Gold Nanoparticle-Decorated Scaffolds Promote NeuronalDifferentiation and MaturationKoby Baranes,†,‡ Michal Shevach,§,∥ Orit Shefi,*,†,‡ and Tal Dvir*,§,∥,⊥

†Faculty of Engineering and ‡Institute of Nanotechnologies and Advanced Materials, Bar Ilan University, Ramat Gan 52900, Israel§The Laboratory for Tissue Engineering and Regenerative Medicine, Department of Molecular Microbiology and Biotechnology,∥Department of Materials Science and Engineering, and ⊥The Center for Nanoscience and Nanotechnology, Tel Aviv University, TelAviv 69978, Israel

*S Supporting Information

ABSTRACT: Engineered 3D neuronal networks are consid-ered a promising approach for repairing the damaged spinalcord. However, the lack of a technological platformencouraging axonal elongation over branching may jeopardizethe success of such treatment. To address this issue we havedecorated gold nanoparticles on the surface of electrospunnanofiber scaffolds, characterized the composite material, andinvestigated their effect on the differentiation, maturation, and morphogenesis of primary neurons and on an immature neuronalcell line. We have shown that the nanocomposite scaffolds have encouraged a longer outgrowth of the neurites, as judged by thetotal length of the branching trees and the length and total distance of neurites. Moreover, neurons grown on the nanocompositescaffolds had less neurites originating out of the soma and lower number of branches. Taken together, these results indicate thatneurons cultivated on the gold nanoparticle scaffolds prefer axonal elongation over forming complex branching trees. Weenvision that such cellular constructs may be useful in the future as implantable cellular devices for repairing damaged neuronaltissues, such as the spinal cord.

KEYWORDS: Axonal elongation, gold nanoparticles, nanocomposite scaffold, neuronal tissue engineering

Injury to the central nervous system (CNS) including thebrain and spinal cord, resulting from physical trauma,

neurodegenerative diseases, or ischemia, is one of the leadingcauses of death and chronic disability in humans.1−4 Postinjury, neurotoxicity, formation of free oxygen radicals, andtissue inflammation often cause secondary damage, which leadsto the expansion of the initial injured zone and death of neuralnetworks.5,6 As the ability of the adult mammalian CNS torepair itself through regeneration of new neurons is limited,7−9

there is a strong interest in cell replacement therapies to restoreactive neuronal cell function and network connections, thusrepairing brain and spinal cord damage.Despite extensive research, there are only limited cell therapy

technologies, which include, among others, injection of externalcells into the injured site and delivery of growth factors topromote regeneration of the existing cells.10−12 The absence ofan in situ supporting three-dimensional (3D) microenviron-ment for cell attachment and guidance may lead to cell death orto the assembly of nonfunctional neuronal circuits.13 The use ofengineered 3D scaffolds for repairing neural damage isconsidered to be a promising approach in neural tissueengineering. These artificial scaffolds provide topographical,biochemical, and mechanical cues for proper organization andtissue formation, biomimicking the native neuronal extracellularmatrix (ECM). Various scaffolds, including electrospun nano-fiber scaffolds, have been examined as candidate platforms for

neuronal differentiation and growth, promoting cell−cell andcell−matrix interactions and altering growth strategy.14−17

Recently, we and others have demonstrated that the neuronalpattern of growth may be affected by nanometric top-ography.18−23 In response to these nanometric cues, the motilegrowth cone at the tip of the axonal branch is fixed to theextracellular translocations in the substrate via adhesionreceptors.24 For example, metal nanoparticles, serving asanchoring sites for the small filopodial projections on 2Dsurfaces, were shown to improve neurite−substrate inter-actions, leading to controlled growth of the neuronalprocesses.20,25 In other studies, carbon nanotubes were usedto affect the morphology of cultured cells and even to alter theirelectrical activity.26 The combination of well-defined andspecific nanoscale cues with 3D scaffolds may provide anadvantageous supporting microenvironment and therefore amore relevant therapeutic approach.In a series of studies we have previously shown that gold

nanowires and nanoparticles embedded within variousbiomimetic scaffolds promote the assembly of cardiac cellsinto elongated and aligned tissues, able to generate a strong,anisotropic contraction force.27−30 Here, we sought toinvestigate the effects of these matrices on other electrogenic

