Remote Tuning of Built-In Magnetoelectric Microenvironment to Promote Bone Regeneration by Modulating Cellular Exposure to Arginylglycylaspartic Acid Peptide

FULL PAPER m-journal.d updates Remote Tuning of Built-In Magnetoelectric Microenvironment to Promote Bone Regeneration by Modulating Cellular Exposure to Arginylglycylaspartic Acid Peptide Wenwen Liu, Fengyi Zhang, Yuanyang Yan, Chenguang Zhang, Han Zhao, Boon Chin Heng, Ying Huang, Yang Shen, Jinxing Zhang, Lili Chen, Xiufang Wen, and Xuliang Deng 1. Introduction Mimicking the endogenous physical microenvironment is a promising strategy for biomaterial-mediated tissue regeneration. However, precise A major strategy in tissue engineerin control of physical cues such as electric/magnetic fields within extracellular is to mimic the biophysical properties of environments to facilitate tissue regeneration remains a formidable challenge target tissues and various critical aspects of the extracellular environment to Here, remote tuning of the magnetoelectric microenvironment is achieved modulate cell function via celk-material by a built-in CoFe, O4/poly(vinylidene fluoridetrifluoroethylene)[P(VDF-TrFE) nteractions H Recently, mimicking the magnetoelectric membrane for effective bone regeneration. The magneto. endogenous physical microenvironment electric microenvironment from the nanocomposite membranes promotes via implanted biomaterials has emerged as osteogenic differentiation of bone marrow mesenchymal stem cells(BI a new strategy for recapitulating the extr cellular microenvironment at the wound/ MSCs) and enhances bone defect regeneration by increasing cellular expo injury site to facilitate tissue regenera- sure and integrin binding to arginylglycylaspartic acid peptide, as predicte tion 2 Endogenous magnetoelectric prop- by molecular dynamics simulations. Moreover, BM-MSCs are directed to the erty is an integral aspect of the natural osteogenic lineage by osteoimmuomodulation which involves accelerating microenvironment of native transition from an initial inflammatory immune response to a pro- healing tissues l and extensive research has dem regenerative immune response. This work offers a strategy to mimic the onstrated that the osteogenic differentia ion, chondrogenic differentiation, PI and magnetoelectric microenvironment for achieving precise and effective tissue erogenic differentiation 6 of mesen- regenerative therapies, as well as provides fundamental insights into the chymal stem cell (MSC)can be modulated biological effects driven by the built-in magnetoelectric membrane, which can by the application of an external electro- be remotely tuned to precisely modulate osteogenesis in situ magnetic field. The magnetoelectric field provided by extracorporeal devices often Dr. W. Liu, Dr. F. Zhang, D Dr. Y. Huang, Prof Prof B. C. he NMPA Key Laboratory for l ials National E Central Laboratory Laboratory for Digital and Peking University School and Hospital of Stomatology Beijing 100081, P R China eking University School an f Stomatal Beijing 100081, P.R. China E- mail: kad Department of Materials Science and Engineering Dr Y Yan. Prof. X Wen The School of Chemistry and Chemical Engineering Beijing 100084, China South China University of Technology Guangzhou Prof. .Zhal Guangzhou 510640, P.R. China E-mail:xiwen@scut.edu.cn Dr. C. Zhang China cal Hospital Prof. L Chen Department of Stomatology Union Hospital Tongji Medical College Guangzhou 510280, China Huazhong University of Science and Technology he ORCID identification number(s)for the author(s)of this article Wuhan 430022, P. R. China canbefoundunderhttps://doi.org/0.1002/adfm.202006226 Do:10.002/adfm202006226 Adv Funct. Mater. 2020. 2006226 2006226of o 2020 Wiley-VCH GmbH

www.afm-journal.de 2006226 (1 of 11) © 2020 Wiley-VCH GmbH Full Paper Remote Tuning of Built-In Magnetoelectric Microenvironment to Promote Bone Regeneration by Modulating Cellular Exposure to Arginylglycylaspartic Acid Peptide Wenwen Liu, Fengyi Zhang, Yuanyang Yan, Chenguang Zhang, Han Zhao, Boon Chin Heng, Ying Huang, Yang Shen, Jinxing Zhang, Lili Chen, Xiufang Wen,* and Xuliang Deng* Mimicking the endogenous physical microenvironment is a promising strategy for biomaterial-mediated tissue regeneration. However, precise control of physical cues such as electric/magnetic fields within extracellular environments to facilitate tissue regeneration remains a formidable challenge. Here, remote tuning of the magnetoelectric microenvironment is achieved by a built-in CoFe2O4/poly(vinylidene fluoridetrifluoroethylene) [P(VDF-TrFE)] magnetoelectric membrane for effective bone regeneration. The magneto￾electric microenvironment from the nanocomposite membranes promotes osteogenic differentiation of bone marrow mesenchymal stem cells (BM￾MSCs) and enhances bone defect regeneration by increasing cellular expo￾sure and integrin binding to arginylglycylaspartic acid peptide, as predicted by molecular dynamics simulations. Moreover, BM-MSCs are directed to the osteogenic lineage by osteoimmuomodulation which involves accelerating transition from an initial inflammatory immune response to a pro-healing regenerative immune response. This work offers a strategy to mimic the magnetoelectric microenvironment for achieving precise and effective tissue regenerative therapies, as well as provides fundamental insights into the biological effects driven by the built-in magnetoelectric membrane, which can be remotely tuned to precisely modulate osteogenesis in situ. DOI: 10.1002/adfm.202006226 Dr. W. Liu, Dr. F. Zhang, Dr. H. Zhao, Dr. Y. Huang, Prof. X. Deng NMPA Key Laboratory for Dental Materials National Engineering Laboratory for Digital and Material Technology of Stomatology Department of Geriatric Dentistry Peking University School and Hospital of Stomatology Beijing 100081, P. R. China E-mail: kqdengxuliang@bjmu.edu.cn Dr. Y. Yan, Prof. X. Wen The School of Chemistry and Chemical Engineering South China University of Technology Guangzhou Guangzhou 510640, P. R. China E-mail: xfwen@scut.edu.cn Dr. C. Zhang Stomatological Hospital Southern Medical University Guangzhou 510280, China Prof. B. C. Heng Central Laboratory Peking University School and Hospital of Stomatology Beijing 100081, P. R. China Prof. Y. Shen State Key Laboratory of New Ceramics and Fine Processing Department of Materials Science and Engineering Tsinghua University Beijing 100084, China Prof. J. Zhang Department of Physics Beijing Normal University Beijing 100875, China Prof. L. Chen Department of Stomatology Union Hospital Tongji Medical College Huazhong University of Science and Technology Wuhan 430022, P. R. China 1. Introduction A major strategy in tissue engineering is to mimic the biophysical properties of target tissues and various critical aspects of the extracellular environment to modulate cell function via cell–material interactions.[1] Recently, mimicking the endogenous physical microenvironment via implanted biomaterials has emerged as a new strategy for recapitulating the extra￾cellular microenvironment at the wound/ injury site to facilitate tissue regenera￾tion.[2] Endogenous magnetoelectric prop￾erty is an integral aspect of the natural biophysical microenvironment of native tissues,[3] and extensive research has dem￾onstrated that the osteogenic differentia￾tion,[4] chondrogenic differentiation,[5] and neurogenic differentiation[6] of mesen￾chymal stem cell (MSC) can be modulated by the application of an external electro￾magnetic field. The magnetoelectric field provided by extracorporeal devices often The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/adfm.202006226. Adv. Funct. Mater. 2020, 2006226

SCENCNNEWS A www.advancedsciencenews.com ww.afm-journalde cannot work specifically and efficiently to achieve intended electrical charge being able to be controlled simultaneously via tissue regeneration. 7 Moreover, accurate control of the dosage application of a remote magnetic field, u thereby effectively pro- and effective working area of the external magnetoelectric viding a controllable local magnetoelectric environment. Cel timulation is difficult to achieve Is Therefore recapitulating lular behavior is known to be influenced by diurnal variations the natural magnetic and /or electric microenvironment at the in the magnetoelectric environment, which have been studied wound/injury site by implanted biomaterials is an alternative by simulating diurnal geomagnetic variations in geomagnetic strategy to modulate cellular and tissue biological processes for storms. 2 Hence, to mutually couple electric and magnetic achieving optimal bone regeneration outcome, and to precisely field effects and control their diurnal shifting are critical con control the in situ microenvironment for conducive tissue siderations for developing built-in biomaterials to mimic the regenerationI9 Biomaterials such as magnetic nanoparticles or in situ magnetoelectric microenvironment for the purpose of piezoelectric microfibers can separately provide the local mag. promoting tissue regeneration. Previous studies have shown letic or electrical field independently for tissue regeneration. sa that the surface potential of piezoelectric materials will typically Nevertheless, this limits the synergy of electrical and mag. decline by 40-50% after implantation. I] Through application of netic stimuli, which is necessary to optimize natural cellular a remote direct current(DC)magnetic field, the magnetoelec function. n1o ric materials can theoretically adjust and maintain the surface Magnetoelectric materials exhibit a unique combination of potential after implantation.(Figure la) Beside maintaining magnetism and electric polarization, with magnetization and the surface potential, the application of an external remote dC a Bone defect D MSCs RGD--FN-lIl . New bone K Osteogenic differentiation of MSCs e. M2 macrophage P/VDF- TrFE)matrix Negative charge stration of magnetoelectric effects and biological effects of CFO /P(VDF-TrFE) magnetoelectric nanocomposite membranes. a)Once a to the nanocomposite membrane, strain in the magnetostrictive phase is induced. ransmitted to the piezoelectric change in electrical polarization. b) Remote tuning of CFO/P(VDF-TrFE)magneto anocomposite membranes vi macrophage polarization, and Adw. Funct Mater. 2020. 2006226 2006226(2of1) g 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (2 of 11) © 2020 Wiley-VCH GmbH cannot work specifically and efficiently to achieve intended tissue regeneration.[7] Moreover, accurate control of the dosage and effective working area of the external magnetoelectric stimulation is difficult to achieve.[8] Therefore, recapitulating the natural magnetic and/or electric microenvironment at the wound/injury site by implanted biomaterials is an alternative strategy to modulate cellular and tissue biological processes for achieving optimal bone regeneration outcome, and to precisely control the in situ microenvironment for conducive tissue regeneration.[9] Biomaterials such as magnetic nanoparticles or piezoelectric microfibers can separately provide the local mag￾netic or electrical field independently for tissue regeneration.