Formulated Chinese Medicine Shaoyao Gancao Tang Reduces Tau Aggregation and Exerts Neuroprotection through Anti-Oxidation and Anti-Inflammation

Misfolded tau proteins induce accumulation of free radicals and promote neuroinflammation by activating microglia-releasing proinflammatory cytokines, leading to neuronal cell death. Traditional Chinese herbal medicines (CHMs) have been widely used in clinical practice to treat neurodegenerative diseases associated with oxidative stress and neuroinflammation. This study examined the neuroprotection effects of formulated CHMs Bai-Shao (made of Paeonia lactiflora), Gan-Cao (made of Glycyrrhiza uralensis), and Shaoyao Gancao Tang (SG-Tang, made of P. lactiflora and G. uralensis at 1 : 1 ratio) in cell model of tauopathy. Our results showed that SG-Tang displayed a greater antioxidative and antiaggregation effect than Bai-Shao and Gan-Cao and a stronger anti-inflammatory activity than Bai-Shao but similar to Gan-Cao. In inducible 293/SH-SY5Y cells expressing proaggregant human tau repeat domain (ΔK280 tauRD), SG-Tang reduced tau misfolding and reactive oxygen species (ROS) level in ΔK280 tauRD 293 cells and promoted neurite outgrowth in ΔK280 tauRD SH-SY5Y cells. Furthermore, SG-Tang displayed anti-inflammatory effects by reducing nitric oxide (NO) production in mouse BV-2 microglia and increased cell viability of ΔK280 tauRD-expressing SH-SY5Y cells inflamed by BV-2 conditioned medium. To uncover the neuroprotective mechanisms of SG-Tang, apoptosis protein array analysis of inflamed tau expressing SH-SY5Y cells was conducted and the suppression of proapoptotic proteins was confirmed. In conclusion, SG-Tang displays neuroprotection by exerting antioxidative and anti-inflammatory activities to suppress neuronal apoptosis in human tau cell models. The study results lay the base for future applications of SG-Tang on tau animal models to validate its effect of reducing tau misfolding and potential disease modification.


Introduction
Neurodegenerative diseases including Alzheimer's disease (AD) and tauopathy are characterized by the presence of hyperphosphorylated, insoluble, and filamentous tau protein, which leads to neuronal dysfunction and loss [1]. Tau is an ubiquitously distributed microtubule-associated protein that promotes and stabilizes microtubule assembly. Aside from helping microtubule assembly, tau also interacts with other cytoskeleton components to play a role in axonal transport [2]. Tau is encoded by MAPT (microtubule-associated protein tau) gene located on chromosome 17q21, containing 16 exons [3]. By alternative splicing, tau proteins exist as six different isoforms in human central nervous system (CNS). Exons 9-12 encode four C-terminal microtubule binding motifs which are imperfect copies of an 18-aminoacid tau repeat domain (tau RD ). Different point mutations found in tau RD reduced the ability of tau to promote microtubule assembly [4] and accelerated aggregation of tau into filaments [5]. In addition, a single amino acid deletion (ΔK280) was found in patients with frontotemporal dementia and AD [6][7][8]. ΔK280 is extremely fibrillogenic and frequently used to model tau aggregation [9][10][11].
Emerging evidence has shown protein aggregation as a trigger for inflammation and neurodegeneration [12]. Activated microglia are found in the postmortem brain tissues of human tauopathy, and microglial burden correlated with tau burden in most of the pathologically afflicted areas [13,14]. Chronic activation of microglia may enhance the hyperphosphorylation of tau and the subsequent development of neurofibrillary tangles [15]. Activated microglia contribute to neurofibrillary pathology in AD through production of interleukin (IL)-1 and activation of neuronal p38-MAPK (mitogen-activated protein kinase 1) in vitro [16] and in vivo [17]. In transgenic mice that develop both tau and amyloid pathologies (3 × Tg-AD line), lipopolysaccharide-(LPS-) induced activation of glia exacerbates tau pathology [18]. Tau oligomers colocalize with astrocytes and microglia to induce inflammation, leading to neuronal damage and eventual cell death [19]. Being a critical component in pathogenesis, neuroinflammation provides an attractive therapeutic target in the treatment and prevention of AD and other tauopathy [20,21].
