Introduction

Parkinson’s disease (PD) is a multifaceted neurodegenerative disease with a variety of symptoms including motor difficulties such as tremors, bradykinesias, rigidity and postural instability [1]. Recent interventions have been targeted to abrogate several pathological characteristics of PD, including the accumulation of neurotoxic aggregates of α-synuclein and associated neuroinflammatory degeneration of dopaminergic neurons in the substantia nigra [1,2,3]. Interestingly, clinical observations suggest that circulating plasma levels of sex hormones, including estrogen, may affect the incidence, progression, and intervention response for PD [4, 5]. For instance, not only is the prevalence of PD higher in males but the risk of developing PD post-menopause is significantly higher in females [4, 5], and therefore this presents an opportunity to explore estrogen modulators as potential therapeutics for PD.

Phytoestrogens, especially soy derived isoflavones including genistein (GEN) and daidzein (DAI) have been extensively studied for diverse health promoting effects via estrogen dependent and independent mechanisms [6]. Importantly, several pre-clinical studies [7,8,9] and some human clinical trials [10,11,12] have described a neuroprotective effect of isoflavones with potential therapeutic applications against neurodegenerative pathologies, but, their mechanisms of action are not well understood. Moreover, several studies have indicated that the microbial biotransformation of isoflavones in the gut has a significant effect on their clinical therapeutic efficacy including neuroprotection. For example, it has been shown that equol (EQL), a gut microbial metabolite of DAI, has more potent anti-neuroinflammatory effects in the LPS-stimulated murine microglial cells than either GEN or DAI [13]. Additionally, EQL was also more effective in reducing LPS-induced cytotoxicity in neuroblastoma N2a cells and increasing nerve growth in astrocytes C6 cells [13]. However, to date, the neuroprotective effects of EQL against PD have not been reported. In a previously reported study by our group evaluating the blood–brain barrier (BBB) permeability and anti-neuroinflammatory effects of several polyphenol microbial metabolites, EQL exhibited high permeability (for both the gut and BBB) and reduced pro-inflammatory cytokines in murine microglial BV2 cells [14]. Herein, we aimed to evaluate the comparative neuroprotective effects of EQL, DAI and GEN in murine microglial BV2 cells and human and neuroblastoma SH-SY5Y cells. In addition, EQL’s neuroprotective effects against PD in an in vivo Caenorhabditis elegans model are reported.

Materials and Methods

Chemicals

Dimethylsulfoxide (DMSO), lipopolysaccharide (LPS), 17β-estradiol (EST), daidzein (DAI), and genistein (GEN), 6-hydroxydopamine (6-OHDA), and 1-methyl-4-phenylpyridinium (MPP+) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). S-equol (EQL) was purchased from Chembest (Shanghai, China). DMEM/F-12, phenol red-free DMEM medium and trypsin-versene were purchased from Life Technologies (Grand Island, NY, USA).

Cell Culture

Murine microglia (BV2 cells) were generously gifted by Dr. Grace Y. Sun (University of Missouri at Columbia, MO, USA). Human neuroblastoma (SH-SY5Y cells) were obtained from American Type Culture Collection (ATCC, VA, USA). Cells were maintained in DMEM/F-12 high glucose (4.5 g/L) and supplemented with 10% inactivated FBS and 1% P/S (100 U/mL penicillin, 100 mg/mL streptomycin; Life Technologies, Gaithersburg, MD, USA) at 37 °C in 5% CO2. Cell viability of SH-SY5Y cells was assessed with the CellTiter-Glo 2.0 luminescent cell viability assay (CTG; Promega, Madison, WI, USA).

