C57BL/6 male mice at 12 weeks of age and weighing 22.5 ± 1.5 g received care according to the An-Najah National University ethical guidelines. All animal protocols were approved by the institutional animal care ethical committee (Ref: Med. Oct/2018/59).

Liver injury mice models were induced using carbon tetrachloride (CCl₄; Sigma, C-5331) introduced by i.p injections of 0.5 μL pure CCl₄/g body weight (one to nine dilution in corn oil) twice a week for 2and 4 weeks as acute and advanced chronic liver injury. During the liver injury duration, 2 weeks in the chronic model and 1 week in the acute model, mice were i.p injected with testosterone (Merck; T1500; purity ≥98%) in the concentration of 100 µg/mouse [4 mg/kg mouse body weight] twice a week for the remaining weeks. In all experiments, mice were sacrificed 2 days after the final CCl₄ injection. To this end, the animals were weighed and anesthetized with inhaled 5% isoflurane for 10 s before cervical dislocation.

The following mice groups were included: 1) Naive mice (mice untreated with neither CCl₄ nor testosterone), 2) mice group treated with testosterone only, 3) CCl₄-treated mice—acute liver injury mode (2-week injections), 4) CCl₄-treated mice—acute liver injury and treated with testosterone, 5) CCl₄-treated mice—chronic liver injury mode (4-week injections), and 6) CCl₄-treated mice—chronic liver injury and treated with testosterone. Each experimental group included six mice and the experiment was repeated three times (a total of 108 mice).

The posterior third of prostate and liver tissues were fixed in 3% formalin overnight and then embedded in paraffin in an automated tissue processor. Sections (7 μm) were stained with H&E to assess steatosis, area regions of necroinflammation, and apoptotic bodies, and with 0.1% Sirius red F3B in a saturated picric acid stain (Abcam, ab150681) to visualize connective tissue. A veterinary pathologist assessed all histopathological findings and reported assessments and the grade of the assessment.

Peripheral blood from the heart that was collected on the sacrifice day was centrifuged at 5,000 rpm for 15 min at 4°C to obtain the serum. Serum ALT (Abcam; ab285263), AST (Biocompare; MBS2019147), Fasting blood sugar (Biocompare; MBS7200879), C-peptide (Biocompare; MBS007738), cholesterol (Abcam; ab285242), and triglycerides (Biocompare; MBS726589) were determined using ELISA kits according to the manufacturer’s protocols.

RNA was obtained from liver tissue using trizol buffer (Bio-Lab; Cat# 90102331). Liver tissues were homogenized at RT, and 0.2 mL chloroform (Bio Lab; Cat# 03080521) was added. The samples were then incubated for 15 min at room temperature and centrifuged (1,400 rpm) for 15 min at 4°C. For RNA precipitation, the supernatant in each sample was transferred to a new micro-centrifuge tube, and 0.5 mL of isopropanol (Bio Lab; Cat# 16260521) was added, followed by 10 min incubation at 25°C. The tubes were then centrifuged (12,000 rpm) for 10 min at 4°C, the supernatants were removed, and 1 mL of 75% ethanol was added to the pellet, followed by centrifugation (7,500 rpm) for 5 min. The pellets were air-dried at room temperature for 15 min, 50 μL of DEPC was added, and the samples were heated for 10 min at 55°C. RNA purification from NK cells was assessed using RNeasy plus mini kit (CAT# 74034) according to the manufacturer’s guidelines. cDNA was obtained using a High-Capacity cDNA Isolation Kit (R&D; Cat# 1406197). RT-PCR reactions were performed using TaqMan Master Mix (Applied Biosystems; Cat# 4371130) to quantify αSMA and collagen III mRNA levels. Results were normalized to gapdh as a housekeeping gene and analyzed using QuantStudio™ 5 Real-Time PCR System.

