Long-term administration of resveratrol and MitoQ stimulates cavernosum antioxidant gene expression in a mouse castration model of erectile dysfunction

Abstract

Aims:

Erectile dysfunction is a common complication within many pathological conditions associated with low testosterone. Testosterone deficiency increases oxidative stress in the penile tissue that contributes to endothelial dysfunction and subsequent erectile dysfunction. Current therapies do not ameliorate oxidative stress so targeting oxidative stress may improve erectile dysfunction. Resveratrol and MitoQ are two prospective drugs that have antioxidant-like properties and may be useful to improve erectile dysfunction induced by androgen deprivation.

Materials and methods:

We castrated 12-week-old male C57BL/6 mice and performed an eight-week intervention with oral delivery of resveratrol or MitoQ at low and high doses. We assessed vascular reactivity of the corpus cavernosum and internal pudendal arteries (IPA) through dose-dependent responses to vasodilatory, vasocontractile, and neurogenic stimuli in a myograph system. We performed qRT-PCR to measure expression changes of 18 antioxidant genes in the corpus cavernosum.

Key findings:

Castration significantly impaired erectile function via impaired endothelial-dependent and-independent relaxation, and increased constriction of the corpus cavernosum, and induced severe endothelial dysfunction of the IPA. Castration decreased expression of 8 of the antioxidant genes investigated. Resveratrol and MitoQ were ineffective in reversing the effects of androgen deprivation on vascular reactivity, however high-dose resveratrol treatment upregulated several key antioxidant genes, including Cat, Sod1, Gstm1, and Prdx3.

Significance:

Our findings suggest that oral resveratrol and MitoQ treatment may provide protection to the corpus cavernosum under androgen deprived conditions by stimulating endogenous antioxidant systems. However, they may need to be paired with vasoactive drugs to reverse erectile dysfunction under androgen deprived conditions.

1. Introduction

Testosterone declines in men at a rate of approximately 1% each year after 40 years of age,[1] which is associated with an increased prevalence in erectile dysfunction (ED).[2] Approximately 52% of men experience some symptoms of ED throughout their life.[3] Current research has demonstrated that erectile function is highly dependent on testosterone.[4–6] Beyond the age-related decline of testosterone production (termed hypogonadism), there are several pathological conditions associated with decreased testosterone production, such as obesity and diabetes mellitus.[7] All these conditions share the common complication of organic ED[8] which may present in hypogonadal men younger than 40 years of age.

Patients treated for prostate cancer also frequently suffer from organic ED. This patient population typically undergoes chemical castration (androgen deprivation therapy) as a form of treatment to shrink and minimize metastasis of the prostate cancer.[9] This is an effective treatment since the prostate is an androgen-dependent organ, however ED is developed as an unintended consequence.[10] Studies show testosterone therapy can improve male sexual function, including erectile function in hypogonadal men,[5] however this is contraindicated for prostate cancer patients.[6] While several studies suggest promising results from testosterone therapy, controversy remains surrounding its use.[11] First line ED therapies including oral phosphodiesterase 5 (PDE5) inhibitors are ineffective due in part to the severe oxidative stress in erectile tissue resulting from castration.[12] Finding alternative therapies that can counteract some of the effects of testosterone deficiency may allow advances in ED treatment.

Oxidative stress has been causatively linked to ED pathogenesis resulting from a variety of pathologies, including those associated with diminished testosterone production. Oxidative stress arises when reactive oxygen species (ROS) production outweighs antioxidant clearance of ROS.[13] Under normal conditions, production and clearance are balanced, however, under pathological conditions associated with oxidative stress, ROS production may be increased and/or ROS clearance may be impaired. ROS clearance is primarily modulated by antioxidant and detoxification enzymes. The primary modulator of the antioxidant defense system is the nuclear factor E2-related factor 2 (Nrf2).[14] Nrf2 is a transcription factor that translocates into the nucleus where it then binds to the antioxidant response elements (ARE) on DNA which leads to increases in the transcription of many downstream antioxidant/detoxification genes e.g., glutamate-cysteine ligase catalytic subunit (Gclc), NADPH quinone dehydrogenase 1 (Nqo1), and heme oxygenase 1 (Hmox1). Many of the genes stimulated by Nrf2 are components of the glutathione and thioredoxin antioxidant systems or function to regenerate components of these systems. Upregulation of antioxidant defense through the stimulation of Nrf2 can occur using antioxidant-like drugs such as resveratrol and Mitoquinol Mesylate (MitoQ). Resveratrol indirectly stimulates Nrf2 through sirtuin 1 (Sirt1) activation.[15] It is unknown whether MitoQ directly or indirectly activates Nrf2, but its antioxidant benefits seem to be mediated, in part, by Nrf2 activation.[16,17] The benefits of these drugs have been elicited in several cardiovascular disease models that possess severe oxidative stress and endothelial dysfunction. Treatment with resveratrol or MitoQ in these models demonstrate reduced oxidative stress levels through decreases in ROS levels and improved antioxidant system defense.[13,17,18] Mitigation of oxidative stress in these cases has coincided with restoration of endothelial function. Resveratrol has been shown to improve erectile function in rodent models of ED resulting from Type 1 diabetes, hypercholesterolemia, and unpredictable chronic mild stress.[12,19–21] These therapies may therefore be beneficial in the treatment of ED resulting from diminished testosterone production. Thus far, no research has been conducted on the effects of any antioxidant-like compounds on erectile function following androgen deprivation. Therefore, the aim of this study was to determine the effect of resveratrol and MitoQ on endothelial function and antioxidant defense in a mouse model deprived of testicular androgen production.

