin is shown at bottom. Values are given as means ± SEM of five rats in each group. *P < 0.05 vs control.

Figure 6.

Representative immunofluorescence of Cx43 in media and endothelium of the mesenteric artery of control and DHT rats revealed differences in the amount and intensity of immunolabeling. Original magnification of all pictures: ×200. Relative fluorescence intensity in the endothelium and vascular smooth muscle cell are presented. Values are given as means ± SEM of four rats in each group. *P < 0.05 vs control.

Dihydrotestosterone decreases Cx43 gene expression levels in cultured mesenteric artery smooth muscle cells

Based on the effects of hyperandrogenism on Cx43 expression and relaxation in mesenteric arteries, we next assessed the direct effects of DHT stimulation on Cx43 expression using an in vitro cell culture model. Cx43 mRNA levels in cultured mesenteric artery smooth muscle cells were dose-dependently decreased by DHT treatment (Figure 7A, P < 0.05, n = 4). DHT treatment also significantly decreased Cx43 protein levels (Figure 7B, P < 0.05, n = 4), indicating that androgens directly decrease Cx43 expression.

Figure 7.

DHT downregulates Cx43 expression in cultured mesenteric artery smooth muscle (MASM) cells. Concentration-dependent DHT-induced decrease in Cx43 (A) mRNA, and (B) protein in cultured MASM cells. The primary MASM cells were treated with vehicle or DHT for 5 days with fresh medium along with DHT replaced every day. Cell extracts were prepared and subjected to Cx43 expression analysis using qRT-PCR and western blot. Quantitation of vascular Cx43 mRNA and protein expression was normalized relative to β-actin. Data represent mean of four independent experiments. *P < 0.05 vs controls.

Discussion

The major finding of this study is the identification of a novel androgen-mediated mechanism that controls the expression of Cx43 in mesenteric artery smooth muscle cells by negatively regulating Cx43 transcript and protein levels. Regulation of Cx43 expression by androgens has significant functional consequences, as we determined that androgens exert significant control by blunting EDHF-mediated endothelial relaxation and inducing hypertension. Therefore, we suggest that decreases in vascular Cx43-stimulated responses may mediate the development and maintenance of hypertension induced by hyperandrogenism in adult females.

The model was designed to create a physiological state of hyperandrogenism that mimics androgen levels observed in women with PCOS. DHT, a nonaromatizable androgen, was selected to avoid the confounding effects of estrogen. DHT treatment increased plasma DHT levels by 2-fold compared to controls, which is consistent with the available data that DHT levels remain relatively stable across the menstrual cycle and are approximately 2-fold higher in women with PCOS compared with those without PCOS (140–160 pg/mL vs 70–80 pg/mL, respectively) [56,57].

The DHT-treated rats have higher blood pressure compared to control rats, which is consistent with the reports of increased blood pressure in hyperandrogenic PCOS women [42–44] and suggests that the mechanisms controlling blood pressure in adult females are perturbed by elevated androgens. The effect on arterial pressure increase is without any changes in heart rate, indicating that DHT does not affect the sympathetic activity.

Kidneys play an important role in long-term control of blood pressure [58]. To dissect out the effect of DHT on vascular function, we examined EDHF function in isolated mesenteric arteries. The current study shows that EDHF contributes substantially to mesenteric arterial relaxation, supporting earlier findings [59,60]. The most striking finding of our study is that EDHF-mediated mesenteric vasodilation was significantly decreased in DHT rats compared with controls, suggesting that elevated androgens blunt endothelial EDHF function. It has been shown that EDHF plays a more important role in vasodilation in females than in males [61–63], and thus suppression of such a critical regulator of vascular tone by androgens could contribute to increased blood pressure in DHT females. Mesenteric vasorelaxation to levcromakalim was not different between control and DHT females, suggesting that it is not the smooth muscle vasodilating capability that is reduced in DHT females, but some function related to the initiation and propagation of hyperpolarization signals.