Serum levels of testosterone and estradiol were assessed using abcam: ab285350 and Creative diagnostics: DEIA04927, respectively. Moreover, intracellular IL-6 and IFN-γ concentrations were assessed using Human IL-6 Quantikine ELISA Kit (R&D; D6050) and Human IFN-γ Quantikine ELISA Kit (R&D; 285-IF) according to the manufacturer’s protocols.

The livers were extracted and placed in Petri dishes containing 10 mL of DMEM medium (Biological Industries; Cat# 01–055-1A). The liver tissue was thoroughly dispersed using a stainless-steel mesh, and the cells were collected along with the medium and transferred to 50 mL tubes containing 10 mL of DMEM. Subsequently, the cells were cautiously moved to new tubes containing Ficoll (Abcam; Cat# AB18115269) and subjected to centrifugation at 1,600 rpm for 20 min at 20°C. The resulting supernatant from each tube was transferred to fresh tubes and centrifuged again at 1,600 rpm for 10 min at 4°C. Following the second centrifugation, the cell pellet in each tube was resuspended in 1 mL of DMEM to isolate and purify NK cells using the Stem Cells kit (Cat# 19665).

The trNK cells isolated from harvested mice livers were diluted to a concentration of 1 million cells per milliliter in a saline buffer supplemented with 1% bovine albumin (Biological Industries; Cat# 02–023-5A). Subsequently, the cells were labeled with the following antibodies. Anti-mouse NK1.1 (murine NK cell marker) (Biogems; Cat# 83712–70), anti-CD49a (MACS; Lot# 5150716246), anti CD49b (MACS; Lot# 5150716256), anti-mouse lysosomal-associated membrane protein-1 (CD107a; NK1.1 cells cytotoxicity marker, eBioscience, Cat# 48–1071), and anti-IL-6 R (R&D; Cat# 48–1044) were used. All antibodies were incubated for 40 min at 4°C. pHSCs (106 cells/mL) were stained with rabbit anti-mice αSMA (R&D; IC1420P). The cells were washed with 0.5 mL staining buffer and fixed with 20 mL 2% paraformaldehyde. All stained cells were analyzed with a flow cytometer (BD LSR Fortessa™, Becton Dickinson, Immunofluorimetry systems, Mountain View, CA).

Statistical differences were analyzed with a two-tailed unpaired Student’s t-test (for comparisons between two groups) or one-way or two-way analysis of variance (ANOVA; one-way ANOVA with the Newman‒Keuls post hoc test for comparisons among multiple groups) with GraphPad Prism 9.0 (GraphPad Software, La Jolla, CA). A t-test of p-value ≤0.05 was considered statistically significant and was calculated as the difference in means between two variables. A Mann-Whitney U test was performed to evaluate whether the mice metabolic panel elements (ALT, AST, Cholesterol, Triglyceride, and FBS levels) were altered following testosterone treatment in both the acute and chronic liver injury groups. The correlation coefficient r test and normality test [the Shapiro-Wilk test] were used (p-value ≤ 0.05 is considered statistically significant). The experiment was repeated three times, with each repetition consisting of 10 sample replicates. Results are presented as mean ± SD or as average means of experimental replicates ±SD.