2. Methods

2.1. Experimental protocol

Male C57Bl/6NHsd mice (n=104) were purchased from Envigo (Indianapolis, IN, USA) at 12 weeks of age. All mice were housed in individual cages for 10 days with ad libitum access to standard rodent chow (LabDiet #5001, St. Louis, MO, USA) and water. Mice were randomized into six groups of equal body weight (bw). The sham surgery-only (SHAM) group underwent a sham surgery where testicular androgen production remained intact.[22] Castration-only (CAST), resveratrol fed castrated group (CR), and MitoQ fed castrated group (CM) groups underwent bilateral surgical orchiectomies to stop testicular androgen production. Buprenorphine was given to all mice immediately after surgery and again 5 hours post-surgery to reduce post-operative pain. Dietary interventions began three days following castration surgery. The SHAM and CAST groups had ad libitum access to food and water for the duration of the intervention. CR Low was fed a resveratrol enriched diet with a concentration of 0.08 mg/g of food (~10 mg/kg mice bw) and CR High was fed a resveratrol enriched diet with a concentration of 0.8 mg/g (~100 mg/kg mice bw), whereby resveratrol was mixed into the standard rodent chow (5001 PMI Laboratory Rodent Diet, Envigo Teklad Diets, Madison, WI, USA). These doses convert to approximately 1 and 10 mg/kg, respectively, when using the Human effective dose (HED) formula.[23] Resveratrol doses were chosen based on the lowest and maximal doses that have shown beneficial effects following treatment.[24,25] Resveratrol (manufactured by Chengdu Wagott Bio-Tech Co., Ltd.) was kindly provided by Barrington Nutritionals (Layton, UT, USA). CM Low consumed MitoQ at a concentration of 250 μM (~20 mg/kg mice bw, HED≈2 mg/kg) and CM High at 500 μM (~45 mg/kg mice bw, HED≈4 mg/kg) dissolved in the drinking water.[23] 250 μM MitoQ has previously been shown to restore carotid artery endothelial function in aged mice [13], while 500 μM MitoQ has been demonstrated to be safe with no observed toxicity.[26] MitoQ was kindly provided by MitoQ Limited (Auckland, New Zealand). All mice had ad libitum access to food and water. Food and water levels were replaced and recorded every two and three days, respectively. The intervention lasted for eight weeks to mimic an extended period of androgen deprivation. Following the intervention, mice were euthanized in the freely fed state by double pneumothorax and exsanguination of the vena cava under deep anesthesia following an intraperitoneal injection of 90 mg/kg ketamine and 10 mg/kg xylazine. Tissues were then extracted and used immediately for ex vivo physiological experimentation or were harvested for quantitative real time polymerase chain reaction (qRT-PCR) analysis. For harvesting corpus cavernosum (CC) tissue, penile tissue was exposed by freeing the tissue of skin and fascia. The penile shaft was then isolated by cutting at the penile base and the proximal glans. The corpus spongiosum, dorsal vein, and connective tissues were stripped from the penile shaft. The remaining CC tissue was placed on a Kimwipe, and blood was expelled by gently rolling over the tissue with a curved forceps. The tissue was rinsed in ice-cold phosphate buffered saline, patted dry, and snap frozen in liquid nitrogen and stored at −80°C until processing. The Animal Care and Use Committee at Florida State University approved these procedures (protocol #1831).

2.2. RNA Extraction and qRT-PCR

CC tissues were homogenized in 500 μl of buffer RLT provided in Qiagen RNeasy Mini Kit (Qiagen Inc., Hilden, Germany). RNA was isolated using the same RNeasy Kit according to the manufacturer’s instructions. RNA quantity and purity was determined with a NanoDrop One microvolume spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA). cDNA was synthesized from 300 ng of total RNA using Ready-To-Go You-Prime First Strand Beads (Cytiva, Marlborough, MA, USA). qRT-PCR was conducted on a QuantStudio3 (ThermoFisher Scientific) RT-PCR thermal cycler using TaqMan Fast Advanced Master Mix (ThermoFisher Scientific). The conditions for TaqMan Fast Advanced Master mix included an initial 2 min cycle at 50°C and 20 sec at 95°C, followed by 60 cycles that included a 1 sec denature step at 95°C, a 20 sec annealing step at 60°C, and a 1 min extension step at 72°C within each cycle. Each experimental group included samples from 8–12 animals and each sample was performed in triplicate. Relative expression levels of all genes were normalized using the delta delta Ct method.[27] 18S was used as the internal control as 18S expression was not affected by either castration or treatment. Primer sequences and functions for all TaqMan gene expression assays are listed in Table 1.

Table 1.

Gene information and Taqman primers used for qRT-PCR.

General Biochemical Function	Gene	Description	Assay ID	Amplicon Size	Accession Number
Glutathione Antioxidant System	Gclc	Glutamate-Cysteine Ligase Catalytic Subunit	Mm00802658_m1	78	NM_010295
	Gpx1	Glutathione Peroxidase 1	Mm00656767_g1	134	NM_008160
	Gpx4	Glutathione Peroxidase 4	Mm00515041_m1	103	NM_008162
	Gclm	Glutamate-Cysteine Ligase Modifier Subunit	Mm00514996_m1	76	NM_008129
Thioredoxin Antioxidant System	Txnip	Thioredoxin Interacting Protein	Mm01265659_g1	71	NM_001009935 NM_023719
	Txn1	Thioredoxin 1	Mm00726847_s1	113	NM_011660
	Txnrd1	Thioredoxin Reductase 1	Mm00443675_m1	63	NM_001042513 NM_001042514 NM_001042523 NM_015762
	Prdx3	Peroxiredoxin 3	Mm00545848_m1	133	NM_007452
	Prdx5	Peroxiredoxin 5	Mm00465365_m1	70	NM_012021
Detoxification	Nqo1	NADPH quinone dehydrogenase 1	Mm01253561_m1	81	NM_008706
	Gstm1	Glutathione S-Transferase 1	Mm00833915_g1	177	NM_010358
	Sod1	Superoxide Dismutase 1	Mm01344233_g1	71	NM_011434
	Sod2	Superoxide Dismutase 2	Mm01313000_m1	67	NM_013671
	Sod3	Superoxide Dismutase 3	Mm00448831_s1	97	NM_011435
	Cat	Catalase	Mm00437992_m1	64	NM_009804
Vasodilation	Nos1	Nitric Oxide Synthase 1 (Neuronal NOS, nNOS)	Mm00435175_m1	69	NC_000071
	Nos3	Nitric Oxide Synthase 3 (Endothelial NOS, eNOS)	Mm00435217_m1	71	NM_008713
Antioxidant System Regulation	Sirt1	Sirtuin 1	Mm01168521_m1	94	NM_001159589 NM_019812
	Sirt3	Sirtuin 3	Mm00452131_m1	68	NM_022433 NM_001127351 NM_001177804
Heme and Iron Metabolism	Hmox1	Heme Oxygenase 1	Mm00516005_m1	69	NM_010442
Inflammatory System	Nos2	Nitric Oxide Synthase 2 (Inducible NOS, iNOS)	Mm00440502_m1	66	NM_010927
	Tnf	Tumor Necrosis Factor	Mm00443258_m1	81	NM_001278601 NM_013693
	Nlrp3	NLR Family Pyrin Domain Containing 3	Mm00840904_m1	84	NM_145827
qRT-PCR Standards	18S	18S Ribosomal RNA	Hs99999901_s1	187	X03205