Traditional Chinese herbal medicines (CHMs) have accumulated several lines of beneficial evidence in the treatment of AD [22][23][24]. However, treatment approaches addressing inflammatory processes in tauopathy have not been well investigated. Bai-Shao and Gan-Cao are formulated CHMs prepared from herbs Paeonia lactiflora (P. lactiflora) and Glycyrrhiza uralensis (G. uralensis), respectively. Total glucosides of paeony extracted from P. lactiflora may exert anti-inflammatory activities that contribute to its analgesic effect through modulating production of proinflammatory cytokines from macrophage-like synoviocytes [25]. In addition, ethanol extracts of G. uralensis possess inhibitory effects against NF-κB-mediated inflammatory response and strong activation of the Nrf2-ARE-antioxidative stress signaling pathways [26]. In this study, Bai-Shao, Gan-Cao, and Shaoyao Gancao Tang (SG-Tang), a formulated CHM made of P. lactiflora and G. uralensis at 1 : 1 ratio, were tested in a tau aggregation model [27] to reveal underlying pathogenesis and develop therapeutic strategy targeting neuroinflammation in tauopathy.

Preparation of Formulated CHMs.
Bai-Shao (Code: 5722), Gan-Cao (Code: 5536), and SG-Tang (Code: 0703H) were provided by Sun Ten Pharmaceutical Co. Ltd. (New Taipei City, Taiwan). To prepare the CHM stock solution, 5 g powder was dissolved in 10 ml ddH 2 O, vortexed to mix well, and then centrifuged at 4000 rpm for 10 min at room temperature. The supernatant was collected and used for further experiments.
2.5. DPPH Assay. The DPPH radical-scavenging activity was measured in a reaction mixture containing 0.1 ml of 0.2 mM DPPH radical solution and 0.1 ml of each tested formulas (100~1000 μg/ml). The solution was rapidly mixed and incubated for 30 min at 25°C. The scavenging capacity was measured by monitoring the absorbance at 517 nm with a microplate reader (Multiskan GO, Thermo Scientific, Waltham, MA, USA). The half maximal effective concentrations (EC 50 ) were calculated using the interpolation method.
2.6. Detection of Inflammatory Mediators. Murine RAW 264.7 macrophage cells were seeded in DMEM containing 1% FBS and pretreated with tested formulas (0.5~2 mg/ml) or celecoxib (50 μM) for 8 h followed by LPS (1 μg/ml) stimulation. The release of NO was evaluated by Griess assay according to the manufacturer's protocol (Sigma-Aldrich). The levels of tumor necrosis factor-(TNF-) α, IL-1β, and IL-6 were determined using a mouse enzymelinked immunosorbent assay (ELISA) system (R&D Systems, Minneapolis, MN, USA) following the manufacturer's protocol. The optical density at 450 nm was detected using a microplate reader (ELISA Reader: SpectraMAX340PC; Molecular Devices, Sunnyvale, CA, USA). In addition, the immortalized murine microglial BV-2 cells, an alternative model system for primary microglia, were used. BV-2 cells were seeded in DMEM containing 1% FBS. Next day, cells were pretreated with SG-Tang for 8~24 h, stimulated with LPS (1 μg/ml) for 20 h, and released of NO in the media determined.

Cell Viability/Cytotoxicity Assays of Inflamed SH-SY5Y
Cells. Previously, cell-free media obtained from LPS/IFN-γexposed microglia-like cells resulted in the highest toxicity on cell viability of SH-SY5Y cells [28]. To prepare conditioned medium (CM) with inflammatory factors, BV-2 cells were stimulated with a combination of LPS (1 μg/ml) and IFN-γ (100 ng/ml) for 24 h. After morphology examination, the BV-2 CM were collected, pooled, and centrifuged to remove cell debris. The induced inflammation was confirmed by release of NO, TNF-α, IL-1β, and IL-6 in the media and increased Iba1 expression in the cell lysate.
