Abstract
The increasing prevalence of obesity adds another dimension to the pathophysiology of testosterone (TEST) deficiency (TD) and potentially impairs the therapeutic efficacy of classical TEST replacement therapy. We investigated the therapeutic effects of selective androgen receptor modulation with trenbolone (TREN) in a model of TD with the metabolic syndrome (MetS). Male Wistar rats (n=50) were fed either a control standard rat chow (CTRL) or a high-fat/high-sucrose (HF/HS) diet. After 8 weeks of feeding, rats underwent sham surgery or an orchiectomy (ORX). Alzet miniosmotic pumps containing either vehicle, 2-mg/kg·d TEST or 2-mg/kg·d TREN were implanted in HF/HS+ORX rats. Body composition, fat distribution, lipid profile, and insulin sensitivity were assessed. Infarct size was quantified to assess myocardial damage after in vivo ischaemia reperfusion, before cardiac and prostate histology was performed. The HF/HS+ORX animals had increased sc and visceral adiposity; circulating triglycerides, cholesterol, and insulin; and myocardial damage, with low circulating TEST compared with CTRLs. Both TEST and TREN protected HF/HS+ORX animals against sc fat accumulation, hypercholesterolaemia, and myocardial damage. However, only TREN protected against visceral fat accumulation, hypertriglyceridaemia, and hyperinsulinaemia and reduced myocardial damage relative to CTRLs. TEST caused widespread cardiac fibrosis and prostate hyperplasia, which were less pronounced with TREN. We propose that TEST replacement therapy may have contraindications for males with TD and obesity-related MetS. TREN treatment may be more effective in restoring androgen status and reducing cardiovascular risk in males with TD and MetS.
There is a marked shift in global population demographics, with both an increase in the incidence of obesity (1, 2) and the average age of the global population (3). Our understanding of the etiology of age-associated pathologies is now complicated by obesity which is not an isolated medical condition but is associated with several comorbidities (ie, dyslipidaemia, hypertension, insulin resistance/diabetes), which all impact on cardiovascular health. In males, obesity-associated metabolic syndrome (MetS) and testosterone (TEST) deficiency (TD) are strongly interrelated (4). Pathological TD occurs in approximately 1 in 200 men (5) and appears most prevalently in obese males with obesity-associated MetS (6, 7). Together these pathologies present a high risk for cardiovascular disease, and increase morbidity and mortality, particularly in ageing males (8). Although there are additional criteria that have been suggested for the diagnosis of the MetS, it is consistently characterized by increased visceral adiposity, dyslipidaemia, and insulin resistance (9–11) and increases cardiovascular risk by 2-fold (12). TD impairs myocardial tolerance to ischemia reperfusion (I-R) both directly (13) and indirectly through myocardial desensitization to insulin (14, 15).
Research evidence suggests that the antiadiposity effects of TEST are mediated by the androgen receptor (AR). Male mice lacking ARs develop visceral obesity (16, 17). Also, inherited androgen resistance in men decreases AR-mediated gene transcription and is characterized by increased visceral adiposity (18). AR signaling in adipose tissue also plays a critical role in insulin sensitivity independently of changes in adiposity (19) and may provide a novel therapeutic target for the treatment of insulin resistance.
Treatment of TD with TEST replacement therapy (TRT) leads to improvements in body composition, serum lipid profile and glucose management (20–23), which may translate into improved outcomes after myocardial I-R insult. This growing scientific support for TRT helped promote the 12-fold increase in prescription of TEST between the years 2000 and 2011 (24). However, limitations inherent in the use of TRT have been identified, especially in overweight or obese populations, due to TEST's susceptibility to bioconversion to its active metabolites dihydrotestosterone and estradiol by the 5α-reductase (25) (expressed in the myocardium, prostate, and skin) (13, 26, 27) and aromatase (28) (expressed in adipose and other tissues; see Ref. 29).
The TEST metabolite dihydrotestosterone has an approximately 3-fold greater affinity for the AR than TEST and is not only responsible for the androgenic effects of TRT but also causes impaired myocardial tolerance to I-R (13) and localized, albeit predominantly benign, prostate hyperplasia (30). Pitt's Unified Theory of Prostate Cancer (31), however, supports a critical role for aromatase and its TEST metabolite estradiol (28) in the etiology of prostate cancer, independently of 5α-reductase activity.
Unlike TEST, 17β-hydroxyestra-4,9,11-trien-3-one (trenbolone [TREN]) is a selective AR modulator and does not undergo 5α-reduction or aromatization (32, 33). Additionally, TREN causes a more potent reduction in retroperitoneal/visceral fat pad mass than TEST in hypogonadic male rats (34, 35). More recently, we have shown that TREN improves multiple components of the MetS and improves myocardial tolerance to I-R in normogonadic rats (36).
In obese males, androgens are more readily converted to estradiol by adipose tissue aromatase (37, 38). A so-called “testosterone-estradiol” shunt has been implicated as a major driver for the vicious hypogonadal-obesity cycle and may also explain the failures of traditional TRT for the restoration of circulating TEST in obesity (39). In TEST-deficient animals, expression of 5α-reductase and aromatase appears to increase in the myocardium (13). We believe that these enzyme profile characteristics of obesity and TD warrant further investigation into the therapeutic efficacy of compounds modulating the AR without interaction with these enzymes, particularly in relevant pathological models.
In the present study, we investigated the systemic and cardiovascular effects of TREN treatment in TEST-deficient rats with the MetS. We hypothesized that: 1) TREN improves body composition, lipid profile, and insulin sensitivity (hallmarks of the MetS); and 2) TREN improves myocardial tolerance to I-R more effectively than TEST.
