High-dose testosterone supplementation disturbs liver pro-oxidant/antioxidant balance and function in adolescent male Wistar rats undergoing moderate-intensity endurance training

Introduction

Endurance training (EndTr) plays a key role in various endurance sports, e.g., cross-country skiing, long-distance running, and cycling. However, it can also be a vital addition to both bodybuilding (M. Barroso. 8 tips for balancing bodybuilding and endurance training; for the URL, see Supplemental File) and strength training, particularly in team sports (Coffey & Hawley, 2017). It also gets growing attention from the general public as a health-promoting factor when used in moderation (Fikenzer et al., 2018; Ruegsegger & Booth, 2018); the positive effects of the training also concern liver health (Shephard & Johnson, 2015).

Some adolescents and young adults, mostly males (Kanayama & Pope Jr, 2018; Johnston et al., 2019), combine endurance-oriented physical training with the use of testosterone and/or its synthetic derivatives termed anabolic-androgenic steroids (AAS). The main goal of this tactic is a faster reduction of body fat and boosting muscularity with the idea of improving physical performance and appearance (Hartgens & Kuipers, 2004; Kanayama & Pope Jr, 2018), and hence self-esteem. A positive correlation of the bodily effects with dosage inspires AAS abuse at high doses. Some endurance athletes reported the effectiveness of AAS for accelerating the recovery after intense physical exercises (Hartgens & Kuipers, 2004). AAS were also reported to improve running endurance in male rats (Van Zyl, Noakes & Lambert, 1995; Georgieva & Boyadjiev, 2004), but no positive AAS effect on endurance or blood serum markers recovery was seen in other rat studies (Delgado, Saborido & Megias, 2010) or healthy men (Baume et al., 2006).

Xenobiotics and excess endobiotics are mostly processed and removed by the liver; this situation renders this organ the critical site of steroid toxicity (Russmann, Kullak-Ublick & Grattagliano, 2009; Vinken et al., 2013). The abuse of AAS and their ‘prototypic’ testosterone is associated with a variety of severe adverse health effects (Hartgens & Kuipers, 2004; Van Amsterdam, Opperhuizen & Hartgens, 2010; Vanberg & Atar, 2010), including a number of those directly linked to liver injury (Russmann, Kullak-Ublick & Grattagliano, 2009; Bond, Llewellyn & Van Mol, 2016; Solimini et al., 2017). A growing body of data links this damage to a variety of genomic and nongenomic actions of these drugs, including enhanced generation of reactive oxygen species and lipid peroxidation, and the related activation of cellular stress-signaling pathways (Russmann, Kullak-Ublick & Grattagliano, 2009; Cerretani et al., 2013; Bond, Llewellyn & Van Mol, 2016). Disruption of redox homeostasis is a well-established event in both drug hepatotoxicity and various liver diseases (Cichoż-Lach & Michalak, 2014; Li et al., 2015; Arauz, Ramos-Tovar & Muriel, 2016). The primary sources of reactive oxygen species, the noxious mediators of oxidative stress, are cytochrome P450 enzymes of the endoplasmic reticulum and mitochondria (Guengerich, 2008; Jones, 2008) and several enzymes present in peroxisomes that abound in the liver (Cerretani et al., 2013; Cichoż-Lach & Michalak, 2014).

Hepatocytes carry several systems capable of preventing or limiting the adverse effects of enhanced oxidative stress (Jones, 2008; Frankenfeld et al., 2014; Sies, 2015). These systems include various antioxidant enzymes, e.g., glutathione peroxidase (GPx), glutathione reductase (GR), superoxide dismutase (SOD) and catalase (CAT), and low molecular weight antioxidants, e.g., tocopherols and reduced glutathione (GSH) that act mostly in lipophilic and hydrophilic milieus, respectively. Data on the effects of AAS abuse combined with physical exercise training on the liver antioxidant barrier in adolescent subjects are scarce (Molano et al., 1999; Pey et al., 2003). Adolescence is a period when some people, mainly boys, become familiar with AAS as drugs of abuse. It is also a period of essential changes in androgen catabolism in men (Horst, Bartsch & Dirksen-Thiedens, 1977; Belgorosky & Rivarola, 1987a; Belgorosky & Rivarola, 1987b; Stárka, Pospísilová & Hill, 2009), which condition may significantly modify the outcome of androgen action (Mantovani & Fucic, 2014).