[9a] Nevertheless, this limits the synergy of electrical and mag￾netic stimuli, which is necessary to optimize natural cellular function.[10] Magnetoelectric materials exhibit a unique combination of magnetism and electric polarization, with magnetization and electrical charge being able to be controlled simultaneously via application of a remote magnetic field,[11] thereby effectively pro￾viding a controllable local magnetoelectric environment. Cel￾lular behavior is known to be influenced by diurnal variations in the magnetoelectric environment, which have been studied by simulating diurnal geomagnetic variations in geomagnetic storms.[12] Hence, to mutually couple electric and magnetic field effects and control their diurnal shifting are critical con￾siderations for developing built-in biomaterials to mimic the in situ magnetoelectric microenvironment for the purpose of promoting tissue regeneration. Previous studies have shown that the surface potential of piezoelectric materials will typically decline by 40–50% after implantation.[13] Through application of a remote direct current (DC) magnetic field, the magnetoelec￾tric materials can theoretically adjust and maintain the surface potential after implantation. (Figure 1a) Beside maintaining the surface potential, the application of an external remote DC Figure 1. Illustration of magnetoelectric effects and biological effects of CFO/P(VDF-TrFE) magnetoelectric nanocomposite membranes. a) Once a magnetic field is applied to the nanocomposite membrane, strain in the magnetostrictive phase is induced, which is transmitted to the piezoelectric constituent, leading to a change in electrical polarization. b) Remote tuning of CFO/P(VDF-TrFE) magnetoelectric nanocomposite membranes via application of an external DC magnetic field promotes bone regeneration by enhancing FN adsorption, RGD exposure, macrophage polarization, and osteogenic differentiation of MSCs. Adv. Funct. Mater. 2020, 2006226

SCENCENEWS m-journal de magnetic field can readily mimic the diurnal shifting magne- Table 1. Number of adsorbed residues of FN-l17-10 on membranes with toelectric environment. I4 Altogether, these studies point to different CFO contents(MD simulations) promising applications of magnetoelectric materials in tissue regeneration. In this study, we fabricated the CoFe2O4 (CFO)/ 5w:% ASN13,ASP15, THR16,CLY17,VAL18, LEU19,GLY46, ASN47,SER48, 44 poly(vinylidene fuoridetrifluoroethylene)[P(VDF-TrFE)) mag U49, GLU50, GLU51, VAL52, VAL53, HIS54 ASP56, GLN57, CYS60 netoelectric nanocomposite membranes, which can be regu THR61, PHE62, ASP63, ASN64,SER66, LEU69, ASP81 ASP122 lated by application of a remote DC magnetic field to generate LEU123, THR124, ASN125, GLU141, SER145, PRO146, SER147, ASP148 a built-in magnetoelectric microenvironment. Molecula TYR170, GLUI7I, GLN172, PRO289, PRO338, GLY339, ASP341 TYR366.ARG367.THR368 content of CFO nanoparticles for attaining the greatest argi- 10 wt% GLUT, ASN13, PRO14 ASPI5, HR16, GLYT/, VALIS, LEU19, THR20 nylglycylaspartic acid(RGD) sites exposure. The magneto- EU49, GLU51, VAL52, CAL53, HIS54, ALA55, ASP56, GLN57, SER59 electric microenvironment provided by the magnetoelectric CYS60, THR61, PHE62, ASP63, ASN64, LEU65, SER66, PRO67-ASP122 membranes can enhance osteogenesis and bone regenera GLU141SER145SER147ASP196ILE197.GLU223.GLY249THR250 tion within the bone defect area. The bone regeneration can GLU251, GLN271, SER273, THR274, VAL275, SER276ASP277, VAL278 PRO289, GLU312, THR313, GLY314, GLY315, ASN316, SER317, LYS337 not only be attributed to the direct osteogenic effect of bone RO338, GLY339, VAL340, ASP341, TYR366, ARG367, THR368 marrow mesenchymal stem cells(BM-MSCs), but also to the improved osteoimmunomodulatory microenvironment. 15 wt% ASP15, GLYI, VAL18, THR32, LEU49, GLU50,GLU51,VAL52,VAL53,31 HIS54, ALA55, ASP56, GLN57, SER58, SER59, CYS60, THR61, PHE62 The interaction between BM-MSCs and macrophages enhance ASP63, ASN64, ASP80, GLU137, GLU141, ASP277, PRO289, GLU312, bone repair by accelerating the transition from inflammatory immune response to regenerative immune response within the 20 wt% ASP56, CLN57 ASP63 ASP122, LU137, ASP138 GLU141 SER147, 19 bone defect areas( Figure 1b) GLLUI71, PRO289, THR313, GLY314, ASN316, PRO338, GLY33 VAL340, ASP341, ARG367, THR368 2. Results and Discussion RGD peptide is an archetypical ligand in the 10th type Ill 2.1. MD Simulation of RGD Exposure on CFO/P(VDF-TrFE) domain of FN, which mediates cell adhesion through specific anes interactions with various integrin receptors. [I5I The simulation predicted that the RGD sequence on 10 wt% CFO content mem- To predict the biological properties of the built-in magneto- branes were exposed towards the solution phase(Figure 2a) electric microenvironment provided by our fabricated magne- which are conducive for integrin binding. The root mean toelectric nanocomposite membrane, MD simulations were square deviation values!(0. 193 for 5 wt%, 0. 211 for 10 wt%, performed. The CFO/P(VDF-TrFE) magnetoelectric nanocom- 0.160 for 15 wt%, 0.148 for 20 wt%)of RGD indicated the high posite membrane was composed of ferromagnetic CFO nano- interdomain elasticity and flexibility of the RgD configuration particles embedded within a ferroelectric P(VDF-TrFE) matrix in the 10 wt% CFO content membranes. The enhanced RGD that provide excellent flexibility for the membrane. In this elasticity and flexibility could in turn facilitate RGD-integrin study, nanocomposite membranes with 5-20 wt% CFO content binding. RGD peptide was reported to facilitate cell spreading were selected for the MD simulations(Figure Sla, Supporting and motility, I stem cell differentiation, IS) and nanoparticle Information). Within the simulation models(Figure S1b, Sup- internalization, ib, 19 possibly through the assembly of clus porting Information), 10 wt% CFO content membranes exhib- ters of adhesion proteins. 120 Overall, the MD simulations indi ited the minimum distance between the FN-Ill7-10 protein cated that the 10 wt% CFO content membranes could enhance and nanocomposite membrane(Figure SIc, Supporting Infor- FN-I1I7-10 adsorption and optimize RGD domain exposure to mation). Additionally, the 10 wt% CFO content membranes promote cell adhesion. 21) According to previous studies, the exhibited strong interaction between the fibronectin (FN) FN adsorption capacity is an important extracellular environ nodule and the surface of the membrane, which possessed mental factor, which directly effects initial cell attachment and the greatest numbers of adsorbed FN-lll7-10 residues within proliferation, 2 Cells adhesion and spreading area was also 0.35 nm(Table 1). The enhanced FN-1ll7-10 protein adsorption positively correlated with increased surface RGD density. 2) The by 10 wt% CFO content membranes was also supported by the strongest van der Waal interactions and the highest value of Table 2. The interac interaction energy(Table 2). Under the influence of long-range different CFO contents MD By between FN-I117-10 and membranes with electrostatic interactions, the protein gradually moved to the surface. As the protein gets closer to the surface, the short. dw)(k)me ange van der Waals interaction also starts to affect the protein o 702725 481.366 221.359 Therefore, the protein continuously adjusts its conformation and finally adsorbs stably on the surface, under the synergistic effects of van der waals and electrostatic interactions, There. 15%6 701.053 488.772 212.281 fore in our simulation, 10 wt% CFo content membranes were 20 wt% -543.027 134935 predicted to have the most favorable surface for FN-1117-10 pro- E interaction energy: bEee electrostatic interaction energy, E van de tein adsorption Adv Funct. Mater. 2020. 2006226 2006226日3of1) o 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (3 of 11) © 2020 Wiley-VCH GmbH magnetic field can readily mimic the diurnal shifting magne￾toelectric environment.[14] Altogether, these studies point to promising applications of magnetoelectric materials in tissue regeneration. In this study, we fabricated the CoFe2O4 (CFO)/ poly(vinylidene fluoridetrifluoroethylene) [P(VDF-TrFE)] mag￾netoelectric nanocomposite membranes, which can be regu￾lated by application of a remote DC magnetic field to generate a built-in magnetoelectric microenvironment. Molecular dynamics (MD) simulation was used to predict the optimal content of CFO nanoparticles for attaining the greatest argi￾nylglycylaspartic acid (RGD) sites exposure. The magneto￾electric microenvironment provided by the magnetoelectric membranes can enhance osteogenesis and bone regenera￾tion within the bone defect area. The bone regeneration can not only be attributed to the direct osteogenic effect of bone marrow mesenchymal stem cells (BM-MSCs), but also to the improved osteoimmunomodulatory microenvironment. The interaction between BM-MSCs and macrophages enhance bone repair by accelerating the transition from inflammatory immune response to regenerative immune response within the bone defect areas. (Figure 1b) 2. Results and Discussion 2.1. MD Simulation of RGD Exposure on CFO/P(VDF-TrFE) Composite Membranes To predict the biological properties of the built-in magneto￾electric microenvironment provided by our fabricated magne￾toelectric nanocomposite membrane, MD simulations were performed. The CFO/P(VDF-TrFE) magnetoelectric nanocom￾posite membrane was composed of ferromagnetic CFO nano￾particles embedded within a ferroelectric P(VDF-TrFE) matrix that provide excellent flexibility for the membrane. In this study, nanocomposite membranes with 5–20 wt% CFO content were selected for the MD simulations (Figure S1a, Supporting Information). Within the simulation models (Figure S1b, Sup￾porting Information), 10 wt% CFO content membranes exhib￾ited the minimum distance between the FN-III7-10 protein and nanocomposite membrane (Figure S1c, Supporting Infor￾mation). Additionally, the 10 wt% CFO content membranes exhibited strong interaction between the fibronectin (FN) module and the surface of the membrane, which possessed the greatest numbers of adsorbed FN-III7-10 residues within 0.35 nm (Table 1). The enhanced FN-III7-10 protein adsorption by 10 wt% CFO content membranes was also supported by the strongest van der Waal interactions and the highest value of interaction energy (Table 2). Under the influence of long-range electrostatic interactions, the protein gradually moved to the surface. As the protein gets closer to the surface, the short￾range van der Waals interaction also starts to affect the protein. Therefore, the protein continuously adjusts its conformation and finally adsorbs stably on the surface, under the synergistic effects of van der Waals and electrostatic interactions. There￾fore in our simulation, 10 wt% CFO content membranes were predicted to have the most favorable surface for FN-III7-10 pro￾tein adsorption. RGD peptide is an archetypical ligand in the 10th type III domain of FN, which mediates cell adhesion through specific interactions with various integrin receptors.