2.3. Ex vivo vascular reactivity

Penile tissue was removed under a dissection microscope and placed in ice-cold Krebs solution of the following composition (in mM): NaCl 130, KCl 4.7, KH₂PO₄ 1.18, MgSO₄ 1.18, NaHCO₃ 14.9, dextrose 5.6, CaCl₂ 1.56, and EDTA 0.03 dissolved in distilled water. All components of the Krebs solution were purchased from Sigma Aldrich (St. Louis, MO, USA). The penile shaft was separated from the glans penis then the urethra, dorsal vein, and connective tissues were carefully excised from the penile shaft in the chilled Krebs solution under dissection microscope. Two segments of the CC were obtained by cutting along the septum and mounted in a DMT 820MS muscle strip myograph system (Danish Myotechnology A/S (DMT), Aarhus, Denmark). Distal segments of the internal pudendal arteries (IPA) were dissected out as previously described by our lab and mounted in a DMT 620M wire myograph system using 25 μm tungsten wire.[28] Myograph systems were coupled to a PowerLab 8/30 data acquisition system (AD Instruments, Dunedin, New Zealand) for continuous recording of isometric tension measurement with LabChart Pro software (AD Instruments). Tissue segments were bathed in Krebs solution maintained at 37°C and continuously bubbled with a 95% O₂ and 5% CO₂ mixture. CC tissue was allowed to equilibrate for 1 h, then stretched to a resting tension of 4 mN, followed by an additional 1 h of equilibration. The length-tension relationship of the IPA segments was tested by gradually increasing the diameter until the transmural pressure exceeded 100 mmHg, and the vessel diameters were set to 90% of the internal circumference that elicited 100 mmHg transmural pressure calculated by DMT Normalization software (LabChartPro, AD Instruments). Tissue and IPA viability and contractile function were tested with high potassium (120 mM) Krebs solution, with KCl substituted for NaCl. Following successive washes with Krebs solution to achieve a stable resting tension, segments were constricted with cumulative dose-response applications of 0.001 μM-10 μM of the α₁-adrenoreceptor agonist phenylephrine (PE), 0.001 μM-3 μM of the prostaglandin/thromboxane A₂ agonist U-46619, and electrical field stimulation (EFS). EFS was applied with a CS8 stimulator (Danish Myo Technology) controlled by DMT MyoPULSE software. The following parameters were used for EFS: frequencies of 1–32Hz, 10 second stimulations, 2 millisecond pulse width at 20 V, and 2 minutes between each stimulation. Vasodilatory capacity was tested following 1 μM PE (IPA) or 10 μM PE (CC) pre-constriction using 0.001 μM-3 μM nitric oxide (NO) donor sodium nitroprusside (SNP), 0.001 μM-10 μM acetylcholine (ACh), and non-adrenergic, non-cholinergic nerve (NANC) mediated relaxation. NANC mediated relaxation was tested following a 30-minute incubation with 1 μM atropine and 30 μM guanethidine followed by EFS. Atropine acts as a competitive antagonist for muscarinic receptors and inhibits cholinergic signaling pathway thus preventing acetylcholine-mediated relaxation. Guanethidine inhibits the release of norepinephrine from adrenergic fibers. With these drugs, NANC fibers are isolated and relaxation from these fibers can be measured. NANC-mediated relaxation was then repeated following a 30-minute incubation with 5 μM Mito-TEMPO in the organ bath. Mito-TEMPO is a mitochondria-targeted superoxide scavenger. All relaxations to vasodilators were normalized as a percentage restoration to the resting tension from the PE pre-constricted value. Contractile responses were expressed as a percentage of maximal KCl-induced constriction. All drugs were dissolved in distilled water except for atropine, which was dissolved in 100% ethanol. All drugs were purchased from Sigma Aldrich except for U-46619 and acetylcholine, which were purchased from Enzo Biochem (New York, NY, USA).

2.5. Statistical Analysis

For gene expression analysis, the Brown-Forsythe test was applied to test for equal variances between the groups, after which one-way analysis of variance (ANOVA) was used to compare differences of the means between groups. For ex vivo vascular reactivity studies, two-way repeated measures ANOVA was used to compare differences between groups. All data are presented as mean ± standard error of the mean (SEM). Tukey’s post hoc test was used to correct for multiple comparisons and to determine differences between individual groups where significant differences in the ANOVA model were observed. All analysis was completed using GraphPad Prism Software v9 (La Jolla, CA, USA). Significance was set at p≤0.05 for all analyses.

3. Results

3.1. Animal Model Characteristics

The initial body weight of all groups was similar. At the end of the eight-week intervention, all castrate groups had a lower body weight than the SHAM group (p<0.05). There were no differences in weight between the resveratrol and MitoQ treated groups compared to CAST. SHAM weight was 32.86±0.73, CAST was 29.86±0.92, CR Low was 27.83±0.64, CR High was 28.68±0.59, CM Low was 28.42±0.54 and CM High was 28.86±0.43 (means ± SEM). CR Low consumed 9.9±0.98 mg resveratrol per kg body weight per day while CR High consumed a daily amount of 114.73 ± 10.59 mg/kg of resveratrol (means ± standard deviation). CM Low consumed an average of 20.20±1.91 mg/kg of MitoQ, and CM High drank a daily amount of 43.67 ± 4.43 mg/kg of MitoQ.

