From Natural Compound Screening to Myelin-Associated Effects: Identification of Morusin as a Potent Promoter of Oligodendrocyte Differentiation

Abstract

Myelination is essential for rapid axonal conduction and neuronal integrity, and its loss in demyelinating diseases such as multiple sclerosis (MS) leads to progressive neurological impairment. Despite advances in immunomodulatory therapies, effective strategies that promote remyelination remain limited. Here, we identify Morusin, a prenylated flavonoid natural compound, as a potent enhancer of oligodendrocyte (OL) differentiation and myelination-associated outcomes. Using a fluorescence-based screen of diverse flavonoids in primary rat oligodendrocyte progenitor cells (OPCs), it was found that Morusin markedly increased myelin basic protein (MBP) expression. To enable cross-species validation, we established a SOX10-inducible human OPC differentiation system, which shortened differentiation time and allowed functional screening in human cells. In this platform, Morusin enhanced OL maturation and induced a transcriptional profile enriched for myelination- and axon ensheathment-related genes, including MBP, PLP1, MAG, and SIRT2. Furthermore, in myelin oligodendrocyte glycoprotein (MOG)₃₅–₅₅-induced experimental autoimmune encephalomyelitis (EAE) mice, Morusin improved myelination-associated histological features and functional recovery, comparable to the benchmark compound Benztropine. Collectively, these findings identify Morusin as a promising natural compound with pro-myelinating activity across multiple experimental systems and highlight the potential of rationally guided natural compound screening for regenerative therapy in demyelinating diseases.

1. Introduction

Myelination, a tightly orchestrated developmental process in the central nervous system (CNS), is essential for saltatory conduction of action potentials and for maintaining long-term axonal integrity [1,2,3,4]. Disruption of myelin sheaths, as occurs in demyelinating disorders such as multiple sclerosis (MS), results in progressive neurological deficits and remains a major therapeutic challenge. Although current disease-modifying therapies effectively mitigate inflammation, no existing intervention restores lost myelin, underscoring the urgent need for strategies that actively promote remyelination and durable functional recovery [2,5,6].

Endogenous oligodendrocyte progenitor cells (OPCs) persist within demyelinated lesions and retain the potential to differentiate into myelinating oligodendrocytes (OLs); however, this process is often inefficient or arrested, particularly in chronic disease stages [2,5,7,8]. Consequently, there is growing interest in therapeutically activating endogenous OPCs to stimulate remyelination without relying on exogenous cell transplantation. While transplantation of stem cell-derived OL lineage cells has demonstrated promising outcomes in preclinical models, translation to clinical application is constrained by challenges in scalability, manufacturing reproducibility, safety, and regulatory approval [6,7,9]. Small-molecule modulation of OPC fate offers an attractive, clinically tractable alternative for promoting controlled OL differentiation and myelin repair. Compared with genetic or cell-based approaches, small molecules allow reversible and dose-dependent control of OPC behavior with high reproducibility, enabling systemic delivery and facilitating large-scale application [8]. Several compounds, such as benztropine and miconazole, have been shown to promote OL differentiation and functional remyelination in animal models [6,8]. Nonetheless, most studies have been limited to rodent systems, and few have established robust pipelines that integrate human-derived OPC platforms with in vivo validation to accelerate translational discovery [9].

Among various bioactive chemical classes, flavonoids—plant-derived polyphenolic compounds—have attracted attention for their neuroprotective, anti-inflammatory, and potentially pro-myelinating properties [10]. For instance, hesperetin ameliorates myelin damage in lysolecithin-induced optic nerve injury [11], apigenin exerts neuroprotective and anti-inflammatory effects in neuroinflammatory models, and synthetic flavone analogs have been reported to enhance remyelination and functional recovery [12]. Despite these promising findings, systematic investigations of flavonoid structure–activity relationships in OPC differentiation remain limited. In this study, we conducted a fluorescence-based screen of 100 structurally diverse flavonoids using primary rat OPCs and a human SOX10-inducible OPC differentiation platform.

