Concentration-dependent effect of silymarin on concanavalin A-stimulated mouse spleen cells in vitro

Numerous phytochemicals produced by plants as secondary metabolites are potent immunomodulators and their biological activity was demonstrated in vitro on isolated cells and in vivo on animal models of various diseases. One natural compound with a very long history of use in traditional medicine is silymarin (SIL), isolated from the seeds of Silybum marianum L. (Asteraceae). It contains a mixture of flavonolignans of which silybin A/B, silychristin A, silydianin, isosilybin A, and 2,3-dehydrosilybin form approximately 70%–80% and other components are flavonoids, fatty acids, and some undefined molecules. SIL is subject to intensive research and has received great attention due to its wide range of pharmacological effects. SIL displays free radical scavenging and strong antioxidant abilities, which are associated with inducing superoxide dismutase activities and increasing cellular glutathione content (Kwon et al., 2013; Surai, 2015). A growing number of studies have shown that SIL and its main component, silybin A/B, exhibit hepatoprotective, anticarcinogenic, immunomodulatory, and antiangiogenic activities (Saller et al., 2001; Fraschini et al., 2002; Gažák et al., 2007; Esmaeil et al., 2017), suggesting multiple molecular mechanisms regulated by SIL's components. Anticancer activities involve suppression of the inflammatory process that leads to neoplastic transformation, hyperproliferation, and downregulation of progression of carcinogenesis and angiogenesis (Ramasamy and Agarwal, 2008). Reports on immunomodulatory activities of SIL and silybin indicate that stimulation or suppression of inflammatory reactions in animal models of diseases is dependent on the concentration in medium as well as administered dose in vivo. It has been shown that SIL can modulate human T-lymphocyte proliferation and secretion of cytokines. In vitro treatment of peripheral blood mononuclear cells from β-thalassemia patients with low SIL concentrations up to 20 μg/ml led to the restoration of glutathione levels and stimulated suppressed lymphocyte proliferation (Alidoost et al., 2006). On the contrary, in vitro treatment with higher doses (80 μM/ml) of SIL resulted in suppressed expression of T-cell activation-and exhaustion markers on CD4+ and CD8+ T cells from chronically infected HIV-positive subjects and also modulated functions of other human primary cells (Lovelace et al., 2017). SIL exerted significant inhibitory effect on proliferation of CD3-activated mouse spleen T-helper (CD4+) cells in vitro at a concentration of 50 μM/ml or higher (Gharagozloo et al., 2010). Studies showed that anti-inflammatory activities of SIL are dependent on the inhibition of translocation of transcription nuclear factor kappa B (NF-kB) from the cytoplasm into the nucleus (Manna et al., 1999; Gharagozloo et al., 2010) where it regulates transcription of many genes for inflammatory mediators and cytokines.

Activation of T lymphocytes during an immune response is mediated by various T-lymphocyte receptors and triggers a series of programmed gene regulations, proliferation, and effector functions. Concanavalin A (CoA), a lectin isolated from the seeds of Canavalia ensiformis (the Jack bean), is a very potent mitogen for lymphocytes and can induce rapid cell proliferation, preferentially of T lymphocytes, including both helper and suppressor subsets (Dwyer and Johnson, 1981). It was shown that lymphocytes possess about 10⁶ CoA receptors per cell on the cell membrane and bind to various glycosyl proteins and to a-d-mannose residues on glycolipids (Wang et al., 1971). CoA also stimulates the energy metabolism of T cells via modulation of cell respiration localized in mitochondria and the overall effects are mostly mediated by modulation of ATP turnover (Krauss et al., 1999). Research accumulated in the past decade has revealed that immune cells change their metabolic programs which are tightly regulated via mitochondrial dynamics, to support activation and differentiation into individual subsets (O’Neill et al., 2016). Mitochondria are central organelles of metabolism that provide energy during the differentiation and maintenance of immune cell phenotypes (Ozay et al., 2018). Moreover, alterations of mitochondrial membrane potential are important in apoptosis and may precede nuclear signs of apoptosis in several different cell types. Kulkarni et al. (1998) showed that CoA-induced apoptosis in human fibroblasts is associated with breakdown of mitochondrial membrane potential and loss of calcium homeostasis in mitochondria. Immunomodulatory activities of SIL have been evaluated in various cell models in vitro, but it is not clear whether SIL treatment of CoA-activated mouse spleen cells involves modulation of cytokines secretion and alterations in mitochondrial homeostasis and whether it can prevent cell apoptosis.