For SH-SY5Y cell viability assay, DMEM-F12 was then mixed with two times volume of BV-2 CM (a final FBS concentration at 10%) and added to undifferentiated ΔK280 tau RD -DsRed SH-SY5Y cells for 2 days to induce inflammation. Cell viability was determined by MTT assay as described. For SH-SY5Y cytotoxicity assay, neuronaldifferentiated ΔK280 tau RD -DsRed SH-SY5Y cells were treated with BV-2 CM for 5 days as described and media were collected. 100 μl of supernatant from each sample was transferred to 96-well plate to examine the release of lactate dehydrogenase (LDH) by using LDH cytotoxicity assay kit (Cayman, Ann Arbor, MI, USA). The absorbance was read at 490 nm with a microplate reader (Multiskan GO, Thermo Scientific).
2.11. Human Apoptosis Antibody Array. Protein samples from ΔK280 tau RD -DsRed SH-SY5Y cells with different treatments (Dox uninduced/induced, CM unstimulated/ stimulated, and SG-Tang unpretreated/pretreated) were prepared and incubated with apoptosis antibody array membranes (RayBiotech, Norcross, GA, USA). The relative levels of 43 apoptosis-related proteins in human cell lysates were measured with the array. The detected changes in protein levels were confirmed by Western blot or caspase 3 activity assay. 2.13. Caspase 3 Activity Measurement. Cells were lysed in 1 × lysis buffer by repeated cycles of freezing and thawing. Caspase 3 activity was measured with the caspase 3 assay kit according to the manufacturer's instructions (Sigma-Aldrich).
2.14. Statistical Analysis. For each set of values, data are represented as mean ± SD of three independent experiments. Differences between groups were evaluated by two-tailed Student's t-test or ANOVA (one-way and two-way) with post hoc LSD test where appropriate. p values < 0.05 were considered significant.

Formulated CHMs and Cytotoxicity. Three formulated
CHMs, Bai-Shao, Gan-Cao, and SG-Tang were studied. To examine the cytotoxicity of these CHM formulas, MTT assay was performed on HEK-293 or SH-SY5Y cells after treatment with the tested formulas for 24 h. As shown in Figure 1(a), Bai-Shao, Gan-Cao, and SG-Tang exhibited very low cytotoxicity in HEK-293 and SH-SY5Y cells.
Next, the amounts of active constituents, paeoniflorin and ammonium glycyrrhizinate, in these CHM formulas were analyzed by full-spectrum analytic HPLC. As shown in Figure 1(b), chromatographic patterns showed peaks at 230 and 250 nm corresponding to the retention time compatible with paeoniflorin and ammonium glycyrrhizinate, respectively. The amounts of active constituents in these CHM formulas (0.5 g/ml) were 4.06% (42.25 mM) for paeoniflorin in Bai-Shao, 5.78% (34.41 mM) for ammonium glycyrrhizinate in Gan-Cao, and 2.81% (29.33 mM) for paeoniflorin and 2.43% (14.52 mM) for ammonium glycyrrhizinate in SG-Tang.
Misfolded tau may increase the production of reactive oxygen species (ROS) [31]. To examine whether these CHM formulas display antioxidative effects, ROS level was evaluated in Tet-On ΔK280 tau RD -DsRed 293 cells. As Figure 3(d) shows, pretreatment with Congo red (10 μM, a positive control) or formulas (100 μg/ml) significantly reversed the ROS level elevated by misfolded tau production compared to no treatment (88~95% vs. 100%, p = 0 045 <0.001). These data showed the anti-oxidative effects of Bai-Shao, Gan-Cao, and SG-Tang, and SG-Tang possesses a greater anti-oxidative effect than Bai-Shao or Gan-Cao.

Effects of SG-
Our results demonstrate that SG-Tang could protect cells from cell death, increase neurite outgrowth, and reduce hyperphosphorylation of tau in inflamed misfolded tauexpressing ΔK280 tau RD -DsRed cells.