3.2. Quantitative gene expression analysis.

Quantitative gene expression of several antioxidant, detoxification, and inflammatory-related genes were evaluated using qRT-PCR and represented in Figures 1–3. Figure 1A–F. shows detoxication related genes. Castration significantly decreased Nqo1, superoxide dismutase 2 (Sod2), and Sod3 (fig. 1 A, E, F; p≤0.05) while there were no changes with glutathione s-transferase 1 (Gstm1), Sod1, or catalase (Cat) gene expression (fig. 1 B, C, D). Sod1 breaks down superoxide to hydrogen peroxide and has been classified as an androgen-resistant gene, where neither castration nor testosterone treatment affects its expression in prostate tissue.[9] Hydrogen peroxide is then converted to water and oxygen by Cat. Cat has been considered to have a complex androgen relationship.[9] In this case, castration has no effect on expression, but testosterone treatment increases its expression in prostate tissue. In the present study, high-dose resveratrol treatment significantly increased Sod1 and Cat expression compared to CAST (p≤0.05). These two enzymes are critical for antioxidant regulation. Like Cat and Sod1, Gstm1 expression was not affected with castration but was significantly elevated with high-dose resveratrol treatment. Gstm1 performs detoxification in conjugation with glutathione. High-dose resveratrol and MitoQ were trending to increase Nqo1 and Sod2 compared to the low-dose groups, however they were not significant compared to the castration-only group.

Figure 1.

The expression of genes related to detoxification in mouse corpus cavernosum was tested by qRT-PCR. Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered either a low-dose (CR Low) or high-dose (CR High) resveratrol or low-dose (CM Low) or high-dose (CM High) MitoQ. (A) NADPH dehydrogenase quinone 1 (Nqo1); (B) Glutathione s-transferase 1 (Gstm1); (C) Catalase (Cat);(D) Superoxide dismutase 1 (Sod1); (E) Superoxide dismutase 2 (Sod2); (F) Superoxide dismutase 3 (Sod3). Gene expression was normalized to 18S expression. Values represent means ± SEM for n = 8–12 animals per group. *p≤0.05 compared to SHAM. †p≤0.05 compared to CAST. ‡p≤0.05 compared to same treatment group (low vs high).

Figure 3.

The expression of genes related to antioxidant system regulation, inflammatory markers, and vasodilation in mouse corpus cavernosum was tested by qRT-PCR. Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered either a low-dose (CR Low) or high-dose (CR High) resveratrol or low-dose (CM Low) or high-dose (CM High) MitoQ. Inflammatory markers: (A) Tumor necrosis factor (Tnf); (B) Inducible NOS (Nos2); (C) NLR family pyrin domain containing 3 (Nlrp3). Antioxidant: (D) Heme oxygenase 1 (Hmox1). Antioxidant system regulation: (E) Sirtuin 1 (Sirt1); (F) Sirtuin 3 (Sirt3). Vasodilation: (G) Neuronal NOS (Nos1); (H) Endothelial NOS (Nos3). Gene expression was normalized to 18S expression. Values represent means ± SEM for n = 8–12 animals per group. *p≤0.05 compared to SHAM. †p≤0.05 compared to CAST. ‡p≤0.05 compared to same treatment group (low vs high).

Glutathione and thioredoxin antioxidant system-related genes are represented in Figure 2. Gclc, glutamate-cysteine ligase modifier subunit (Gclm), glutathione peroxidase 1 (Gpx1), and Gpx4 are a part of the glutathione antioxidant system (Fig 2A–D). Gclc and Gclm are responsible for glutathione biosynthesis while Gpx1&4 are directly involved in redox reactions. This system is impaired with castration as seen with significant decreases in Gclc and Gpx1 (p<0.05). Neither resveratrol nor MitoQ treatment significantly improved the genes in this system however, they were trending to increase when comparing the high-dose to low-dose treatment. High-dose MitoQ treatment elevated expression of Gclm to levels comparable to SHAM.

Figure 2.

The expression of genes related to glutathione and thioredoxin antioxidant system in mouse corpus cavernosum was tested by qRT-PCR. Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered either a low-dose (CR Low) or high-dose (CR High) resveratrol or low-dose (CM Low) or high-dose (CM High) MitoQ. Glutathione antioxidant system: (A) Glutamate-cysteine ligase catalytic subunit (Gclc); (B) Glutamate-cysteine ligase modifier subunit (Gclm); (C) Glutathione peroxidase 1 (Gpx1); (D) Glutathione peroxidase (Gpx4). Thioredoxin antioxidant system: (E) Thioredoxin interacting protein (Txnip); (F) Thioredoxin 1 (Txn1); (G) Thioredoxin reductase 1 (Txnrd1); (H) Peroxiredoxin 3 (Prdx3); (I) Peroxiredoxin 5 (Prdx5). Gene expression was normalized to 18S expression. Values represent means ± SEM for n = 8–12 animals per group. *p≤0.05 compared to SHAM. †p≤0.05 compared to CAST. ‡p≤0.05 compared to same treatment group (low vs high).

To examine the thioredoxin antioxidant system, thioredoxin interacting protein (Txnip), thioredoxin 1 (Txn1), thioredoxin reductase 1 (Txnrd1), peroxiredoxin 3 (Prdx3), and Prdx5 were assessed (Figure 2E–I). Txnip is a prooxidant marker that enzymatically inhibits Txn1 which is a major regulator for redox reactions. Like Txn1, Txnrd1, Prdx3, and Prdx5 are all involved in and perform redox reactions. Castration had no effects on Txnip and Txnrd1 expression but significantly reduced Prdx3&5 expression when compared to SHAM (p<0.05). The high-dose resveratrol group had significantly elevated Txnip expression compared to all other groups. For Txn1, we see elevated expression after castration (p<0.05), which was not significantly impacted by resveratrol or MitoQ treatment. High-dose resveratrol and MitoQ were trending to increase the expression of Txnrd1 and Prdx5 when compared to CAST and the low-dose groups. However, both high-dose treatments significantly elevated Prdx3 expression compared to low-dose treatment groups.