[15] The simulation predicted that the RGD sequence on 10 wt% CFO content mem￾branes were exposed towards the solution phase (Figure 2a), which are conducive for integrin binding. The root mean square deviation values[16] (0.193 for 5 wt%, 0.211 for 10 wt%, 0.160 for 15 wt%, 0.148 for 20 wt%) of RGD indicated the high interdomain elasticity and flexibility of the RGD configuration in the 10 wt% CFO content membranes. The enhanced RGD elasticity and flexibility could in turn facilitate RGD-integrin binding. RGD peptide was reported to facilitate cell spreading and motility,[17] stem cell differentiation,[18] and nanoparticle internalization,[17b,19] possibly through the assembly of clus￾ters of adhesion proteins.[20] Overall, the MD simulations indi￾cated that the 10 wt% CFO content membranes could enhance FN-III7-10 adsorption and optimize RGD domain exposure to promote cell adhesion.[21] According to previous studies, the FN adsorption capacity is an important extracellular environ￾mental factor, which directly effects initial cell attachment and proliferation.[22] Cells adhesion and spreading area was also positively correlated with increased surface RGD density.[23] The Table 1. Number of adsorbed residues of FN-III7-10 on membranes with different CFO contents (MD simulations). Surface Adsorbed residues Total 5 wt% ASN13,ASP15,THR16,GLY17,VAL18,LEU19,GLY46,ASN47,SER48, LEU49,GLU50,GLU51,VAL52,VAL53,HIS54,ASP56,GLN57,CYS60, THR61,PHE62,ASP63,ASN64,SER66,LEU69,ASP81,ASP122, LEU123,THR124,ASN125,GLU141,SER145,PRO146,SER147,ASP148, TYR170,GLU171,GLN172,PRO289,PRO338,GLY339,ASP341, TYR366,ARG367,THR368 44 10 wt% GLU11,ASN13,PRO14,ASP15,THR16,GLY17,VAL18,LEU19,THR20, LEU49,GLU51,VAL52,CAL53,HIS54,ALA55,ASP56,GLN57,SER59, CYS60,THR61,PHE62,ASP63,ASN64,LEU65,SER66,PRO67,ASP122, GLU141,SER145,SER147,ASP196,ILE197,GLU223,GLY249,THR250, GLU251,GLN271,SER273,THR274,VAL275,SER276,ASP277,VAL278, PRO289,GLU312,THR313,GLY314,GLY315,ASN316,SER317,LYS337, PRO338,GLY339,VAL340,ASP341,TYR366,ARG367,THR368 58 15 wt% ASP15,GLY17,VAL18,THR32,LEU49,GLU50,GLU51,VAL52,VAL53, HIS54,ALA55,ASP56,GLN57,SER58,SER59,CYS60,THR61,PHE62, ASP63,ASN64,ASP80,GLU137,GLU141,ASP277,PRO289,GLU312, PRO338,GLY339,ASP341,ARG367,THR368 31 20 wt% ASP56,GLN57,ASP63,ASP122,GLU137,ASP138,GLU141,SER147, GLLU171,PRO289,THR313,GLY314,ASN316,PRO338,GLY339, VAL340,ASP341,ARG367,THR368 19 Table 2. The interaction energy between FN-III7-10 and membranes with different CFO contents (MD simulations). Surface Einta) [kJ mol−1 ] Eeleb) [kJ mol−1 ] Evdwc) [kJ mol−1 ] 5 wt% −702.725 −481.366 −221.359 10 wt% −773.174 −431.619 −341.555 15 wt% −701.053 −488.772 −212.281 20 wt% −543.027 −408.092 −134.935 a)Eint, interaction energy; b)Eele, electrostatic interaction energy; c)Evdw, van der Waals interaction energy. Adv. Funct. Mater. 2020, 2006226

SCENCNNEWS www.advancedsciencenews.com 10wt% RGD d The range of v o Immersion with DC magnetic field o Immersion without DC magnetic field steoinductive surface 士壬 MD simulation and characterization of the CFO/P(VDF-TrFE) magnetoelectric nanocomposite membrane a) Snapshots of the simulations, showing that the RGD site was oriented towards the solution phas he 10 wt% CFO/P(VDF-TrF )SEM images of 10 wt% CFO/P(VDF-TrFE)membranes. c)The magnetic-field-induced surface potential of CFO/P(VDF- TrFE)me ferent CFO content. Vr ctrical communication-endogenous voltage gradients across the plasma membrane. ( Typical nv). 4al d) Zeta potential of 10 wt% CFO/P(VDF- TrFE)membranes without immersion and membranes immersed in culture mediu exposure to a remote DC magnetic field after 1, 7, 14 days increased FN adsorption which lead to increased RGD expo. CFO nanoparticles content(Figure S2e, Supporting Informa- ure can facilitate integrin binding to enhance focal adhesion tion). In hysteresis loop tests, the maximum magnetization (FA) formation and effect MSC and macrophage responses to value of the different composite membranes was approximately biomaterials. 16. 4 These MD simulations predict that the mag. proportional to the amount of CFO nanoparticles within the matrix( Figure S2f, Supporting Information).The netoelectric microenvironment provided by 10 wt% CFO con- P(VDF-TrE y 594 9 tent composite membranes could induce optimal biological magnetoelectric effect of the nanocomposite membranes is fte due to an elastic coupling interaction between electrical polari- zation and particle concentration within the co-polymer 2.2. Characterizing the Magnetoelectric Microenvironment matrix might favor arrangement of polar conformations and Provided by the CFO/P(VDF-TrFE)Nanocomposite Membranes therefore lead to enhanced ferroelectric and piezoelectric responses. 2 Based on the MD simulation, the CFO/P(VDF-TrFE)magneto. We next evaluated the magnetoelectric effects of membranes. electric nanocomposite membranes with different CFO nano- The magnetoelectric effect, defined as the variation of the elec- article contents(5, 10, 15, 20 wt%)( Figure S2a, Supporting trical polarization of a material in the presence of an applied Information) were fabricated. Scanning electron microscope magnetic field, or as the induced magnetization in the presence (SEM) imaging revealed extensive agglomeration of CFo of an applied electric field, can be seen as a bridge between anoparticles at contents >15 wt%( Figure 2b; Figure S2b, the electric and magnetic properties of matter. I The results upporting Information). Homogenous dispersion of CFo demonstrated that 10 wt% CFO content membranes exhib- anoparticles within the piezoelectric matrix is a key prereg. ited the largest range of magnetic-field-induced surface poten- isite for achieving a significant magnetoelectric effect 25) tial(Figure 2c) among all groups after corona poling at room With increasing content of CFO nanoparticles, the content of temperature. The magnetic-field-induced surface potential of the B-phase within the P(VDF-TrFE) matrix decreases, as evi- 10 wt% CFO content membranes could be tuned from0 to denced by Fourier transform infrared spectroscopy imaging 91.15 mv by increasing the remote DC magnetic field from (Figure S2c, Supporting Information) and X-ray diffraction 0 to 3000 Oe. Our preliminary results showed that the surface patterns(Figure S2d, Supporting Information). The B-phase potential of around 54 mV is most favorable for osteogenesis 126 is closely correlated with piezoelectric properties, as it is an However, the retention period of the surface potential induced electrically active phase. 26 Consistent with the B-phase data, by piezoelectric materials is not sufficient for optimal osteo- the piezoelectric coefficients (d33) decreased with increasing genesis. For the magnetoelectric composites, the surface Adw. Funct Mater. 2020. 2006226 20062264of) g 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (4 of 11) © 2020 Wiley-VCH GmbH increased FN adsorption which lead to increased RGD expo￾sure can facilitate integrin binding to enhance focal adhesion (FA) formation and effect MSC and macrophage responses to biomaterials.[16,24] These MD simulations predict that the mag￾netoelectric microenvironment provided by 10 wt% CFO con￾tent composite membranes could induce optimal biological effects. 2.2. Characterizing the Magnetoelectric Microenvironment Provided by the CFO/P(VDF-TrFE) Nanocomposite Membranes Based on the MD simulation, the CFO/P(VDF-TrFE) magneto￾electric nanocomposite membranes with different CFO nano￾particle contents (5, 10, 15, 20 wt%) (Figure S2a, Supporting Information) were fabricated. Scanning electron microscope (SEM) imaging revealed extensive agglomeration of CFO nanoparticles at contents >15  wt% (Figure  2b; Figure S2b, Supporting Information). Homogenous dispersion of CFO nanoparticles within the piezoelectric matrix is a key prereq￾uisite for achieving a significant magnetoelectric effect.[25] With increasing content of CFO nanoparticles, the content of the β-phase within the P(VDF-TrFE) matrix decreases, as evi￾denced by Fourier transform infrared spectroscopy imaging (Figure S2c, Supporting Information) and X-ray diffraction patterns (Figure S2d, Supporting Information). The β-phase is closely correlated with piezoelectric properties, as it is an electrically active phase.[26] Consistent with the β-phase data, the piezoelectric coefficients (d33) decreased with increasing CFO nanoparticles content (Figure S2e, Supporting Informa￾tion). In hysteresis loop tests, the maximum magnetization value of the different composite membranes was approximately proportional to the amount of CFO nanoparticles within the P(VDF-TrFE) matrix (Figure S2f, Supporting Information). The magnetoelectric effect of the nanocomposite membranes is due to an elastic coupling interaction between electrical polari￾zation and magnetostrictive components.[27] Hence, appro￾priate CFO nanoparticle concentration within the co-polymer matrix might favor arrangement of polar conformations and therefore lead to enhanced ferroelectric and piezoelectric responses.[28] We next evaluated the magnetoelectric effects of membranes. The magnetoelectric effect, defined as the variation of the elec￾trical polarization of a material in the presence of an applied magnetic field, or as the induced magnetization in the presence of an applied electric field, can be seen as a bridge between the electric and magnetic properties of matter.[11] The results demonstrated that 10 wt% CFO content membranes exhib￾ited the largest range of magnetic-field-induced surface poten￾tial (Figure  2c) among all groups after corona poling at room temperature. The magnetic-field-induced surface potential of 10 wt% CFO content membranes could be tuned from 0 to 91.15  mV by increasing the remote DC magnetic field from 0 to 3000 Oe. Our preliminary results showed that the surface potential of around 54 mV is most favorable for osteogenesis.[26] However, the retention period of the surface potential induced by piezoelectric materials is not sufficient for optimal osteo￾genesis.[13] For the magnetoelectric composites, the surface Figure 2. MD simulation and characterization of the CFO/P(VDF-TrFE) magnetoelectric nanocomposite membrane a) Snapshots of the molecular dynamic simulations, showing that the RGD site was oriented towards the solution phase on the 10 wt% CFO/P(VDF-TrFE) magnetoelectric mem￾brane. b) SEM images of 10 wt% CFO/P(VDF-TrFE) membranes. c) The magnetic-field-induced surface potential of CFO/P(VDF-TrFE) membranes with different CFO content. Vmem: Bioelectrical communication-endogenous voltage gradients across the plasma membrane. (Typical value: −60 to −100 mV).[48] d) Zeta potential of 10 wt% CFO/P(VDF-TrFE) membranes without immersion and membranes immersed in culture medium with or without exposure to a remote DC magnetic field after 1, 7, 14 days. Adv. Funct. Mater. 2020, 2006226

