Modulating Lineage Specification in Stem Cell Differentiation via Bioelectrical Stimulation Intensity Matching

Development and regeneration in biological tissues are fundamentally affected by stem‐cell‐fate commitment. Bioelectricity is heterogeneous between different tissues and crucially regulates cell behaviors, including cell differentiation. However, the effects of heterogeneous bioelectricity on stem‐cell differentiation remain poorly understood. Herein, it is shown that providing stem cells with electrical stimulation matching the endogenous membrane potentials of cells derived from different tissues (osteogenic‐related: −55.05 ± 4.22 mV, neurogenic‐related: −84.8 ± 7.48 mV) can induce their osteogenic or neurogenic lineage commitment. Molecular dynamics simulations indicated that the osteogenic‐related surface potential favors the adsorption of fibronectin, while the neurogenic‐related surface potential enhances the adsorption of FGF‐2. These different protein adsorptions trigger either downstream Wnt or Erk signaling, which direct stem‐cell differentiation. Surface‐potential‐mediated lineage‐specification of stem cells using bioelectrical intensity has enormous potential application value in tissue regenerative therapy.


Introduction
Organogenesis during embryonic development is influenced by cell communication via the exchange of various types of signals. [1]As well as well-established biochemical cues, biophysical cues such as electrical signals also serve as key regulators in cell communication. [2]hese endogenous electrical signals are heterogenous in different organs during embryonic development. [3]Thus, understanding how electrical signals participate in the regulation of embryonic development is imperative.The differentiation of stem cells strongly impacts organogenesis. [4]Changing the intensity of electrical signals may strongly impact the behavior of stem cells, which in turn affects cell-fate patterning. [5]Although previous studies have demonstrated the importance of electrical stimulation in stem-cell differentiation, how the intensity of this electrical stimulation affects the pattern of stem-cell-fate commitment remains unclear.
Increasing evidence indicates that externally provided electrical signals can exert an influence on cell differentiation toward processes such as osteogenesis, [6] neurogenesis, [7] cardiogenesis, [8] and myogenesis. [9]Immortalized cell lines are commonly used in research into electrically mediated cell differentiation. [10]These studies have preliminarily revealed the promoting or inhibiting effects of electrical signals on cell differentiation.However, these studies were not designed to simulate cell-fate commitment during the early stages of embryonic development.Stem cells exhibit higher pluripotency, making them more suitable for exploring the diversification of cell-fate commitment patterns as mediated by electric signals during embryo development. [11]his type of research also requires power sources that achieve the kind of electrical stimulation suitable for the investigation of electrically mediated stem-cell-fate commitment.Piezoelectric materials that can generate electrical signals without an external device have been widely applied in biomedical engineering. [12]oly(vinylidene fluoridetrifluoroethylene) [P(VDF-TrFE)] is a piezoelectric polymer that exhibits excellent biocompatibility and flexibility. [13]Accordingly, it is considered to be a promising implantable electrical-potential generator.To investigate how stemcell-fate commitment is affected by the intensity of electrical signals, the provided electrical intensity should match those of endogenous bioelectrical microenvironments.The surface potential of P(VDF-TrFE) can be controlled within the typical range of endogenous electrical potentials in well-differentiated tissues (−40 to −90 mV), [14] making it ideal for mimicking the endogenous bioelectrical microenvironments of different tissues and investigating how the intensity of electrical stimulation mediates stem-cell differentiation.
In this study, we investigated the effect of electrical-stimulation intensity on cell-fate commitment.We demonstrated that by matching the surface potential of a P(VDF-TrFE) membrane with those of different tissues, stem cells attached to the membrane can be induced toward specific lineages.The P(VDF-TrFE) membrane with the surface potential of −55.05 ± 4.22 mV, which matches the membrane potential of osteoblasts, induced osteogenic differentiation of stem cells; while setting the membrane potential to −84.8 ± 7.48 mV, which matches that of neural cells, effected the neural differentiation of stem cells (Figure 1).We have also demonstrated that providing a microenvironment favorable for the adsorption of particular differentiation-related proteins may be the initial factor by which heterogeneous surface potential mediates the determination of stem-cell lineage.Accordingly, this study provides a theoretical basis for the potential application of bioelectrical intensity in regenerative medicine.