Materials and Methods
Fifty male Wistar rats (12 wk of age; 300 ± 30 g) were acquired from the Australian Research Centre. Rats were housed under an artificial 12-hour light, 12-hour dark cycle at a constant temperature of 21°C (40% humidity) and provided with ad libitum access to standard rat chow and water. All animal experimentation was approved and performed in accordance with the guidelines of the Animal Ethics Committee of Griffith University (permit number MSC/01/11) and the Australian code of practice for the care and use of animals for scientific purposes.
Experimental design
One week after arrival at the animal facility, 50 rats (13 wk of age) were randomly assigned to one of 5 groups (n = 10). The animals received either control standard rat chow (CTRL) (n = 10) or a high-fat/high-sucrose (HF/HS) diet (n = 40). After 8 weeks of feeding, CTRL (n = 10) and HF/HS (n = 10) rats underwent sham surgery and the remaining HF/HS animals (n = 30) underwent orchiectomy (ORX) followed by a 4-week recovery period. Alzet miniosmotic pumps were then implanted for drug or vehicle delivery for a further 8 weeks. Pumps contained vehicle (CTRL, n = 10; HF/HS, n = 10; HF/HS+ORX, n = 10), TEST (HF/HS+ORX+TEST, n = 10), or TREN (HF/HS+ORX+TREN, n = 10).
HF/HS diet
Rats assigned to the HF/HS groups were provided ad libitum access to a HF/HS diet for 20 weeks. Similar diets have been shown to induce hyperphagia (40) and increases visceral fat content (41) in rats. The macronutrient composition of the control (CTRL) diet was 69.9% carbohydrate (36.3% sugar), 22.9% protein, and 7.3% fat, whereas the HF/HS was 69.0% carbohydrate (54.6% sugar), 13.8% protein, and 17.3% fat (per 100-g dry weight).
Surgical procedures
Orchiectomies
Animals were anesthetized (2.5% isoflurane in 1 L min−1 100% medical grade O2) before external and periscrotal incisions were performed to access and remove both testes. The periscrotal cavity was closed using 6–0 prolene sutures, and the external scrotal incision was closed using 4–6 stainless steel wound clips. Animals designated as controls underwent a sham operation where an external scrotal and periscrotal incision was made before closing both wounds as described above.
Infusion pump implantation
Four weeks after the ORX/sham operations, Alzet infusion pumps (model 2004; Alza Corp) were implanted sc. Pumps were prepared with vehicle (45% wt/vol 2β-hydroxypropyl cyclodextrin), TREN (suspended in 45% wt/vol 2β-hydroxypropyl cyclodextrin), or TEST (suspended in 45% wt/vol 2β-hydroxypropyl cyclodextrin). The dosage delivery rate of each steroid was 2.0 mg/kg·d in accordance with previous work by ourselves using this delivery method and demonstrating a therapeutic effect on body composition (36). All 2β-hydroxypropyl cyclodextrin suspensions were prepared in Milli-Q H2O and passed through a 0.45-μm filter (Millex). Pumps were replaced after 4 weeks, in accordance with the lifespan of the device, to allow for an 8-week continuous treatment.
After all surgical procedures, burprenorphine (10 μg/kg·d, im) and enrofloxacin (5 mg/kg, ip) were administered for 3 days after surgery for the management of postoperative pain and infection, respectively; and all efforts were made to minimize animal suffering.
Body composition assessment
Body composition
Total, lean and fat mass were determined by dual-energy x-ray absorptiometry (DXA) (XR-800; Norland Medical Systems, Inc, host software version Illuminatus version 4.2.4) at 19 weeks of age. To perform scans, rats were sedated with ketamine (Ketamil; Troy Laboratories) and xylazine (ilium xylazil-20; Troy Laboratories) ip. Scans were performed at 1.5 × 1.5-mm resolution at 60 mm/s in small animal mode.
After animal killing, retroperitoneal, epididymal, and visceral fat pads were excised to quantify visceral fat content and fat distribution in each animal. Subcutaneous fat mass was determined by calculating the difference between total fat mass (quantified by DXA) and visceral (excised) fat mass.
In vivo myocardial tolerance to ischemia-reperfusion assessments
At 32 weeks of age, rats were anesthetized (sodium pentobarbital, 60 mg/kg ip) in the morning at 8 am ± 1 hour, intubated and ventilated (Model 683; Harvard Instruments) before being placed on an adjustable heating pad to maintain a core temperature of 36°C–37°C, as monitored by rectal thermometer (Model 52II; Fluke Corp). Maintenance of anesthesia was confirmed by assessing pedal withdrawal reflexes at 10-min intervals. A thoracotomy and reversible left anterior descending coronary artery ligation were performed as described previously (42). The left anterior descending coronary artery was occluded for 45 minutes followed by 120 minutes of reperfusion initiated by ligature release. At the end of reperfusion, hearts were rapidly excised, mounted on a Langendorff perfusion system, stained with Evan's Blue dye, and frozen overnight before tritetrazolium chloride counterstaining to delineate viable and necrotic myocardium, as outlined previously (42). Tissue areas were quantified by volumetric planimetry using a flat-bed scanner for image capture and computer software for image analysis (UTHSCSA Image Tool, V3), with infarct size (area) expressed as percentage of area at risk (ischemic area).
At killing, whole blood was collected from each animal and kept on ice at 4°C for 30 minutes before centrifugation at 1200g for 10 minutes. Serum was transferred to Eppendorf tubes for storage at −80°C for later analysis. Levels of serum cardiac-specific creatine kinase (CK-MB), were quantified using an automated clinical chemistry analyzer unit operated by a trained technician (COBAS INTEGRA 400; Roche). Concentrations of CK-MB were quantitated using calibrator for automated systems/lipids (CFAS/L) (Roche Diagnostics) in addition to validation using commercially prepared quality control (QC) specimens (Roche Diagnostics).