Our earlier study on the effects of long-term testosterone treatment on the liver antioxidant barrier and some blood markers of liver injury in sedentary adolescent male rats showed some signs of enhanced liver oxidative stress and toxicity but no potential lasting harm (Sadowska-Krępa et al., 2017). Here, we tested the possible harmful effects of this treatment in a situation aimed to model androgen abuse to aid physical exercise training. The studied constituents of hepatic antioxidant defense systems included selected antioxidant enzymes (SOD, CAT, GPx, and GR) and nonenzymatic low molecular weight antioxidants (α- and γ-tocopherols and GSH); liver thiobarbituric acid-reactive substances (TBARS) content was assayed as an index of oxidative stress. Our choice of testosterone enanthate (TE) was mainly based on the fact that testosterone was for many years among the most frequently abused doping drugs (Handelsman, 2006), and testosterone formulations (chiefly esters) are a cornerstone of most oral and injectable AAS stacking regimens (for the respective URLs see Supplemental File). However, we also intended to avoid hepatotoxicity related to chemical alteration (mostly 17α-alkylation) of the sterane core in most AAS (Hartgens & Kuipers, 2004; Russmann, Kullak-Ublick & Grattagliano, 2009; Büttner & Thieme, 2010).

Materials & Methods

Results

Discussion

The main finding of this work is that testosterone supplementation enhanced hepatic stress and adversely altered liver pro-oxidant/antioxidant balance in EndTr adolescent male rats. These effects were evidenced by the increased occurrence of elevated blood GGT, AST and ALT activities, worsened blood lipid profile, reduced liver SOD and CAT activities and GSH content, and elevated liver TBARS content. The higher TE dose significantly lowered BW gain and LW as compared to those in TE-untreated EndTr rats, but caused no substantial shift in the LW/BW ratio (−5%, p = 0.66) and thus no apparent hepatotoxicity in EndTr rats. Earlier, we found a significantly reduced LW/BW ratio in high-dose TE-treated adolescent UTr rats (Sadowska-Krępa et al., 2017), which may have been due to the start of TE treatment at a younger age.

The training was performed at an intensity that allows maximal fat oxidation (Purdom et al., 2018), and the mean BW gains were less in all three EndTr groups than in the respective UTr groups, see Sadowska-Krępa et al. (2017). Notably, mean BW gain was identical for the TE-untreated and low-dose TE-treated groups in both the EndTr and the Utr rat cohort, indicating no interference of this treatment with food intake and somatic growth. The lowered BW gain in the EndTr+TE80 rats may also have resulted from reduced appetite and food intake. Such effects causing inadequate compensation of energy expenditure were commonly found in treadmill-trained male rats (Harpur, 1980; Molano et al., 1999; Foletto et al., 2015). However, reduced food intake was seen as well in long-term stanozolol-treated sedentary male rats (Yu-Yahiro et al., 1989). Another cause of reduced weight gain could be androgen-induced loss of adipose tissue (mostly visceral) and liver fat through various mechanisms (Yu-Yahiro et al., 1989; De Pergola, 2000; Hoyos et al., 2012).

Elevated serum AST, ALT, ALP, and GGT are well-known markers of AAS hepatotoxicity (Urhausen, Torsten & Wilfried, 2003; Ozer et al., 2008; Singh, Bhat & Sharma, 2011). However, serum ALT and AST may also come from injured skeletal muscles (Ozer et al., 2008). Exercise training alone can elevate serum ALT and AST, but not GGT, in both men (Pettersson et al., 2008; Romagnoli et al., 2014) and male rats (Pey et al., 2003; Chang et al., 2013). We found no significant difference between our TE-untreated EndTr rats and matching UTr rats in any of the four markers but AST activity that was higher in the former (p = 0.024, Student’s t-test); for the UTr rats data, see Sadowska-Krępa et al. (2017). In the present study, raised serum AST, ALT, and GGT but not ALP activities were found in a sizable subset of the EndTr+TE8 rats and most of the EndTr+TE80 rats. However, the mean AST and ALT activities in the latter were only higher by 1/3 than those in the TE-untreated UTr rats (Sadowska-Krępa et al., 2017). Even smaller and nonsignificant were the relative increases in the mean AST (+6%) and ALT activity (+23%) in the EndTr+TE80 rats as compared to those in their TE-untreated counterparts, while the respective increase in the mean GGT activity was fairly robust (+57%). Our data on the effects of the lower TE dose are in line with those reported by others. Namely, no sizable increase in serum AST, ALT, ALP, or GGT was found in adolescent male rats given 12 weeks of moderate-intensity EndTr and five intragastric doses weekly of 2 mg/kg_BW of fluoxymesterone, methylandrostanolone, or stanozolol during the last eight weeks (Molano et al., 1999; Pey et al., 2003). Notably, these AAS are considerably more hepatotoxic than testosterone (Hartgens & Kuipers, 2004; Russmann, Kullak-Ublick & Grattagliano, 2009; Büttner & Thieme, 2010).