Non-Contact co-Culture Model and Cellular Viability

A non-contact co-culture model with LPS-BV2-conditioned media and SH-SY5Y cells was used as previously described [15]. Briefly, BV2 cells were pretreated with test samples (DAI, GEN, and EQL;10 and 20 μM) or positive control, estradiol (EST; 10 and 20 μM), for 1 h, and then stimulated with LPS (1 mg/mL) for 23 h. Next, the conditioned media were removed from BV2 cells and placed with SH-SY5Y cells seeded in white walled 96-well plates at 100,000 cells/mL. SH-SY5Y cells were then incubated for 24 h and their cell viability was evaluated using the CTG (CellTiter-Glo 2.0) assay. Next, the neuroprotective effects of EST, DAI, GEN, and EQL against neurotoxins induced toxicity were evaluated in SH-SY5Y cells. SH-SY5Y cells were seeded in white walled 96-well plates at 100,000 cells/mL for 24 h. Cells then were directly treated with test samples (DAI, GEN, and EQL; 10 and 20 μM) or positive control EST (10 and 20 μM) for 2 h in serum free media. Next, neurotoxins including 6-OHDA (100 μM) or MPP+ (2 mM) were added to SH-SY5Y cells for 24 h, followed by the measurement of cellular viability using the CTG assay [16].

Caenorhabditis elegans Assay

C. elegans were maintained as previously described and used for the evaluation of neuroprotective effects against neurotoxin MPP+ induced toxicity [17]. Briefly, wild type (N2) C. elegans were maintained on normal nematode growth media at 20 °C. nematodes were age synchronized and worms were plated in a 96-well microplate at L1 stage (approximately 20 worms/well). Worms were fed with Escherichia coli OP50 (5 mg/mL) and treated with MPP+ (750 μM) and EQL (17.5 and 30 μM). After 48 h incubation, the surviving worms were counted every 12 h until no live worms remained.

Statistical Analysis

All data are reported as mean ± standard errors of at least three independent samples. Analysis of all cellular data were conducted by ANOVA followed by Dunnett’s test for multiple comparisons of group means. Significance for the group treated with toxic agent as compared to the control group is presented as p ≤ 0.001 (###) and p ≤ 0.0001 (####). Significance for the group treated with test samples as compared to the group treated with toxic agent was defined as: p ≤ 0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***) and p ≤ 0.0001 (****). Lifespan analysis for C. elegans identified using the Kaplan–Meier, tested for statistical significance using the log rank test (Mantel Cox). Drosophila data were analyzed by unpaired one-tailed t-tests. GraphPad Prism software 8.0 (GraphPad Software, Inc., San Diego, CA) was used to create all graphs and for all statistical analysis calculations.

Results and Discussions

Equol (EQL) Reduces the Cytotoxicity of SH-SY5Y Cells in a Non-Contact co-Culture Model

Test samples including EST, DAI, GEN, and EQL were evaluated for cytoprotective effects in a non-contact co-culture model with LPS-BV2-conditioned media and SH-SY5Y cells. At the highest test concentration (20 μM), all of the samples did not show significant toxicity in SH-SY5Y cells as compared to the control group (cell viability >90%; Fig. 1 A). Therefore, concentrations of test samples at 10 and 20 μM were selected for further evaluations. GEN, DAI, and EQL have been previously reported by our group to exert anti-neuroinflammatory effects by reducing the secretion of proinflammatory cytokines including interleukin 6 (IL-6) and tumor necrosis factor alpha (TNF-α), and decreasing the production of inflammatory biomarkers, such as nitric oxide (NO), in BV2 cells [14]. Therefore, herein, we further evaluated the cytoprotective effect of GEN, DAI, and EQL with LPS-BV2-conditioned media and SH-SY5Y cells. The LPS-BV2-conditioned media reduced the viability of SH-SY5Y cells by 56.9% as compared to the control group treated with BV2-conditioned medium. EST, a phytoestrogen used as a reference test sample, only showed a trend of increasing the viability of SH-SY5Y cells (Fig. 1 B). DAI (at 10 μM) exhibited neuroprotective effects against LPS-BV2-conditioned media induced cytotoxicity by increasing the viability of SH-SY5Y cells by 29.2% (Fig. 1 B). In addition, EQL (both at 10 and 20 μM), a gut microbial metabolite of DAI, increased the viability of SH-SY5Y cells by 27.2 and 32.4%, respectively, as compared to the group treated with LPS-BV2-conditioned media (Fig. 1 B). This is not surprising as our group has previously reported that EQL showed neuroprotective effects by downregulating the secretion of IL-6 and TNF-α in LPS-stimulated BV2 microglia [14]. In addition, EQL showed higher protective activity in the non-contact co-culture model as compared to isoflavones DAI and GEN which was in agreement with a published study showing a similar trend [13]. Taken together, it is possible that EQL may exert neuroprotective effects by alleviating inflammatory stress induced cytotoxicity in the SHSY5Y neuroblastoma cells.