2. Results

2.1. Establishment and Characterization of Primary Rat OPC Cultures

To identify natural compounds (NCs) with pro-myelinating effects, we first established a primary rat oligodendrocyte progenitor cell (OPC) culture system that enables efficient expansion and reliable differentiation for low-variability screening (Figure 1A). OPCs were isolated from postnatal day 1 rats and enriched by fluorescence-activated cell sorting (FACS) based on co-expression of PDGFRα and A2B5, followed by in vitro expansion under defined culture conditions [13]. All subsequent screening experiments were performed using OPC cultures isolated with the same A2B5⁺PDGFα⁺ FACS gating strategy to ensure consistency across conditions. Phase-contrast imaging revealed a progressive increase in bipolar OPC morphology (Figure 1B), and immunofluorescence confirmed that the majority of cultured cells expressed canonical OPC markers, including A2B5, NG2, PDGFRα, OLIG2, and SOX10 (Figure 1C) [14]. Upon treatment with triiodothyronine (T3), a well-established inducer of oligodendrocyte (OL) differentiation through thyroid hormone signaling, cells displayed enlarged somata and elaborated, radially branching processes (Figure 1D) and expressed mature OL markers, including O4, MBP, and MAG (Figure 1E). These results demonstrate that our OPC culture system provides a robust and reproducible platform for downstream differentiation and screening of NCs with potential pro-myelinating activity.

2.2. Natural Compounds Screening Using Rat OLs

Building upon the established differentiation system (Figure 1), we performed a fluorescence-based screen of 100 structurally diverse flavonoids—bioactive NCs—to identify small molecules that promote OL maturation (Figure 2A). MBP expression intensity (Figure 2B) and the proportion of MBP⁺ cells (Figure 2C) were quantified following fixation and immunostaining of NC-treated cultures to assess OL differentiation efficiency. Among all compounds assessed, K26 (Morusin) and K27 (Morusinol) exhibited the most consistent and potent enhancement of MBP expression (Figure 2D). Notably, these compounds share similar structural motifs. Concentration–response analysis confirmed that K26-treated cultures showed markedly increased MBP⁺ OLs, with peak activity at 1–2 μM—comparable to T3-treated positive controls (Figure 2E,F). Based on its reproducible and dose-responsive enhancement of OL differentiation, Morusin (1.0 μM) was selected as a lead compound for further validation in human OL systems.

2.3. Establishment of a SOX10-Inducible Human OL Differentiation System for Compound Screening

Differentiation of human OLs typically requires prolonged culture exceeding two months, posing challenges for compound screening [15]. Recent studies have demonstrated that transient exogenous SOX10 expression accelerates OL differentiation and enhances efficiency [16]. To leverage this, we examined whether inducible SOX10 expression could facilitate differentiation from previously established OLIG2⁺NKX2.2⁺ pre-OPCs (Figure 3A) [17]. Pre-OPCs were transduced with a Dox-inducible SOX10 lentiviral construct and treated with Dox for four days. O4⁺ cells emerged as early as four days after induction, and mature OL markers were detected by day 10 following Dox withdrawal, indicating successful progression to myelin-producing OLs in the absence of continued exogenous SOX10 expression (Figure 3B). Using this optimized system, we screened 10 NCs—including Morusin—identified from the rodent assay (Figure 2). Pre-OPCs were treated from day 5 to day 14 without Dox (Figure 3A). Immunocytochemistry revealed that several compounds, including T3 and Morusin, promoted OL differentiation relative to DMSO controls (Figure 3C), with Morusin exerting the strongest effect, consistent with results in rodent cells (Figure 3D). To explore the underlying mechanism, we performed RNA sequencing on DMSO- and Morusin-treated cells. Transcriptomic analysis identified 1161 differentially expressed genes (DEGs) (Figure 3E), with significant upregulation of myelination-associated genes such as SIRT2, MAG, MBP, and PLP1 (Figure 3F). Gene Ontology (GO) enrichment further revealed terms related to “axon ensheathment” and “myelination” (Figure 3G). Collectively, transcriptomic profiling followed by pathway enrichment and functional annotation analyses revealed that Morusin treatment was associated with regulation of gene networks involved in oligodendrocyte differentiation, myelin-related processes, and relevant signaling pathways (Figure 3). GO enrichment of Morusin-upregulated genes highlighted MAPK-related terms, including “MAPK cascade” and “regulation of MAP kinase activity” (Supplementary Figure S1A,B). Pharmacological inhibition of MEK using PD0325901 attenuated the Morusin-associated increase in MBP⁺ cells, suggesting involvement of the MEK–ERK axis in Morusin-induced oligodendrocyte progenitor maturation (Supplementary Figure S1C).