Fabrication and Characterization of P(VDF-TrFE) Membranes with Different Surface Potentials
We fabricated P(VDF-TrFE) membranes with different surface potentials by adjusting the -phase content. [15]Figure S1 (Supporting Information) illustrates the preparation of the P(VDF-TrFE) membranes.Figure S2a (Supporting Information) evidences the well-established flexibility of P(VDF-TrFE) membranes.The -phase content can be controlled by simply changing the annealing temperature. [16]In this study, annealing temperatures of 90 and 120 °C were used to construct P(VDF-TrFE) membranes.The scanning electron microscopy (SEM) images in Figure S2b (Supporting Information) show the increased presence of rod-like structures on the surface of the P(VDF-TrFE) membrane treated at 120 °C.These rod-like structures are attributed to -phase crystallinities. [17]Furthermore, the X-ray diffraction (XRD) pattern for this membrane shows an intense peak characteristic of the  phase (Figure 2a).
To verify the surface potential properties of the P(VDF-TrFE) membranes, their remnant polarization, piezoelectric coefficient (d 33 ), and zeta potential values were measured.As shown in Figure 2b, the 120 °C treated membrane shows the highest maximal polarization and remnant polarization.The d 33 value was determined after corona poling at room temperature.The d 33 values for the P(VDF-TrFE) membranes are in line with the -phase data.The d 33 value is ≈21 pC N −1 for the 120 °C−24 kV group, while that for the 90 °C−20 kV group is ≈9 pC N −1 (Figure 2c).Similarly, the zeta surface potential for the 90 °C−20 kV group is ≈−55 mV, while that for the 120 °C−24 kV group is ≈−85 mV (Figure 2d).The surface potentials of the P(VDF-TrFE) membranes are within those typical of endogenous electric potentials (−40 to −90 mV).The 120 °C−24 kV group exhibits the lower relative surface potential (Figure 2e), meaning it bears more negative charge.These results indicate that P(VDF-TrFE) membranes can provide electrical microenvironments suitable for simulating the endogenous electrical potentials of different tissues.
The water-contact angles, roughnesses, and mechanical properties for the P(VDF-TrFE) membranes with different surface potentials showed no significant differences (Figures S3 and S4, Supporting Information).Therefore, these P(VDF-TrFE) membranes allow us to evaluate the effect of different electricalstimulation intensity on multidirectional cell differentiation independently of the other physical and chemical properties of the membranes.
To investigate the influence of different surface potential on cell activity, cell counting kit 8 (CCK8) analysis (Figure S5, Supporting Information) of stem cells adhered to the P(VDF-TrFE) membranes was performed.The result shows no significant differences between P-55, P-85, and P-0, and the viable cell counts increase from day 1 to 7.These results indicate that the P(VDF-TrFE) membranes are not cytotoxic.Furthermore, cytoskeletonimmunostaining images of cells seeded on the P(VDF-TrFE) membranes with different surface potential were captured after 12 and 24 h cultivation (Figure S6, Supporting Information).P-55 cells show polygonal morphology and more extensive spreading than the other groups.The increased spreading area and specific morphology is related to their potential osteogenic differentiation. [18]

Osteogenic/Neurogenic Differentiation of Bone Marrow Mesenchymal Stem Cells on P(VDF-TrFE) Membranes with Different Surface Potentials
Since bone marrow mesenchymal stem cells (BMSCs) have the potential of multidirectional differentiation, they provide an excellent platform for studying the effects of electrical microenvironments on cell differentiation (Figure 3a).Accordingly, we employed immunofluorescence chemistry to explore the differentiation phenotyping of BMSCs.After 3 and 7 days of incubation, P-55 shows the brightest signal for the osteogenic-related marker RUNX2 (Figure 3b).After 14 days of incubation, P-85 shows an increased immunofluorescence intensity for the neuronal marker TUJ1 (Figure 3c) and the presence of neuron-like somas and neurites (Figure 3c; Figure S7, Supporting Information), indicating that the neuronal differentiation of BMSCs might be correlated to lower surface potential.
Osteogenic differentiation of P-55 was corroborated by osteogenic gene expression, alkaline phosphatase (ALP) staining, and Alizarin Red staining analyses.P-55 upregulates the expression of osteogenic-related genes Runx2, Collagen I, and Bmp2 (Figure 3d).After 7 days of incubation, P-55 shows increased ALP production (Figure 3e).In addition, more mineralization nodules are observed for P-55 after 21 days incubation (Figure 3f).These results indicate that the differentiation of BMSCs is directed by heterogenic surface potential, with osteogenic differentiation being facilitated by higher surface potential and neurogenic differentiation being facilitated by lower surface potential.
It has been reported that the membrane potential of osteoblasts is between −42 and −58 mV, [19] which matches the surface potential of group P-55 in our study.Furthermore, the membrane potential of neural cells such as neurons and gila range between −70 and −90 mV, [20] which approximates the surface potential of the group P-85 in our study.Therefore, we propose that the intensity matching surface potential is a key regulator in stem-cell-fate commitment.