SCENCENEWS m-journal de potential could be sustained for as long as required by applying of the cell-binding domain of RGD on adsorbed Fn was fur- a magnetic field in a non-contact manner. For instance, the ther evaluated. On CFO 10-E/M, the RGD site combined with nanocomposite membrane with 10 wt% of CFO content has the HFNZ1 antibody 52 displayed the highest intensity and big appropriate concentration for optimal magnetoelectric effect. gest area among all groups( Figure S4d, Supporting Informa The surface potential of CFO/P(VDF-TrFE)membranes were tion). These results thus showed that CFO 10-E/M can enhance decreased after 14 days immersion in culture medium without the FN adsorption and RGD site exposure. The enhanced FN application of a remote DC magnetic field( Figure 2d). The adsorption and RGD exposure could promote cell adhesion, surface potential of 10 wt% CFO content membranes could migration, and spreading. 245) Consistent with the prediction be tuned to around 54 mV with a remote DC 23000e mag. of MD simulation, the magnetoelectric microenvironment pre netic field, which were utilized for subsequent biological vided by CFo 10-E/M can modulate cellular behavior assays(Figure 2c). Upon application of a remote DC mag. The biological performance of the CFO/P(VDF-TrFE)mag- netic field, the surface potential of the membranes were main- netoelectric nanocomposite membranes was further investi tained throughout 14 days of immersion in culture medium gated. Assessment of proliferation through the cell counting Figure 2d). Since the diurnal geomagnetic variation has a pro. kit-8(CCK-8 )assay demonstrated that there were no significar nounced circadian rhythmicity that exerts biological effects, 4 differences between groups(Figure S4e, f, Supporting Infor a remote DC magnetic field mimicking the natural magnetic mation). However, the cytoskeleton immunostaining images field with 12 h shifting was selected to control the magneto- showed that BM-MSCs on CFO 10-E/M exhibited the largest electric microenvironment in this study. The surface rough- spreading area and the most abundant FA formation among all ness and water contact angle measurements yielded different groups( Figure S5a-f, Supporting Information). The increased values according to the different nanoparticle contents of the FA and spreading area are correlated with enhanced RGD expo- various groups( Figure S3a, b, Supporting Information). Some sure, and might also be correlated to the osteogenic potential of researchers have reported that roughness difference, which is BM-MSCs 1.3 a microroughness level of around 100-500 nm, exert negligible effects on the cell adhesion and biological functions of larger cell types, such as BM-MSCs, osteoblasts, and neurons. 2 2.4. Biomimetic Magnetoelectric Microenvironment Enhances The surface wettability study also showed that there is no dif- Osteogenesis of BM-MSCs through RGD Exposure-Mediated ference in cell adhesion when the surface contact angles are Cellular Mechanotransduction round 60 to 80 degrees after 180 min of incubation. ol Hence, the roughness and contact angle values differences between The pro-osteogenic effects of the magnetoelectric microen- groups should have no biological effects in this study. Tensile vironment provided by CFO/P(VDF-TrFE) membranes were strength and the elastic modulus of membranes decreased evaluated with BM-MSCs, under exposure to a remote DC along with increased CFO concentrations. However, based on magnetic field. The quantitative analysis of alizarin red staining our previous study, changes to the elastic modulus and tensile demonstrated significantly higher mineralization values strength in this scale will have negligible effects on osteogen- CFO 10-E/M compared to membranes with different CFO s31(Figure S3c, d, Supporting Information). Moreover, the contents(Figure S6a, Supporting Information). CFO 10-E/M tensile strength and elastic modulus values showed that the exhibited the strongest fluorescence intensity of RUNX2 immu- fexibility of the membranes could be suitable for clinical appli- nostaining after 3 days of incubation(Figure 3a; Figure S6d, cations, such as being utilized as non-resorbable membranes Supporting Information). Osteogenic gene expression and pro- to prevent the ingrowth of non-osteogenic tissues and main- tein production were further evaluated and the results showed tain the capacity for cell occlusion 13,31 After corona poling, that CFO 10-E/M upregulated the expression of osteogenesis 10 wt% CFO content membranes combined with application of related genes(RUNX2, ALP, Collagen I, OCN, OSX)and a remote DC magnetic field can provide an optimal controllable proteins(OPN, BMP2, RUNX2) in the absence of osteogenic magnetoelectric microenvironment. supplements( Figure 3b: Figure S6b, c, f, Supporting Informa tion). After 7 days of culture without osteogenic supplements. higher ALP activity was detected on CFO 10-E/M(Figure S6e 2.3. Biomimetic Magnetoelectric Microenvironment Promotes Supporting Information). These results thus demonstrated that FN Adsorption and RGD Exposure That Leads to Enhanced CFO 10-E/M provides an optimal magnetoelectric microenv Cell Adhesion ronment for BM-MSC osteogenic differentiation To explore the mechanisms by which osteogenesis of BM In biological assays, the membranes were grouped as follows: MSCs is promoted by enhanced FN adsorption and RGD expo CFO 0, CFO 5, CFO 10, CFO 15, and CFO 20, which represent sure in the CFO 10-E/M group, the expression of integrin was 0, 5, 10, 15, and 20 wt% of CFO content within membranes, analyzed and it was found that the gene expression and pro- represents group exposed to up with polarization treatment. M tein production levels of integrin as were significantly high a remote DC magnetic field. Nc in CFO 10-E/M versus other groups(Figure 3c, d; Figure S7a, represents blank group without membrane Supporting Information). Integrin a5 has been reported to par To assay FN adsorption in vitro, quantitative evaluation and ticipate in MSC osteogenesis. 242, 34 Integrins connect the cell immunofluorescence images were used. The results dem. cytoskeleton to the microenvironment and serve as sensors of onstrated that CFO 10-E/M had the greatest FN adsorption mechanical signals. 35 Increased integrin clustering can lead capacity(Figure S4a-C, Supporting Information). The exposure to increased adhesion maturation, which in turn can modulate Adv Funct. Mater. 2020. 2006226 20062265of1) o 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (5 of 11) © 2020 Wiley-VCH GmbH potential could be sustained for as long as required by applying a magnetic field in a non-contact manner. For instance, the nanocomposite membrane with 10 wt% of CFO content has the appropriate concentration for optimal magnetoelectric effect. The surface potential of CFO/P(VDF-TrFE) membranes were decreased after 14 days immersion in culture medium without application of a remote DC magnetic field (Figure  2d). The surface potential of 10 wt% CFO content membranes could be tuned to around 54  mV  with a remote DC 2300Oe  mag￾netic field, which were utilized for subsequent biological assays (Figure  2c). Upon application of a remote DC mag￾netic field, the surface potential of the membranes were main￾tained throughout 14 days of immersion in culture medium (Figure 2d). Since the diurnal geomagnetic variation has a pro￾nounced circadian rhythmicity that exerts biological effects,[14] a remote DC magnetic field mimicking the natural magnetic field with 12  h shifting was selected to control the magneto￾electric microenvironment in this study. The surface rough￾ness and water contact angle measurements yielded different values according to the different nanoparticle contents of the various groups (Figure S3a,b, Supporting Information). Some researchers have reported that roughness difference, which is a microroughness level of around 100–500 nm, exert negligible effects on the cell adhesion and biological functions of larger cell types, such as BM-MSCs, osteoblasts, and neurons.[29] The surface wettability study also showed that there is no dif￾ference in cell adhesion when the surface contact angles are around 60 to 80 degrees after 180 min of incubation.[30] Hence, the roughness and contact angle values differences between groups should have no biological effects in this study. Tensile strength and the elastic modulus of membranes decreased along with increased CFO concentrations. However, based on our previous study, changes to the elastic modulus and tensile strength in this scale will have negligible effects on osteogen￾esis[13] (Figure S3c,d, Supporting Information). Moreover, the tensile strength and elastic modulus values showed that the flexibility of the membranes could be suitable for clinical appli￾cations, such as being utilized as non-resorbable membranes to prevent the ingrowth of non-osteogenic tissues and main￾tain the capacity for cell occlusion.[13,31] After corona poling, 10 wt% CFO content membranes combined with application of a remote DC magnetic field can provide an optimal controllable magnetoelectric microenvironment. 2.3. Biomimetic Magnetoelectric Microenvironment Promotes FN Adsorption and RGD Exposure That Leads to Enhanced Cell Adhesion In biological assays, the membranes were grouped as follows: CFO 0, CFO 5, CFO 10, CFO 15, and CFO 20, which represent 0, 5, 10, 15, and 20 wt% of CFO content within membranes, respectively. E represents group with polarization treatment. M represents group exposed to a remote DC magnetic field. NC represents blank group without membrane. To assay FN adsorption in vitro, quantitative evaluation and immunofluorescence images were used. The results dem￾onstrated that CFO 10-E/M had the greatest FN adsorption capacity (Figure S4a–c, Supporting Information). The exposure of the cell-binding domain of RGD on adsorbed FN was fur￾ther evaluated. On CFO 10-E/M, the RGD site combined with HFN7.1 antibody[32] displayed the highest intensity and big￾gest area among all groups (Figure S4d, Supporting Informa￾tion). These results thus showed that CFO 10-E/M can enhance the FN adsorption and RGD site exposure. The enhanced FN adsorption and RGD exposure could promote cell adhesion, migration, and spreading.[24b] Consistent with the prediction of MD simulation, the magnetoelectric microenvironment pro￾vided by CFO 10-E/M can modulate cellular behavior. The biological performance of the CFO/P(VDF-TrFE) mag￾netoelectric nanocomposite membranes was further investi￾gated. Assessment of proliferation through the cell counting kit-8 (CCK-8) assay demonstrated that there were no significant differences between groups (Figure S4e,f, Supporting Infor￾mation). However, the cytoskeleton immunostaining images showed that BM-MSCs on CFO 10-E/M exhibited the largest spreading area and the most abundant FA formation among all groups (Figure S5a–f, Supporting Information). The increased FA and spreading area are correlated with enhanced RGD expo￾sure, and might also be correlated to the osteogenic potential of BM-MSCs.