The inflammatory markers, tumor necrosis factor (Tnf), inducible nitric oxide synthase (Nos2), and NLR family pyrin domain containing 3 (Nlrp3), are represented in figure 3(A–C). Tnf is an inflammatory cytokine, and, in our study, there were no significant changes to its expression. Nlrp3 modulates inflammation, immune response, and apoptosis.[29] The treated castrated groups had elevated Nlrp3 expression compared to the SHAM group. Nos2 is an inflammatory associated nitric oxide synthase (NOS) and was elevated within all the castration groups. Next in Figure D, Hmox1 was significantly reduced after castration. High-dose groups partially restored Hmox1 expression back to levels comparable to SHAM. Resveratrol performs many of its antioxidant functions through activation of Sirt1. Exploring Sirt1 and Sirt3 expression post-castration and treatment yield no significant changes between all the groups (fig E–F). Neuronal NOS (Nos1, nNOS) and endothelial NOS (Nos3, eNOS) expression did not significantly change post castration. However, high-dose resveratrol did significantly elevate Nos3 expression compared to SHAM and the low-dose group (p<0.05).

3.3. Ex vivo vascular reactivity

The endothelial-dependent and endothelial-independent relaxation of the CC and IPA to ACh and SNP are represented in Figure 4. Castration significantly reduced the relaxation response to ACh and SNP for both CC and IPA. Neither resveratrol nor MitoQ treatment significantly affected endothelium-dependent or -independent relaxation of either tissue.

Figure 4.

Assessment of endothelial-dependent relaxation and endothelial-independent relaxation of the corpus cavernosum (CC) and internal pudendal arteries (IPA). Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered a high-dose resveratrol or MitoQ treatment (CR/CM High). Endothelial-dependent and endothelial-independent relaxation was tested with a cumulative dose-response to acetylcholine (ACh) and the NO-donor sodium nitroprusside (SNP), respectively. (A) CC relaxation response to ACh; (B) IPA relaxation response to ACh; (C) CC relaxation response to SNP; (D) IPA relaxation response to SNP. Values represent means ± SEM for n = 10 animals per group. *SHAM different from all other groups (p≤0.05).

Castration significantly impaired NANC-mediated relaxation of the CC but not the IPA (fig. 5A–B). We observed no treatment effects of resveratrol or MitoQ for the CC, however, there was a significant main effect of MitoQ treatment on NANC-mediated relaxation of the IPA relative to the CAST group, as well as significantly increased IPA relaxation at the 16 Hz frequency. EFS-mediated constriction is represented in figure 5C–D. Castration significantly increased the contractile response of CC but there were no differences in IPA constriction. Treatment did not affect the EFS-constriction response of the CC. In figure 6, the contractile responses to PE and U-46619 are represented. At sub-maximal concentrations of PE, all the castrated groups had greater contractile responses of the CC than the SHAM group. Next, all castrated groups had greater CC constriction to U-46619 than SHAM, while there were no treatment effects observed. There was a significant two-way ANOVA main effect of increased U-46619-mediated constriction of the IPA in the CAST relative to the SHAM group (p=0.010), while there was a significant main effect of reduced U-46619-mediated constriction of the IPA with MitoQ treatment relative to the CAST group (p=0.023).

Figure 5.

Assessment of NANC-mediated relaxation and nerve-stimulated contraction of the corpus cavernosum (CC) and internal pudendal arteries (IPA). Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered a high-dose resveratrol or MitoQ treatment (CR/CM High). NANC-mediated relaxation was tested by pre-constricting IPA with 1 μM PE and CC with 10 μM of PE and the addition of 1 μM Atropine and 30 μM Guanethidine, followed by electrical field stimulation (EFS). Nerve-stimulated contraction was tested by electrically stimulating the CC and IPA for 10 seconds at increasing frequencies (1–32 Hz). (A) CC relaxation response to EFS; (B) IPA relaxation response to EFS; (C) CC contraction to EFS; (D) IPA contraction to EFS. Values represent means ± SEM for n = 10 animals per group. *p≤0.05 compared to SHAM. † p≤0.05 CM High compared to CAST. †† ≤0.05 significant two-way ANOVA main effect CM High compared to CAST.

Figure 6.

Assessment of vasoconstriction of the corpus cavernosum (CC) and internal pudendal arteries (IPA). Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered high-dose resveratrol or MitoQ treatment (CR/CM High). Constriction was tested with 0.001 μM-10 μM phenylephrine (PE) and 0.001 μM-3 μM U-46619. (A) CC contraction response to PE; (B) IPA contraction response to PE; (C) CC contraction response to U-46619; (D) IPA contraction response to U-46619. Values represent means ± SEM for n = 10 animals per group. * SHAM different from all other groups (p≤0.05). Significant two-way ANOVA main effect † p≤0.05 CAST compared to SHAM and †† ≤0.05 CM High compared to CAST.

The CC response to Mito-TEMPO is represented in figure 7. This figure compares the NANC-mediated relaxation response of each group with and without the mitochondria targeted antioxidant Mito-TEMPO added directly to the organ bath. For SHAM and CM, there were no significant differences following Mito-TEMPO incubation. However, Mito-TEMPO significantly improved relaxation of the CC for the CAST and CR groups. This suggest that mitochondrial ROS production/oxidative stress is impairing relaxation responses in the CC and shows that directly targeting mitochondrial ROS with antioxidants may improve erectile function even though it is still significantly lower than SHAM.

Figure 7.

Assessment of NANC-mediated relaxation of the corpus cavernosum (CC) without Mito-TEMPO (Veh) or with Mito-TEMPO. Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered a high-dose resveratrol or MitoQ treatment (CR/CM). NANC-mediated relaxation was tested by pre-constricting CC with 10 u μM of PE, 1 μM Atropine, 30 μM Guanethidine, and 5 μM Mito-TEMPO, followed by electrical field stimulation. (A) SHAM relaxation response w/o Mito-TEMPO; (B) CAST relaxation response w/o Mito-TEMPO; (C) CR relaxation response w/o Mito-TEMPO; (D) CM relaxation response w/o Mito-TEMPO. Values represent means ± SEM for n = 10 animals per group. *p≤0.05 significant two-way ANOVA main effect of Mito-TEMPO treatment.