Received: October 4, 2015Revised: November 30, 2015Published: December 16, 2015

Letter

pubs.acs.org/NanoLett

© 2015 American Chemical Society 2916 DOI: 10.1021/acs.nanolett.5b04033Nano Lett. 2016, 16, 2916−2920

cells. Three-dimensional composite scaffolds were fabricatedusing electrospinning, decorated with gold nanoparticles(AuNPs) and characterized for particle size, distribution, andtopographical properties. Next, the effect of the designed cueson neuronal growth and maturation was investigated (Figure1).Fibers with diameters of several hundreds of nanometers

have been previously shown to promote beneficial effects inneurons, including proliferation and differentiation in neuronalstem cells.14,16 Therefore, we have fabricated electrospun fiberscaffolds from polycaprolactone (PCL)-gelatin with an averagefiber diameter of 260 ± 70 nm (Figure 2A and SupplementaryFigure 1). The fiber diameter may be controlled by changingthe electrospinning parameters, such as polymer concentration,applied voltage, velocity of polymer injection, etc.31 In addition,scaffold’s porosity may be further tuned by coelectrospinning of

the scaffold’s polymer with a sacrificial polymer that later oncan be dissolved during cell seeding with culture medium.32

While PCL provided mechanical strength to the fibers,33 thebioinductive gelatin contributed to the fibers’ biological activity,promoting cell adhesion and spreading.34,35 We next sought toincorporate AuNPs to the electrospun fiber scaffolds byevaporation of gold (Figure 2B−D). The functional groups inthe gelatin were utilized as binding sites for the evaporated goldNPs.29,36 Next, EDX analysis was performed in order toconfirm the AuNP presence (Figure 2C). Environmentalscanning electron microscope (ESEM) images revealed that10 nm nominal thickness AuNPs were homogeneously anduniformly deposited on the fibers (Figure 2D). Fiber coverageby the NPs and the composite topography were also evaluatedby a transmitting electron microscope (TEM) and an atomicforce microscope (AFM), respectively (Figure 2E−G). As

Figure 1. Schematic overview of the study. Polymeric nanofibers from PCL/gelatin are fabricated by electrospinning (A). Gold nanoparticles (10nm) are evaporated on the fibers (B) to create nanocomposite scaffolds (C). Neuronal cells seeded within the nanocomposite scaffolds growelongated axons and form 3D networks (D).

Figure 2. AuNPs electrospun fiber scaffolds. (A) Fibrous scaffolds with average fiber diameter of 260 nm were fabricated by electrospinning of PCLand gelatin. Scale bar = 10 μm. (B) AuNPs were evaporated on the fibers to create the nanocomposite scaffolds. (C) EDX analysis confirming theexistence of AuNPs on the fibers. (D) ESEM image of AuNPs on the surface of the fibers. Scale bar = 500 nm. (E) TEM image of the fibers revealedc.a. 10 nm islands of AuNPs (dark) on the surface of the fibers (light). Scale bar = 5 nm. (F, G) Roughness of the pristine (F) and AuNP fibers (G)as imaged by AFM.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b04033Nano Lett. 2016, 16, 2916−2920

2917

shown, the c.a. 10 nm particles were dispersed on the fibers,creating gold islands rather than a continuous layer of the metal(Figure 2E). This allows the cells to interact with the PCL-gelatin fibers and the AuNPs, providing both biological motifsand topographical and electrical cues.20,28,29 AFM resultsprovided further evidence for modification of the surface ofthe fibers with nanoscale roughness providing topographic cues(Figure 2F,G).Next we sought to evaluate the effect of the engineered

