Cytokines influence the development and function of the adrenals, testes, and ovaries, and further modulate steroidogenesis [1, 2]. Inflammatory cytokines are elevated in the male reproductive tract during both acute and chronic inflammatory diseases, contributing to the development of male hypogonadism [3, 4, 5]. Several cytokines, including the pro-inflammatory cytokine tumor necrosis factor-alpha (TNF-$\alpha$), are secreted locally by leukocytes (such as macrophages and lymphocytes) in the male reproductive tract [3]. Sertoli cells and Leydig cells also synthesize and secrete various cytokines, including TNF-$\alpha$ [4].

Originally discovered as a protein that mediates tumor necrosis with a role in cancer cachexia [6], TNF-$\alpha$ is now considered a pleiotropic cytokine with a central role in acute and chronic inflammatory responses [7, 8]. In murine and human testes, TNF Receptor-1 (TNFR1) is the main TNF receptor expressed by Leydig cells, as well as in interstitial macrophages and lymphocytes. TNF Receptor-2 (TNFR2) has been identified in rat Sertoli cells and human peritubular cells [3]. High levels of TNF-$\alpha$ inhibit Leydig cell steroidogenesis, resulting in reduced testosterone, impaired spermatogenesis, and reduced steroidogenic enzyme gene expression [9, 10]. In healthy human males, acute administration of TNF-$\alpha$ transiently increases serum luteinizing hormone (LH) and decreases testosterone, with no change in follicle-stimulating hormone (FSH) and sex hormone binding globulin (SHBG) [11]. Yet, the mechanisms by which these changes occur remain unclear. However, the induction of oxidative stress and apoptosis may play a role as potential mechanisms for TNF-$\alpha$ induced toxic effects on Leydig cells [12].

Several clinically approved protein-based TNF-$\alpha$ antagonists are used in the treatment of inflammatory diseases, such as etanercept, infliximab, adalimumab, recombinant human TNF receptor, Fc-fusion protein, and golimumab [13]. Although effective, these drugs have a series of systemic side effects as they have been shown to increase the risk of lymphoma and infections, cause cardiovascular and neurologic problems, and induce auto-antibodies and lupus-like syndromes [14, 15, 16]. Therefore, some focus has been redirected to herbal extracts containing different chemical classes (polyphenols, alkaloids, steroids, terpenoids, fatty acids, macrolides) which can act as TNF-$\alpha$ antagonists or potentially protect against mechanisms of TNF-$\alpha$ induced cytotoxicity [13, 16, 17]. Natural extracts can surpass synthetic compounds as they are more likely to be safer, less toxic, and cost-effective [17]. In male hypogonadism associated with inflammation, natural products and plant extracts that are immune regulating and act as antioxidants may be useful in Leydig cell dysfunction [18].

Lessertia frutescens (Linnaeus (L.)) Goldblatt and Manning) (abbreviated as Lf in this manuscript), indigenous to Southern Africa, is traditionally considered an important medicinal herb [19, 20]. Common local names for Lf are Sutherlandia or Cancer bush (English), Kankerbos (Afrikaans), Umnwele (isiXhosa), Insiswa (isiZulu), and Phetola (Sotho) [19]. Numerous constituents of Lf extractions from leaves have been identified, including free amino acids (asparagine, proline, and L-arginine), non-protein amino acids (L-canavanine and $\gamma$-aminobutyric acid), pinitol, cycloartane glycosides (sutherlandiosides A, B, C and D), flavonol glycosides (sutherlandins A, B, C and D), hexadecanoic acid, $\gamma$-sitosterol. Furthermore, various polysaccharides make up to 35% of the dry weight of the aqueous extract [21, 22, 23, 24].

Lf is traditionally used for blood purification (bitter tonic), as an anti-stress remedy, and in the prevention and treatment of cancer [19, 25, 26]. Using in vitro and in vivo experimental models, extractions from Lf have been shown to be hypoglycaemic and antidiabetic [27, 28, 29, 30, 31, 32], immune modulating [31, 33, 34, 35, 36, 37], antioxidant [38, 39, 40], analgesic [31], anti-bacterial [38] and anti-human immunodeficiency virus [41]. Lf has also been classified as an adaptogen, promoting adaptability and resilience to environmental stressors [42] and promoting an anti-stress response [43, 44, 45]. Furthermore, Lf is antiproliferative and pro-apoptotic in cancer cell lines [21, 46, 47, 48, 49, 50, 51, 52, 53].