Figure 8 compares the NANC-mediated relaxation response of the IPA of each group with and without Mito-TEMPO added directly to the organ bath. The addition of 5 μM Mito-TEMPO significantly augmented NANC-mediated relaxation in the SHAM and CAST IPAs, however it did not improve relaxation in the CR or CM groups. As there were no deficiencies in NANC-mediated relaxation of the IPA associated with castration, the enhancement of this response in both untreated groups implicates a possible suppressive role of mitochondrial ROS on vasodilation of this artery.

Figure 8.

Assessment of NANC-mediated relaxation of the internal pudendal arteries (IPA) without Mito-TEMPO (Veh) or with Mito-TEMPO. Mice underwent a sham-surgery (SHAM) or were surgically castrated (CAST). Additional sets of castrated mice were administered high-dose resveratrol or MitoQ treatment (CR/CM). NANC-mediated relaxation was tested by pre-constricting IPA with 1 μM of PE, 1 μM Atropine, 30 μM Guanethidine, and 5 μM Mito-TEMPO, followed by electrical field stimulation. (A) SHAM relaxation response w/o Mito-TEMPO; (B) CAST relaxation response w/o Mito-TEMPO; (C) CR relaxation response w/o Mito-TEMPO; (D) CM relaxation response w/o Mito-TEMPO. Values represent means ± SEM for n = 10 animals per group. *p≤0.05 significant two-way ANOVA main effect of Mito-TEMPO treatment.

4. Discussion

In the present study, we investigated the effects of androgen deprivation via castration on CC and IPA vasoreactivity as an assessment of erectile function. Upon castration, there was an increase in neurogenic (EFS), PE, and thromboxane A₂-mediated constriction in the CC. ACh-mediated, SNP-mediated, and NANC-mediated relaxation of the CC were all markedly decreased. IPA demonstrated increased thromboxane A₂-mediated constriction and decreased relaxation to ACh and SNP following castration. There were no changes in PE constriction, neurogenic constriction, or NANC-mediated relaxation of the IPA. Based on these results, castration led to the development of ED. Antioxidant treatment with resveratrol and MitoQ did not fully reverse the loss of erectile function, however, MitoQ did decrease thromboxane A₂-mediated constriction in the IPA and augmented NANC-mediated relaxation of the IPA. Additionally, acute treatment with Mito-TEMPO improved NANC-mediated relaxation of the CC and IPA in the untreated castrated mice. Next, we examined expression changes for genes associated with antioxidant defenses, detoxification, and inflammation. Several of these genes, including Nrf2 targets: Gclc, Nqo1, and Hmox1, were markedly decreased following castration. Although Sod1, Gstm1, and Cat gene expression did not decrease following castration, there were notable increases following high-dose resveratrol treatment. Lastly, our results demonstrate that resveratrol and MitoQ treatment has dose dependent effects on Cat, Sod1, Prdx3, Txnip, Hmox1, and Nos3 gene expression. There were significant improvements with several genes following high-dose treatment that was not present following low-dose treatment.

ED pathogenesis is complex and can result from a multitude of sources. Normal erectile function is dependent on smooth muscle relaxation of the CC and relaxation of the IPA to allow the inflow of blood into the CC. Decreased vasodilation of the IPA is accepted as a major cause of ED.[30,31] We observed markedly impaired endothelium-dependent and -independent relaxation of CC eight weeks following castration. The decrement in ACh-mediated relaxation of the CC was more pronounced than the decrement in SNP-mediated relaxation, indicating that there likely was some degree of endothelial dysfunction present in the CC. However, the impaired vasodilatory response to NO at the smooth muscle level clearly impacted the ACh response. This response is consistent with the reported decline in penile smooth muscle quality following castration.[32] While we also observed impairments in endothelium-dependent and -independent relaxation of the IPA following castration, the impairment in the smooth muscle response to NO was more subtle and only apparent in the middle of the SNP dose-response curve. In contrast, the impaired vasodilatory response to ACh in the IPA following castration was drastic, indicating severely impaired endothelial function in the IPA of these animals. These results are consistent with a prior study that examined IPA functionality in castrated rats. Four weeks following castration, Alves-Lopes et al. found that castrated rats had lower ACh relaxation and no relaxation differences in SNP-relaxation. Contrary to our study, castration did significantly reduce NANC-mediated relaxation of the IPA. Additionally, histomorphometry of the IPA showed castration increased vascular remodeling.[30] Interestingly, a recent study demonstrated that castration reduced IPA relaxation to ACh but not for the systemic vasculature like the aorta and mesenteric arteries.[33] In all, castration may lead to localized alterations in IPA function and inward hypotrophic vascular remodeling.[30]

In addition to the severe suppression of multiple vasodilatory pathways, we also observed castration to upregulate multiple contractile pathways. Castration increased neurogenic constriction of the CC but not the IPA. Additionally, it increased contraction of CC from the stimulation of α₁-adrenoreceptors with phenylephrine, although, not to the same extent as U-46619. There was a subtle increase in PE constriction sensitivity of the CC in castrated groups, whereas U-46619 induced markedly greater CC constrictions in all castrated groups across the dose-response curve. U-46619 stimulates the prostaglandin/thromboxane A₂ receptor, which has been shown to stimulate vasoconstriction in penile arteries and the CC via Rho kinase-mediated changes in Ca²⁺ sensitivity.[34] Indeed, increased U-46619-stimulated vasoconstriction of penile arteries of obese Zucker rats has been attributed to enhanced Ca²⁺ sensitivity.[35] It appears likely that these mechanisms are enhanced in the CC following castration. We observed a more subtle increase in U-44619-mediated constriction in the IPA following castration, which was abrogated by chronic MitoQ treatment. Suppression of the heightened post-castration thromboxane A₂ sensitivity of the IPA may be one avenue through which MitoQ positively influences erectile function.