Discussion
In this study, we demonstrated neuroprotection, antioxidative and anti-inflammatory effects of formulated CHM SG-Tang. Our results showed that SG-Tang displayed a greater antioxidative and antiaggregation effect than Bai-Shao and Gan-Cao and a stronger anti-inflammatory activity than Bai-Shao but similar to Gan-Cao (Figures 2 and 3). Moreover, SG-Tang showed neuroprotective effect of promoting neurite outgrowth probably by ameliorating tau misfolding and oxidative stress in our tauopathy model (Figures 3 and  4). The anti-inflammatory effects of SG-Tang were further demonstrated by using LPS-stimulated BV-2 microglia ( Figure 5). Targets identified from human apoptosis protein array indicate SG-Tang may suppress the expression levels of proapoptotic proteins in inflamed ΔK280 tau RD -DsRed SH-SY5Y cells and thus elevate the cell viability ( Figures 6 and 7). In human tauopathy, substantial activated microglia are found in regions of phosphorylated tau accumulation [35]. In tau P301S transgenic mice, prominent glial activation precedes tangle formation and the pattern of activated glia correlates closely with the distribution and density of NFTs [36]. As neuroinflammation is linked to the progression of tauopathy, anti-inflammatory strategy may be effective at reducing tau-related pathology. Indeed, FK506 attenuates tau pathology and increased lifespan in tau P301S mouse model [36]. Treatment of 3xTg-AD mice with anti-inflammatory drug ibuprofen reduces tau phosphorylation and memory impairment [37]. Administration of potent anti-inflammatory minocycline reduces the development of disease-associated tau species in the htau mouse model [38] by reducing several inflammatory factors [39]. In the present study, we applied BV-2 conditioned medium to proaggregant ΔK280 tau RD -DsRed 293/SH-SY5Y cells to mimic neuroinflammation. The study results reveal that CHM formula SG-Tang displays neuroprotection by exerting anti-inflammatory and antiapoptotic activities.
Inflammation is a double-edged sword. Inflammatory response could lead to activation of immune system and elimination of pathogens thereby reducing further cell loss.
Although inflammation might be protective and beneficial to cells, prolonged or dysregulated inflammatory process could also result in production of neurotoxic factors that exacerbate neurodegenerative pathology and cause cell death [12]. Thus, a potential strategy for treating tauopathies is to intervene in microglial activation and neuroinflammation. NSAID has been commonly used as treatment of inflammation and known to be neuroprotective [40]. The mechanism of NSAID has been shown to inhibit the synthesis or activity of inflammatory mediators such as prostaglandin and COX isoforms 1 and 2. Although NSAID could effectively suppress the inflammatory symptoms, these agents may also induce significant side effects such as increased risk of thrombotic cardiovascular and cerebrovascular events [41]. Therefore, more safely, anti-inflammatory drugs need to be explored and developed.
There is a growing interest in natural compounds/products with anti-inflammatory activities which have long been used for treating inflammation-related diseases. In this study, SG-Tang used was formulated with Bai-Shao (P. lactiflora) and Gan-Cao (G. uralensis) and analyzed by HPLC using two main active constituents, paeoniflorin and ammonium glycyrrhizinate (Figure 1). Both paeoniflorin and glycyrrhizinic acid were demonstrated to be able to cross the bloodbrain barrier (BBB) in middle cerebral artery occlusion rats [42]. However, multiplicity of the components in P. lactiflora and G. uralensis contributes to the effects of antioxidation and anti-inflammation. In the root of P. lactiflora, a total of 40 components including 29 monoterpene glycosides, 8 galloyl glucoses, and 3 phenolic compounds were identified [43]. Among them, paeoniflorin, a monoterpene glycoside, is known to possess anti-inflammatory effect and has been applied to cerebral ischemic injury [44]. Paeoniflorin also exhibits neuroprotective effects via inhibiting neuroinflammation in APP/PS1 and in PS2 mutant mice [45,46]. Paeoniflorin and the isomer albiflorin attenuated neuropathic pain by inhibiting the activation of p38 MAPK pathway in spinal microglia and subsequent upregulated IL-1β and TNF-α [47]. Benzoylpaeoniflorin, another  paeoniflorin-related glycoside in P. lactiflora root, protected primary rat cortical cells against H 2 O 2 -induced oxidative stress [48]. In addition to monoterpene glycosides, gallic acid, a phenolic compound in P. lactiflora root, displayed antioxidative effect by scavenging free radicals, inhibiting lipid peroxidation, and protecting against oxidative DNA damage [49]. Paeonol, another phenolic compound in P. lactiflora root, exerted neuroprotective effect in the model of ischemia through reducing proinflammatory receptors/ mediators [50].