[13,33] 2.4. Biomimetic Magnetoelectric Microenvironment Enhances Osteogenesis of BM-MSCs through RGD Exposure-Mediated Cellular Mechanotransduction The pro-osteogenic effects of the magnetoelectric microen￾vironment provided by CFO/P(VDF-TrFE) membranes were evaluated with BM-MSCs, under exposure to a remote DC magnetic field. The quantitative analysis of alizarin red staining demonstrated significantly higher mineralization values in CFO 10-E/M compared to membranes with different CFO contents (Figure S6a, Supporting Information). CFO 10-E/M exhibited the strongest fluorescence intensity of RUNX2 immu￾nostaining after 3 days of incubation (Figure 3a; Figure S6d, Supporting Information). Osteogenic gene expression and pro￾tein production were further evaluated and the results showed that CFO 10-E/M upregulated the expression of osteogenesis￾related genes (RUNX2, ALP, Collagen I, OCN, OSX) and proteins (OPN, BMP2, RUNX2) in the absence of osteogenic supplements (Figure  3b; Figure S6b,c,f, Supporting Informa￾tion). After 7 days of culture without osteogenic supplements, higher ALP activity was detected on CFO 10-E/M (Figure S6e, Supporting Information). These results thus demonstrated that CFO 10-E/M provides an optimal magnetoelectric microenvi￾ronment for BM-MSC osteogenic differentiation. To explore the mechanisms by which osteogenesis of BM￾MSCs is promoted by enhanced FN adsorption and RGD expo￾sure in the CFO 10-E/M group, the expression of integrin was analyzed and it was found that the gene expression and pro￾tein production levels of integrin α5 were significantly higher in CFO 10-E/M versus other groups (Figure  3c,d; Figure S7a, Supporting Information). Integrin α5 has been reported to par￾ticipate in MSC osteogenesis.[24a,34] Integrins connect the cell cytoskeleton to the microenvironment and serve as sensors of mechanical signals.[35] Increased integrin clustering can lead to increased adhesion maturation, which in turn can modulate Adv. Funct. Mater. 2020, 2006226

SCENCNNEWS www.advancedsciencenews.com ww.afm-journalde a「co10 SEMCO10E[coc CFOCFOCFO RK CFO 10-E b 20〓cFo10-/Mmco10E■ CFO O-E■ 平 Day 3 ALP Day 3 RUNX2 Day 14 Collagen Day 14RUNX2 cFo10E/McFo10E■cFo0E■Nc MAPKI MAPK3 Figure 3. Osteoinductive potential of the CFO/P(VDF-TrFE) magnetoelectric nanocomposite membrane and the potential mechanisms by which induces osteogenesis in vitro. a)Immunofluorescence images showed the upregulated expression of RUNX2 in the CFo 10-E/M group 100 um). b)RT-qPCR revealed the upregulation of osteogenic markers(ALP, RUNX2, and Collagen I)in the CFO 10-E/M the upregulation of Itas, FN, MAPK], MAPK3, FAK, and YAP in the CFO 10-E/M group. (*VS CFO 10-E/M, VS NC, p<0.05)d) Western bl indicated that the CFO 10-E/M group exhibited upregulated expression of integrin as/ERK1/2 cascade- related proteins and increased YAP activation e)Immunofluorescence staining showed enhanced nuclear localization of YAP in the CFO 10-E/M group(Scale bars: 50 um) cell downstream signaling proteins such as ERK and nuclear phosphorylated- YAP(Figure 3d), which were consistent with localization mechanosensitive transcription regulators like the immunostaining results. Therefore, we propose that th YAP/TAZ. 565) Next, we evaluated the activation of ERK signaling magnetoelectric microenvironment provided by CFO 10-E/M and YAP nuclear localization to characterize the intracellular induces BM-MSC osteogenic differentiation by increasing RGD signaling pathways. The CFO 10-E/M membrane markedly exposure and initiating sensors of biomechanical signaling, upregulated the gene expression levels of ERK1/2, which were which in turn promotes FA maturation and trigger a series of consistent with increased protein expression(Figure 3c, d ). mechanotransduction-related molecular processes, involving Immunofluorescence staining showed that YAP is concen- the activation of ERK signaling and nuclear localization of YAP trated within the cell nuclei on CFO 10-E/M(Figure 3e). The statistical analysis of fluorescence intensity ratios between the cell's nucleus and cytoplasm demonstrated that nuclear locali- 2.5. Biomimetic Magnetoelectric Microenvironment Modulates zation of YAP is significantly higher on CFO 10-E/M versus the Osteoimmunomodulatory Responses In Vitro other membrane groups(Figure S7b, Supporting Information The protein expression results showed that BM-MSCs cul- Bone regeneration is a highly complex process which involves tured on CFo 10-E/M displayed increased YAP but decreased various different cell types such as ne cells, progenitor Adw. Funct Mater. 2020. 2006226 2006226(6of1) g 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (6 of 11) © 2020 Wiley-VCH GmbH cell downstream signaling proteins such as ERK and nuclear localization mechanosensitive transcription regulators like YAP/TAZ.[36] Next, we evaluated the activation of ERK signaling and YAP nuclear localization to characterize the intracellular signaling pathways. The CFO 10-E/M membrane markedly upregulated the gene expression levels of ERK1/2, which were consistent with increased protein expression (Figure  3c,d). Immunofluorescence staining showed that YAP is concen￾trated within the cell nuclei on CFO 10-E/M (Figure  3e). The statistical analysis of fluorescence intensity ratios between the cell’s nucleus and cytoplasm demonstrated that nuclear locali￾zation of YAP is significantly higher on CFO 10-E/M versus the other membrane groups (Figure S7b, Supporting Information). The protein expression results showed that BM-MSCs cul￾tured on CFO 10-E/M displayed increased YAP but decreased phosphorylated-YAP (Figure  3d), which were consistent with the immunostaining results. Therefore, we propose that the magnetoelectric microenvironment provided by CFO 10-E/M induces BM-MSC osteogenic differentiation by increasing RGD exposure and initiating sensors of biomechanical signaling, which in turn promotes FA maturation and trigger a series of mechanotransduction-related molecular processes, involving the activation of ERK signaling and nuclear localization of YAP. 2.5. Biomimetic Magnetoelectric Microenvironment Modulates Osteoimmunomodulatory Responses In Vitro Bone regeneration is a highly complex process which involves various different cell types such as immune cells, progenitor Figure 3. Osteoinductive potential of the CFO/P(VDF-TrFE) magnetoelectric nanocomposite membrane and the potential mechanisms by which it induces osteogenesis in vitro. a) Immunofluorescence images showed the upregulated expression of RUNX2 in the CFO 10-E/M group (Scale bars: 100 µm). b) RT-qPCR revealed the upregulation of osteogenic markers (ALP, RUNX2, and Collagen I) in the CFO 10-E/M group. c) RT-qPCR revealed the upregulation of Itgα5, FN, MAPK1, MAPK3, FAK, and YAP in the CFO 10-E/M group. (* VS CFO 10-E/M, # VS NC, p < 0.05) d) Western blot analysis indicated that the CFO 10-E/M group exhibited upregulated expression of integrin α5/ERK1/2 cascade-related proteins and increased YAP activation. e) Immunofluorescence staining showed enhanced nuclear localization of YAP in the CFO 10-E/M group (Scale bars: 50 µm). Adv. Funct. Mater. 2020, 2006226

SCENCENEWS m-journal de ells, and MSC. I/ Macrophages are the first cell type to arrive at Supporting Information). These elevated integrin expression the bone defect site, and have long been thought to contribute levels on CFO 10-E/M indicated that the magnetoelectric micro- to the initial inflammation and debridement of the injury loca- environment could enhance macrophage polarization through zation is a key regulator of bone regeneration. /g o phage polari- integrin-related pathways. The western blot results showed that the protein expression levels of integrin pl, phosphoryla To further investigate the osteoimmunomodulatory effects tion levels of phosphoinositide 3-kinase(PI3K) and serine/ of the magnetoelectric microenvironment, we evaluated the threonine kinase Akt(Akt) were increased in the CFO 10-E/M polarization of macrophages and osteogenic differentiation group(Figure S12a, Supporting Information). The CFO 10-E/M of BM-MSCs within a co-culture system. Macrophages were group exhibited the strongest fluorescence intensity of phos eeded on the nanocomposite membranes, while BM-Msc phorylated Akt immunostaining( Figure S12b, Supporting were seeded in the upper chambers of transwell culture dishes. Information). Furthermore, the western blot results showed Macrophages cultured on CFO 10-E/M expressed high expres- that the CFO 10-E/M group had significantly reduced nuclear sion levels of CD206, which is the M2 marker(Figure 4a, b; Fig. factor kappa B(NF-xB/p65)levels compared to other groups ures S8a and S1a, Supporting Information). Meanwhile, there (Figure S12a, Supporting Information). Protein adsorption and is no significant change in the expression of the MI marker integrin binding interactions have been demonstrated to modu CCR7(Figure S8b, c, Supporting Information) by macrophages late inflammation. 24C.32. 4) On CFO 10-E/M, the enhanced FN cultured on CFO 10-E/M, as compared with the other groups. adsorption resulted in adoption of an active conformation that The results proved that CFO 10-E/M could promote M2 polari- led to more RGD binding site exposure and increased macro- zation of macrophages. The immunofluorescence staining phage integrin Bl binding Increased integrin Bl binding acti- showed that expression levels of integrin a5, integrin Bl, and vates PI3K/Akt signaling, leading to the inhibition of NF-xB vinculin in macrophages were increased( Figures S9-S11, activation and subsequent anti-inflammatory polarization(M2) CFO 0- cFo10E/M‖cFo10E CFO O-E 目喝 叫 CFO 10-E 56.0 38.2 35.4 =:o10E1=e Day14 RUNX2 , o10E/M■co0E CFO 10-E Macrophage polarization and osteogenesis of BM-MSC on CFO/P(VDF-TrFE) magnetoelectric nanocomposite membranes in vitro. a)Immu- #f VS NC, p<0.05) Adv Funct. Mater. 2020. 2006226 2006226卩7of1) o 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (7 of 11) © 2020 Wiley-VCH GmbH cells, and MSC.[37] Macrophages are the first cell type to arrive at the bone defect site, and have long been thought to contribute to the initial inflammation and debridement of the injury loca￾tion.[38] Mounting evidence has shown that macrophage polari￾zation is a key regulator of bone regeneration.[39] To further investigate the osteoimmunomodulatory effects of the magnetoelectric microenvironment, we evaluated the polarization of macrophages and osteogenic differentiation of BM-MSCs within a co-culture system. Macrophages were seeded on the nanocomposite membranes, while BM-MSC were seeded in the upper chambers of transwell culture dishes. Macrophages cultured on CFO 10-E/M expressed high expres￾sion levels of CD206, which is the M2 marker (Figure 4a,b; Fig￾ures S8a and S13a, Supporting Information). Meanwhile, there is no significant change in the expression of the M1 marker CCR7 (Figure S8b,c, Supporting Information) by macrophages cultured on CFO 10-E/M, as compared with the other groups. The results proved that CFO 10-E/M could promote M2 polari￾zation of macrophages. The immunofluorescence staining showed that expression levels of integrin α5, integrin β1, and vinculin in macrophages were increased (Figures S9–S11, Supporting Information). These elevated integrin expression levels on CFO 10-E/M indicated that the magnetoelectric micro￾environment could enhance macrophage polarization through integrin-related pathways. The western blot results showed that the protein expression levels of integrin β1, phosphoryla￾tion levels of phosphoinositide 3-kinase (PI3K) and serine/ threonine kinase Akt (Akt) were increased in the CFO 10-E/M group (Figure S12a, Supporting Information). The CFO 10-E/M group exhibited the strongest fluorescence intensity of phos￾phorylated Akt immunostaining (Figure S12b, Supporting Information). Furthermore, the western blot results showed that the CFO 10-E/M group had significantly reduced nuclear factor kappa B (NF-κB/p65) levels compared to other groups (Figure S12a, Supporting Information). Protein adsorption and integrin binding interactions have been demonstrated to modu￾late inflammation.