topography on cell morphology. Primary neurons from themedicinal leech were seeded on the 3D fibrous scaffolds withAuNPs or on pristine scaffolds, by a single droplet. As thismodel allows a low density culture of neurons, it enables toperform morphological analyses of neuronal processes at thesingle cell level. Since metallic nanoparticles may harm cells, wehave initially shown that the nanocomposite scaffolds are nottoxic (Supplementary Figure 2). To evaluate neuronal growthwithin the scaffolds the cell-seeded constructs were immunos-tained for α-tubulin on day 6. Confocal images revealed thatneurons grown on the pristine scaffolds developed shortneurites which were incorporated close to the cell soma (Figure3A). In contrast, neurons grown on the AuNP scaffolds hadlonger and thinner neurites with dendritic trees spreadingfurther from the cell soma (Figure 3B). Movies of Z-stack

images of the primary neurons revealed outgrowth of theneurites in the 3D space, implying that the extensions havepenetrated into the fibrous scaffold and interacted with thefibers (Supplementary Movies 1 and 2). In most cases, theshort outgrowth of neurites on pristine scaffolds led to nonetwork formation, as compared to neurites growing on AuNPs(Figure 3C,D and Supplementary Table 1). To further evaluatethe interactions of the neurons with the fibrous scaffolds, thecell constructs were fixed and imaged using SEM. As shown,the soma adhered to the top layer of the scaffold, while theneurites were not limited to one level, penetrating into themesh, extending, and adhering to fibers on different layers(Figure 3E and F).During cell dissociation procedure the neurons lose their

original morphology. After seeding, an outgrowth processstarts, where neurites originate from the soma, extend andbranch, developing into a complex dendritic tree. Quickmorphogenesis, including neurite initiation and extensiontoward target sites may be essential for effective regenerationof neuronal tissue post trauma. To quantify the effect of AuNPsscaffolds on the neuronal outgrowth process, several morpho-metric parameters, at the single cell level, were measured. Asshown in Figure 4A, neurons on AuNPs scaffolds developed

into significantly longer branching trees (p = 0.008). Neuronson AuNPs scaffolds also developed to a significantly longerdistance (Figure 4B; p < 0.0001). Moreover, the length of thelongest neurite in each neuron indicated that the AuNPsscaffold may encourage a longer outgrowth (Figure 4C; p <0.0001). On the other hand, neurons on the AuNPs scaffoldhad less neurites originating out of the soma (Figure 4D; p =0.007) and significantly lower total number of branches (Figure4E; p < 0.0001). These morphometric parameters may indicate

Figure 3. Primary neuron growth on the scaffolds. Confocalmicroscopy images of single neurons cultivated on pristine (A) orAuNPs (B) scaffolds and stained against α-tubulin. Bar = 100 μm. (C,D) Neuronal networks formed on pristine (C) or AuNPs (D)scaffolds. Bar = 50 μm. E. SEM images of neurons growing on AuNPsscaffolds. Bar = 10 μm. (F) Neurite growth cones extending minorprocesses between the nanofibers. Bar = 3 μm.

Figure 4. Morphometric analyses of primary neuron growth onpristine and AuNPs scaffolds. (A) Average total branching tree length.(B) Average growth radius. (C) Average longest neurite length. (D)Average number of neurites originating from the soma. (E) Averagenumber of branching points. At least five cell-seeded scaffolds wereanalyzed in each group.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b04033Nano Lett. 2016, 16, 2916−2920

2918

on the tendency to prefer elongation over forming complexbranching trees in the AuNPs scaffolds.37,38 Overall, themorphometric analyses revealed a distinct neuronal behavioron the different scaffolds, indicating that AuNPs scaffolds aremore suitable to promote axon elongation, therefore, moreappropriate for bridging over long gaps.One of the main challenges of neuronal tissue engineering is