2.4. Pro-Myelinating Effects of Morusin in SOX10-Inducible Human Organoids

To assess whether Morusin enhances myelination in a three-dimensional human context, we applied the inducible SOX10 system to generate myelinating cerebral organoids. Human embryonic stem cells (hESCs) were transduced with Dox-inducible SOX10 lentivirus and differentiated into organoids following a modified protocol (Figure 4A,B) [18]. Dox was added for one week starting on day 24 of differentiation, when OLIG2 and NKX2.2 expression were detected (Figure 4C). Subsequent differentiation proceeded for two additional weeks in glial differentiation medium containing DMSO, T3, or Morusin (without Dox). On day 48, O4⁺ OLs were readily detected within organoids (Figure 4D), and the proportion of MBP⁺ cells was significantly higher in Morusin-treated organoids compared with controls (Figure 4E,F). These data demonstrate that Morusin enhances OL maturation and myelination in human organoids.

2.5. Myelination-Associated Effects of Morusin in an Experimental Autoimmune Encephalomyelitis (EAE) Model

Several known pro-myelinating agents have been reported to improve myelination-associated outcomes in the EAE model [8,19,20]. We therefore examined whether Morusin exerts similar effects in MOG_35-55-induced EAE mice, a widely used model of multiple sclerosis (MS) (Figure 5A). At the peak of disease severity (15 days post-immunization), mice were randomized into three groups and treated with DMSO, Benztropine (a positive control), or Morusin. Both Benztropine- and Morusin-treated mice showed a modest improvement in clinical score compared with the DMSO-treated group, with statistical significance observed only at the final experimental day (Figure 5B). Histological analysis revealed increased myelination-associated features in the spinal cords of Morusin- and Benztropine-treated animals compared with the DMSO-treated controls (Figure 5C). Transmission electron microscopy (TEM) further demonstrated the presence of compact myelin sheaths with visible major dense lines in the Morusin- and Benztropine-treated groups, whereas DMSO-treated mice displayed sparse and disorganized myelin structures (Figure 5D).

To complement these qualitative histological observations, image-based quantification was performed on representative spinal cord sections. MBP-positive area coverage and the presence of compact myelinated regions were assessed, indicating a greater presence of myelination-associated features in the Morusin-treated group compared with the DMSO- and Benztropine-treated groups (Figure 5C,D and Table S4).

Collectively, these findings support myelination-associated improvements following Morusin treatment in the EAE spinal cord. However, given the pathological nature of the EAE model, these results should not be interpreted as definitive evidence of remyelination as a repair process (Figure 5E). Consistent with previous reports, while some EAE studies describe a plateau after peak disease, a gradual recovery pattern without a sustained plateau has also been observed in comparable MOG₃₅–₅₅ models [21,22], which aligns with the clinical score trajectory observed in the vehicle-treated group in this study.

3. Discussion

Natural compounds (NCs) have long been recognized for their therapeutic potential across cancer and infectious diseases; however, their applications in neurological disorders—particularly demyelinating diseases—remain underexplored due to challenges in isolation, structural characterization, and functional screening [10,12]. In this study, we identify Morusin, a prenylated flavonoid previously reported to possess antioxidant, anti-inflammatory, and anticancer activities [23], as a potent pro-myelinating compound. In initial rodent OPC screens, it was found that both Morusin and its structural analog Morusinol enhanced OL differentiation, suggesting that shared functional groups may confer myelination-promoting activity. To address the temporal limitations of human OL differentiation [16,17,19], we established a SOX10-inducible system that enables efficient and accelerated maturation. Transient SOX10 expression permitted high-throughput screening without prolonged transcriptional manipulation, allowing the detection of compound-specific effects. Using this system, Morusin markedly enhanced OL differentiation within two weeks in pre-OPCs and within seven weeks in hESC-derived organoids, demonstrating strong translational potential. Benztropine, a centrally acting muscarinic antagonist previously shown to enhance remyelination in MS models, served as a benchmark for efficacy [6].