Fig. 1
figure 1

Neuroprotective effects of estradiol (EST), isoflavones [including genistein (GEN) and daidzein (DAI)] and their gut microbial metabolite (equol; EQL). Evaluation of cytotoxicity with treatment of EST, GEN, DAI, and EQL (at 20 μM) for 24 h in human neuroblastoma SH-SY5Y cells (A). Protective effects of EST, GEN, DAI, and EQL on LPS-BV2-conditioned media induced cytotoxicity of SH-SY5Y cells (B). Data are expressed as mean ± standard error (n ≥ 6). Statistical significance was defined as ## p < 0.01, as compared to the control group; *p < 0.05 and ** p < 0.01 as compared to the conditioned media treated group

Full size image

Equol (EQL) Rescues the Viability of SH-SY5Y Cells Exposed to PD Associated Neurotoxins

The neuroprotective effects of GEN, DAI, and EQL were evaluated in a cell based PD model against neurotoxins including 6-OHDA and MPP+ induced cytotoxicity. Treatments of 6-OHDA and MPP+ decreased the viability of SH-SY5Y cells to 31.5 and 60.0%, respectively, as compared to the control group (Fig. 2 A). GEN did not exhibit neuroprotective effects against 6-OHDA and MPP+ induced toxicity in SH-SY5Y cells, whilst DAI alleviated 6-OHDA induced cytotoxicity by increasing the cell viability by 26.5% (10 μM). DAI also reduced MPP+ induced toxicity by restoring the cell viability by 15.1 and 19.9% (10 and 20 μM), respectively (Fig. 2 B). EQL (at 10 and 20 μM) reduced 6-OHDA and MPP+ induced toxicity by increasing the viability of SH-SY5Y cells by 30.5–34.5% and 17.9–18.9%, respectively. EST only exhibited protective effect by rescuing the cell viability to 29.3% at 10 μM in the model of 6-OHDA induced cytotoxicity. Both MPP+ and 6-OHDA are neurotoxins used in experimental models to trigger PD-like characteristics by producing reactive oxygen species including free radicals in the mitochondria of dopaminergic neurons, which further leads to the neural cell death in the substantia nigra. In a previously reported study, ellagitannins and their gut microbial metabolite, namely, urolithin A, exhibited neuroprotective effects by reducing mitochondrial oxidative stress in a rat PD model [18]. In addition, studies using human neuroblastoma SH-SY5Y cells showed that polyphenol-derived microbial metabolites exerted neuroprotective effects against oxidative stress induced cytotoxicity [19] and protected SH-SY5Y cells from injury induced by neurotoxin 3-morpholinosydnonimine (SIN-1) [20]. These published studies and our current study support that gut microbial metabolites of dietary polyphenols may exert neuroprotective effects against oxidative stress in cell based models. Moreover, EQL has been reported to show neuroprotective effects in models of Alzheimer’s disease [21] and focal brain ischemia [22] by ameliorating oxidative stress. Therefore, the neuroprotective effects of EQL in the cellular PD model may be attributed to its protective effects against mitochondrial oxidative stress [23]. However, further studies are warranted to confirm this.

