nanotopographical surfaces for stem cell fate control: engineering...

Nano Today (2014) 9, 759—784

Available online at www.sciencedirect.com

ScienceDirect

journa l homepage: www.e lsev ier .com/ locate /nanotoday

REVIEW

Nanotopographical surfaces for stem cellfate control: Engineering mechanobiologyfrom the bottom

Weiqiang Chena,b,1,2, Yue Shaoa,b,1, Xiang Lia,b, Gang Zhaoc,Jianping Fua,b,d,∗

a Integrated Biosystems and Biomechanics Laboratory, University of Michigan, Ann Arbor, MI 48109, USAb Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109, USAc Department of Electronic Science and Technology, University of Science and Technology of China,Hefei 230027, PR Chinad Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI 48109, USA

Received 3 September 2014; received in revised form 1 December 2014; accepted 20 December 2014Available online 28 January 2015

KEYWORDSBiomaterials;Nanotopography;Stem cell;Tissue engineering;Regenerativemedicine;Mechanobiology

Summary During embryogenesis and tissue maintenance and repair in an adult organism, amyriad of stem cells are regulated by their surrounding extracellular matrix (ECM) enrichedwith tissue/organ-specific nanoscale topographical cues to adopt different fates and functions.Attributed to their capability of self-renewal and differentiation into most types of somaticcells, stem cells also hold tremendous promise for regenerative medicine and drug screening.However, a major challenge remains as to achieve fate control of stem cells in vitro with highspecificity and yield. Recent exciting advances in nanotechnology and materials science haveenabled versatile, robust, and large-scale stem cell engineering in vitro through developmentsof synthetic nanotopographical surfaces mimicking topological features of stem cell niches.
In addition to generating new insights for stem cell biology and embryonic development, thiseffort opens up unlimited opportunities for innovations in stem cell-based applications. Thisreview is therefore to provide a summary of recent progress along this research direction,with perspectives focusing on emerging methods for generating nanotopographical surfacesand their applications in stem cell research. Furthermore, we provide a review of classi- cal as well as emerging cellular mechano-sensing and -transduction mechanisms underlying
∗ Corresponding author at: Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109, USA. Tel.: +1 734 615 7363.E-mail address: [email protected] (J. Fu).

1 These authors contributed equally to this work.2 Current address: Department of Mechanical and Aerospace Engineering, New York University, New York, NY 10012, USA.

http://dx.doi.org/10.1016/j.nantod.2014.12.0021748-0132/© 2014 Elsevier Ltd. All rights reserved.

dx.doi.org/10.1016/j.nantod.2014.12.002

http://www.sciencedirect.com/science/journal/17480132

www.elsevier.com/locate/nanotoday

http://crossmark.crossref.org/dialog/?doi=10.1016/j.nantod.2014.12.002&domain=pdf

mailto:[email protected]

dx.doi.org/10.1016/j.nantod.2014.12.002

760 W. Chen et al.

stem cell nanotopography sensitivity and also give some hypotheses in regard to how a multitudeof signaling events in cellular mechanotransduction may converge and be integrated into corepathways controlling stem cell fate in response to extracellular nanotopography.© 2014 Elsevier Ltd. All rights reserved.

I

Rgrbi[tcbcoqsstcadtdtplstvbaat

cbeilwEsfiltooaIetilt

[pbgnaimieOcftono

pw[t[amcfoMmasteiet

pppossnccipography. We also provide some speculations as regard tohow these mechanotransductive events may converge with

ntroduction

apid advances in stem cell studies have unveiled thereat potential of stem cells as promising solutions foregenerative medicine, disease modeling, developmentaliology studies, and drug screening [1—6]. Stem cells,ncluding adult stem cells [7], embryonic stem cells (ESCs)8], and induced pluripotent stem cells (iPSCs) [9], sharehe ability of self-renewal and differentiation into specificell lineages, providing invaluable cell sources for variousiomedical and biological applications [6]. The self-renewalapability of stem cells enables rapid cell expansion with-ut losing cell stemness, critical for generating enough celluantities for large-scale cell-based applications. Directingtem cell differentiation into defined lineages with tissue-pecific mature functions is also important for treatingissue-specific degenerative diseases (such as neurologi-al, hepatic, hematopoietic, and diabetic diseases [1,4,6])nd in vitro disease modeling and drug screening. Amongifferent stem cells, adult stem cells possess limited,issue-specific regenerative potential and thus can onlyifferentiate into a few lineages [7]. In contrast, pluripo-ent stem cells (PSCs), including both ESCs and iPSCs,ossess the potential of differentiating into all three germayers, i.e. endoderm, mesoderm, and ectoderm, and sub-equently into any type of somatic cells [1,10]. Althoughogether, both adult and pluripotent stem cells can provideirtually unlimited cell sources for in vitro and in vivo cell-ased applications, a major technical hurdle remains as tochieve large-scale, high-efficiency cell expansion as wells directed differentiation into cell lineages of mature func-ions with high specificity and yield.