RNA-Seq Analysis
To further establish the pattern of cell differentiation as mediated by the intensity of electrical stimulation, RNA-seq was performed (Figure 4a).Comparing P-85 and P-55, GSEA analysis shows that the biological processes of osteogenesis, such as skeletal system development, osteoblast development, ossification, regulation of osteoblast proliferation, and bone trabecula morphogenesis are enriched for P-55 (Figure 4b).Conversely, neurogenesis-related biological processes such as regulation of neurogenesis, regulation of nervous system development, neural crest cell development, neural nucleus development, and regulation of neuron projection development are enriched for P-85 (Figure 4b).
These results further reveal the pattern of cell differentiation as mediated by the intensity of electrical stimulation: the surface potential ≈−55 mV is related to osteogenic differentiation, while the surface potential ≈−85 mV is related to neurogenic differentiation.Furthermore, GSEA analysis revealed that biological processes related to the Wnt signaling pathway are enriched for P-55, while those related to the Erk signaling pathway are enriched for P-85 (Figure 4c).Wnt signaling is recognized as a crucial regulator in the processes of bone formation, growth, and development. [21]It is wellestablished that activation of Wnt signaling can foster osteogenic differentiation of stem cells. [22]This result suggests that a surface potential ≈−55 mV induces osteogenic differentiation of stem cells through the activation of Wnt signaling.Additionally, activation of Erk signaling is associated with neural differentiation of stem cells during the early stages of embryonic development. [23]his suggests that neurogenic differentiation mediated by a surface potential ≈−85 mV may be related to activation of Erk signaling Furthermore, the osteogenic-related genes Col1a1, Postn, Pth1r, Ctsk, and Atf4, and genes in the Wnt signaling pathway Wnt4, Wnt6, and downstream Bmper for P-55 are upregulated on day 21 compared with day 3 (Figure 4d; Figure S8, Supporting Information).Therefore, these results indicate that P-55 plays the stimulative role in guiding BMSCs from the undifferentiation stage into osteogenic lineage.In neural differentiation, P-85 shows higher expression levels of the neurogenesis-related genes Neu1, Gadd45b, Igf1, Mecp2, and Klf9 on day 21 than those on day 3 (Figure 4e).Furthermore, expression of the Erk-related genes Mapk1, Hras, and Nras are downregulated on day 21 (Figure S9, Supporting Information).This result is consistent with previous studies showing that Erk signaling promotes early-stage neural differentiation of stem cells. [24]ollectively, our findings on gene expression, protein expression, and transcript patterns confirm that osteogenic commitment of BMSCs is favored for P-55 while neurogenic commitment is promoted for P-85.