cell differentiation and maturation. As nanomaterials andnanotopographies have been shown to affect neuronaldevelopment,39−41 we next sought to assess whether theAuNPs can affect the differentiation of immature cells. It waspreviously shown that PC12 cells, cultured on collagen-coatedsubstrates with the addition of nerve growth factor (NGF),differentiate into neuron-like cells with properties of matureneurons with excitable membranes.42 We hypothesized that themicroenvironment within our composite scaffolds mayencourage the cells to acquire morphological features of matureneurons. We further speculated that the combination of gelatinand AuNPs could be used as anchoring sites for the cells andthus, replace the essential collagen coating. PC12 cells wereseeded on the fibrous scaffolds with or without 10 nm AuNPs.β-NGF was added to the growth medium to initiate thedifferentiation stage and neuronal growth, and developmentwas evaluated on day 6. As shown in Figure 5A, single neuronscultured on the pristine scaffolds exhibited rounded morphol-ogy with limited-sized neurite extensions. On the contrary,PC12 cells cultured on the AuNP scaffolds revealed extendedneurites with dendritic trees spread farther from the cell soma(Figure 5B). Z-stack imaging revealed neurite outgrowth in-between the fibers (Supplementary Movie 3). Furthermore,while in the pristine scaffolds the neurons could not extendlong neurites and therefore did not form neuronal networks(Figure 5C, and Supplementary Table 1); the neurons on theAuNP scaffolds were able to reach adjacent cells and formedbasic neuronal networks (Figure 5D).As 2D substrates are the main cultivation system for PC12

cells and collagen coating is considered essential for thedifferentiation of these cells, we have compared the 3D cellgrowth on scaffolds, with or without AuNPs and withoutadditional collagen coating, to the growth on such surfaces.

Morphometric analyses of confocal images revealed that theneurons on the AuNP scaffolds exhibited neuronal growthstrategy similar to the primary neurons (as shown in Figure 4).The average total branching tree length, average growth radius,and the average length of longest neurite were higher in theAuNP scaffolds (Figure 5E−G). Furthermore, the neuronalmorphology in the AuNP scaffolds resembled the morphologyon the 2D collagen-treated surfaces, indicating on the samedifferentiation level and emphasizing the importance of theAuNPs for 3D neuronal tissue engineering. Interestingly, theaverage number of branching points in cells grown in AuNPsscaffolds was significantly lower than the number seen inpristine scaffolds and on the 2D collagen-treated surfaces(Figure 5H). Moreover, coating of the AuNP scaffolds withcollagen enhanced the NPs effect in the examined parameters(Supplementary Figure 3). Over all, as in the primary culture,the ability to grow longer neurites with less branched dendritictree points-out on the potential of the composite biomaterial tobridge over longer gaps, as seen in spinal cord injuries and mayprovide a platform for effective tissue regeneration.In summary, we have demonstrated the beneficial effect of

AuNPs on the differentiation, growth, and maturation ofneurons on 3D biomaterial scaffolds. The nanocompositematerial is composed of PCL-gelatin nanofibers, providingmechanical support for the cultured cells. Decoration of thefibers with 10 nm AuNPs provided additional topographicaland anchoring sites for superior morphogenesis. Morphometricanalyses of primary and neuronal cell line behavior onnanocomposite electrospun fibers revealed elaborated neuronalgrowth and axonal elongation, leading to more complexneuronal networks.Future studies should focus on determining the exact role of

the AuNPs in neuronal tissue engineering. For example,investigating whether these nanoparticles promote axonelongation and higher expression of neuronal markers due totopographical cues or due to their conductivity, which allowsbetter transfer of an electrical signal, as seen withcardiomyocytes.27,36 Another interesting study would beinvestigating the effect of nanoparticle size and shape onneuronal tissue assembly. Finally, we envision that the reported

Figure 5. PC12 cell growth. (A−D) Confocal microscopy images of isolated neurons (A, B) and neuronal networks (C, D) immunostained againstα-tubulin, growing within pristine (A, C) and AuNPs scaffolds (B, D). Neuronal network was almost completely absent on pristine scaffolds (C)compared to AuNPs scaffolds (D). Bars A, B = 25 μm, C, D = 50 μm. (E−H) Morphometric analyses of PC12 cell growth on pristine scaffolds,AuNP scaffolds or 2D collagen-treated surfaces. (E) Average total branching tree length. (F) Average growth radius. (G) Average length of longestneurite. (H) Average number of branching points. n.s. denotes for not significant.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b04033Nano Lett. 2016, 16, 2916−2920

2919

approach may be useful in the future as implantable cellulardevices for repairing damaged neuronal tissues, such as thespinal cord.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.nano-lett.5b04033.