It is currently unknown what effect Lf has on Leydig cell function. It is further plausible that Lf may protect against inflammation-mediated Leydig cell dysfunction. However, no investigation has been done in this direction. As the impact of Lf on Leydig cell function and dysfunction remains unclear and poorly investigated, the aim of this study was to investigate the potential protective effect of Lf extraction on Leydig cells exposed to TNF-$\alpha$in vitro.

Since Lf is reported to have modulating effects in vivo and in vitro on immune function [31, 33, 34, 35, 36, 37], oxidative stress [38, 39, 40], steroid hormone synthesis [43, 44, 45], and apoptosis [21, 46, 47, 48, 49, 50, 51, 52, 53], we investigated the effect of aqueous LFE in protecting Leydig cells against the damage induced by TNF-$\alpha$. The results of this study found that TNF-$\alpha$ dose-dependently decreased cell viability across all concentrations tested. This is consistent with a previous study showing that TM3 Leydig cell viability was significantly reduced in vitro over 48 h at similar TNF-$\alpha$ concentrations of 0.1, 1, 10, and 100 ng/mL, although the effect on viability was much more dramatic at 10 and 100 ng/mL compared to the results of this study [54].

The effects of LFE on Leydig cell viability and cytotoxicity have not been previously reported. This study found that LFE increased cell viability at the lower concentrations but decreased viability at the highest one, with a corresponding increase in cytotoxicity. This suggests that the LFE may stimulate Leydig cells at lower concentrations but exerts a cytotoxic effect at higher concentrations used in this study. There are only a few studies on cell viability for LFE in vitro, all of which use higher LFE concentrations compared to this study. SH-SY5Y neuroblastoma cells treated with 20–60 µg Lf for 24 h showed increased cell viability, contrary to reduced viability at a much lower concentration of 100 ng/mL LFE in our study [60]. However, the extraction method is described only as 1 Kg of dried leaves in an unknown volume of boiling water before freeze-drying, and the dosage is provided as an absolute number (20–60 µg) only [60]. In cultured renal LLC-PK1 and MDBK epithelial cells, exposure to an aqueous extraction over 48 h resulted in no significant change in cell viability at concentrations of 0.3–6 mg/mL compared to the reduced viability for LFE at 100 ng/mL, although there was a significant reduction at 12–24 mg/mL [61]. The variation in differences in viability at different dosages across these studies is, however, not clear and may be a response to different cell lines used. This different response of cell lines is further reported, where exposure of normal human lymphocytes to 2.5 mg/mL of ethanolic and aqueous LFE showed an increase in cytotoxicity and decrease in cell viability after 24 h [62], and no effect was observed on LLC-PK1 and MDBK cells at similar dosages [61].

LFE cytotoxicity is more widely reported in cancer cell lines. In prostate cancer in vitro and in vivo, methanolic LFE inhibits cell proliferation [47], similar to what was observed for ethanolic LFE in breast carcinoma and leukemia cells in vitro [21]. MCF-7 breast adenocarcinoma cells have been shown to be more susceptible to cytotoxic effects of 1 mg/mL aqueous LFE (48 h exposure) than non-tumorigenic MCF-12A cells [46]. Similarly, at 1.5 mg/mL, antiproliferative effects and morphological signs of apoptosis were observed in MCF-7 cells after 24 h of exposure [48]. Furthermore, Lf has been shown to cause cytotoxicity in the Salmonella typhimurium strain TA97a, even if this was not evident in strains TA98, TA100, and TA102 [63]. Although our results showed reduced viability and increased toxicity at 100 ng/mL, hot water extracts of Lf leaves did not affect cell viability in Chinese Hamster Ovary (CHO) or Lung Carcinoma Epithelial (A549) cells exposed for 24 h at Lf concentrations up to 500 µg/mL, even if reduced cell viability in HepaRG cells was apparent at concentrations higher than 100 µg/mL [40]. However, it is difficult to compare the results of these different LFE and concentrations on different physiological and pathological cell lines that are currently available in the literature.

Our results suggest that LFE protects cells against TNF-$\alpha$ induced cytotoxicity at lower concentrations, but it may be cytotoxic as well at higher concentrations. Furthermore, the results of this study show that although cytotoxicity was increased for TNF-$\alpha$ concentration of 10 ng/mL, this was not observed for the lower concentrations of LFE when co-cultured with TNF-$\alpha$, thus supporting the protective effect of lower LFE concentrations on Leydig cell toxicity. In a study using 1-methyl-4-phenylpyridinium (MPP+) to reduce viability in SH-SY5Y neuroblastoma cells, pre-treatment with LFE (20 µg) prevented this decrease [60]. Similarly, LFE concentrations between 10 µg/mL and 500 µg/mL protected against reduced cell viability induced by 50 µM t-BHP over a period of 24 h in CHO, A549, and HepaRG human liver cells, which was not apparent for 1000 µg/mL LFE [40].