Testosterone is a key hormone for the development and maintenance of normal penile health and function. We have demonstrated that the loss of androgens promotes a somewhat differential effect on vasoreactivity between the CC and IPA. A similarly asynchronous effect occurs when examining the functional effects of aging on the human CC and penile resistance arteries compared to arteries of the systemic vasculature. Using specimens from organ donors, Assar et al. demonstrated that aging caused dysfunction in endothelium-dependent relaxation and hypercontractility in all vascular types. However, the CC and penile resistance arteries developed dysfunction after 40 years of age, whereas the mesenteric artery and the aorta maintained function until 65 years of age.[36] This may be due in part to the distinct reliance of the CC and penile arteries on testosterone and the age associated decline in testosterone that begins at 40 years of age. It has been implicated that the decline of erectile function from the loss of testosterone is in part due to the decline in endothelial health caused by an increase in oxidative stress.[37,38] Castration models have been useful to elicit the effects of a loss in testosterone on erectile function.[4,37–41] Additionally, it represents a model for prostate cancer patients who receive chemical castration as a form of treatment. In the penis, antioxidant system regulation and capacity are primarily controlled by testosterone, hence the downregulation of Nrf2 targets following castration. A previous study showed that testosterone propionate was able to stimulate the Nrf2/ARE pathway in aged liver cells resulting in increased expression of Nqo1, Hmox1, Cat, and glutathione, demonstrating that testosterone influences these antioxidant systems beyond the pelvis.[42] We examined the propensity for oxidative stress by measuring gene expression changes in the CC for several genes associated with antioxidant defenses, detoxification, and inflammation. Upon castration, we observed a reduction in expression of the antioxidant genes: Gclc, Gpx1, Prdx3, Prdx5, Nqo1, Sod2, Sod3, and Hmox1. To our knowledge, this is the first study to demonstrate these expression changes following castration. To further probe the influence of oxidative stress on erectile function, we assessed NANC-mediated relaxation pre- and post-incubation with Mito-TEMPO, a compound that directly detoxifies ROS produced in the mitochondria. Mito-TEMPO incubation improved the CC relaxation response of the CAST group, whereas no significant effects were observed for the SHAM group. These results may suggest that acute castration-induced mitochondrial ROS inhibited the relaxation response. Similarly, Mito-TEMPO enhanced the CC relaxation response of resveratrol treated animals, but not MitoQ treated animals. It is plausible that chronic treatment with MitoQ sufficiently suppressed mitochondrial ROS such that further quenching of mitochondrial ROS acutely had no effect on function, which was not the case for the resveratrol treated animals. While there were no castration-induced changes in NANC-mediated relaxation of the IPA observed, acute Mito-TEMPO exposure augmented the response in the SHAM and CAST groups, while Mito-TEMPO had no effect on the IPAs of mice chronically treated with either antioxidant compound. Interestingly, it appears that targeting mitochondrial ROS either chronically with in vivo MitoQ or acutely with ex vivo Mito-TEMPO increases NANC-mediated relaxation of the IPA. As we are unaware of any prior studies that have utilized either of these mitochondria-targeted antioxidants in the context of erectile function, this is an area that should be further investigated.

The use of resveratrol as a therapeutic drug to limit oxidative damage and improve endothelial health has been investigated extensively in cardiovascular disease models, however in vivo resveratrol administration has been shown to induce beneficial effects on erectile function in rodent ED models of type-1 diabetes, hypercholesterolemia, and chronic stress.[19–21] Cardiovascular disease models have been useful to outline the mechanism at which resveratrol performs its antioxidant functions. Resveratrol can act as an antioxidant directly, however, it primarily performs its antioxidant functions through Sirt1 activation. Sirt1 regulates many downstream processes that affect NO production and antioxidant defenses.[43] Resveratrol can enhance eNOS expression and activity, and reduce eNOS uncoupling. Expression of eNOS is assumed to be Sirt1 dependent due to downstream changes to forkhead box protein O (FOXO) transcription factors after resveratrol treatment. Resveratrol mimics some actions of testosterone through the regulation of the S1P1/Akt/ FOXO3a pathway. Testosterone can decrease oxidative stress and eNOS uncoupling by upregulating the S1P1/Akt/ FOXO3a pathway.[38] Moreover, resveratrol improves antioxidant defense by indirectly activating Nrf2. Nrf2 is a transcription factor that modulates our primary antioxidant defense system.[14] One study by Ungvari et al. used Nrf2 knockdown human primary coronary artery cells and Nrf2 knockout mice to show the effects of resveratrol on vascular function and antioxidant defense and that these effects were mediated by Nrf2.[18] First, they demonstrated increased transcriptional activity of Nrf2 and upregulation of Nrf2 targets Nqo1, Hmox1, and Gclc expression in cultured wild type but not Nrf2 knockdown human coronary arterial endothelial cells post-resveratrol treatment. Next, they fed one set of wild-type and Nrf2 knockout mice a high fat diet (HFD) and another set a HFD enriched with resveratrol. The primary findings were that resveratrol treatment restored vasodilation and decreased ROS production in the femoral artery of HFD-fed wild-type mice meanwhile the HFD-fed Nrf2 knockout mice only had partial restoration of vasodilation and partial decrease of ROS production.[18]

MitoQ is a mitochondrial-targeting antioxidant and is a novel therapeutic drug that still needs extensive research to support its uses.[44] It is a proficient antioxidant, partly due to Nrf2 stimulation. In a model of intestinal ischemia reperfusion, MitoQ pretreatment attenuated the increases in oxidative stress and improved intestinal barrier injury. Additionally, there was significant upregulation in Nrf2 and subsequent elevations of Nqo1, Hmox1, and Gclc protein expression. Using siRNA to knockdown Nrf2 in intestinal cells, they were able to demonstrate that these beneficial changes were, in fact, partly due to Nrf2.[17] In our study, there were dose-dependent effects on gene expression following MitoQ treatment. Administration of a high dose of MitoQ exerted significant enhancements in Hmox1, Prdx3, and Cat compared to the lower dose treated group. For these genes, there was a significant decrease in expression in the low-dose group compared to SHAM suggesting castration led to impairments. Currently, the dose dependent effects of MitoQ on gene expression are unknown. Several studies demonstrate that MitoQ positive benefits on oxidative stress is mediated via Gclc, Nqo1, and Hmox1 however, contrary to other studies, we did not see a significant change in Gclc or Nqo1 gene expression. Our results demonstrate that MitoQ induced Nrf2 action can stimulate other downstream genes besides Nqo1, Gclc, or Hmox1. These cumulative changes in gene expression following treatment may contribute to the reduced oxidative stress and the endothelial improvements seen in other studies. In a study by Giosca-Ryan et al., using an age-related model of vascular endothelial dysfunction, old mice treated with 250 μM MitoQ infused drinking water had complete restoration of endothelial function of the carotid artery. Additionally, they saw an increase in NO bioavailability and a restoration of the NO component of endothelial dilation.[13] MitoQ treatment also reduced superoxide production and increased resistance to mitochondrial ROS stressors.