We have earlier found significant increases in the mean serum level of AST, ALT, and ALP, but not GGT, in 5-week-old UTr rats given 6-week high-dose TE-treatment (Sadowska-Krępa et al., 2017). In the present study, 60% of the TE-untreated EndTr rats showed elevated serum AST levels indicating an exercise-related leakage of the enzyme from skeletal muscles. Unexpectedly, there was no significant TE-treatment-related increase in mean AST and ALT levels, while the high-dose TE treatment-related increase in mean ALP level was nearly identical to that in the UTr rats. However, contrary to the latter, the EndTr+TE80 rats showed a much higher mean GGT level than their TE-untreated counterparts. Serum ALT and ALP levels correlated positively with TE dosage in both the EndTr (see the Results section) and the UTr study cohort (R_S = 0.70, p < 10⁻³, and R_S = 0.50, p < 10⁻², respectively, N = 37 for both, see Supplemental Files, UTr_cohort_ser_enz_vs_TE_dose.xlsx). In contrast, weekly TE dose positively correlated with GGT but not AST level in the EndTr rats (this study), while the reverse was right in the UTr rats (R_S = 0.06, p = 0.72, N = 34, and R_S = 0.74, p < 10⁻³, N = 37, respectively, see Supplemental Files, UTr_cohort_ser_enz_vs_TE_dose.xlsx). This disparity may relate to the slightly different mean age of the two rat cohorts at the start of TE treatment, or a training-related change in resistance of the circulating enzymes’ sources, or both.

Being fairly ubiquitous (Bataller-Sifre, Guiral-Olivan & Bataller-Alberola, 2011), GGT alone has insufficient specificity as a serum marker of liver injury. In humans, it is usually raised in liver pathologies involving cholestasis and jaundice, including those caused by AAS (Urhausen, Torsten & Wilfried, 2003; Lumia & McGinnis, 2010). In rats, it is supposedly a better cholestasis marker than serum ALP but less reliable than in other species (Ozer et al., 2008). We found elevated serum ALP activity in only two out of 12 EndTr+TE80 rats, but only one of them showed moderately elevated serum GGT activity. Notably, testosterone or its esters rarely cause adverse hepatobiliary effects, except possibly in aging (Nucci et al., 2017).

Cell membrane-bound GGT is a vital part of hepatic antioxidant lines that sustain cellular GSH and cysteine homeostasis (Zhang, Forman & Choi, 2005). Its link with serum GGT is not known. Serum GGT was proposed as an aggregate marker of organismal oxidative stress caused by various diseases and environmental chemicals (Lee & Jacobs Jr, 2009; Chang et al., 2013). The TE treatment-related differences in serum GGT and the occurrence of excessive serum GGT levels were much larger in the EndTr cohort than those found earlier (Sadowska-Krępa et al., 2017) in the UTr cohort. However, there was no significant association between the levels of serum GGT and serum ALP, ALT, or AST. Hence, a significant part of the rise in serum GGT activity might not be due to liver stress or damage. Instead, it may indicate increased oxidative stress occurring in multiple organs.

Mean liver TBARS content was only slightly higher in the TE-untreated EndTr rats than in their UTr counterparts (Sadowska-Krępa et al., 2017); this could be related to the constitutively high liver metabolic activity and the corresponding oxidative stress. However, the relative differences in liver TBARS content between the EndTr+TE8 and EndTr+TE80 rats and their TE-untreated counterparts were modest (+15 and +48%, respectively) compared to those in their sedentary counterparts (+56 and +78%, respectively, see Sadowska-Krępa et al., 2017). Hence, the absolute mean TBARS levels were nearly equal in the corresponding TE-treated UTr and EndTr rat groups. Though the present data confirm the pro-oxidant action of supraphysiological TE doses, they also suggest an attenuation of the added oxidative stress in the EndTr rats.

Decreased activities of liver CAT and SOD in high-dose TE-treated rats and negative correlations of these enzyme activities with weekly TE doses proved a harmful action of massive testosterone supplementation on liver antioxidant enzymes in adolescent EndTr rats. Even more significant relative declines, including those in GPx and GR activities, were found in the heart of EndTr male adolescent rats given testosterone propionate doses that produced much higher serum TT levels. Their associated relative increases in left heart ventricle TBARS content exceeded these in the liver. However, the maximum absolute mean TBARS contents found in the two organs were nearly identical to the respective maximum found in the soleus muscle and far above that in the extensor digitorum longus (Sadowska-Krepa et al., 2013). This similarity may be related to the fact that the soleus muscle, the myocardium, and the liver, but not the extensor digitorum longus, rely almost entirely on oxidative metabolism in exercise.