Molecular Dynamics Simulation of Protein Adsorption
One of the first events following the implantation of a biomaterial is its adsorption of proteins. [25]Accordingly, to better understand the initial relationship between surface potential and stem cells, we performed classical molecular dynamics (MD) simulations to simulate protein adsorption onto a P(VDF-TrFE) membrane (Figure 5a).Adherent proteins on a material surface are strongly associated with cellular behavior on that surface, such as attachment, [26] proliferation, [27] and differentiation. [28]ibronectin (FN) is considered essential for osteogenic differentiation owing to its close affinity for integrins expressed by stem cells. [29]Accordingly, we studied FN adsorption under different electrical microenvironments to better understand the osteogenic differentiation of stem cells.
In the simulation models, P-55 exhibits good adsorbability of FN protein (Figure 5b).The distance between the P(VDF-TrFE) surface and FN for group P-55 is lower than that for the other groups (Figure 5d).Overall, the MD results indicate that P-55 exhibits the most favorable electrical environment for FN adsorption.The enhanced FN adsorption leads to increased integrin binding, [30] which can facilitate Wnt signaling and regulate stemcell differentiation. [31]According to the RNA-seq results, the expression of genes related to Wnt signaling is increased for P-55, which is consistent with the MD simulation of FN adsorption.Thus, it may be speculated that enhanced osteogenesis for P-55 is initiated by favorable FN adsorption and promoted via Wnt signaling.The binding between a neural-differentiation-related protein and P(VDF-TrFE) was also investigated.Fibroblast growth factor-2 (FGF-2), also known as basic FGF, is a multi-functional growth factor. [32]There is extensive evidence that FGF-2 is a crucial growth factor for neural differentiation of stem cells [33] and neurite outgrowth. [34]Thus, the adsorption of FGF-2 was simulated to investigate the effects of different electrical microenvironments on the neural differentiation of stem cells.The snapshot in Figure 5c shows the small distance between the P(VDF-TrFE) surface and FGF-2 protein for P-85.Quantitative analysis shows that P-85 exhibits the shortest distance between the P(VDF-TrFE) surface and FGF-2 (Figure 5e).This result indicates that P-85 has more affinity with FGF-2 than P-55 and P-0.FGF-2 has been shown to activate the Mapk signaling pathway by combining with cell-membrane receptors. [35]The Mapk/Erk pathway is well-known to transmit signals from the cell surface to the nucleus, and Erk signaling is considered an important regulator in cell growth and differentiation. [36]Previous studies have demonstrated that Mapk/Erk signals play a crucial role in FGF-2induced neural differentiation of stem cells. [37]Therefore, based on the MD results for FGF-2 adsorption and the enriched Erk signaling for P-85 revealed by RNA-seq, we propose that the P-85 membrane provides a favorable microenvironment for FGF-2 adsorption, which in turn contributes to neurogenesis via activation of Erk signaling.

Osteogenic/Neurogenic Differentiation in Dental Pulp Stem Cells
Following from our findings for BMSCs, we investigated whether they are also applicable to another type of stem cells.Dental pulp stem cells (DPSCs) are derived from the neural crest and show multilineage differentiation. [38]Furthermore, they can be easily obtained from exfoliated or extracted teeth.These characteristics make DPSCs an ideal seed cell for regenerative medicine. [39]hus, the expression of genes and proteins related to osteogenic and neural differentiation in DPSCs under different electrical microenvironments was investigated (Figure 6a).
To evaluate osteogenic induction under different surface potentials, ALP and Alizarin Red staining were performed.After 7 days incubation, ALP production is significantly higher for P-55 (Figure 6b).Comparing P-85 and P-0, more mineralization nodules are observed for P-55 after 21 days incubation (Figure 6c).
The expression of osteogenic-related proteins was also studied (Figure 6d).Compared with the other groups, protein expression of BMP2, COLLAGEN I, and SP7 on day 7 and COLLAGEN I, ALP, BMP2, SP7, OPN, VINCULIN, OCN on day 14 are elevated for P-55.
For neurogenesis, P-85 exhibits the strongest fluorescence intensity for neuronal marker TUJ1 after 7 days incubation (Figure 6e).Furthermore, the expression of neuraldifferentiation-related genes MUSASHI1, NF-H, TUJ1, and MAP2 are increased for P-85 (Figure 6f).Therefore, our findings further verify that osteogenic differentiation of stem cells is facilitated for P-55, while neuronal differentiation is favored for P-85.
Overall, these findings identify the appropriate surface potential matching as one of the key factors in stem-cell differentiation.