Liver sections from mice with acute and chronic CCl₄-induced liver injury were evaluated for liver injury and phenotypic changes after treatment with testosterone. Representative images of H&E staining (Figure 1A) and Sirius Red staining (Figure 1B) of liver sections depicting acute and chronic liver injury are shown. The H&E staining of CCl₄-treated livers revealed centrilobular hepatocytes that were swollen, along with extensive necrotic areas containing a high number of infiltrating inflammatory cells (white arrows left) and steatosis (white arrows right), indicating the presence of a chronic CCl₄ model. However, in mice treated with testosterone, there was a delay in the appearance of these histological findings, and a significant reduction in both microvascular and macrovascular steatosis was observed specifically in the chronic model. Sirius red staining of livers from the CCl₄ mice exhibited increased collagen deposition in perisinusoidal areas in both the acute and chronic CCl₄ models (black arrows), but the impact of collagen depositions was more pronounced in the chronic model. Treatment with testosterone resulted in a remarkable reduction in the dense fibrous tissue of the stained area compared to the vehicle-treated mice. Figure 1C summarizes a detailed histology scoring system for H&E and fibrosis assessments (Scheuer, 1991; Ishak et al., 1995; Bedossa and Poynard, 1996; Wallace et al., 2015). Biochemical markers were also assessed in our mice groups. Serum inflammatory profiles of ALT (Figure 1D) and AST (Figure 1E) were elevated in both acute and chronic CCl₄ models in favor of the latter one. A significant amelioration in ALT levels of 2.1-folds and 3.6-folds were achieved in the acute and chronic CCl4 model, respectively, following testosterone treatment (p < 0.05). Similar effects of testosterone treatment were achieved only in the chronic CCl4 model of 2.6-folds while no effects were seen in the acute model. To confirm liver fibrosis in CCl₄-induced mice, fibrosis markers were quantified by assessing liver αSMA (Figure 1F) and collagen III (Col III) (Figure 1G) using RT-PCR. The data showed a significant increase in αSMA and Col III levels in both the acute and chronic CCl₄ models compared to the vehicle group, with a 1.2-fold increase in the acute model and a 4.2-fold increase in the chronic model for αSMA, and a 1.2-fold increase in the acute model and a 3.2-fold increase in the chronic model for Col III (p = 0.002). However, mice with liver fibrosis receiving testosterone treatment exhibited significant reductions in αSMA and Col III levels, with a 1.2-fold decrease and a 1.3-fold decrease, respectively, (p < 0.03) in the acute CCl₄ model, and a 2.2-fold decrease and a 2.3-fold decrease (p < 0.03) in the chronic CCl₄ model. The results obtained from both RT-PCR and histology assessments were comparable and clearly indicated an improvement in liver injury, inflammation, and fibrosis in liver sections following treatment with testosterone. To elucidate the effects of testosterone on alleviating histopathological findings of liver sections, we assessed serum testosterone and estradiol levels. Testosterone could be metabolized to estradiol through aromatization (Roncati et al., 2016). In male models, testosterone is the major source of plasma estradiol, the main biologically active estrogen, 20% of which is secreted by the testes. Plasma estrone, 5% of which is converted to plasma estradiol, originates from tissue aromatization, mainly adrenal and androstenedione (Vermeulen et al., 2002).

Supplementary Figure S1 displays serum testosterone (A) and estradiol (B) following testosterone treatment. Low serum testosterone levels were significantly obtained following CCl₄ inductions compared to untreated mice and were positively correlated with the severity of liver fibrosis. Testosterone treatment elevated serum testosterone levels and was comparable in all mice groups including the control group (untreated mice). In parallel, the same pattern of estradiol serum levels was achieved in untreated mice and showed reductions in their levels along severities of liver fibrosis. Testosterone treatment induced elevated estradiol levels and was positively correlated to liver fibrosis severity of chronic CCl₄ inductions (2.3-folds, p = 0.0001). Although estradiol levels increased following testosterone treatment, they remained within the normal range (Ström et al., 2012), highlighting the importance of testosterone in delaying liver fibrosis.

CCl₄ administration in C57BL/6J mice exacerbates high cholesterol levels and induces steatohepatitis changes in the liver. Previous research by Bassi et al. (2005) demonstrated that CCl₄ leads to increased hepatic lipid profiles of cholesterol, fatty acids, and triglycerides in both chronic and acute treatments in rats. Based on these findings, we utilized this model to examine the metabolic outcome markers of lipid and glucose profiles following treatment with testosterone. Our mice models exhibited perturbations in the metabolic profile of CCl₄-induced animals. In Figure 2, we observe elevated serum levels of cholesterol (Figure 2A), triglycerides (Figure 2B), C-peptide (Figure 2C), and fasting blood sugar (FBS) (Figure 2D) in both the acute and chronic treatments of CCl₄ mice, with a more significant impact in the chronic model. However, the liver injury mice treated with testosterone demonstrated lower serum levels of cholesterol, triglycerides, and C-peptide compared to the control groups receiving the vehicle, while also displaying a reduction in FBS levels (Figure 2D).