Current therapies such as PDE5 inhibitors are ineffective in more severe cases of ED including diabetes, obesity, and androgen deprived prostate cancer patients, due in part to excessive oxidative stress. Combination therapies with PDE5 inhibitors may be beneficial for this population. One study examines the use of resveratrol and sildenafil as a combination therapy to improved ED in a streptozotocin-induced model of type-1 diabetes. Sildenafil and resveratrol alone improved erectile function compared to untreated diabetic rats, however, resveratrol in combination with sildenafil was able to improve erectile function greater than either therapy alone.[12] Looking beyond the functional results, resveratrol treatment augmented cavernous eNOS and nNOS protein and mRNA levels, and attenuated cavernous ROS levels, whereas sildenafil had no impact on these parameters. Unsurprisingly, sildenafil had a greater restorative effect on cavernous cyclic guanosine monophosphate (cGMP) levels than resveratrol, however, combination therapy nearly restored cavernous cGMP levels to that of the control rats. Remarkably, they even demonstrated a partial restoration of serum testosterone levels with resveratrol treatment, which may have contributed to the observed effects of resveratrol.[12] Given the stimulatory effect of high-dose resveratrol and MitoQ treatment on several antioxidant genes in the androgen deprived cavernosum observed in this study, utilizing these therapeutics in combination with PDE5 inhibitors may plausibly increase PDE5 inhibitor potency in disease states where they are insufficiently effective. Many antioxidant strategies for clinical treatment of cardiovascular disease have been attempted over the past decades, very few of which have ultimately gained acceptance for broad clinical applications.[45] A myriad of reasons may explain the ineffective therapeutic efficacy, including but not limited to suboptimal dosing and therapeutic duration, ineffective delivery of the compound to the necessary compartment of the cell, targeting of the correct ROS for the particular disease state, compensatory upregulation of ROS producing enzymes, and ineffective modulation of redox-based post-translational modifications.[45] These same barriers apply for redox-based clinical treatment of ED. By the time a patient seeks therapy for ED, and especially if they are refractory to PDE5 inhibitors, it is likely that eNOS- and nNOS-mediated NO production are substantially attenuated, some degree of fibrosis may be present in the CC and hardening of the IPA and penile arteries may be present. In these cases, antioxidant therapy alone may not be powerful enough to substantially augment e/nNOS activity or to reverse structural changes. Indeed, several clinical trials have reported greater effectiveness of ED treatment when various antioxidants are combined with PDE5 inhibitors than with PDE5 inhibitor treatment alone.[46–50] Future clinical application of antioxidant compounds for ED treatment may ultimately include use at the very early stages of disease pathogenesis or in combination with PDE5 inhibitors. Even so, not all antioxidants are the same, and it is reasonable to hypothesize that compounds that promote activation of the endogenous Nrf2-mediated antioxidant defense system hold greater promise than oxidant scavengers or enzymatic inhibitors.

While this study did see enhancements of several detoxification genes, there was no direct measure of oxidative stress markers. Direct measurement of superoxide and other oxidative stress markers such as malondialdehyde levels can confirm the changes seen with castration and post-treatment. Moreover, we report the relative changes in gene expression levels which may not necessarily translate to changes at the protein or activity level. The next limitation deals with Txnip and Txn1 expression. Txnip is a prooxidant marker while Txn1 is directly involved in redox reactions in the thioredoxin antioxidant system. We did not expect Txnip expression to increase with treatment. Similarly, Txn1 levels increased with castration and surprisingly, was trending to decrease with treatment. The increase in Txnip and decrease in Txn1 expression may be counteracting each other, however we are unable to draw an accurate conclusion regarding activity of the thioredoxin system. Further studies should include a thioredoxin activity assay to accurately understand the changes caused by castration and treatment. A common complication following castration is fibrosis of the corpus cavernosum. Histological assessment would help characterize potential changes in tissue morphology arising from therapy. Given the variability of cavernous gene expression levels and the six experimental groups used in this study, we may have been underpowered to detect biologically relevant changes in gene expression. For example, both high-dose resveratrol and MitoQ restored Hmox1 to levels comparable to the SHAM group yet were not statistically different from the CAST group (p~0.09). Next, therapeutic treatment started three days after castration. Due to the myriad function of testosterone, it is likely that wide-ranging detrimental processes have established within three days of the sudden loss of androgens, as a decreased cavernous nerve-stimulated erectile response has been reported as early as one day following castration.[51] A pre-habilitation therapeutic strategy may alleviate some of these effects in this model.

5. Conclusion

Castration induced substantial changes in the vasoreactivity of the corpus cavernosum and the internal pudendal artery that are consistent with severe erectile dysfunction, which coincided with decreased expression of several genes in the cavernosum that are essential for maintenance of the cellular antioxidant defense systems. Neither long-term administration of resveratrol nor MitoQ considerably restored vasoreactivity in the castrated state, although a few modest changes were observed following MitoQ administration that may be marginally beneficial to overall erectile function. However, resveratrol and MitoQ were able to stimulate and improve expression of several antioxidant genes. They may be investigated as an additional therapeutic in conjunction with current remedies to improve treatment of erectile dysfunction. PDE-5 inhibitors lose their effectiveness in more severe cases of ED due in part to rampant oxidative stress. Resveratrol and MitoQ may sufficiently reduce oxidative stress in severe ED to restore PDE5 inhibitors effectiveness. Future research should explore co-treatment in a castrated model of ED.