[24c,32,40] On CFO 10-E/M, the enhanced FN adsorption resulted in adoption of an active conformation that led to more RGD binding site exposure and increased macro￾phage integrin β1 binding. Increased integrin β1 binding acti￾vates PI3K/Akt signaling, leading to the inhibition of NF-κB activation and subsequent anti-inflammatory polarization (M2) Figure 4. Macrophage polarization and osteogenesis of BM-MSC on CFO/P(VDF-TrFE) magnetoelectric nanocomposite membranes in vitro. a) Immu￾nostaining images and b) FC analysis of CD206 expression, which indicated that the CFO 10-E/M group promoted macrophage M2 polarization in the co-culture system (Scale bars: 25 µm). c) Alizarin red staining (Scale bars: 200 µm), d) immunostaining images of RUNX2 expression (scale bars: 100 µm) and e,f) RT-qPCR analysis showing that the CFO 10-E/M group enhanced BM-MSC osteogenesis in the co-culture system. (* VS CFO 10-E/M, # VS NC, p < 0.05). Adv. Funct. Mater. 2020, 2006226

SCENCNNEWS A www.advancedsciencenews.com ww.afm-journalde of macrophages. 2 Hence, our data indicated that the RGD- 10-E/M membrane exhibited the strongest immunostaining integrin binding induced PI3K/Akt activation could be one intensity of CCR(Figure $15, Supporting Informa the potential mechanisms by which the CFO10-E/M group amongst all groups. The immunostaining signals of CD206 mediates M2 polarization of macrophages were weak in all groups(Figure S17, Supporting Information) To evaluate osteogenic differentiation in this co-culture After 4 days of implantation, cells from the interstitial fluid col- Sy'stem, ALP activity(Figure $13b, Supporting Information), lected at the defect region of the CFO 10-EM group showed zarin red staining(Figure 4c), immunofluorescence staining the highest ratio of CDllc-positive cells(MI macrophage ( Figure 4d), and osteogenic gene expression(Figure 4e, f) anal- marker) among all groups(Figure $18a, Supporting Informa is were performed. CFo 10-E/M exhibited the most BM-MSc tion). MI macrophages contribute to an initial acute inflamma osteogenic differentiation among all groups. Macrophages with tory stage in vivo s8 Meanwhile, the protein spectrum results M2 phenotype secrete cytokines, such as IL-10, BMP2, and demonstrated that complement components such as Clr, Cfh VEGF, which in turn modulate mesenchymal progenitor cell C5, Cfb, and C8b were upregulated in the CFo 10-E/M group recruitment, angiogenesis and bone regeneration. H Therefore, Enrichment of immune-related pathways and biological pro- M2 polarization of macrophages is one important mechanism cesses, such as endocytosis, complement, and coagulation cas- by which BM-MSC osteogenic differentiation is promoted by cade were also detected in the CFO 10-E/M group(Figure 5c, the magnetoelectric microenvironment. Figure S16b, Supporting Information). There is increasing evi- dence that the complement system, a crucial arm of the innate immune system, plays an important role in bone homeostasis 2.6. Biomimetic Magnetoelectric Microenvironment Accelerates regeneration, and inflammation. Fa The activated coagulation Bone Regeneration In Vivo and complement products can recruit immune cells to the injury site, which leads to a simultaneous early inflammatory The therapeutic efficacy of the magnetoelectric nanocomposite response. 42 Bone regeneration can be modulated by these membranes on bone defect repair were further investigated in inflammatory molecules. -I MI phenotype macrophages arrived vivo. The magnetoelectric nanocomposite membranes were at the injury site during the initial stage and are involved in mplanted to cover critical-sized (5 mm) calvarial defects in the early infammatory response. +4 In this study, the early M1 mature rats. I26 The P(VDF-TrFE) based membrane was not macrophage response in vivo could not be detected by in vitro sticky to the newly regenerated bone tissue, 1 which facilitates testing, and this might be because the complex osteoimmu- subsequent removal of the membrane and bone defect healing nomodulatory environment of the bone defect area is diffcult without residual materials. Since the removable membrane to mimic in vitro 37b These results thus imply that the CFO can be remotely tuned with an external DC magnetic field, the 10-E/M membrane could trigger the initial immune response removed membranes have the potential to be reusable. His. during the early stage of bone repair. logical analysis of Massons trichrome staining and H&E We next evaluated the of mi to m2 transition taining showed that the CFo 10-E/M led to complete healing, affected by built- in magnetoelectric microenvironment in vivo with flat and consecutive bone structures in 8 weeks. The After 4 days of implantation, assessment of adherent macro- mature osteoid tissue was present in the top center region of phages on the nanocomposite membranes showed that the the defect in the CFO 10-E/M group. These results revealed that immunostaining intensity of CD206 in the CFO 10-E/M group the CFO 10-E/M membrane promoted more new bone forma- was stronger than the other groups(Figure Se)and that the ion than the other groups( Figure S14a, b, Supporting Informa- immunostaining intensity of CCR7 in the CFO 10-E/M group on). In micro-CT tests, the CFO 10-E/M group demonstrated was weaker than the other groups( Figure S15, Supporting the most new bone formation at all-time points. After 4 and Information). After 14 days of implantation, adherent macro- 8 weeks post-surgery, the CFo 10-E/M group demonstrated phages on the nanocomposite membrane and cells from the homogeneous and contiguous regenerated mature bone tissue interstitial fluid collected at the defect regions of the CFO within the defect area. By contrast, in the other groups, new 10-E/M group exhibited the highest ratio of M2 macrophages bone tissue formed mostly at the marginal areas around the(Figures S17 and S18b, Supporting Information)Meanwhile, riginal bone defect( Figure 5a; Figure S14c, Supporting Infor- the adherent macrophages on nanocomposite membranes mation). Quantitative analysis revealed that the CFo 10-E/M demonstrated the weakest CCR7 immunostaining signals in membrane significantly increased regenerated bone volume the CFO 10-E/M group(Figure S16a, Supporting Informa- after 8 weeks of implantation(Figure 5b). These results thus tion). Notably, macrophage M2 polarization happened earlier confirmed that the magnetoelectric microenvironment pro- on the membrane than in the interstitial fuid of the CFO vided by the CFO 10-E/M membrane promotes enhanced bone 10-E/M group, which suggested that the cells adherent on the membrane can sense the magnetoelectric microenvironment directly and further accelerate the transition from the Mi to M2 phenotype. Cells within the membrane proximity will 7. Biomimetic Magnetoelectric Microenvironment Modulates sense the magnetoelectric microenvironment later than cells Osteoimmunomodulatory Responses In Vive that are in direct contact with the membrane since bone is a highly dynamic organ, the fracture healing is affected by the The osteoimmunomodulatory effects of the built-in magi surrounding fracture microenvironment, such as inflamma electric microenvironment were further investigated in tory processes. 4] The interstitial fluid provides the 3D env After 1 day of implantation, adherent macrophages on the ronment for inflammatory response, which is important for Adw. Funct Mater. 2020. 2006226 2006226(8of1) g 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (8 of 11) © 2020 Wiley-VCH GmbH of macrophages.[32] Hence, our data indicated that the RGD￾integrin binding induced PI3K/Akt activation could be one of the potential mechanisms by which the CFO10-E/M group mediates M2 polarization of macrophages. To evaluate osteogenic differentiation in this co-culture system, ALP activity (Figure S13b, Supporting Information), alizarin red staining (Figure 4c), immunofluorescence staining (Figure 4d), and osteogenic gene expression (Figure 4e,f) anal￾ysis were performed. CFO 10-E/M exhibited the most BM-MSC osteogenic differentiation among all groups. Macrophages with M2 phenotype secrete cytokines, such as IL-10, BMP2, and VEGF, which in turn modulate mesenchymal progenitor cell recruitment, angiogenesis and bone regeneration.[41] Therefore, M2 polarization of macrophages is one important mechanism by which BM-MSC osteogenic differentiation is promoted by the magnetoelectric microenvironment. 2.6. Biomimetic Magnetoelectric Microenvironment Accelerates Bone Regeneration In Vivo The therapeutic efficacy of the magnetoelectric nanocomposite membranes on bone defect repair were further investigated in vivo. The magnetoelectric nanocomposite membranes were implanted to cover critical-sized (5  mm)  calvarial defects in mature rats.[26] The P(VDF-TrFE) based membrane was not sticky to the newly regenerated bone tissue,[13] which facilitates subsequent removal of the membrane and bone defect healing without residual materials. Since the removable membrane can be remotely tuned with an external DC magnetic field, the removed membranes have the potential to be reusable. His￾tological analysis of Masson’s trichrome staining and H&E staining showed that the CFO 10-E/M led to complete healing, with flat and consecutive bone structures in 8 weeks. The mature osteoid tissue was present in the top center region of the defect in the CFO 10-E/M group. These results revealed that the CFO 10-E/M membrane promoted more new bone forma￾tion than the other groups (Figure S14a,b, Supporting Informa￾tion). In micro-CT tests, the CFO 10-E/M group demonstrated the most new bone formation at all-time points. After 4 and 8 weeks post-surgery, the CFO 10-E/M group demonstrated homogeneous and contiguous regenerated mature bone tissue within the defect area. By contrast, in the other groups, new bone tissue formed mostly at the marginal areas around the original bone defect (Figure 5a; Figure S14c, Supporting Infor￾mation). Quantitative analysis revealed that the CFO 10-E/M membrane significantly increased regenerated bone volume after 8 weeks of implantation (Figure  5b). These results thus confirmed that the magnetoelectric microenvironment pro￾vided by the CFO 10-E/M membrane promotes enhanced bone regeneration in vivo. 2.7. Biomimetic Magnetoelectric Microenvironment Modulates Osteoimmunomodulatory Responses In Vivo The osteoimmunomodulatory effects of the built-in magneto￾electric microenvironment were further investigated in vivo. After 1 day of implantation, adherent macrophages on the CFO 10-E/M membrane exhibited the strongest immunostaining intensity of CCR7 (Figure S15, Supporting Information) amongst all groups. The immunostaining signals of CD206 were weak in all groups (Figure S17, Supporting Information). After 4 days of implantation, cells from the interstitial fluid col￾lected at the defect region of the CFO 10-E/M group showed the highest ratio of CD11c-positive cells (M1 macrophage marker) among all groups (Figure S18a, Supporting Informa￾tion). M1 macrophages contribute to an initial acute inflamma￾tory stage in vivo.