Serum lipids, insulin, and glucose assessments
Serum samples were thawed and prepared for multianalyte analysis (COBAS INTEGRA 400; Roche). All analytes of interest (total cholesterol, triglycerides and glucose) were quantified within an hour of sample thawing. Concentrations of each analyte were quantitated using CFAS/L (Roche Diagnostics) in addition to validation using commercially prepared QC specimens (Roche Diagnostics).
Serum insulin was quantified using a commercial homogenous time-resolved fluorescence kit (CISBIO) with a lower detection limit of 5 μU/mL and measured sample coefficient of variation (CV) less than 9% across the entire range of quantification.
Impaired insulin sensitivity (ie, insulin resistance) was routinely assessed using the homeostatic model assessment of insulin resistance (HOMA-IR) (43, 44). HOMA-IR values were calculated using the formula: [glucose (mg/dL) × serum insulin (μU/mL)]/405.
Testicular mass assessments and circulating sex hormone quantitation
Immediately after killing of sham-operated animals, each pair of testicles was surgically excised and weighed. A combined weight of each pair of testicles was reported for each animal.
Serum TEST concentrations were quantified by homogenous time-resolved fluorescence kit (CISBIO) with a lower detection limit of 2.6 ng/dL and measured sample CV less than 9% across the entire range of quantification.
Serum TREN concentrations were quantified by commercially available ELISA kit (Bio Scientific) with a lower detection limit of 0.05 ng/mL and measured sample CV less than 7% across the entire range of quantification.
Assessments of left ventricular (LV) mass and tissue morphology
Immediately after killing, the right ventricle and major vessels were removed from each heart and the LV was weighed. A 2-mm-thick transverse tissue biopsy was excised from the base of each rat heart above the ligation site for storage in formalin (10% vol/vol in 0.9% saline). One week after storage, samples were fixed and sectioned for hematoxylin and eosin (H & E) staining and histological assessment. Areas of replacement fibrosis as hallmarks of myocardial infarction (45) were identified as discolored (purple) accumulations of fibroblasts in the myocardium in place of typical cardiac myocytes (pink) and were delineated using computer software (UTHSCSA Image Tool, V3).
Assessments of liver enzyme activity and renal function
Several circulating biomarkers of hepatic activity and renal function were assessed in serum samples collected at killing. The activities of aspartate aminotransferase and alanine aminotransferase (markers of hepatic activity); and serum concentration of urea and total protein (markers of renal function) were quantified using an automated clinical chemistry analyzer unit operated by a trained technician (COBAS INTEGRA 400 Pathology System; Roche). Concentrations of each analyte were quantitated using CFAS/L (Roche Diagnostics) in addition to validation using commercially prepared QC specimens (Roche Diagnostics).
Prostate mass and morphology assessment
Immediately after killing, prostates were surgically excised, weighed, sectioned, and stored in formalin (10% vol/vol in 0.9% saline). Formalin-stored prostates were later sectioned and prepared for H & E staining and histological assessment.
Statistics
Although animals were housed in pairs, each rat was measured and assessed as a single experimental unit for statistical analyses. Two-way ANOVA and Fisher's Least Significant Difference test assessed differences between groups. P < .05 was considered significant.
Results
Body composition
At killing, although total body mass was similar between groups, body composition was greatly affected by the HF/HS diet, ORX, and androgen treatments.
The HF/HS diet increased total fat mass (Table 1) and fat mass expressed relative to body mass (39.5 ± 2.3 vs 31.9 ± 2.4% body mass in CTRL) (Figure 1A; and consequently decreased lean mass relative to body weight (57.6 ± 2.3 vs 65.4 ± 2.4% total mass in CTRLs, P < .05) (Figure 1B. Additionally, HF/HS diet feeding increased both visceral fat mass (Table 1) and visceral fat mass relative to body mass by approximately 30%, compared with CTRLs (8.3 ± 0.4 vs 6.2 ± 0.4% total mass in CTRLs) (Figure 2A.
Body Composition
| CTRL, n = 10 | HF/HS, n = 10 | HF/HS+ORX, n = 10 | HF/HS+ORX+TEST, n = 10 | HF/HS+ORX+TREN, n = 10 | |
|---|---|---|---|---|---|
| Body mass (g) | 730.1 ± 15.7 | 734.2 ± 20.8 | 718.6 ± 32.6 | 696.8 ± 22.3 | 690.8 ± 22.8 |
| Lean mass (g) | 477.1 ± 18.8 | 422.2 ± 18.6 | 367.3 ± 18.9ab | 407.3 ± 16.5a | 430.8 ± 16.6c |
| Fat mass (g) | 233.6 ± 18.0 | 291.1 ± 19.7a | 330.4 ± 19.7a | 268.4 ± 19.2c | 239.1 ± 25.0c |
| Subcutaneous fat mass (g) | 188.2 ± 16.2 | 230.2 ± 16.4 | 268.1 ± 16.0a | 213.6 ± 15.1c | 197.2 ± 21.3c |
| Visceral fat mass (g) | 45.4 ± 2.9 | 60.9 ± 3.9a | 62.4 ± 5.6a | 54.8 ± 4.9a | 41.9 ± 4.1bcd |
| CTRL, n = 10 | HF/HS, n = 10 | HF/HS+ORX, n = 10 | HF/HS+ORX+TEST, n = 10 | HF/HS+ORX+TREN, n = 10 | |
|---|---|---|---|---|---|
| Body mass (g) | 730.1 ± 15.7 | 734.2 ± 20.8 | 718.6 ± 32.6 | 696.8 ± 22.3 | 690.8 ± 22.8 |
| Lean mass (g) | 477.1 ± 18.8 | 422.2 ± 18.6 | 367.3 ± 18.9ab | 407.3 ± 16.5a | 430.8 ± 16.6c |
| Fat mass (g) | 233.6 ± 18.0 | 291.1 ± 19.7a | 330.4 ± 19.7a | 268.4 ± 19.2c | 239.1 ± 25.0c |
| Subcutaneous fat mass (g) | 188.2 ± 16.2 | 230.2 ± 16.4 | 268.1 ± 16.0a | 213.6 ± 15.1c | 197.2 ± 21.3c |
| Visceral fat mass (g) | 45.4 ± 2.9 | 60.9 ± 3.9a | 62.4 ± 5.6a | 54.8 ± 4.9a | 41.9 ± 4.1bcd |
Data presented as mean ± SEM.