Altogether, the above data indicate that testosterone has an anti-fibrotic effect, most probably due to its effects in ameliorating lipid and glucose profiles, both of which are risk factors contributing to fibrogenesis. Thus, our results indicate testosterone is a potential target for delaying and inhibiting liver injury by improving insulin sensitivity. To further explore the mechanism behind the antifibrotic effects of testosterone, we further assessed the inflammatory and immune contribution in alleviating liver injury. We assessed serum IL-6 levels, the activity of isolated liver tissue-resident NK (trNK) cells, and the expression of IL-6 receptors on trNK cells. Testosterone exhibited immune-modulating properties, supported by in vitro evidence suggesting its potential to suppress the expression of proinflammatory cytokines such as TNFα, IL-1β, and IL-6, while enhancing the expression of the anti-inflammatory cytokine IL-10 (Arslan et al., 2016). Furthermore, testosterone displayed anti-inflammatory effects by significantly inhibiting adipose tissue formation and downregulating the expression of various adipocytokines, including leptin, TNF-α, IL-6, and IL-1, while positively correlating with adiponectin levels. Conversely, low testosterone levels were associated with increased expressions of inflammatory markers. Data presented in Figure 3 shows that both naive mice treated and untreated with testosterone had comparable low levels of serum IL-6 of 65 ± 10 pg/mL (p = ns). Serum IL6 showed increased levels within liver injury severities of 180 ± 24 pg/mL and 345 ± 52 pg/mL in the acute and chronic models, respectively (p = 0.002). Testosterone treatments exhibited a significant decrease of 2.4-fold (p = 0.0001) and 2.3-fold (p = 0.0003) in the acute and chronic models, respectively (p = 0.002). Testosterone has an anti-inflammatory effect due to the reduction of inflammatory cytokines (Bianchi, 2019).

The immune system plays a crucial role in the process of tissue healing. Consequently, managing the immune system is essential for effective planning of the healing process. In our study, we examined liver tissue-resident NK cells (trNK) isolated from the different mice groups. NK cells have been shown to possess antifibrotic properties by eliminating activated hepatic stellate cells (HSCs) (Muhanna et al., 2008). However, it is believed that their functionality may be impaired in cases of advanced liver injury (Amer et al., 2018; Salhab et al., 2020). Figure 4A shows an inverse correlation between trNK secretions of IFN-γ and liver injury severity of a 2-fold decrease (p = 0.003). Following testosterone treatment, trNK secretions of IFN-γ showed higher levels of 2.1 and 6.3 folds in the acute and chronic models, respectively. Regarding the correlation coefficient r test, Figure 4A’s Pearson’s correlation coefficient (−0.675) provides evidence for a large inverse strength of association between the variables.

The same data patterns were obtained using NK activation markers of CD107a (Figure 4B). Moreover, to further associate trNK activatory effects following testosterone treatment with IL-6 receptor, flow cytometry analysis was performed as indicated in Materials and Methods. Figure 4C shows a significant reduction of 1.6-fold and 2.5-fold in IL-6 receptor in the acute and chronic liver injury model, respectively, compared to mice receiving the vehicle (p = 0.0001).

Our results undoubtedly indicate the impact of testosterone as a potential therapy as they exhibit antifibrotic and anti-inflammatory effects by reducing αSMA mediated by increased NK activity and reductions in IL-6 receptor and could be of beneficial influence for patients with advanced liver injury.