In our primary screen, benztropine and miconazole were included as previously reported small-molecule inducers of OL differentiation [24]. Under our screening conditions, miconazole produced a weaker response than T3 and was therefore not carried forward into downstream analyses, whereas benztropine consistently showed a strong promyelinating effect and was used as the in vivo reference compound. Benztropine has been linked to muscarinic receptor antagonism [6], while miconazole has been associated with MAPK and thyroid-hormone-related signaling [8]. Although Morusin is structurally distinct from these molecules, its activity may intersect with regulatory pathways that influence OL lineage progression, and further studies will be needed to clarify whether Morusin converges on these mechanisms or represents a mechanistically separate class of pro-myelinating compounds. In this context, it is important to note that the initial compound screen was conducted at a single low micromolar concentration to enable standardized comparison across a structurally diverse compound library while minimizing nonspecific or cytotoxic effects. This approach was intended as an initial filtering step to identify candidate compounds with differentiation-promoting activity, rather than to define optimal dose ranges or rank compounds by maximal efficacy. Consequently, compounds with narrower or shifted effective concentration windows may not have been captured in the primary screen. Subsequent dose–response analyses were therefore performed for selected lead compounds, including Morusin, to validate and refine their activity.

Morusin exhibited comparable improvements in myelination-related outcomes in the EAE model, rather than definitive remyelination, reflecting the interpretative limitations of MOG₃₅–₅₅ EAE in which demyelination and spontaneous repair occur concurrently. In addition, the ability of candidate compounds to access the central nervous system is a critical consideration for translational application. Although direct assessment of blood–brain barrier (BBB) permeability was beyond the scope of this study, Morusin is a prenylated flavonoid with physicochemical properties that have been associated with CNS bioavailability, and related compounds have demonstrated central nervous system activity in vivo [25,26]. Nevertheless, dedicated pharmacokinetic and BBB-penetration studies will be required to definitively establish Morusin’s bioavailability within the CNS. Future studies should elucidate the precise molecular mechanisms through which Morusin promotes OL differentiation and influences myelin repair, including potential interactions with thyroid hormone and muscarinic signaling pathways [6,7,9]. Our RNA-seq-based bioinformatic analyses serve a hypothesis-generating role by highlighting candidate signaling programs that may accompany Morusin-driven oligodendrocyte progenitor maturation, including MAPK-associated signatures [27,28]. However, these MAPK-linked changes could reflect either a driver of differentiation or a secondary consequence of shifts in cellular composition (e.g., relative enrichment of more mature oligodendrocyte states) following Morusin exposure. Accordingly, the directionality between kinase signaling and lineage progression cannot be conclusively established from the current dataset, and future work using time-resolved phospho-signaling and single-cell approaches will be important to distinguish initiating signaling events from downstream state transitions. Despite these limitations, our findings provide compelling evidence that Morusin significantly enhances myelination across rodent and human systems. The observed effects in EAE should therefore be interpreted within the constraints of the model but support continued investigation of Morusin as a promising natural compound for demyelinating disease research.

4. Materials and Methods

4.1. Cell Culture

Primary rat OPCs were established from the cerebral cortices of P1 rat brains as previously described [13] and purified by fluorescence-activated cell sorting (FACS) using PDGFRα (R&D Systems, Minneapolis, MN, USA, AF1062R) and A2B5 (R&D Systems, FAB1416G) antibodies on a BD FACSAria II system (BD Biosciences, San Jose, CA, USA). OPCs used for screening were isolated using identical A2B5 and PDGFRα gating criteria in all experiments. Sorted OPCs were cultured in expansion medium consisting of DMEM/F12 supplemented with 1× B-27 supplement without vitamin A (Thermo Fisher Scientific, Waltham, MA, USA), 1× N-2 supplement (Thermo Fisher Scientific), 1% penicillin–streptomycin, 1% L-glutamine, 1% non-essential amino acids, 20 ng/mL FGF2 (Peprotech, Cranbury, NJ, USA), and 20 ng/mL PDGF-AA (Peprotech). H9 human embryonic stem cells (hESCs; WiCell Research Institute, Madison, WI, USA) were maintained in E8 medium (Thermo Fisher Scientific, A1517001) on Matrigel-coated dishes (BD Biosciences).