P < .05 compared with CTRL.
P < .05 compared with HF/HS.
P < .05 compared with HF/HS+ORX.
P < .05 compared with HF/HS+ORX+TEST.
Body Composition
| CTRL, n = 10 | HF/HS, n = 10 | HF/HS+ORX, n = 10 | HF/HS+ORX+TEST, n = 10 | HF/HS+ORX+TREN, n = 10 | |
|---|---|---|---|---|---|
| Body mass (g) | 730.1 ± 15.7 | 734.2 ± 20.8 | 718.6 ± 32.6 | 696.8 ± 22.3 | 690.8 ± 22.8 |
| Lean mass (g) | 477.1 ± 18.8 | 422.2 ± 18.6 | 367.3 ± 18.9ab | 407.3 ± 16.5a | 430.8 ± 16.6c |
| Fat mass (g) | 233.6 ± 18.0 | 291.1 ± 19.7a | 330.4 ± 19.7a | 268.4 ± 19.2c | 239.1 ± 25.0c |
| Subcutaneous fat mass (g) | 188.2 ± 16.2 | 230.2 ± 16.4 | 268.1 ± 16.0a | 213.6 ± 15.1c | 197.2 ± 21.3c |
| Visceral fat mass (g) | 45.4 ± 2.9 | 60.9 ± 3.9a | 62.4 ± 5.6a | 54.8 ± 4.9a | 41.9 ± 4.1bcd |
| CTRL, n = 10 | HF/HS, n = 10 | HF/HS+ORX, n = 10 | HF/HS+ORX+TEST, n = 10 | HF/HS+ORX+TREN, n = 10 | |
|---|---|---|---|---|---|
| Body mass (g) | 730.1 ± 15.7 | 734.2 ± 20.8 | 718.6 ± 32.6 | 696.8 ± 22.3 | 690.8 ± 22.8 |
| Lean mass (g) | 477.1 ± 18.8 | 422.2 ± 18.6 | 367.3 ± 18.9ab | 407.3 ± 16.5a | 430.8 ± 16.6c |
| Fat mass (g) | 233.6 ± 18.0 | 291.1 ± 19.7a | 330.4 ± 19.7a | 268.4 ± 19.2c | 239.1 ± 25.0c |
| Subcutaneous fat mass (g) | 188.2 ± 16.2 | 230.2 ± 16.4 | 268.1 ± 16.0a | 213.6 ± 15.1c | 197.2 ± 21.3c |
| Visceral fat mass (g) | 45.4 ± 2.9 | 60.9 ± 3.9a | 62.4 ± 5.6a | 54.8 ± 4.9a | 41.9 ± 4.1bcd |
Data presented as mean ± SEM.
P < .05 compared with CTRL.
P < .05 compared with HF/HS.
P < .05 compared with HF/HS+ORX.
P < .05 compared with HF/HS+ORX+TEST.
Body composition. A, Total fat mass percentage. B, Lean mass percentage. Data presented as mean ± SEM. a, P < .05 compared with CTRL; b, P < .05 compared with HF/HS; c, P < .05 compared with HF/HS+ORX.
Fat distribution. A, Visceral fat mass percentage. B, Subcutaneous fat mass percentage. Data presented as mean ± SEM. a, P < .05 compared with CTRL; b, P < .05 compared with HF/HS; c, P < .05 compared with HF/HS+ORX; d, P < .05 compared with HF/HS+ORX+TEST.
ORX HF/HS diet-fed animals had increased total fat mass compared with the CTRL but not HF/HS group (Table 1). However, HF/HS+ORX animals had greater fat mass relative to body mass compared with the HF/HS group (45.9 ± 1.6 vs 31.9% body mass) (Figure 1A. The HF/HS+ORX animals also had reduced lean mass (Table 1) and lean mass relative to body mass (51.2 ± 1.6 vs 57.6 ± 2.3% body mass in HF/HS) (Figure 1B. Although ORX did not affect visceral fat mass or sc fat mass (compared with sham-operated HF/HS animals), HF/HS+ORX animals had increased sc fat mass relative to body mass when compared with HF/HS animals (37.26 ± 1.4 vs 31.25 ± 2.0% body mass in HF/HS) (Figure 2B.