TNF-$\alpha$ is known to regulate apoptosis in many cell types through TNFR1 [9]. Following receptor binding, intracellular mediators of TNF-$\alpha$-induced apoptosis include activation of caspase-8, Bid (a Bcl-2 family protein), Bax-Bak, and cytochrome c, which results in a protein complex consisting of apoptotic protease activating factor-1, cytochrome c, and caspase-9, that then activates caspase-3 as a major execution caspase that initiates cell death [9]. With exposure to toxicants, the nuclear factor of activated T-cells promotes FasL transcription in Leydig cells [64]. Although numerous toxins may induce oxidative stress and apoptosis in Leydig cells in vitro, including di(2-ethylhexyl)phthalate [65], hydrogen peroxide [66, 67, 68, 69], triptolide [70], it is not known whether TNF-$\alpha$ can induce Leydig cell apoptosis. The results of this study found that TNF-$\alpha$ increases caspase 3/7 activation across all concentrations of TNF-$\alpha$ tested. In similar studies where Leydig cells were exposed to toxins, caspase-3 activation was required for ethane dimethanesulfonate-induced apoptosis [71]. Corticosterone also induced apoptosis in isolated rat Leydig cells [72], mediated through activation of Fas/FasL and caspase-3, alongside reduced mitochondrial membrane potential and increased reactive oxygens species (ROS) generation [73]. Di(2-ethylhexyl)phthalate induced oxidative stress and apoptosis in TM3 Leydig cells, thus increasing caspase-3 and -9 alongside Bax/Bcl2 ratio [65]. Further, hydrogen peroxide-induced oxidative stress in TM3 Leydig cells via increased ROS, Bax, and caspase-3 and reduced Bcl-2, PI3K, and p-AKT [68]. Similarly, activation of caspase-3 in Leydig cell apoptosis is reported in vitro for lead acetate [74] and alcohol [75].

Previously, it has been shown that high concentrations of LFE can be toxic. In cultured renal LLC-PK1 and MDBK epithelial cells, exposure for 48 h to an aqueous LFE significantly increased caspase-3 and -7 activity [61]. In human lymphocytes, 24 h exposure to 2.5 mg/mL of ethanolic and aqueous LFE each induced necrosis, reduced ATP, increased DNA fragmentation, and inhibited caspase 3/7 [62]. However, the apoptotic effect has been more widely reported in cancer cell lines. LFE (2.5 and 5 mg/mL) induced apoptosis in an oesophageal cancer cell line, with increased Annexin V/propidium iodide and caspase-3 and -7 [50]. In SiHa cervix carcinoma cells, a methanol LFE (50 µg/mL–200 µg/mL) reduced cell proliferation and increased caspase 3/7 activity after exposure for 4–24 h [76]. Ethanolic LFE extracts reduced proliferation and induced apoptosis in CaCo2 colon cancer cells, with down-regulation of PI3-K and Akt phosphorylation pathways and decreased forkhead transcription factor (FKHR) phosphorylation [53]. However, this increase in caspase 3/7 with LFE exposure is not always consistent. Ethanolic LFE (0.15–2.5 mg/mL) inducing apoptosis at higher dosages in various melanoma cell lines showed no activation of caspase 3/7, -8, and -9 over a period of 24 h, although this increased after 48 and 72 h of exposure [52]. However, the use of a pan-caspase inhibitor suggested that apoptosis may also occur independently of caspase activation with LFE exposure [52].

In our study, it appears that LFE is protective against TNF-$\alpha$ induced early apoptosis in Leydig cells. In SH-SY5Y neuroblastoma cells exposed to MPP+, pre-treatment with LFE reverted an increase in caspase 3/7 and -9 activity [60]. Additionally, different herbal extracts have been shown to improve toxin-induced apoptosis in Leydig cells, including taurine [65], astaxanthin [66], melatonin [77], resveratrol [67], rutin [68], quercetin [70] and extracts from Morinda officinalis [69].

The apoptotic effect of TNF-$\alpha$ on Leydig cells can also be mediated by intracellular ROS production, predominantly from the mitochondria. ROS may promote cytochrome-C release, leading to caspase-3 cleavage and apoptosis [9]. However, this was not observed in our study. Based on our results, it appears that the protective effect of LFE on Leydig cell cytotoxicity and early apoptosis may not be mediated through antioxidant inhibition of superoxide over a period of 24 h. LFE has, however, been shown to both reduce and increase cellular ROS in different experimental models. This includes protection against toxin-induced ROS production in CHO, A549, and HepaRG cells [40], primary rat cortical neurons and microglial cells [2], neutrophils [39], and macrophages [21, 37] and neuroblastoma cells [60]. However, ROS has been increased by LFE in neuroblastoma cells [51] and macrophages [36].