[38] Meanwhile, the protein spectrum results demonstrated that complement components such as C1r, Cfh, C5, Cfb, and C8b were upregulated in the CFO 10-E/M group. Enrichment of immune-related pathways and biological pro￾cesses, such as endocytosis, complement, and coagulation cas￾cade were also detected in the CFO 10-E/M group (Figure 5c,d; Figure S16b, Supporting Information). There is increasing evi￾dence that the complement system, a crucial arm of the innate immune system, plays an important role in bone homeostasis, regeneration, and inflammation.[42a] The activated coagulation and complement products can recruit immune cells to the injury site, which leads to a simultaneous early inflammatory response.[42] Bone regeneration can be modulated by these inflammatory molecules.[43] M1 phenotype macrophages arrived at the injury site during the initial stage and are involved in the early inflammatory response.[44] In this study, the early M1 macrophage response in vivo could not be detected by in vitro testing, and this might be because the complex osteoimmu￾nomodulatory environment of the bone defect area is difficult to mimic in vitro.[37b] These results thus imply that the CFO 10-E/M membrane could trigger the initial immune response during the early stage of bone repair. We next evaluated the process of M1 to M2 transition affected by built-in magnetoelectric microenvironment in vivo. After 4 days of implantation, assessment of adherent macro￾phages on the nanocomposite membranes showed that the immunostaining intensity of CD206 in the CFO 10-E/M group was stronger than the other groups (Figure  5e) and that the immunostaining intensity of CCR7 in the CFO 10-E/M group was weaker than the other groups (Figure S15, Supporting Information). After 14 days of implantation, adherent macro￾phages on the nanocomposite membrane and cells from the interstitial fluid collected at the defect regions of the CFO 10-E/M group exhibited the highest ratio of M2 macrophages. (Figures S17 and S18b, Supporting Information) Meanwhile, the adherent macrophages on nanocomposite membranes demonstrated the weakest CCR7 immunostaining signals in the CFO 10-E/M group (Figure S16a, Supporting Informa￾tion). Notably, macrophage M2 polarization happened earlier on the membrane than in the interstitial fluid of the CFO 10-E/M group, which suggested that the cells adherent on the membrane can sense the magnetoelectric microenvironment directly and further accelerate the transition from the M1 to M2 phenotype. Cells within the membrane proximity will sense the magnetoelectric microenvironment later than cells that are in direct contact with the membrane. Since bone is a highly dynamic organ, the fracture healing is affected by the surrounding fracture microenvironment, such as inflamma￾tory processes.[45] The interstitial fluid provides the 3D envi￾ronment for inflammatory response, which is important for Adv. Funct. Mater. 2020, 2006226

ADVANCED m-journal de CFO O-E/M CFO 10-E/M CFO O-E CFO 10-E e CFO 10-E/M CFO 10-E CFO O-E O O-M CFO 10-E/M 10-E/M VSI Pathway Enrichm d Figure 5. CFO/P(VDF-TrFE)magnetoelectric nanocomposite membrane mediated bone regeneration and immune response in vivo. a)Representa- tive CT images of bone regeneration within rat cranial defects at 8 weeks after membrane implantation, with the CFo 10-E/M group exhibiting the st abundant new bone formation. Yellow arrows indicate enhanced bone regeneration in the CFo 10-E/M group. Yellow triangles denote new bone denote the boundary between nascent bone and host bone. b)Quantitative analysis of the total volume of newly formed bone ted in the ceo 1o-E/M group. Blue arrow indicates the enriched genes in the cop at some immune-related proteins and pathways were ssue.(* VS CFO 10-E/M, p<0.05)c)KEGG pathway analysis, d )heat map demonstrating t ment and coagulation cascades pathway of the CFo 10-E/M group. e)Immunofluorescence images of CD206 expression, indicating that the CFo 10-E/M membrane promoted adherent macrophage M2 polarization(Scale bars: 50 um) bone regeneration. After initial inflammation, the phenotype 3. Conclusion was transitioned to M2 at the remodeling stage. M2 macro- phages are known as the pro- healing phenotype. 46 Many This study developed a flexible and reusable magnetoelec biomaterials have been designed to promote tissue regenera- tric nanocomposite membrane for bone regeneration which tion by activating M2 polarization of macrophages, I/Taken can be regulated by a remote dC magnetic field to mimic the together, these results thus indicated that the CFO 10.E/M natural magnetoelectric microenvironment. Based on MD membranes could activate the initial immune response and simulations together with biological evaluation, the 10 wt% accelerate the transition from M1 to M2 phenotype to further CFO/P(VDF-TrFE) magnetoelectric nanocomposite mem- promote bone regeneration. Therefore, our findings sug. branes were confirmed to be the optimal group for promoting gested that the CFo 10-E/M membrane could provide a mag. bone regeneration by increasing RGD exposure. Moreover, the netoelectric microenvironment with 12h remote DC magnetic magnetoelectric microenvironment provided by the magneto- field shifting to enhance bone regeneration by activating the electric nanocomposite membrane not only directly enhanced me response and accelerating the transition from the BM-MSC osteogenic differentiation, but also regulated the oste- Immune ammation stage to bone healing stage. iMmuno modulatory environment to improve bone regenera The remotely tuned magnetoelectric microenvironment pro- tion. The osteoimmunomodulatory microenvironment wit vided by the removable and easily shaping membrane have the the bone defect area triggered initial inflammation and then potential to be utilized clinically for space maintenance and subsequently promoted MI to M2 transition of macrophages Our research thus provides a novel strategy of remote tuning Adv Funct. Mater. 2020. 2006226 20062269of1) o 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (9 of 11) © 2020 Wiley-VCH GmbH bone regeneration. After initial inflammation, the phenotype was transitioned to M2 at the remodeling stage. M2 macro￾phages are known as the pro-healing phenotype.[46] Many biomaterials have been designed to promote tissue regenera￾tion by activating M2 polarization of macrophages.[47] Taken together, these results thus indicated that the CFO 10-E/M membranes could activate the initial immune response and accelerate the transition from M1 to M2 phenotype to further promote bone regeneration. Therefore, our findings sug￾gested that the CFO 10-E/M membrane could provide a mag￾netoelectric microenvironment with 12h remote DC magnetic field shifting to enhance bone regeneration by activating the immune response and accelerating the transition from the acute inflammation stage to bone healing stage. The remotely tuned magnetoelectric microenvironment pro￾vided by the removable and easily shaping membrane have the potential to be utilized clinically for space maintenance and bone regeneration. 3. Conclusion This study developed a flexible and reusable magnetoelec￾tric nanocomposite membrane for bone regeneration which can be regulated by a remote DC magnetic field to mimic the natural magnetoelectric microenvironment. Based on MD simulations together with biological evaluation, the 10 wt% CFO/P(VDF-TrFE) magnetoelectric nanocomposite mem￾branes were confirmed to be the optimal group for promoting bone regeneration by increasing RGD exposure. Moreover, the magnetoelectric microenvironment provided by the magneto￾electric nanocomposite membrane not only directly enhanced BM-MSC osteogenic differentiation, but also regulated the oste￾oimmunomodulatory environment to improve bone regenera￾tion. The osteoimmunomodulatory microenvironment within the bone defect area triggered initial inflammation and then subsequently promoted M1 to M2 transition of macrophages. Our research thus provides a novel strategy of remote tuning Figure 5. CFO/P(VDF-TrFE) magnetoelectric nanocomposite membrane mediated bone regeneration and immune response in vivo. a) Representa￾tive CT images of bone regeneration within rat cranial defects at 8 weeks after membrane implantation, with the CFO 10-E/M group exhibiting the most abundant new bone formation. Yellow arrows indicate enhanced bone regeneration in the CFO 10-E/M group. Yellow triangles denote new bone. Yellow dotted lines denote the boundary between nascent bone and host bone. b) Quantitative analysis of the total volume of newly formed bone tissue. (* VS CFO 10-E/M, p < 0.05) c) KEGG pathway analysis, d) heat map demonstrating that some immune-related proteins and pathways were upregulated in the CFO 10-E/M group. Blue arrow indicates the enriched genes in the complement and coagulation cascades pathway of the CFO 10-E/M group. e) Immunofluorescence images of CD206 expression, indicating that the CFO 10-E/M membrane promoted adherent macrophage M2 polarization (Scale bars: 50 µm). Adv. Funct. Mater. 2020, 2006226