In this study, we aimed to investigate the concentration-dependent effect of SIL on proliferation of CoA-stimulated splenocytes, viability of cells, and secretion of IFN-g and IL-4 cytokines ex vivo in relation to gene expressions of transcription factors Foxp3 and NF-kB. In addition, metabolic changes in cells were evaluated by means of the mitochondrial membrane potential and apoptosis.

SIL is a natural product widely studied for its numerous beneficial effects on health and prevention of many diseases (Gažák et al., 2007). It is considered very safe as after oral administration the 50% lethal dose was 10 g/kg in rats (Fraschini et al., 2002). Studies on immunomodulatory activities indicate concentration-dependent effect at various in vitro and in vivo experimental conditions mostly reporting on suppression of inflammatory mediators (Esmaeil et al., 2017). In this study, we aimed to compare the effects of lower doses (5 and 10 μg/ml) and higher doses (20 and 40 μg/ml) of SIL preparation (Sigma) in which flavonolignans silybin A/B formed about 52%. We focused on spleen T lymphocytes isolated from healthy Balb/c mice following activation by plant mitogen CoA. We first investigated the dose-dependent effect on T-cell proliferation showing that low dose of 5 μg/ml can enhance proliferation and significant suppression was observed at a concentration of 40 μg/ml. Generally, upon priming of T cells with different stimuli, heterogeneous population is generated since naïve T cells have stem-like properties and can differentiate into virtually all different types of effector, memory, or regulatory cells. Higher sensitivity to suppressive effect of SIL was found when purified mouse spleen CD4+ T cells were used. Namdari et al. (2018) found that concentrations of 20 μg/ml of SIL and higher significantly reduced cell proliferation. By using CD4+ splenocytes from C57/Bl6 mice, proliferation assay revealed that SIL, at 50 μM concentration (approximately 25 μg/ml), significantly inhibited CD4+ cell proliferation (Gharagozloo et al., 2010). Suppressive effect of higher concentrations of SIL was not due to the direct induction of cell death, which was confirmed in our study by trypan blue test and in the study of Namdari et al. (2018) by propidium iodide staining. Moreover, we revealed that reduced proliferation was not due to decreased cell viability as the proportions of live/dead cells remained similar as in untreated cells.

Reports on concentration-dependent effect of SIL administration to mice present various findings. Johnson et al. (2003) showed that the absolute numbers of splenic CD3+ T lymphocytes were reduced after intraperitoneal administration of 10 and 50 mg/kg of SIL; however, PHA-induced T-lymphocyte proliferation ex vivo was increased in 10 mg/kg of body weight-treated group. Wilasrusmee et al. (2002) studied the standardized milk thistle extract containing SIL and other phytochemicals in murine lymphocyte proliferation tests using CoA as the mitogen. They showed that the extract profoundly increased lymphocyte proliferation at doses of 250 μg/ml and only moderately at lower doses. This effect of milk thistle was associated with an increase in IFN-γ, IL-4, and IL-10 cytokines. Since SIL was found to form 70%–80% in the whole extract, the stimulatory activity after oral administration could be comparable with the results of another study on mice in which the purified SIL was used (Karimi et al., 2018). The authors showed that treatment of healthy mice with low dose of SIL (50 mg/kg) for 14 consecutive days stimulated both cellular and humoral immune functions. It increased the proliferation of phytohemagglutinin-A-stimulated spleen T-cell proliferation and IFN-γ secretion ex vivo. Pharmacokinetic studies on tissues distribution of silybin, the major active constituent of SIL, have revealed that in mice orally fed with silybin at dose of 50 mg/kg the peak levels were observed at 0.5 h after administration, and concentrations in tissues were the following: 8.8 μg/g (liver), 4.3 μg/g (lung), 123 μg/g (stomach), and 5.8 μg/ml (pancreas) (Zhao and Agarwal, 1999). These low concentrations presented in these organs, except stomach, could support the stimulatory activities of SIL found in our in vitro study and in vivo studies mentioned earlier.

We further investigated the effect on cytokine secretion by CoA-stimulated cells and showed that only the highest concentration of SIL (40 μg/ml) leveled down the amount of IFN-γ and IL-4, probably due to diminished cell numbers as a consequence of reduced proliferation. The higher susceptibility to SIL was found for CD4+ T-helper cells activated with monoclonal anti-CD3/anti-CD28 antibodies, where SIL at concentrations of 20 and 30 μg/ml significantly decreased IFN-γ gene expression and upregulated Foxp3 transcription factor in Treg cells. Foxp3 factor plays a key role in the maintenance of lymphoid homeostasis in a number of immune circumstances, where Tregs guide immunosuppressive reactions (Feuerer et al., 2009).