Fig. 2
figure 2

Neuroprotective effects of EST, GEN, DAI, and EQL on 6-OHDA (A) and MPP+ (B) induced neurotoxicity of SH-SY5Y cells. Data are expressed as mean ± standard error (n ≥ 6). Statistical significance was defined as #### p < 0.0001, as compared to the control group; *p < 0.05, ** p < 0.01, and ***p < 0.001 as compared to the neurotoxins treated group

Full size image

Equol (EQL) Increases the Lifespan of MPP+-Exposed C. elegans

Given that EQL showed superior activity to GEN and DAI in the aforementioned assays, it was selected for further evaluation of its neuroprotective effects against MPP+ induced neurotoxicity in a PD model of C. elegans. Worms were first exposed to MPP+ (750 μM) and then treated with two concentrations of EQL (17.5 and 35 μM) which were non-toxic to the worms. As shown in Fig. 3 A, a non-treated group was used as control and the MPP+ exposure to worms caused high lethality. The median and maximum survival of worms in MPP+ treated group was 72 h whereas EQL at 17.5 μM significantly prolonged the median and maximum lifespan by 1.5-fold (108 h) and 2.0-fold (144 h), respectively (Fig. 3 B and Table 1). In addition, treatment with EQL at 35 μM increased the median and maximum lifespan of worms by 1.5-fold (108 h) and 2.2-fold (156 h), respectively (Fig. 3 C and Table 1). These results indicated that EQL was able to ameliorate the lethality of the worms and extended their lifespan in the presence of MPP+. Although EQL has been reported to show neuroprotective effects in several in vitro [13, 24] and rodent [22, 25] models of neurodegenerative diseases, to date, this is the first report on the evaluation of its anti-PD effects in a model of C. elegans.

Fig. 3
figure 3

Neuroprotective effects of EQL against MPP+ induced neurotoxicity in a C. elegans model. Effect of MPP+ (750 μM) on the lifespan of C. elegans (A) and effect of EQL (17.5 and 35 μM) on the lifespan of C. elegans exposed to MPP+ (B and C)

Full size image
Table 1 Evaluation of the neuroprotective effects of EQL against PD in an in vivo model. Median and maximum survival time (h) of C. elegans worms treated with EQL (17.5 and 35 μM) post MPP+ induction of neurotoxicity at 25 °C. Significance for the group treated EQL as compared to the group treated with MPP+ was defined as: p ≤ 0.01 (**) and p ≤ 0.001 (***)
Full size table

Biological evaluations of EQL’s neuroprotective effects using cell based in vitro assays and invertebrate animal based models (e.g., C. elegans) may not accurately reflect its physiologically relevant effects. This is because the concentrations of EQL tested in these assays were usually in the range of micromolar levels, which can be significantly higher than its reported bioavailable plasma levels (nanomolar) from human clinical studies [26, 27]. Therefore, further biological evaluations of EQL at physiological relevant conditions using proper experimental models are critical to elucidate EQL’s neuroprotective effects. Moreover, gut microbial metabolites, including EQL, can undergo phase II metabolism, which conjugates EQL with charged species including glutathione, sulfate, glycine, or glucuronic acid. It has been reported that, after consumption of isoflavones enriched diet, EQL can enter into human plasma circulation, and its conjugated phase II metabolites, such as EQL-glucuronide, EQL-sulfate, and EQL-sulfoglucuronide, are excreted in the urine at low concentrations [28]. Therefore, phase II metabolites of EQL, at their bioavailable levels, may also exert potential biological effects, which can contribute to the overall neuroprotective effect of EQL. However, further biological evaluations of EQL phase II metabolites are warranted to confirm this.

In summary, EQL, a BBB penetrable gut microbial metabolite of the isoflavone, DAI, showed neuroprotective effects by increasing the viability of human neuroblastoma SH-SY5Y cells in a non-contact co-culture model with LPS-BV2-conditioned media. In addition, EQL alleviated PD-related chemically induced neurotoxicity in SH-SY5Y cells and C. elegans. Findings from this study support the neuroprotective effects of EQL against PD but further studies are warranted to elucidate the mechanisms of action.