While we found decreased hepatic GSH content in our TE-treated EndTr rats, stanozolol treatment was reported to increase hepatic GSH content in endurance-trained rats (Pey et al., 2003). The cause of the difference may be distinct pharmacological profiles of stanozolol and testosterone (Fernández et al., 1994; Hartgens & Kuipers, 2004; Russmann, Kullak-Ublick & Grattagliano, 2009; Büttner & Thieme, 2010), but also younger age and thus different reactivity of our rats at the beginning of androgen treatment. The present data contrast also with those from adolescent UTr rats, in which the same TE doses significantly elevated liver SOD activity (Sadowska-Krępa et al., 2017). It is likely that long-term TE treatment of adolescent male EndTr rats perturbed their development and/or boosted oxidative stress beyond the capacity of hepatic defenses. Of note, mean liver activities of SOD, CAT and GPx were higher (p ≤ 0.035, one-tailed t-test) in the TE-untreated EndTr rats than in their UTr counterparts (Sadowska-Krępa et al., 2017), and a similar tendency was found for GR (p = 0.093, one-tailed t-test). These findings implied that the training alone boosted the antioxidant defenses. A similar effect on liver SOD but not CAT and GPx activities was found in male rats of similar age, which were given 16-week EndTr of similar intensity (Song, Igawa & Horii, 1996). In our study, the positive effect of the training on liver SOD and CAT, but not GR and GPx activities, was dose-dependently reduced by concurrent TE treatment. This decline is consistent with enhanced liver oxidative stress in adolescent UTr rats (Sadowska-Krępa et al., 2017) and prompt deactivation of CAT and SOD by oxygen free radicals (Salo et al., 1990; Escobar, Rubio & Lassi, 1996).

GSH is the most abundant antioxidant and the key scavenger of reactive oxygen and nitrogen species in the liver (Russmann, Kullak-Ublick & Grattagliano, 2009; Li et al., 2015). In our studies, its content did not differ between the TE-untreated EndTr and UTr rats. However, in contrast to what we found in the UTr rats study (Sadowska-Krępa et al., 2017), TE treatment suppressed hepatic GSH content in EndTr rats, suggesting increased GSH use. GSH is also vital for sustaining reduced forms of some exogenous antioxidants, including the essential lipophilic antioxidant vitamin E (Scholz et al., 1989). Neither hepatic α-tocopherol nor γ-tocopherol pool was significantly affected by TE treatment in EndTr rats, showing the efficacy of GSH in maintaining appropriate levels of these protectants against free-radical mediated liver damage (Leo, Rosman & Lieber, 1993). Interestingly, liver TBARS content negatively correlated with hepatic α-tocopherol (R_S =  − 0.48, p < 0.01, N = 31) but not γ-tocopherol content (R_S =  − 0.30, p = 0.11, N = 31) within the entire EndTr cohort. This link may reflect the use of α-tocopherol for scavenging reactive oxygen species formed, e.g., due to β-oxidation of fatty acids, and hence for breaking chain propagation and maintaining an adequate redox balance (Niki, 2014).

The EndTr-related shifts in blood lipid profile found in this study are usually linked to reduced atherogenicity in both humans (Shephard & Johnson, 2015; Ruegsegger & Booth, 2018) and rats (Burneiko et al., 2006; Kazeminasab et al., 2017). The changes in blood lipid profile caused by concurrent TE treatment evidenced a reversal of the beneficial effects of the training and worsening of some training-unaffected characteristics except for the tendency of TG level and TG/HDL ratio to drop with the high TE dosage. The latter was likely the result of long-term action of supraphysiological serum TT level on male adipose tissue metabolism and the ensuing decreased body and liver fat contents and availability.

Conclusions

The present results show that long-term systemic high-dose testosterone treatment harms liver antioxidant defense systems and function in adolescent male rats undergoing endurance training. Namely, it abolishes or markedly attenuates most studied metabolic benefits from the training, causing a negative shift in liver pro-oxidative/antioxidative balance evidenced by reduced SOD and CAT activities, raised hepatic TBARS level, and elevated serum GGT activity. These changes suggest increased oxidative stress that likely occurs in other organs as well and may increase morbidity later in life. The same may occur in human male adolescents using massive testosterone or other AAS supplementation as a shortcut to improved sports performance and/or a more muscular physical appearance.

Supplemental Information