Conclusion
In this study, we fabricated P(VDF-TrFE) membranes with different surface potentials matching those of cells derived from different tissues to investigate surface-potential-mediated lineage specification in stem cells.
Our results demonstrate that P(VDF-TrFE) membranes with high surface potential matching that of osteoblasts induce osteogenic lineage commitment of stem cells.Conversely, P(VDF-TrFE) membranes with low surface potential matching that of neural cells direct stem cells toward neurogenic differentiation.Furthermore, MD simulations indicated that the initial factor influencing surface-potential-mediated lineage commitment could whether the surface is favorable to FN or FGF-2 adsorption.The enhanced adsorption of FN or FGF-2 triggers the activation of Wnt or Erk signaling, respectively inducing osteogenic or neurogenic differentiation of stem cells.
This study provides a theoretical basis for understanding organogenesis modulated by bioelectric microenvironment and guiding the future application of artificial organs in regenerative medicine.
Fabrication and Characterization of P(vdf-trfe) Membranes with Different Surface PotentialsP(VDF-TrFE) powders (70/30 mol.%VDF/TrFE) were dissolved in N,N-dimethylformamide under magnetic stirring for 12 h.Then, the fully dissolved solution was spread on a glass substrate and dried at 55 °C before annealing in a vacuum oven for 1 h at 90 or 120 °C according to the method of a previous study. [16]For polarization, P(VDF-TrFE) membranes were subjected to corona poling under a DC electric field (20 or 24 kV) at room temperature for 1 h.The distance between the electrode needle and the membrane was ≈1.5 cm.
Morphology and structure of P(VDF-TrFE) membranes were observed by field-emission scanning electron microscopy (FE-SEM; S-4800, HI-TACHI, Japan) and X-ray diffraction spectroscopy (XRD; Rigaku D/max 2500VB2t/PC, Japan).The surface roughnesses of the P(VDF-TrFE) membranes were characterized by atomic force microscopy (AFM; Bruker, Santa Barbara, CA, USA) in contact mode.Water-contact angles were assessed by a contact-angle instrument (KRÜSS, Germany).Mechanical properties were measured using a universal mechanical testing apparatus (INSTRON-1121, USA).Piezoelectric coefficients (d 33 ) of the polarized samples were determined using a quasi-state d 33 -meter (ZJ-3A, Institute of Acoustics, Chinese Academy of Science, China).The zeta surface potentials of the polarized membranes were determined using a Zeta Sizer Nano-ZS Instrument (Malvern Instruments, Worcestershire WR, UK) at room temperature.The relative surface potentials of the polarized membranes were determined by Kelvin probe force microscopy in tapping mode.Polarization-electric field (P-E) loops were generated using a ferroelectric analyzer setup (TF1000, aixACCT Systems GmbH, Germany) at a frequency of 10 Hz.