The main bioactive components of G. uralensis are triterpene saponins and various types of flavonoids, including glycyrrhetinic acid, glycyrrhizic acid, liquiritigenin, isoliquiritigenin, liquiritin, and licochalcone A [51]. Glycyrrhizin and related compounds were found to show anti-inflammatory activity in vitro [52] and in vivo [53]. Although diammonium glycyrrhizinate rescues neurotoxicity in Aβ 1-42induced mice [54], its effect in tauopathy models is not known. Isoliquiritigenin, isoliquiritin, and liquiritigenin significantly suppressed iNOS, TNF-α, and IL-6 expression in IL-1β-treated rat hepatocytes [55]. Interestingly, the purified glycyrrhiza polysaccharides increased the pinocytic activity, the production of NO, IL-1, IL-6, and IL-12 in macrophages of mice [56]. Glycyrrhetinic acid, liquiritigenin, isoliquiritigenin, and liquiritin were also found to be all potent NRF2 inducers [57]. Moreover, Calzia et al. has shown that polyphenolic phytochemicals displayed a potent antioxidant action by modulating the ectopic F 0 F 1 -ATP synthase activity of the rod outer segments of the retina and prevented the induction of apoptosis [58]. Therefore, polyphenolic compounds from Bai-Shao and Gan-Cao may also exert antioxidative activities not only in but also outside of mitochondria. Given that multiple different compounds in both Bai-Shao and Gan-Cao are exerting effects on different pathways, the combination of Bai-Shao and Gan-Cao may thus have additive protection effects than each alone, which is supported by our study results.
The anti-inflammatory effect of Jakyakgamcho-tang, a formulated P. lactiflora and G. uralensis in Korea, has been shown by inhibiting the NF-κB signaling pathway in keratinocytes [59]. Aberrant activation of NF-κB signaling may lead to apoptosis and cell death [60]. We found that several proapoptotic proteins including BAD, BID, CASP3, CASP8, and CYCS were induced by misfolded tau expression and/ or caused by LPS/IFN-γ-stimulated BV2 microglia. BAD protein is a proapoptotic member of the Bcl-2 gene family involved in initiating apoptosis [61]. BID is also a proapoptotic protein which plays a role as a sentinel for protease-mediated death signals [62]. Caspases are well-studied important mediators of apoptosis. CYCS is known to be released from mitochondria into cytosol to stimulate cell apoptosis [63]. Administration of SG-Tang decreased the production of these proapoptotic proteins, indicating that SG-Tang may target on inhibiting proapoptotic proteins to protect neuron cells from inflammatory damage.
Finally, pretreatment of SG-Tang reversed abnormal hyperphosphorylation at tau Ser202 and Thr231 in inflamed misfolded tau-expressing SH-SY5Y cells ( Figure 6). Tau function is regulated by phosphorylation at specific sites, and tau phosphorylation plays both physiological and pathological roles in the cells. Tau Ser199/202 and Thr205 were found to be locally phosphorylated along the nascent axon during axonogenesis [64]. Phosphorylation of tau Thr231 inhibited tau to bind and stabilize microtubules [65]. Both Ser202 and Thr231 are hyperphosphorylated in degenerating AD brain [66]. Among kinases that regulate tau Ser202 and Thr231 phosphorylation, cyclic AMP-dependent protein kinase (PKA) and cyclin-dependent kinase 2 (CDC2) might be the potential targets of SG-Tang, and SG-Tang treatment may result in activity suppression of these two kinases [66,67]. The exact mechanism for PKA or CDC2 regulation by SG-Tang remains to be determined in our future work.

Conclusions
Plant-derived natural medications have been used for centuries and becoming more popular because of their low side effects. Despite the fact that natural compounds are relatively safe, the complexity of natural products makes nutraceutical preparations difficult to be appropriately designed. In this study, we showed antioxidative and anti-inflammatory effects of SG-Tang as a potential agent for treatment or prevention of neuroinflammation-associated tauopathy. In future, studies of main active compounds paeoniflorin and ammonium glycyrrhizinate in SG-Tang, separately or in combination, in tauopathy cell model are warranted to provide a novel avenue for protection against tauopathy.