The TEST and TREN treatments each reduced total fat mass (Table 1) and fat mass relative to body mass compared with HF/HS+ORX animals (38.3 ± 2.1 and 34.1 ± 2.8 vs 45.9 ± 1.6% body mass in HF/HS+ORX, respectively) (Figure 1A. A similar effect was observed with regards to sc fat mass (30.6 ± 1.8 and 28.1 ± 2.5 vs 37.3 ± 1.4% body mass in HF/HS+ORX, respectively) (Figure 2B. As a result, the TEST and TREN treatments each increased lean mass relative to body mass (58.6 ± 2.1 and 62.9 ± 2.9 vs 51.2 ± 1.6% body mass in HF/HS+ORX, respectively) (Figure 1B but only TREN increased lean mass compared with HF/HS+ORX animals (Table 1). Additionally, only TREN reduced visceral fat mass (Table 1) and visceral fat mass relative to body mass in all HF/HS-fed groups (6.0 ± 0.4 vs 8.3 ± 0.4, 8.6 ± 0.5 and 7.8 ± 0.5% body mass in HF/HS, HF/HS+ORX, and HF/HS+ORX+TEST, respectively) (Figure 2A.
Circulating lipids
The HF/HS group had circulating cholesterol levels similar to CTRLs (Figure 3A. However, circulating triglyceride levels were increased with HF/HS feeding (2.46 ± 0.18mM vs 1.75 ± 0.25mM in CTRLs) (Figure 3B.
Serum lipid measurements. Data presented as mean ± SEM. a, P < .05 compared with CTRL; b, P < .05 compared with HF/HS; c, P < .05 compared with HF/HS+ORX, d, P < .05 compared with HF/HS+ORX+TEST.
The HF/HS+ORX group had increased circulating cholesterol levels compared with sham-operated HF/HS animals (1.95 ± 0.09mM vs 1.54 ± 0.11mM in HF/HS) (Figure 3A. Circulating triglycerides were similar between HF/HS and HF/HS+ORX groups (P > .05).
Both TEST and TREN treatments reduced circulating cholesterol levels in ORX HF/HS diet-fed rats (1.27 ± 0.11mM and 1.11 ± 0.11mM vs 1.95 ± 0.09mM in HF/HS+ORX, respectively) (Figure 3A. However, only TREN, not TEST (P > .05), reduced circulating triglyceride levels (1.76 ± 0.15mM vs 2.49 ± 0.22mM in HF/HS+ORX) (Figure 3B.
Homeostatic model assessment of insulin resistance
Serum glucose levels were similar in all groups (Figure 4A. Serum insulin levels were similar between CTRL and HF/HS; and HF/HS and HF/HS+ORX groups (Figure 4B. Insulin increased almost 2.4-fold in HF/HS+ORX animals compared with CTRLs (89.4 ± 16.9 vs 37.8 ± 5.4 μU/mL in CTRLs). Insulin levels were partly normalized in HF/HS+ORX+TEST animals compared with CTRLs, however, insignificant. However, only TREN treatment significantly reduced serum insulin levels (56.1 ± 2.9 μU/mL) compared with untreated HF/HS+ORX animals (Figure 4B.
HOMA-IR. a, P < .05 compared with CTRL; b, P < .05 compared with HF/HS; c, P < .05 compared with HF/HS+ORX; d, P < .05 compared with HF/HS+ORX+TEST.
The HOMA-IR value for the CTRL group was 3.4 ± 0.8 which almost doubled in the HF/HS group (6.1 ± 1.8, not significant) and increased to 8.5 ± 1.6 in HF/HS+ORX animals (Figure 4C. TEST treatment partially improved HOMA-IR values (5.2 ± 1.2) to values comparable with CTRLs. However, only TREN treatment significantly reduced HOMA-IR values (2.6 ± 0.4) compared with untreated HF/HS+ORX animals (Figure 4C.
Assessments of liver enzymes and renal function
No significant changes were observed between any group in levels of either urea, total protein, aspartate aminotransferase, or alanine aminotransferase.
Circulating androgen levels
The HF/HS diet had no effect on circulating TEST levels. However, ORX reduced circulating TEST by 85% (22 ± 2 vs 143 ± 3 ng/dL in HF/HS) (see figure 5 below). HF/HS+ORX+TEST animals had TEST levels that were almost double that of CTRLs (273 ± 14 vs 155 ± 4 ng/dL in CTRLs). Treatment with TREN did not affect circulating TEST levels (14 ± 5 ng/dL) compared with those of HF/HS+ORX animals (see figure 5 below). Circulating TREN levels were 202 ± 2 ng/dL in TREN-treated animals (see Figure 5 below).
Serum hormone levels. A, TEST. B, TREN. Data presented as mean ± SEM. a, P < .05 compared with CTRL; b, P < .05 compared with HF/HS; c, P < .05 compared with HF/HS+ORX; d, P < .05 compared with HF/HS+ORX+TEST.
Prostate size and morphology
Prostate mass was similar in the HF/HS and CTRL groups (559 ± 47 vs 606 ± 22 mg in CTRLs) (Figure 6). However, prostate mass was reduced by 80% (71 ± 3 mg) in HF/HS+ORX compared with sham-operated HF/HS animals.
Prostate mass and morphology. A, Prostate mass measurements. B, Prostate histology. OM, original magnification. a, P < .05 compared with CTRL; b, P < .05 compared with HF/HS; c, P < .05 compared with HF/HS+ORX; d, P < .05 compared with HF/HS+ORX+TEST.
Both androgen treatments increased prostate mass compared with CTRL, HF/HS, and HF/HS+ORX groups. The prostates from HF/HS+ORX+TEST animals weighed 1333 ± 71 mg and were 2.2-fold heavier than those of CTRLs. The prostates from HF/HS+ORX+TREN animals weighed 1017 ± 39 mg and were 1.7-fold heavier than those from CTRLs. Notably, treatment with TREN increased prostate mass to only approximately 75% of that seen in TEST-treated animals (1017 ± 39 vs 1333 mg in HF/HS+ORX+TEST).