Biocompatibility of P(VDF-TrFE) Membranes: Cell proliferation DPSCs
(5 × 10 4 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 12-well plates.After 1, 3, 5, and 7 days of cultivation, CCK-8 kits were used to evaluate the proliferation of DPSCs according to the manufacturer's instructions.The culture medium was replaced with culture medium mixed with 10% (v/v) CCK-8 solution and incubated at 37 °C for 1 h.The optical densities of the solutions were determined at 450 nm.
Cell spreading DPSCs (1 × 10 5 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 6-well palates.After 12 and 24 h of incubation, cells were treated with 4% paraformaldehyde (PFA) for 15 min.Then, the cells were washed with phosphate-buffered saline (PBS) three times.Phalloidin (1:200, CA1610, Solarbio, China) was diluted with PBS solution for cytoskeletal staining.The phalloidin solution was added to the sample, which was then incubated at room temperature for 45 min.Then, DAPI (diluted 1:1000 with PBS; Sigma) was used for staining the nuclei.Samples were observed under a confocal laser scanning microscope (Lecia).
Osteogenic/Neurogenic Differentiation of BMSCs on P(VDF-TrFE) Membranes: BMSCs (1 × 10 5 cells mL −1 ) were seeded on P(VDF-TrFE) membranes with different surface potentials in 6-well plates.After 3, 7, or 14 days cultivation, the cells were fixed with 4% PFA for 15 min.The residual PFA was rinsed away with PBS.Then, the samples were incubated with Triton X-100 (1:1000, diluted with PBS) at room temperature for 10 min to permeabilize the cell membrane before being incubated with block solution (3 wt% BSA diluted with PBS) at room temperature for 1 h.Polyclonal rabbit anti-RUNX2 (1:100, diluted with 1 wt% BSA solution; Abcam) or polyclonal rabbit anti-BETA III TUBULIN (1:100, diluted with 1 wt% BSA solution; Abcam) were added to the samples, which were then incubated at 4 °C overnight.The primary antibodies were removed, and the samples were washed with PBS three times.Goat anti-rabbit IgG H&L Alexa Fluor 488 (1:500, diluted with 1 wt% BSA solution; Abcam) was added to the samples and incubated at room temperature for 1 h for fluorescent staining.Then, samples were treated with phalloidin solution (1:200, diluted with PBS; Solarbio, China) and incubated at room temperature for 45 min to stain the cytoskeleton.Nuclei were stained with DAPI (1:1000, diluted with PBS solution; Sigma).Images were captured using a confocal laser scanning microscope (Lecia).
BMSCs (1 × 10 5 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 6-well plates.Following 7 or 14 days of cultivation, the total RNA of the BMSCs was extracted using TRIzol Reagent (15 596 026, Invitrogen, USA) according to the manufacturer's instructions.The RNAs were reverse-transcribed into cDNA using a reverse transcription kit (RR037A, Takara Bio Inc., Japan).The rt-PCR was performed by FastStart Universal SYBR Green Master Mix system with QuantStudio Design & Analysis Desktop Software (Thermo Fisher Scientific).The primer sequences are shown in Table S1 (Supporting Information).The housekeeping gene GAPDH served as the internal control.
BMSCs (1 × 10 5 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 6-well plates.After 21 days incubation, the cells were fixed with 4% PFA for 15 min.The residual PFA was rinsed away with PBS.Then, samples were treated with 1 wt% Alizarin Red S (pH 4.2; Sigma) to evaluate calcium deposition.Pictures were captured using a microscope (SZX7, Olympus, Japan).
RNA-Seq of BMSCs on P(VDF-TrFE) Membranes: BMSCs (1 × 10 5 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 6-well plates.Following 3, 7, 14, or 21 days of cultivation, the total RNA of the BMSCs was extracted using TRIzol Reagent.The RNA-seq were tested by Novogene Co., Ltd.Briefly, 1 μg RNA per sample was used as the input material for RNA sample preparation.Sequencing libraries were generated using NEBNext Ultra RNA Library Prep Kit for Illumina (NEB, USA) following the manufacturer's recommendations, and index codes were added to attribute sequences to each sample.The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumia) according to the manufacturer's instructions.After cluster generation, the library preparations were sequenced on an Illumina Hiseq platform and 125 bp/150 bp paired-end reads were generated.
MD Simulation: Structure of FN-III7-10 (PDB id 1FNF), FGF-2 (1BFG) were obtained from the Protein Data Bank.Antechamber [40] was used to obtain the general AMBER force field (GAFF) of the P(VDF-TrFE) film.The B3LYP-D3 functional with 6-31G(d) basis set was applied to calculate restrained electrostatic potential (RESP) charge by Gaussian 16. [41] The solvation model density (SMD) [42] with water was selected to simulate polar environment.Three layers of the P(VDF-TrFE) membrane model (VDF/TrFE = 7:3) was built with a dimension of 12.0 × 10.5 × 18.5 nm 3 .The electric fields with different intensity (−100, −50, 0 mV nm −1 ) were applied to the system to mimic the different potential on the P(VDF-TrFE) membrane.All the proteins on P(VDF-TrFE) membrane configuration were simulated by Gromacs 2020.3. [43]Visual molecular dynamics (VMD) software [44] was employed for the trajectory images between P(VDF-TrFE) membrane and proteins.FN-III7-10 or FGF-2 were set closely to the surface of P(VDF-TrFE) membrane by using the steered molecular dynamics (SMD) under the constant force (1000 kJ•mol −1 •nm −2 ).The system was solvated with SPC water and neutralized by adding 14 sodium ions in FN-III7-10 system and ten chloride ions in FGF-2 system, respectively.After the steering process, MD simulation of 10 ps was performed to elimination of any unreasonable contacts within the system.Then, MD simulations were carried out for another 10 ns to study the proteins adsorption details on P(VDF-TrFE) membrane.The time step of simulations was set to 2 fs.All the simulation was carried out with NVT by Nosé-Hoover thermostat at 300 K.The cut-off of electrostatic and Van der Waals interaction was set as 1.2 nm.We applied LINCS algorithm [45] to deal with the bond constrains.The distance was also calculated between geometric center of the protein and P(VDF-TrFE) membrane at the direction perpendicular to the P(VDF-TrFE) membrane, which is aiming for illustrating the differences of adsorption.
DPSCs (1 × 10 5 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 6-well plates.After 21 days incubation, cells were fixed with 4% PFA for 15 min.The residual PFA was washed off with PBS.Then, samples were treated with 1 wt% Alizarin Red S (pH 4.2; Sigma) to visualize calcium deposition.Pictures were captured using a microscope (SZX7, Olympus, Japan).
DPSCs (1 × 10 5 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 6-well plates.Following 7 or 14 days of cultivation, the total RNA of the DPSCs was extracted using TRIzol Reagent (15 596 026, Invitrogen, USA) according to the manufacturer's instructions.The attracted RNAs were reverse-transcribed into cDNA using a reverse-transcription kit (RR037A, Takara Bio Inc., Japan).rt-PCR was performed using a FastStart Universal SYBR Green Master Mix system with QuantStudio Design & Analysis Desktop Software (Thermo Fisher Scientific).The primer sequences are shown in Table S1 (Supporting Information).The housekeeping gene GAPDH served as the internal control.
DPSCs (1 × 10 5 cells mL −1 ) were seeded onto P(VDF-TrFE) membranes with different surface potentials in 6-well plates.Following 7 or 14 days of cultivation, the total proteins were extracted using RIPA buffer (P0013B, Beyotime, China) with a protease inhibitor cocktail (Thermo Fisher Scientific, USA).The concentrations of the proteins were determined using BCA protein assay kits.Then, 30 μg of extracted protein was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to poly(vinylidene fluoride) (PVDF) membranes.The membranes were blocked using 5% skimmed milk (diluted with TBST) for 1 h at room temperature.Then, the membranes were incubated with