TNF-$\alpha$ has been suggested to negatively affect the hypothalamus-pituitary-testis (HPT) axis, where in vivo (including humans) administration is associated with a decrease in serum testosterone [1, 11]. However, although the results of this study found that TNF-$\alpha$ decreased testosterone across all concentrations, this was significant only for 0.1 ng/mL TNF-$\alpha$. Testosterone and progesterone synthesis were previously decreased in TM3 Leydig cells in a dose-dependent manner after incubating for 48 h with 0.1–100 ng/mL TNF-$\alpha$, suggesting a downregulation of steroidogenesis cascades [54]. However, this is not always consistent as it has also been reported that TNF-$\alpha$ increases testosterone concentrations in adult rat Leydig cell cultures over periods between 7 and 72 h [78]. Although the mechanism for the downregulation of testosterone by TNF-$\alpha$ appears to be through the downregulation of steroidogenic enzymes, it still remains unclear. In rats, intratesticular injection of TNF-$\alpha$ inhibited basal or hCG-stimulated steroidogenesis, mediated through a downregulation of steroidogenic acute regulatory (StAR) protein [79]. This mechanism is supported in cultured porcine Leydig cells, where TNF-$\alpha$ 20 ng/mL over a period of 72 h inhibited LH/hCG-steroidogenesis through decreased StAR [80]. The same concentration tested in this study for co-exposure with LFE concentrations (TNF-$\alpha$ 10 ng/mL) was previously shown to upregulate Leydig cell IL-6 production and nitric oxide production, upregulate COX-2 mRNA expression, and downregulate the expression of Sirt1 mRNA (Sirt1 is known to promote steroidogenesis through upregulation of SF-1) [56]. Similarly, in R2C Leydig cells in vitro, this same concentration of 10 ng/mL TNF-$\alpha$ over 12 h downregulated steroidogenesis through activation of NF-k$\beta$ and subsequent inhibition of Nur77 and SF-1 [58]. In mouse Leydig cells, TNF-$\alpha$ at concentrations of 0.1 and 10 ng/mL inhibited cytochrome P450 side chain cleavage (CYP11A1) and 17$\alpha$-hydroxylase/17,20 lyase (CYP17A1) mRNA [57] and downregulated 3$\beta$-hydroxysteroid dehydrogenase (3$\beta$-HSD) expression [81]. LC-540 Leydig cells treated with TNF-$\alpha$ (10 ng/mL) for 12 h, 24 h, and 48 h showed suppressed histone H3 acetylation and methylation, with increased total histone deacetylases activity and upregulation of HDAC7 [82]. In a similar model, TNF-$\alpha$ upregulated DAX-1 mRNA and protein through the JNK/ERK MAP kinase pathway [83].

There are no previous studies in the literature investigating the effects of LFE on testosterone synthesis in Leydig cells. However, our results showed an increase in testosterone at high concentrations of aqueous LFE (4 mg/mL) orally administered in healthy Wistar rats significantly increased corticosterone, with no effect on testosterone after 28 days [44]. Studies in adrenal steroidogenesis, however, suggested mechanisms for modulating corticosterone and potentially testosterone in these cells. LFE and sutherlandioside B (10 and 30 µM) inhibited steroidogenic enzymes and modulated adrenal hormone biosynthesis in COS-1 and adrenal H295R cells [45]. LFE reduced total steroid production, including cortisol, deoxycortisol, and mineralocorticoid, and inhibited progesterone conversion by CYP17A1 and cytochrome P450 21$\alpha$-hydroxylase (CYP21A2). Sutherlandioside B decreased cortisol and androgen precursors, inhibiting CYP17A1 and 3$\beta$-HSD but not CYP21A2 [45]. In ovine adrenocortical microsomes and mitochondria, aqueous extracts of LFE inhibited substrate binding to cytochrome P450 21-hydroxylase (CYP21) and cytochrome P450 11$\beta$-hydroxylase (CYP11B1), limiting the conversion of progesterone and pregnenolone [43]. Furthermore, chloroform and methanol LFE extractions showed inhibitory effects on deoxycorticosterone binding to CYP11B1 and on progesterone binding to CYP21 [43]. However, it is unclear how LFE may have increased testosterone production in the TM3 Leydig cells. We could speculate that this effect may be due to the short (24 h) exposure in the experiments and other unknown effects on testosterone metabolizing enzymes, including 5$\alpha$-reductase and aromatase.