In our study, SIL at the dose of 20 and 40 μg/ml stimulated expression of Foxp3 transcription factor in CoA-stimulated T lymphocytes, as well as mRNA levels for IFN-γ cytokine, suggesting that CoA probably induced a heterogeneous population of T cells with different transcriptional profiles and cytokine secretion. Stimulation of gene for NF-κB p65 subunit on mRNA level with increasing concentrations of SIL was also observed in our study. In the cytoplasm, NF-κB dimer is present in inactive form bound to inhibitory kappa B (IκB) protein and upon activation of cells the NF-κB is released from IκB and translocates to the nucleus. It was shown that SIL can block the translocation of NF-κB to nucleus via suppression of IκB degradation (Manna et al., 1999). We assume that although mRNA and probably also protein levels of NF-κB subunit were significantly elevated after 40 μg of SIL, simultaneous dose-dependent inhibition of IκB degradation occurred with increasing concentrations of SIL.

Interaction of CoA (concentration 3 μg/ml) with cell surface glycoproteins can induce apoptosis in fibroblasts and other nonimmune cells, and intrinsic pathway of apoptosis can be initiated by imbalance in the mitochondrial membrane potential Ym (Kulkarni et al., 1998). Mitochondrial control and guidance of cellular activities of T cells (Chao et al., 2017; O'Sullivan and Pearce, 2015) were recognized as a key factor in many diseases (Shinohara and Tsukimoto, 2018). In our study, cultivation of unstimulated splenocytes for 70 h decreased Ym and induced transition of naive splenocytes to early apoptosis (approximately 86%) and late apoptosis (approximately 11%). Activation with CoA stimulated Ym in T cells, which correlated with the decreased proportion of the early apoptotic cells (79%) indicating the metabolic reprogramming in activated cells. It was found that CoA bound to the cell membrane is internalized and accumulated primarily onto the mitochondria as early as 1 h posttreatment, and gradually increased mitochondrial membrane permeability. The increased mitochondrial membrane permeability would then lead to staining with Annexin V, the marker labeling the early stage, but no typical apoptosis was observed (Shinohara and Tsukimoto, 2018). We further showed that low concentrations of SIL (up to 10 μg/ml) positively modulated the mitochondrial dynamics in activated T cells which probably accounted for observed reduced proportions of the early apoptotic and late apoptotic/dead cells. Interestingly, higher concentrations (20 and 40 μg/ml) were not pro-apoptotic, in agreement with results of the study of Gharagozloo et al. (2010) on activated CD4+ T cells, supporting our hypothesis that antiproliferative effect of 40 μg/ml SIL (approximately 100 μM/ml) was not due to the cytotoxic effects on activated T cells. Detailed examination of the effect of SIL on proliferation revealed a significant G1 arrest in the cell cycle of CD3+ CD28+ activated T lymphocytes after 96 h of incubation with 100 μM SIL (approximately 50 μg/ml) without causing cell death (Gharagozloo et al., 2013). It is possible that the stimulated cells had an impaired respiratory capacity with a decreased mitochondrial membrane potential due to excess of produced reactive oxygen species. Addition of SIL to stimulated cells increased the mitochondrial membrane potential, suggesting that in CoA-stimulated cells the electron transfer, respiratory activity of the cells, and activity of the enzymes involved in oxidative phosphorylation have increased. Moreover, SIL is likely to prevent peroxidative damage of phospholipids, thereby preventing the initiation of apoptosis through cytochrome c release (Yuan, 2006; Robertson and Orrenius, 2000).

In conclusion, our study demonstrated that mouse spleen T lymphocytes activated by lectin CoA show different susceptibilities to low (£10 μg/ml) and higher (20 and 40μg/ml) SIL concentrations after 70 h of incubation in vitro. Treatment with low concentrations resulted in increased proliferation, cytokine secretion, and mitochondrial membrane potential and reduced transition of T cells to apoptosis. SIL at high concentration had the opposite effect without exerting significant cytotoxicity and upregulated genes for cytokines and transcription factors on mRNA level. It is possible that individual subpopulations of T cells induced by CoA were differentially affected by the various SIL concentrations and profound suppression of T lymphocyte functions was associated with the dose of 40 μg/ml. This correlated with the highest expression of Foxp3 factor indicating that this dose preferentially promoted differentiation to Tregs lymphocytes.