4.2. Natural Compound Screening

A library of natural compounds (NCs; 0.5 or 1 μM) was kindly provided by Dr. Dongho Lee. Rat OPCs were plated onto Matrigel-coated 96-well plates and treated for three days as previously described with individual compounds diluted in OL differentiation medium composed of DMEM/F12 supplemented with 1× B-27 supplement without vitamin A, 1× N-2 supplement, 1% penicillin–streptomycin, 1% L-glutamine, 1% non-essential amino acids, 50 μg/mL ascorbic acid (Sigma-Aldrich, Burlington, MA, USA), and 10 μM forskolin (Millipore, Burlington, MA, USA). Triiodothyronine (T3; 60 ng/mL; Sigma-Aldrich) served as a positive control. For the primary screen, compounds were tested at a single low micromolar concentration (0.5 μM), which is commonly used in cell-based natural compound screens, as it allows standardized comparison while minimizing nonspecific or cytotoxic effects [29,30]. Three days was selected as the treatment period, as previously described in screening studies [31,32,33,34]. Screening outcomes were quantified based on MBP fluorescence intensity and the proportion of MBP⁺ cells following fixation and immunostaining. For the primary screening analysis, MBP fluorescence intensity was quantified from independent biological replicates, and statistical significance was assessed using one-way ANOVA as described in the Section 4.10. The Morusin concentration used in cell, organoid, and animal experiments was chosen based on dose–response validation to ensure reproducible differentiation effects while minimizing potential off-target or cytotoxic effects.

4.3. Generation of High-Titer Lentivirus

High-titer lentiviruses encoding FUW-TetON-SOX10 (Addgene, Watertown, MA, USA #115242) or FUW-M2rtTA (Addgene #20342) were produced in 293FT cells (Invitrogen, Carlsbad, CA, USA R70007) by co-transfection with psPAX2 (Addgene #12260) and pMD2.G (Addgene #12259) using Lipofectamine 2000 (Invitrogen). Viral supernatants were collected at 24, 48, 72, and 96 h post-transfection, filtered through a 0.45-μm membrane, and concentrated by ultracentrifugation (Beckman Avanti J-E; Beckman Instruments, Brea, CA, USA) at 20,000× g for 2 h. Pelleted viral particles were resuspended in 2 mL of cold DMEM and incubated overnight at 4 °C. Freshly prepared viruses were used immediately without long-term storage. A complete list of the 97 flavonoids used in this screen is provided in Supplementary Table S1.

4.4. Human OL Differentiation

For human OL differentiation, OLIG2⁺NKX2.2⁺ pre-OPCs were transduced with doxycycline (Dox)-inducible SOX10 and rtTA lentiviruses prior to differentiation. Dox (1 μg/mL) was administered for the first 4 days of differentiation and withdrawn for the remaining 10 days in OL differentiation medium. The ten most potent NCs identified from the rat OPC screen were applied under identical conditions. T3 (60 ng/mL) was used as a positive control.

4.5. Generation of Cerebral Organoids Harboring Myelination

For organoid generation, H9 hESCs were transduced with lentiviruses encoding SOX10 and rtTA. Cells were dissociated with Accutase (EMD Millipore, Burlington, MA, USA) on day 0 and seeded onto ultra-low attachment plates (Corning, Corning, NY, USA) in E8 medium for four days to form embryoid bodies (EBs). At day 4, EBs were cultured in neural induction medium consisting of DMEM/F12 supplemented with 1× N-2 supplement, 1% penicillin–streptomycin, 1% L-glutamine, 1% non-essential amino acids, 1 μg/mL heparin (Sigma-Aldrich, H3149), and 0.5 μM purmorphamine (Tocris, Bristol, UK) until embedding. Neurospheres larger than 200 µm were embedded in Matrigel droplets on day 10 as described previously. Embedded organoids were cultured in basal medium (DMEM/F12 supplemented with 543 μg/mL sodium bicarbonate, 1× N-2 supplement, 1× B-27 supplement without vitamin A, 1% penicillin–streptomycin, 1% L-glutamine, 1% non-essential amino acids, 5 μg/mL insulin, and 0.5 μM purmorphamine) for two weeks (until day 24). The medium was then replaced with glial differentiation medium containing 10 ng/mL PDGF-AA, 10 ng/mL IGF-1, 10 ng/mL NT-3, 5 ng/mL HGF (all from Peprotech), 10 μM forskolin, and 20 μg/mL ascorbic acid for an additional 24 days. Dox was administered for one week beginning on day 24, followed by compound or T3 treatment until day 48.