Histological assessment of prostates from CTRL and HF/HS animals revealed unremarkable glandular and stromal morphology (Figure 6B. Atrophic changes in glandular morphology were revealed in prostates of ORX (HF/HS+ORX) animals. Prostates of HF/HS+ORX+TEST animals revealed increased glandular proliferation consistent with nodular hyperplasia (Figure 6B. Prostates of TREN-treated animals exhibited only mild glandular hyperplastic changes. There was no definitive evidence of carcinoma in prostate tissue of any experimental group.
Cardiac assessments of LV mass and myocardium morphology
LV mass was similar in HF/HS and CTRL groups (1142 ± 38 vs 1173 ± 20 mg in CTRLs) (Figure 7A. However, ORX animals had reduced LV mass (1063 ± 35 mg) compared with CTRL but not HF/HS animals. Treatment of HF/HS+ORX animals with TEST or TREN protected against this reduction in LV mass (1140 ± 35 and 1143 ± 30 mg, respectively) (Figure 7A.
LV mass and morphology. A, LV weights. B, Cardiac histology. a, P < .05 compared with CTRL.
Histological assessment of the LVs of CTRL, HF/HS, and HF/HS+ORX animals revealed unremarkable myocardial morphology (Figure 7B. Histology of LVs from TEST-treated animals revealed widespread patches of replacement fibrosis consistent with repair after chronic tissue injury. Myocardium of TREN-treated rats also revealed limited fibrosis. No atypical cellular morphology was observed in the myocardium of rats from any experimental group (Figure 7B.
Assessments of myocardial tolerance to I-R
Serum CK-MB levels were similar between CTRL (1246 ± 166 IU/L) and HF/HS (925 ± 272 IU/L) groups (Figure 8A. However, HF/HS-fed animals had reduced infarct sizes compared with controls (19.5 ± 1.2 vs 25.8 ± 1.7% area at risk in CTRLs) (Figure 8B. ORX HF/HS-fed animals had increased CK-MB levels (2514 ± 452 IU/L) and infarct sizes (41.2 ± 1.4% area at risk) compared with both CTRLs and sham-operated HF/HS animals.
Assessments of myocardial ischemia-reperfusion damage. A, Serum CK-MB measurements. B, Infarct size; values expressed as percentage of area at risk (AAR). a, P < .05 compared with CTRL; b, P < .05 compared with HF/HS; c, P < .05 compared with HF/HS+ORX; d, P < .05 compared with HF/HS+ORX+TEST.
Both TEST and TREN treatment significantly reduced CK-MB levels (1138 ± 362 and 653 ± 189 IU/L, respectively) and infarct size (23.3 ± 0.9 and 17.5 ± 1.5% area at risk, respectively) to values similar to both CTRL and HF/HS groups. Additionally, only TREN reduced CK-MB levels compared with CTRLs, and infarct sizes compared with CTRLs and HF/HS+ORX+TEST animals (Figure 8, A and B).
Discussion
The findings of the present study describe the cardiometabolic effects of TREN treatment on body composition, lipid profile-associated insulin sensitivity in a rodent model of the MetS with TD. Additionally, our findings demonstrate the beneficial effects of TREN on myocardial tolerance to I-R injury in this model. TEST, the current gold standard for the clinical treatment of TD, was used as a positive control in all assessments.
Our data demonstrate that TREN treatment has more marked cardioprotective effects that are partially mediated by the significant reductions in visceral adiposity, circulating triglycerides, and insulin resistance, in a model of TEST-deficient MetS. Compared with TEST treatment, which did not improve any of these measures, TREN also exhibited attenuated adverse effects on heart muscle (fibrosis) and prostate (hyperplasia) morphology.
Characteristics of the TEST-deficient MetS model
Our findings confirm that TEST-deficient MetS is characterized by increased sc and visceral adiposity, sarcopenia, hypercholesterolaemia, hypertriglyceridaemia, insulin resistance, low serum TEST, LV atrophy, and increased myocardial susceptibility to I-R injury. As we have demonstrated previously with similar models (35, 46), we propose that the HF/HS+ORX model used in the present study is an appropriate representation of males with TEST-deficient MetS.
Effects of androgens on body composition
Sex hormones play a central role in the management of body composition. Androgens promote the differentiation of precursors from the central pool of mesenchymal stem cell to increase lean (skeletal muscle and bone) mass at the expense of adipogenesis (8, 47). Idiopathic TD (48), androgen deprivation (49), and androgen blockade (50) increase fat mass and decreases lean mass. In contrast, androgen replacement therapy reduces adiposity in men with low TEST (48, 51) via assumed inhibition of mesenchymal stem cell adipogenesis (52) and interactions with the AR, inhibiting triglyceride uptake and lipoprotein lipase activity and causing a more rapid turnover of triglycerides in adipose tissue (53).
In the present study, both TEST and TREN treatment demonstrated favorable effects on body composition in TEST-deficient rats with the MetS. The increases in fat mass observed in HF/HS+ORX animals were normalized with both TEST and TREN treatment. Unlike TEST, TREN treatment additionally reduced visceral fat mass and protected animals against lean mass (muscle) loss. These results support previous findings which highlight TREN's ability to more effectively reduce retroperitoneal and visceral adiposity in male rats (36) than TEST (34). The antiobesity effects of androgens are directly mediated by the AR, which is more densely expressed in visceral than sc adipocytes (54, 55). We propose that the potent reductions in visceral adiposity and increases in lean body mass observed with TREN are due to its unique resistance to enzymatic conversion to estradiol in the adipose tissue of our obese rodent model. Both its increased bio-availability and its high selectivity for the AR, which is characteristic of its 17-β-hydroxylated steroidal structure (32, 33, 56), explain its improved efficacy in visceral adipose tissue reduction than TEST.