According to our results, LFE may protect against a decline in testosterone at higher concentrations. However, there are very few studies in the literature investigating the effect of LFE on testosterone production. There was no change in testosterone concentration in rats undergoing stress immobilization, although there was an increase in corticosterone in the stressed rats. The administration of aqueous LFE (4 mg/mL) had no significant effect on testosterone concentrations but did reduce corticosterone levels significantly [44]. Although animal studies suggest LFE may inhibit progesterone synthesis from pregnenolone through down-regulated CYP17A1 in adrenal cells [45], the effects on testosterone synthesis in Leydig cells may not be affected. Further investigations into the effect of LFE on testosterone synthesis within the broader context of steroidogenesis in different cell types are required.

Due to the exploratory design of the study, there was a limitation in the number of variables investigated. Only one concentration of TNF-$\alpha$ was selected for co-exposure experiments with LFE, and the experiments were limited to 24 h. The interpretation of the results within the existing literature should be done with caution, as there are no similar studies available. Furthermore, the current literature on both TNF-$\alpha$ and LFE is relatively poor and varies widely in experimental models, study design, and outcomes. There is also a lack of information about which specific compounds from LFE might play a more critical role in the regulation of Leydig cell function. Several phytochemicals have been identified in LFE, including amino acids (alanine, asparagine, proline, gamma amino butyric acid), flavonoids, pinitol, and triterpenoids, whose anticancer, antiproliferative, antidiabetic, and antioxidant activities have been demonstrated in different models [84]. Despite the above-mentioned limitations, our study represents the first attempt to investigate the LFE potential of protecting against inflammation-mediated Leydig cell dysfunction, which represents one of the possible contributing factors to male infertility. In fact, considering the physiological role of the androgens, Leydig cell dysfunction may result in inadequate masculinization and spermatogenic failure [85]. This is also related to older men, where increased inflammation and androgen decline are observed [86]. Hence, it is important to investigate the possible use of natural extracts, which are more likely to be safer, less toxic, and cost-effective than synthetic compounds, through their action as immune regulatators and antioxidants. In the future, it would be interesting to further investigate our findings through comparison with an additional treatment group consisting of TNF-$\alpha$ inhibitors, which are currently used in clinical practice, and perform immunocytochemistry experiments to further investigate Leydig cell-specific markers and observe their changes under experimental conditions.

This study reports that TNF-$\alpha$ decreases Leydig cell viability and increases cytotoxicity with early apoptosis and downregulation of testosterone. LFE protected against TNF-$\alpha$-induced cytotoxicity and apoptosis, except at the highest experimental concentrations, where it was cytotoxic. Although TNF-$\alpha$ did not significantly decrease testosterone in most concentrations, co-culture with LFE significantly increased testosterone at higher concentrations. These effects were not mediated through a change in intracellular superoxide. Therefore, our results suggest that aqueous LFE may be useful in the protection of inflammation-induced Leydig cell dysfunction demonstrated by increased cytotoxicity and apoptosis. Further studies are recommended to understand the most optimal LFE concentration and the role of specific LFE isolates that may modulate Leydig cell function.

Lf, Lessertia frutescens; LFE, Lessertia frutescens extract; TNF-$\alpha$, tumour necrosis factor-alpha; hCG, human chorionic gonadotrophin.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

All authors meet the requirements of ICMJE for authorship. KL designed the research study. KL, RH, RF, and CZB performed the research. RF and CZB analyzed the data. KL, RF, and CZB wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

As an in vitro study, ethical clearance was not required and informed consent is not applicable.

Mr. Lilburn Cyster (Department of Biodiversity and Conservation Biology, University of the Western Cape) to help during the freeze-drying preparation of Lessertia frutescens extract. Dr James Campbell (School of Natural Medicine, University of the Western Cape, Bellville, South Africa) for donation of the Lessertia frutescens obtained from Parceval Pharmaceuticals (Wellington, South Africa) for research purposes. Mr. Frans Weitz (Curator of the Herbarium at the University of the Western Cape) for authentication of Lessertia frutescens.

This study was financially supported by the National Research Foundation (NRF) in Pretoria, South Africa (TTK14051367261).

The authors declare no conflict of interest. The National Research Foundation (NRF) did not influence the collection, analysis, and interpretation of data. Opinions expressed and conclusions arrived at are those of the authors and are not necessarily to be attributed to the NRF.