4.6. Immunostaining

Cells were fixed with 4% paraformaldehyde for 10 min at room temperature, washed three times with PBS, and permeabilized for 15 min in PBS containing 0.2% Triton X-100. Samples were blocked with 5% normal donkey serum and 0.01% Triton X-100 for 1 h, then incubated with primary antibodies overnight at 4 °C. After three PBS washes, cells were incubated with Alexa Fluor-conjugated secondary antibodies (1:500; Thermo Fisher Scientific) for 1 h at room temperature and counterstained with DAPI (1 μg/mL; Sigma-Aldrich). Images were captured using an Olympus IX81 inverted fluorescence microscope (Olympus, Tokyo, Japan). For organoid sections, samples were embedded in O.C.T. compound (Tissue-Tek, Sakura Finetek, Torrance, CA, USA), cryosectioned at 10 μm, blocked with 2% donkey serum and 0.2% Triton X-100 for 1 h, and incubated overnight with primary antibodies at 4 °C. Sections were then labeled with Alexa Fluor 488- or 594-conjugated secondary antibodies and counterstained with DAPI. Confocal images were acquired using an Olympus laser scanning microscope (Olympus, Tokyo, Japan). All antibodies are listed in Supplementary Tables S2 and S3.

4.7. RNA-Seq Data Analysis

Total RNA from DMSO- or Morusin-treated human OLs was extracted using TRIzol (Thermo Fisher Scientific) and treated with DNase I. RNA integrity was confirmed using an Agilent 2100 Bioanalyzer (Agilent Technologies Inc, Santa clara, CA, USA) (RIN > 8). cDNA libraries were prepared with the TruSeq RNA Library Prep Kit (Illumina, San Diego, CA, USA) and sequenced on an Illumina HiSeq 2500 platform. Gene expression levels were quantified with Cufflinks v2.1.1 using the Ensembl annotation database. Differentially expressed genes (DEGs) were identified using Cuffdiff with a false discovery rate (FDR) < 0.05 and a ≥2-fold change threshold.

4.8. Experimental Autoimmune Encephalomyelitis (EAE)

Eight-week-old female C57BL/6 mice (OrientBio, Gyeonggi-do, Korea) were acclimated for one week before induction. Mice were immunized subcutaneously with MOG₃₅–₅₅ peptide emulsified in complete Freund’s adjuvant (CFA) and injected intraperitoneally with pertussis toxin (PTX) on the day of immunization and again 24 h later (Hooke Kit™ MOG₃₅–₅₅/CFA Emulsion PTX). Clinical scores were assessed daily using a standard 0–5 scale: 0, no symptoms; 0.5, tail tip weakness; 1, limp tail; 1.5, tail and hindlimb inhibition; 2, partial hindlimb weakness; 2.5, hindlimb dragging; 3, paralysis of one hindlimb; 3.5, paralysis of both hindlimbs; 4, hindlimb and partial forelimb paralysis; 4.5, near-complete paralysis; 5, moribund or death. At peak disease (~day 14–15, clinical score ≈ 3.5), mice were randomized into three treatment groups (DMSO, Benztropine, or Morusin; n = 10 per group). Clinical evaluation was performed daily by blinded experimenters. Two weeks after treatment initiation, animals were sacrificed and spinal cords were collected for histological and ultrastructural analysis. All procedures were approved by the Institutional Animal Care and Use Committee of Korea University (approval no. KUIACUC-2018-0018).

4.9. TEM Sample Preparation and Analysis

EAE mice were anesthetized and perfused with normal saline followed by 2% paraformaldehyde/2.5% glutaraldehyde in 0.1 M phosphate buffer (PB, pH 7.4). Spinal cords were post-fixed overnight at 4 °C, rinsed three times with PB, and then treated with 1% osmium tetroxide in PB for 90 min. Samples were dehydrated through graded ethanol and embedded in epoxy resin. Ultrathin sections (70 nm) were prepared using a UC7 ultramicrotome (Leica Microsystems, Wetzlar, Germany), mounted on 200-mesh grids, and stained with uranyl acetate and lead citrate. TEM images (15,000–60,000×) were obtained using a Hitachi H-7650 electron microscope (Tokyo, Japan) at 80 kV.

4.10. Statistical Analysis

All data are presented as mean ± standard deviation (SD) from at least three independent experiments. Statistical significance was assessed using unpaired two-tailed Student’s t-tests or one-way ANOVA with GraphPad Prism 7 (GraphPad Software, San Diego, CA, USA). Differences were considered significant at p < 0.05. Significance levels are indicated as p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****).