Details of experimental procedures and supplementaryresults (PDF)Movie 1 (AVI)Movie 2 (AVI)Movie 3 (AVI)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: orit.shefi@biu.ac.il.*E-mail: tdvir@post.tau.ac.il.Author ContributionsK.B. and M.S. contributed equally.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSO.S. acknowledges partial support from the Israel ScienceFoundation Individual grant 1403/11. T.D acknowledgespartial support from the European Union FP7 program(Marie Curie, CIG), Alon Fellowship, and the Israel ScienceFoundation Individual grant 700/13.

■ REFERENCES(1) Adekoya, N.; Thurman, D. J.; White, D. D.; Webb, K. W.MMWRSurveill. Summ. 2002, 51 (10), 1−14.(2) University of Alabama, National spinal cord injury statisticalcenter. 2014 Annual Statistical Report; Birmingham, AL, 2014.(3) Stabenfeldt, S. E.; Garcia, A. J.; LaPlaca, M. C. J. Biomed. Mater.Res., Part A 2006, 77 (4), 718−725.(4) Prang, P.; Muller, R.; Eljaouhari, A.; Heckmann, K.; Kunz, W.;Weber, T.; Faber, C.; Vroemen, M.; Bogdahn, U.; Weidner, N.Biomaterials 2006, 27 (19), 3560−9.(5) Anthony, D. C.; Couch, Y. Exp. Neurol. 2014, 258, 105−11.(6) Fitch, M. T.; Silver, J. Exp. Neurol. 2008, 209 (2), 294−301.(7) Bjorklund, A.; Lindvall, O. Nat. Neurosci. 2000, 3 (6), 537−44.(8) Horner, P. J.; Gage, F. H. Nature 2000, 407 (6807), 963−970.(9) Fry, E. J. Clin. Exp. Pharmacol. Physiol. 2001, 28 (4), 253−8.(10) Hejcl, A.; Lesny, P.; Pradny, M.; Michalek, J.; Jendelova, P.;Stulik, J.; Sykova, E. Physiol. Res. 2008, 57 (3), S121−32.(11) Aurand, E. R.; Wagner, J.; Lanning, C.; Bjugstad, K. B. J. Funct.Biomater. 2012, 3 (4), 839−63.(12) Kabu, S.; Gao, Y.; Kwon, B. K.; Labhasetwar, V. J. ControlledRelease 2015, 219, 141−54.(13) Barthes, J.; Ozcelik, H.; Hindie, M.; Ndreu-Halili, A.; Hasan, A.;Vrana, N. E. BioMed Res. Int. 2014, 2014, 921905.(14) Yang, F.; Murugan, R.; Wang, S.; Ramakrishna, S. Biomaterials2005, 26 (15), 2603−10.(15) Koh, H. S.; Yong, T.; Chan, C. K.; Ramakrishna, S. Biomaterials2008, 29 (26), 3574−82.(16) Gnavi, S.; Fornasari, B. E.; Tonda-Turo, C.; Ciardelli, G.;Zanetti, M.; Geuna, S.; Perroteau, I. Mater. Sci. Eng., C 2015, 48, 620−31.(17) Yao, L.; O’Brien, N.; Windebank, A.; Pandit, A. J. Biomed. Mater.Res., Part B 2009, 90 (2), 483−91.