The enlarged visceral adipose depot has been identified as a key correlate of the altered cardiometabolic risk profile observed among individuals with the MetS (57). Therefore, we propose that the reductions in visceral adiposity observed with TREN treatment of HF/HS+ORX animals may translate into reduced cardiometabolic risk in TEST-deficient males with the MetS.
Effects of androgens on serum lipid profile
Increases in visceral adiposity are associated with elevated serum triglyceride levels (58) and impaired postprandial triglyceride clearance (59). Circulating lipid levels correlate with changes in energy use from the visceral fat depot, and reductions in adiposity lead to concomitant reductions in serum lipids, described elsewhere (15).
We recently demonstrated that TREN reduces high-density lipoprotein, low-density lipoprotein, and triglyceride levels in lean, normogonadic male rats (36), which is likely a consequence of reductions in visceral fat mass also induced by TREN in that study. Our current data provide further evidence of TREN's cholesterol and triglyceride reducing effects in a model of TEST-deficient MetS. Furthermore, in the present study, TEST treatment elicited a similar protection against hypercholesterolaemia to that of TREN, which suggests that serum cholesterol levels may be governed by overall body composition, independent of changes in visceral adiposity. Notably, TEST did not affect circulating triglyceride levels, providing evidence that the reductions in circulating triglycerides seen with TREN treatment are likely secondary to the reductions in visceral adipose tissue in these animals.
Dyslipidaemia is a result of visceral obesity (60), as well as a consequence to diet-induced deposition of lipids in other tissues (61–63), and is strongly implicated in the etiology of atherosclerosis (9, 64, 65) and increases the risk of cardiovascular events (specifically ischemic stroke) in patients with and without diabetes (66). We propose that the TREN-induced corrections of dyslipidaemia (increased circulating cholesterol and triglycerides) seen our HF/HS+ORX animals may translate into improved cardiovascular risk profile in TEST-deficient males with the MetS.
Effects of androgens on insulin sensitivity
Impaired fatty acid metabolism contributes to the insulin-insensitivity seen in individuals with visceral obesity with the surgical removal of visceral adipose tissue improving glucose tolerance (67), which is likely mediated by reductions in hepatic triglyceride production. Besides obesity, there is strong evidence to suggest that low TEST also promotes insulin resistance (68). Our data suggest that TREN induced modulation of the AR reduces visceral adipose tissue and consequently improves triglyceride metabolism and insulin sensitivity in TEST-deficient male rats with the MetS.
Insulin also plays a key role in the regulation of circulating triglyceride levels through its interactions with visceral adipose tissue (69, 70). Briefly, hypertrophied intraabdominal adipocytes are characterized by a hyperlipolytic state, which is resistant to the antilipolytic effect of insulin (71). Because the visceral adipose tissue depot enlarges, the release of free fatty acids and triglyceride formation increases. We propose that the improvements in insulin sensitivity seen with TREN treatment in our study may partly explain associated improvements in serum lipid profile, described earlier.
Although it remains unclear as to whether the accumulation of triglycerides in adipose, hepatic, and muscle tissue is a cause or consequence of the dyslipidaemic profile and insulin resistance, these factors appear to be strongly related and independently implicated in the etiology of cardiovascular pathology (72–74).
Effects of androgens on LV mass
For decades, male TD has been known to alter cardiac structure (75) and function (76), which is restored with TEST treatment. Our data from ORX, viscerally obese HF/HS-fed animals confirm these observations, with reductions in LV mass in these animals. In the current study, both TEST and TREN treatment protected against the atrophy observed in HF/HS+ORX animals. These findings may reflect hypertrophic changes mediated by androgens in males (77).
In contrast to our findings, TEST has been shown to correlate with modest atrophic remodeling of the LV in middle-aged men (78). It should, however, be noted that these observations were based on acute measurements of TEST within a population of primarily normogonadic individuals who were presumed to have normal physiological enzyme profiles. Rubio-Gayosso et al (13) recently reported an increased expression of 5α-reductase, and potentially also aromatase, in the myocardium of hypogonadic animals. However, it remains unclear whether expression negatively correlates with circulating TEST levels per se, or whether expression is down-regulated only in response to hypothalamic-pituitary-gonadal axis dysfunction.
Data from this study also provide some evidence to suggest that cardiac fibrosis results from chronic androgen treatment in TEST-deficient male rats with the MetS. Similar observations have been reported in autopsies of young men engaging in androgen abuse (79). The aetiologies of this fibrosis is beyond the scope of this study and are discussed elsewhere (80). Although the mechanisms remain unclear, the fibrotic response to androgen treatment may be driven by localized disruption to redox homeostasis in the cardiac myocyte (81).
Notably, the replacement fibrosis observed with TREN treatment was relatively modest when compared with that seen in TEST-treated animals and was only revealed in a single section of sampled myocardium. In contrast, the fibrosis observed in the hearts of TEST-treated animals displayed widespread replacement-type fibrosis which may have had more severe long-term consequences for cardiac remodeling. Evidently, the mechanisms responsible for this fibrosis and potential abrogative therapies clearly require further investigation.
It is worth noting that H & E staining is not traditionally considered the standard method for assessing fibrosis. The fibrotic lesions identified in the present study, which resulted from TEST and TREN treatment, does, however, justify the further investigation into this potential complication of androgen treatment using quantitative techniques.