In the systemic circulation, the principal EDHF components that mediate initiation and propagation of hyperpolarization responses and vasodilation are Cx37, Cx40, and Cx43. We analyzed the effect of DHT on the EDHF function with the specific role of Cx37, Cx40, and Cx43. The current study shows that Cx37 contributes substantially to mesenteric arterial relaxation as in earlier studies [64–66], but the presence of elevated DHT does not affect Cx37-mediated relaxation. Consistently, elevated DHT did not alter Cx37 mRNA and protein levels. Studies have shown that Cx40 plays an important role in mediating vascular relaxation in mesenteric artery of male Wistar rats [67]; however, its contribution in the mesenteric arterial relaxation in females, and in general to systemic blood pressure, is minimal [34,68.69], which is consistent with our findings. Furthermore, we find that the Cx40-mediated relaxation was not affected by the presence of elevated DHT. This observation is also supported by unchanged Cx40 mRNA and protein levels in the mesenteric arteries of control and DHT rats.

The Cx43 contributed to EDHF relaxation in controls, and its lack of effect on EDHF relaxation in mesenteric arteries of DHT rats suggests that androgen exposure resulted in a loss of the regulatory role of Cx43 in initiating or propagating hyperpolarization. It has been suggested that in females there is a relative predominance of Cx43-dependent vasodilation [34,47,48] that may increase its susceptibility to the effect of elevated androgens. Our data also suggest that the loss of Cx43’s regulatory role in mesenteric arteries of DHT rats resulted chiefly from reduced gap junctional activities due to suppressed expression of this Cx. At present, the mechanisms by which androgens downregulate expression and function of Cx43 in mesenteric arteries specifically in the vascular smooth muscle cells of DHT rats are not clear. The finding that DHT exposure to cultured mesenteric artery smooth muscle cells downregulated Cx43 mRNA suggests that Cx43 is a physiological target for DHT and that androgens can directly regulate Cx43 expression, possibly at the transcriptional level. This ability of androgens to suppress Cx43 expression and activity could contribute to lower levels of Cx43 in male than female organs, like in the heart [70]. This notion is supported by the fact that castration induced increased Cx43 mRNA and protein levels in rat prostate tissue which was abolished by androgen replacement [71]. Further studies are needed to examine the exact mechanism by which androgen regulates Cx43 transcription. It is important to emphasize the following cautionary remarks regarding the aforementioned interpretations. First, although we hypothesize that the decreased Cx43 vascular function in DHT rats could contribute to the observed increase in arterial pressure; further studies using genetic approaches such as the Cx43 knockout mice will be required to demonstrate a causal role for Cx43. Second, although widely used, the pharmacological specificity of Cx mimetic peptides should be carefully examined in future studies. The failure to detect a role for Cx37 and Cx40 could be due to the delivery or pharmokinetic issues of Cx37 and Cx40 mimetic peptides, pointing again for the need for genetic approaches.

In conclusion, this study demonstrates that elevated levels of androgens at concentrations found in clinical conditions induce reduced EDHF-mediated vasodilation and hypertension in females rats, possibly via downregulating Cx43-mediated signaling, providing a molecular mechanism linking elevated androgens and increased risk of hypertension in females. Although caution should be always observed in extrapolating the findings of animal studies directly to humans, the present finding has translational potential and provides a mechanistic understanding worthy of investigation in humans. Strategies targeting attenuation of excessive androgen or upregulation of Cx43 action in the systemic circulation could have important therapeutic potential in treatment of hypertension complicated by elevated androgens.

Supplementary data

Supplementary data are available at BIOLRE online.

Supplementary Table S1. Antibodies used for Western blot and immunofluorescence studies.

Supplementary Figure S1. Full-length western blot showing Cxs protein expression in mesenteric arteries of control and DHT rats.

Supplemental material

Supplementary data are available at BIOLRE online.