(18) Baranes, K.; Chejanovsky, N.; Alon, N.; Sharoni, A.; Shefi, O.Biotechnol. Bioeng. 2012, 109 (7), 1791−7.(19) Baranes, K.; Kollmar, D.; Chejanovsky, N.; Sharoni, A.; Shefi, O.J. Mol. Histol. 2012, 43 (4), 437−47.(20) Alon, N.; Miroshnikov, Y.; Perkas, N.; Nissan, I.; Gedanken, A.;Shefi, O. Int. J. Nanomed. 2014, 9 (1), 23−31.(21) Ferrari, A.; Faraci, P.; Cecchini, M.; Beltram, F. Biomaterials2010, 31 (9), 2565−73.(22) Johansson, F.; Carlberg, P.; Danielsen, N.; Montelius, L.; Kanje,M. Biomaterials 2006, 27 (8), 1251−8.(23) Xie, C.; Hanson, L.; Xie, W.; Lin, Z.; Cui, B.; Cui, Y. Nano Lett.2010, 10 (10), 4020−4.(24) Suter, D. M.; Forscher, P. J. Neurobiol. 2000, 44 (2), 97−113.(25) Gilles, S.; Winter, S.; Michael, K. E.; Meffert, S. H.; Li, P.;Greben, K.; Simon, U.; Offenhausser, A.; Mayer, D. Small 2012, 8(21), 3357−67.(26) Lovat, V.; Pantarotto, D.; Lagostena, L.; Cacciari, B.; Grandolfo,M.; Righi, M.; Spalluto, G.; Prato, M.; Ballerini, L. Nano Lett. 2005, 5(6), 1107−10.(27) Dvir, T.; Timko, B. P.; Brigham, M. D.; Naik, S. R.; Karajanagi,S. S.; Levy, O.; Jin, H.; Parker, K. K.; Langer, R.; Kohane, D. S. Nat.Nanotechnol. 2011, 6 (11), 720−5.(28) Fleischer, S.; Shevach, M.; Feiner, R.; Dvir, T. Nanoscale 2014, 6(16), 9410−4.(29) Shevach, M.; Maoz, B. M.; Feiner, R.; Shapira, A.; Dvir, T. J.Mater. Chem. B 2013, 1 (39), 5210.(30) Tian, B.; Liu, J.; Dvir, T.; Jin, L.; Tsui, J. H.; Qing, Q.; Suo, Z.;Langer, R.; Kohane, D. S.; Lieber, C. M. Nat. Mater. 2012, 11 (11),986−94.(31) Fleischer, S.; Miller, J.; Hurowitz, H.; Shapira, A.; Dvir, T.Nanotechnology 2015, 26 (29), 291002.(32) Zander, N. E.; Orlicki, J. A.; Rawlett, A. M.; Beebe, T. P., Jr. J.Mater. Sci.: Mater. Med. 2013, 24 (1), 179−87.(33) Sepulveda, B.; Angelome, P. C.; Lechuga, L. M.; Liz-Marzan, L.M. Nano Today 2009, 4 (3), 244−251.(34) Ma, Z.; He, W.; Yong, T.; Ramakrishna, S. Tissue Eng. 2005, 11(7−8), 1149−58.(35) Jaiswal, M.; Koul, V.; Dinda, A. K.; Mohanty, S.; Jain, K. G. J.Biomed. Mater. Res., Part B 2011, 98 (2), 342−50.(36) Shevach, M.; Fleischer, S.; Shapira, A.; Dvir, T. Nano Lett. 2014,14 (10), 5792−6.(37) Shefi, O.; Golebowicz, S.; Ben-Jacob, E.; Ayali, A. J. Neurobiol.2005, 62 (3), 361−8.(38) van Pelt, J.; Uylings, H. B. Network 2002, 13 (3), 261−81.(39) Ferrari, A.; Cecchini, M.; Dhawan, A.; Micera, S.; Tonazzini, I.;Stabile, R.; Pisignano, D.; Beltram, F. Nano Lett. 2011, 11 (2), 505−11.(40) Brunetti, V.; Maiorano, G.; Rizzello, L.; Sorce, B.; Sabella, S.;Cingolani, R.; Pompa, P. P. Proc. Natl. Acad. Sci. U. S. A. 2010, 107(14), 6264−9.(41) Kuffler, S. W.; Nicholls, J. G. From neuron to brain: a cellularapproach to the function of the nervous system; Sinauer Associates:Sunderland, MA, 1976; p xiii, 486.(42) Greene, L. A.; Tischler, A. S. Proc. Natl. Acad. Sci. U. S. A. 1976,73 (7), 2424−8.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.5b04033Nano Lett. 2016, 16, 2916−2920

2920