Androgen effects on myocardial tolerance to I-R
TEST treatment exerts direct cardioprotective effects in hypogonadic (82), but not normogonadic (13), male rats. These observations imply that TD impairs myocardial tolerance to I-R. However, the cardioprotection elicited by TEST treatment is predominantly ascribed to its metabolites and their effects in the myocardium. In brief (13), TEST induces cardioprotection in TEST-deficient rats. Second, dihydrotestosterone paradoxically increases I-R injury in both intact and ORX rats through increased affinity for the AR, which may limit the cardioprotective effects of TEST. Third, inhibition of aromatase abrogates the cardioprotective effect of TEST treatment in ORX rats, which suggests that estradiol may be partly responsible for the protection with TEST.
This study is the first to demonstrate this effect in a model of TEST-deficient MetS. Our findings support those reported by other groups (82, 83), which demonstrate a significant cardioprotective effect of TEST in lean TEST-deficient rats. Importantly, this study demonstrates a greater cardioprotective efficacy of TREN (a nonaromatizable and non-5α-reducible selective AR modulator) when compared with TEST. These novel findings suggest that cardioprotection may in fact be conferred directly via modulation of the AR either: 1) independently of estrogenic activity; or 2) through undescribed cross talk/cross-reactivity between TREN and estradiol's receptor(s) in the myocardium. Further investigation into the intracellular steroid receptor interactions and cardioprotective signaling in response to sex hormone replacement is warranted.
In addition to the direct consequences of TD in the myocardium, obesity, dyslipidaemia, and insulin resistance are also implicated in the impaired myocardial tolerance to I-R. The myocardium's metabolic shift to free fatty acid dependency (away from glucose use) (84) in obesity could be detrimental to myocardial efficiency, particularly under ischemic conditions (85), due to the increased metabolic demand for oxygen to facilitate β-oxidation. Dyslipidaemia-associated fatty acid-induced uncoupling of glucose oxidation from glycolysis also results in an accumulation of H+ ions, which cause myocardial damage and appears to affect postischemic function (86–88).
Lastly, hearts from animal models of obesity and insulin resistance (89), isolated insulin resistance (90), or diabetes (91) have a reduced tolerance to I-R and suffer increased damage due to the myocardium's dependency on ATP generation from anaerobic glucose metabolism during ischemia, reviewed by ourselves elsewhere (15).
We propose that the cardioprotective effects of TREN treatment in HF/HS+ORX rats were likely mediated through both direct androgenic activity in the myocardium and indirectly via TREN-induced improvements in body composition, lipid profile, and insulin sensitivity.
Androgen effects in prostate tissue
The 5α-reductase enzymes expressed in prostate tissue (25) are responsible for the bioconversion of circulating TEST to dihydrotestosterone, which appears to play a key role in the development of benign prostate hyperplasia (30). Aromatase expression is also evident in proliferative stromal cells of the prostate, especially in regions surrounding hyperplastic glands (92), and may play a role in the induction or development of prostate disorders. However, TEST's metabolites are independently implicated in the etiology of both benign prostate hyperplasia and malignancy. Both TEST and dihydrotestosterone mediate their effects on the prostate via the AR (93).
In the present study, the hyperplastic effect observed in the prostates of rats treated with TEST was markedly reduced in prostates of TREN-treated animals. This is likely due to the inability of the prostate to convert TREN to dihydrotestosterone. In comparison, the selective AR modulation conferred by TREN treatment resulted in a more modest hyperplastic effect in the prostate, likely mediated through similar yet reduced AR affinity for TREN compared with dihydrotestosterone. Additionally, the assumed increases in estradiol bioavailability (via aromatization) in TEST-treated rats likely subjects prostates to increased oncogenesis and neoplastic growth, according to Pitt's Unified Theory of Prostate Cancer (31). We propose that TREN treatment exhibits favorable effects on prostate size and morphology relative to TEST therapy due to its selective modulation of the AR in this tissue. This study provides some evidence of a lack of interaction between TREN and either 5α-reductase or aromatase, at least in prostate tissue.
Conclusions
TREN treatment elicits favorable changes in body composition (specifically, in the reduction of visceral adiposity), lipid profile, insulin sensitivity, and myocardial tolerance to I-R in TEST-deficient male rats with the MetS. In comparison with TREN, the therapeutic effects of TEST were consistently less pronounced in these animals. Additionally, widespread replacement fibrosis in the myocardium and prostate hyperplasia were identified as consequences of TEST treatment and were markedly less pronounced in TREN-treated animals. Neither the androgen-deficient state nor the treatment with TEST or TREN affected the hepatic or renal activity in these animals. We propose that TRT may have contraindications in TEST-deficient males with obesity-related MetS, and that selective AR modulation with TREN may elicit a preferable benefit to risk ratio in these populations.
Acknowledgments
We thank the invaluable contributions of technical support offered by Melissa Leung and Rossana Nogueira for processing the study's high resolution histological and DXA images, respectively.
This work was supported by funding from the Menzies Research Institute Queensland, Griffith University, Australia.
Disclosure Summary: The authors have nothing to disclose.
Abbreviations
- AR
androgen receptor
- CFAS/L
calibrator for automated systems/lipids
- CK-MB
cardiac-specific creatine kinase
- CTRL
control standard rat chow
- CV
coefficient of variation
- DXA
dual-energy x-ray absorptiometry
- H & E
hematoxylin and eosin
- HF/HS
high fat/high sucrose
- HOMA-IR
homeostatic model assessment of insulin resistance
- I-R
ischemia reperfusion
- LV
left ventricular
- MetS
metabolic syndrome
- ORX
orchiectomy
- QC
quality control
- TD
TEST deficiency
- TEST
testosterone
- TREN
trenbolone
- TRT
TEST replacement therapy.