Figure 3 .
Figure 3. Osteogenic/neurogenic differentiation of BMSCs on P(VDF-TrFE) membranes with different surface potentials.a) Schematic of P(VDF-TrFE)membrane-induced osteogenic/neurogenic differentiation in BMSCs.b) Immunofluorescence images showing the upregulated expression of osteorelated protein RUNX2 in P-55 after 3 and 7 days incubation.c) Immunofluorescence images showing the upregulated expression of neuro-related protein TUJI in P-85 after 14 days incubation.d) RT-qPCR results showing the upregulation of osteogenic markers (Runx2, Collagen I, Bmp2) in P-55.(p < 0.05).e) ALP staining and f) Alizarin Red staining showing the enhanced osteogenic differentiation of BMSCs on the P-55 membrane.

Figure 4 .
Figure 4. RNA-Seq results for BMSCs cultured on P(VDF-TrFE) membranes with different surface potentials.a) Schematic showing RNA-Seq for BMSCs cultured on P(VDF-TrFE) membranes with different surface potentials.b) GSEA analysis showing the enrichment of biological processes related to osteogenesis for P-55 and neurogenesis for P-85.c) GSEA analysis showing the enrichment of Wnt signaling for P-55 and Erk signaling for P-85.d) RNA-Seq showing the expression of osteogenic-related genes for P-55 and e) neurogenic-related genes for P-85 after 3 and 21 days of incubation.

Figure 5 .
Figure 5. MD simulations of FN and FGF-2 adsorption on P(VDF-TrFE) membranes.a) Schematic of MD simulations of protein adsorption on P(VDF-TrFE) membranes.b) Snapshots of MD simulations (0/10 000 ps) for FN protein on P-55 and c) FGF-2 protein adsorption on P-85.d) Quantitative analysis of the distance between the surface of P(VDF-TrFE) and the FN protein or e) FGF-2 protein.

Figure 6 .
Figure 6.Verification of the induction pattern for stem-cell differentiation mediated by different surface potentials.a) Schematic of P(VDF-TrFE)membrane-induced osteogenic/neurogenic differentiation of DPSCs.b) ALP staining and c) Alizarin Red staining showing that P-55 enhances DPSC osteogenic differentiation.d) Western blot results showing the increased expression of osteogenic-differentiation-related proteins for P-55.e) Immunofluorescence images showing the upregulated expression of neuro-related protein TUJI for P-85 after 7 days incubation.f) RT-qPCR results showing the upregulation of neurogenic markers for P-85.(p < 0.05).