SCENCNNEWS www.advancedsciencenews.com of the magnetoelectric microenvironment for precisely control- [7 J.J. Cook, N ). Summers, E A. Cook, Clin. Podiatr. Med. Surg. 2015 bone regeneration in situ, which holds much promise for ving efficient bone repair in the clinic. 18]C. Daish, R. Blanchard, K. Fox, P. Pivonka, E. Pirogova, Ann. Biomed. Eng. 2018, 46, 525 [9] a)H. M. Yun, S.). Ahn, K.R. Park, M.J. Kim, J.J. Kim, G. Z. in, H. W. Kim, E. C. Kim, Biomaterials 2016, 85, 88; b)S. M. Dadfar, 4. Experimental Section K. Roemhild N. I. Drude. S. von Stillfried. R. Knuchel. F Kiessling Details of the materials and experimental methods used are availabl T. Lammers, Adv. Drug Delivery Rev. 2019, 138, 302; c)C. Ning, Z. Zhou, G. Tan, Y. Zhu, C. Mao, Prog. Polym. Sci. 2018, 81, 144 no]M. Cifra, Z. Fields, A. Farhadi, Prog. Biophys. Mol. Biol. 2011, 10 P Martins, S. Lanceros-Mendez, Adv. Funct. Mater. 2013, 23, 3371 Supporting Information [2] a) V Supporting Information is available from the Wiley Online Library or M. G. Talikina, Y. V. Chebotareva, Y. G. lzyumov, A.A.Batrakova from the author V.A. Nepomnyashchikh, Biofizika 2017, 62, 675; b)V. V. Krylov, N. P. Kantserova, L. A. Lysenko, E. A. Osipova, IntJ.Bio 2019,63,241 [3]X. Zhang, C. Zhang, Y. Lin, P. Hu, Y. Shen, K. Wang, S. Meng. Acknowledgements Y. Chai, X.Dai, X Liu, Y Liu, X Mo, C. Cao, S. Li, X Deng, L Chen, ACS Nano2016,10,7279 W.L. and F.Z. contributed equally to this work. This work was [4]V. V. Krylov, Bioelectromagnetics 2017, 38, 497 by the National Key R&D Program of China(2018YFCl [5] a)M. D. Pierschbacher, E. Ruoslahti, Nature 1984 2018YFE0194400), National Natural Science Foundation b)E. Ruoslahti, Annu. Rev. Cell Dev. Biol. 1996, 12, 697 (81600905,81991505, 31670993), Peking University Medicine Fund [16]Y. Wei, S Jiang, M. Si, X. Zhang, ) Liu, Z. Wang, C. Cao, J of Fostering Young Scholars' Scientific Technological Innovation H. Huang, L Chen, S. Wang, C. Feng, X. Deng, L Jiang, Ad supported by the Fundamental Research Funds for the Central 2019,31,1900582 Foundation (182185), and Key Science and Technology Program of [7 a) M. Schwartzman, M. Palma, ). Sable, ).Abramson, X.Hu with the approval of the Animal Care and Use Committee of Peking T. Segura, Biomaterials 2013, 34, 3938 niversity(IACUC number: LA2019161) 8] T Re'em, O. Tsur-Gang, S Cohen, Biomaterials 2010, 31, 6746 9 H. Kong, S. Hsiong, D. ) Mooney, Nano Lett. 2007, 7, 161 [20] a)H. Shroff, C.G. Galbraith, ). A. Galbraith, H. White, ). Gillette S. Olenych, M. W. Davidson, E. Betzig, Proc. Natl. Acad. Sci. U.S.A Conflict of Interest 2007, 104, 20308; b)R. Changed, H. Cai, S. Wind, M. P. Sheetz, Nat.Mate:2019,18,1366 The authors declare no conflict of interest [21] a)X Ren, Y. Feng, ). Guo, H. Wang, Q. Li, J. Yang, X. Hao, ). Lv, N. Ma, W. Li, Chem. Soc. Rev. 2015, 44, 5680; b)E. M. Pierschbacher. Science 1987. 238. 491 K hae,J- U. Lee, G. H. Kim, Colloids Surf, B 2019. b)H Ta gan, E. Sadroddiny, Biomed. Pharmacother 2017, 88, 956. arginylglycylaspartic acid, bone regeneration, magnetoelectric [23)G Kocer, P. Jonkheijm, Adv. Healthcare Mater. 2017, 6, 1600862. microenvironments, nanocomposite membranes, remote tuning [24] a)Z. Hamidouche, O. Fromigue, ). Ringe, T. Haupl, P. Vaud vne. P. Received: July 23, 2020 Sci. U. S.A. 2009, 106, 18587: b)P. Chatakun, R. Nunez. Toldra Revised: September 15, 2020 Diaz Lopez, C. Gil-Recio, E. Martinez.Sarra, F. Hernandez.Alfaro, Published online E. Ferres-Padro, L Giner-Tarrida, M. Atari, Cell. Mol. Life Sci. 2014 71, 113; c)T. D. Zaveri, ) .S. Lewis, N. V Dolgova, M. ). Clare-Salzler B G. Keselowsky, Biomaterials 2014, 35, 3504. [25] P. Martins, R. Goncalves, S. Lanceros-Mendez, A. Lasheras, []S.W. Benoit, P. S. Danielle, A.R. D. Michael, K. S Anseth, Nat. 1. Gutierrez, M. Barandiaran, Appi Sci.2014,313,215 Mater.2008,7,816 [26] C. Zhang, W Liu, C. Cao, F. Zhang, Q. Tang, S. Ma, I.). Zhao, [2] a) W. Li, Z. Yan, ). Ren, X. Qu, Chem. Soc. Rev. 2018, 47, 8639: L Hu. Y. Shen, L. Chen. Adv. Healthcare Mater. 2018.7. 1701466. G. Thrivikraman, G. Madras, B. Basu, Biomaterials 2016, 77, 26. [27 a)Y. Zong, Z. Yue, P. Martins, ) Zhuang, Y Du, S. Lanceros-Mendez [3] A Maziarz, B Kocan, M. Bester, S. Budzik, M. Cholewa, T. Ochiya, M.J. Higgins, Nanoscale 2018, 10, 17370: b)B. Tang, ). Zhuang A. Banas. Stem Cell Res. Ther. 2016. 7.54 L Wang, B. Zhang, S. Lin, F Jia, L Dong, Q. Wang, K. Cheng, [4] a)P. Fomby, A.J. Cherlin, A. Hadjizadeh, C. ). Doillon, W. Weng, ACS Appl. Mater. Interfaces 2018, 10, 7841 V. Sueblinvong, D.). Weiss, ). H. T. Bates, T. Gilbert, W.C. Liles, [28 P. Martins, A. Lasheras, ). Gutierrez, M. Barandiaran, I. Orue C. Lutzko, Rajagopal, D ) Prockop, D. Chambers, A. Giangreco S. Lanceros-Mendez, Phys. D: Appl. Phys. 2011, 44, 495303 A Keating, D. Kotton, P L Lelkes, D. E. Wagner, D. ). Prockop, Ann. [29]a)H.. Chang, Y. Wang, in Regenerative Medicine and Tissue Eng Yan, Y. Ma, H Wu, C. Liu, Stem Cell Res. Ther. 2019, 1 /E.H. Li Am. Thorac. Soc. 2010, 12, 181; b)). Chen, C. Tu, X.Tan ring.Cells and Biomaterials, (Ed: D. Eberli), In Tech, London 2011 993-1001; b)). M. Rice, ) . A. Hunt, ). A. Gallagher, P. Hanarp J C. Tu, Y Xiao, Y. Ma, H. Wu, M. Song, Stem Cell Res. Ther. 2018, 9, 215. D.S.Sutherland, ). Gold, Biomaterials 2003, 24, 4799; c)).Yang, [6] a)Y Seong, ). Moon, ). Kim, Life Sci. 2014, 102, 16; b)H).Kim, Y. Zhou. F Wei Y. Xiao. Clin. Oral Im Jung, ). H. Park, ). H. Kim, K N. Ko, C. W. Kim, Exp. Biol. Med. [30] a)). Wei, T. Igarashi, N. Okumori, T. T. Maetani. B. Liu 2013.238,923 M. Yoshinari. Biomed. Mater. 2009 02; b)G. Perumal, Adw. Funct Mater. 2020. 2006226 2006226(0of1) g 2020 Wiley-VCH GmbH

www.advancedsciencenews.com www.afm-journal.de 2006226 (10 of 11) © 2020 Wiley-VCH GmbH of the magnetoelectric microenvironment for precisely control￾ling bone regeneration in situ, which holds much promise for achieving efficient bone repair in the clinic. 4. Experimental Section Details of the materials and experimental methods used are available in the Supporting Information. Supporting Information Supporting Information is available from the Wiley Online Library or from the author. Acknowledgements W.L. and F.Z. contributed equally to this work. This work was supported by the National Key R&D Program of China (2018YFC1105303/04, 2018YFE0194400), National Natural Science Foundation of China (81600905, 81991505, 31670993), Peking University Medicine Fund of Fostering Young Scholars’ Scientific & Technological Innovation supported by the Fundamental Research Funds for the Central Universities (BMU2018PY005), Beijing Municipal Natural Science Foundation (7182185), and Key Science and Technology Program of Hunan Province (2017WK2041). All animal experiments were conducted with the approval of the Animal Care and Use Committee of Peking University (IACUC number: LA2019161). Conflict of Interest The authors declare no conflict of interest. Keywords arginylglycylaspartic acid, bone regeneration, magnetoelectric microenvironments, nanocomposite membranes, remote tuning Received: July 23, 2020 Revised: September 15, 2020 Published online: [1] S. W.  Benoit, P. S.  Danielle, A. R. D.  Michael, K. S.  Anseth, Nat. Mater. 2008, 7, 816. [2] a) W.  Li, Z.  Yan, J.  Ren, X.  Qu, Chem. Soc. Rev. 2018, 47, 8639; b) G. Thrivikraman, G. Madras, B. Basu, Biomaterials 2016, 77, 26. [3] A. Maziarz, B. Kocan, M. Bester, S. Budzik, M. Cholewa, T. Ochiya, A. Banas, Stem Cell Res. Ther. 2016, 7, 54. [4] a) P.  Fomby, A. J.  Cherlin, A.  Hadjizadeh, C. J.  Doillon, V.  Sueblinvong, D. J.  Weiss, J. H. T.  Bates, T.  Gilbert, W. C.  Liles, C. Lutzko, J. Rajagopal, D. J. Prockop, D. Chambers, A. Giangreco, A. Keating, D. Kotton, P. I. Lelkes, D. E. Wagner, D. J. Prockop, Ann. Am. Thorac. Soc. 2010, 12, 181; b) J.  Chen, C.  Tu, X.  Tang, H.  Li, J. Yan, Y. Ma, H. Wu, C. Liu, Stem Cell Res. Ther. 2019, 10, 1. [5] C. Tu, Y. Xiao, Y. Ma, H. Wu, M. Song, Stem Cell Res. Ther. 2018, 9, 215. [6] a) Y.  Seong, J.  Moon, J.  Kim, Life Sci. 2014, 102, 16; b) H. J.  Kim, J. Jung, J. H. Park, J. H. Kim, K. N. Ko, C. W. Kim, Exp. Biol. Med. 2013, 238, 923. [7] J. J. Cook, N. J. Summers, E. A. Cook, Clin. Podiatr. Med. Surg. 2015, 32, 45. [8] C.  Daish, R.  Blanchard, K.  Fox, P.  Pivonka, E.  Pirogova, Ann. Biomed. Eng. 2018, 46, 525. [9] a) H. M. Yun, S. J. Ahn, K. R. Park, M. J. Kim, J. J. Kim, G. Z. Jin, H. W.  Kim, E. C.  Kim, Biomaterials 2016, 85, 88; b) S. M.  Dadfar, K. Roemhild, N. I. Drude, S. von Stillfried, R. Knüchel, F. Kiessling, T.  Lammers, Adv. Drug Delivery Rev. 2019, 138, 302; c) C.  Ning, Z. Zhou, G. Tan, Y. Zhu, C. Mao, Prog. Polym. Sci. 2018, 81, 144. [10] M. Cifra, J. Z. Fields, A. Farhadi, Prog. Biophys. Mol. Biol. 2011, 105, 223. [11] P. Martins, S. Lanceros-Méndez, Adv. Funct. Mater. 2013, 23, 3371. [12] a) V. V.  Krylov, E. A.  Osipova, N. A.  Pankova, M. G.  Talikina, Y. V.  Chebotareva, Y. G.  Izyumov, A. A.  Batrakova, V. A.  Nepomnyashchikh, Biofizika 2017, 62, 675; b) V. V.  Krylov, N. P. Kantserova, L. A. Lysenko, E. A. Osipova, Int. J. Biometeorol. 2019, 63, 241. [13] X.  Zhang, C.  Zhang, Y.  Lin, P.  Hu, Y.  Shen, K.  Wang, S.  Meng, Y. Chai, X. Dai, X. Liu, Y. Liu, X. Mo, C. Cao, S. Li, X. Deng, L. Chen, ACS Nano 2016, 10, 7279. [14] V. V. Krylov, Bioelectromagnetics 2017, 38, 497. [15] a) M. D.  Pierschbacher, E.  Ruoslahti, Nature 1984, 309, 30; b) E. Ruoslahti, Annu. Rev. Cell Dev. Biol. 1996, 12, 697. [16] Y. Wei, S. Jiang, M. Si, X. Zhang, J. Liu, Z. Wang, C. Cao, J. Huang, H. Huang, L. Chen, S. Wang, C. Feng, X. Deng, L. Jiang, Adv. Mater. 2019, 31, 1900582. [17] a) M.  Schvartzman, M.  Palma, J.  Sable, J.  Abramson, X.  Hu, M. P.  Sheetz, S. J.  Wind, Nano Lett. 2011, 11, 1306; b) J.  Lam, T. Segura, Biomaterials 2013, 34, 3938. [18] T. Re’em, O. Tsur-Gang, S. Cohen, Biomaterials 2010, 31, 6746. [19] H. J. Kong, S. Hsiong, D. J. Mooney, Nano Lett. 2007, 7, 161. [20] a) H.  Shroff, C. G.  Galbraith, J. A.  Galbraith, H.  White, J.  Gillette, S. Olenych, M. W. Davidson, E. Betzig, Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 20308; b) R. Changede, H. Cai, S. J. Wind, M. P. Sheetz, Nat. Mater. 2019, 18, 1366. [21] a) X.  Ren, Y.  Feng, J.  Guo, H.  Wang, Q.  Li, J.  Yang, X.  Hao, J.  Lv, N.  Ma, W.  Li, Chem. Soc. Rev. 2015, 44, 5680; b) E.  Ruoslahti, M. Pierschbacher, Science 1987, 238, 491. [22] a) S. J.  Chae, J. U.  Lee, G. H.  Kim, Colloids Surf., B 2019, 181, 408; b) H. T. Aiyelabegan, E. Sadroddiny, Biomed. Pharmacother. 2017, 88, 956. [23] G. Koçer, P. Jonkheijm, Adv. Healthcare Mater. 2017, 6, 1600862. [24] a) Z.  Hamidouche, O.  Fromigué, J.  Ringe, T.  Häupl, P.  Vaudin, J. C.  Pagès, S.  Srouji, E.  Livne, P. J.  Marie, Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 18587; b) P.  Chatakun, R.  Núñez-Toldrà, E. J. Díaz López, C. Gil-Recio, E. Martínez-Sarrà, F. Hernández-Alfaro, E. Ferrés-Padró, L. Giner-Tarrida, M. Atari, Cell. Mol. Life Sci. 2014, 71, 113; c) T. D. Zaveri, J. S. Lewis, N. V. Dolgova, M. J. Clare-Salzler, B. G. Keselowsky, Biomaterials 2014, 35, 3504. [25] P.  Martins, R.  Gonçalves, S.  Lanceros-Mendez, A.  Lasheras, J. Gutiérrez, J. M. Barandiarán, Appl. Surf. Sci. 2014, 313, 215. [26] C.  Zhang, W.  Liu, C.  Cao, F.  Zhang, Q.  Tang, S.  Ma, J. J.  Zhao, L. Hu, Y. Shen, L. Chen, Adv. Healthcare Mater. 2018, 7, 1701466. [27] a) Y. Zong, Z. Yue, P. Martins, J. Zhuang, Y. Du, S. Lanceros-Mendez, M. J.  Higgins, Nanoscale 2018, 10, 17370; b) B.  Tang, J.  Zhuang, L.  Wang, B.  Zhang, S.  Lin, F.  Jia, L.  Dong, Q.  Wang, K.  Cheng, W. Weng, ACS Appl. Mater. Interfaces 2018, 10, 7841. [28] P.  Martins, A.  Lasheras, J.  Gutierrez, J. M.  Barandiaran, I.  Orue, S. Lanceros-Mendez, J. Phys. D: Appl. Phys. 2011, 44, 495303. [29] a) H.-I.  Chang, Y.  Wang, in Regenerative Medicine and Tissue Engi￾neering - Cells and Biomaterials, (Ed: D. Eberli), InTech, London 2011, pp. 993–1001; b) J. M. Rice, J. A. Hunt, J. A. Gallagher, P. Hanarp, D. S.  Sutherland, J.  Gold, Biomaterials 2003, 24, 4799; c) J.  Yang, Y. Zhou, F. Wei, Y. Xiao, Clin. Oral Impl. Res. 2016, 27, 1031. [30] a) J.  Wei, T.  Igarashi, N.  Okumori, T.  Igarashi, T.  Maetani, B.  Liu, M.  Yoshinari, Biomed. Mater. 2009, 4, 045002; b) G.  Perumal, Adv. Funct. Mater. 2020, 2006226