In the physiological stem cell niche, stem cells areonstantly challenged by both soluble cues and insolu-le, physical stimuli dynamically regulated in the localxtracellular matrix (ECM) [11,12]. The stem cell—ECMnterface is composed of structural units of nanometerength scales, which in turn regulate stem cell fate alongith other physical factors [13—16]. Specifically, in vivoCM is enriched with hierarchical fibers and fibrils con-isting of filamentous proteins such as collagen, elastin,bronectin, vitronectin, and laminin, presenting adhesive

igands on a structured landscape with spatial organiza-ions and characteristic dimensions of a few to hundredsf nanometers [17]. The helical surface topographical peri-dicity of individual ECM fibrils (e.g., collagen I) is alsophysical cue that can dictate stem cell behaviors [18].

n direct contact with the ECM, the cell membrane isnriched with adhesive molecules and protrusive struc-ures with characteristic nanometer length scales. For
nstance, integrin, the transmembrane receptor directlyinking ECM ligands to intracellular adaptor proteins andhe actin cytoskeleton (CSK), has a dimension of 20—50 nm
cfr

13,19]. Nanoscale filopodia (‘‘nanopodia’’), cell membranerotrusions containing bundled actin filaments, also haveeen shown in cellular probing of extracellular nanotopo-raphical features [20]. Cellular sensing of extracellularanotopographical cues through nanoscale architecturend dynamics of cell-ECM adhesions initiates downstreamntracellular mechanotransductive events, resulting in aultitude of nanotopography-sensitive cellular behaviors,

ncluding cell adhesion, morphology, proliferation, genexpression, self-renewal, and differentiation [16,21—27].wing to its potent role in regulating stem cell fate, extra-ellular nanotopography recently attracts much attentionrom bioengineers and materials scientists in an efforto achieve stem cell fate control using synthetic nan-topographical surfaces generated from different novelanofabrication technologies and material synthesis meth-ds [17,25].

A few key molecular players have emerged in accom-any with several principal mechanotransductive path-ays for regulating stem cell nanotopography sensitivity

19,28]. Specifically, existing evidence has suggestedhe involvement of integrin-mediated adhesion signaling29], CSK contractility (tension) and integrity [30,31],nd nuclear mechanics [32,33] in mechanosensing andechanotransduction of extracellular nanotopographical

ues [19]. However, it remains unclear how these dif-erent mechanotransductive cellular machineries functionr collaborate differently in different stem cell systems.oreover, it remains elusive how mechanoregulation atultiple levels (genetic and epigenetic, transcriptional,

nd post-transcriptional including microRNA) and timecales are integrated into a core regulatory networko control stem cell nanotopography sensitivity. Futurexploration of nanotopography-sensitive pathways will helpmprove rational designs of functional biomaterials fornhancing their performance in stem cell-based applica-ions.

A major goal of this review is to offer a summary of recentrogress on the new trend of engineering synthetic nanoto-ographical surfaces for controlling stem cell fate. We firstrovide a review of state of the art nanofabrication meth-ds for generating functional nanotopographical surfaces fortem cell studies, with a focus on those applicable for large-cale stem cell culture. We then highlight applications ofanotopographical surfaces in recent investigations for theontrol of stem cell fate. We discuss a few important intra-ellular mechanotransductive mechanisms that have beenmplicated in cellular responses to extracellular nanoto-

lassical signal transduction pathways to control stem cellate. We conclude by offering some perspectives on futureesearch directions and opportunities for leveraging stem

Nanotopography for stem cell fate control 761

Table 1 Comparison of methods for generating nanotopography.

Fabrication technique Cost Throughput Controllability offeature shape

Controllability offeature size

Materialcompatibility

Photolithography ** *** *** * ***

E-beam lithography *** * *** *** *

Colloid lithography ** *** * * *

Nanoimprinting *** *** *** *** **

Replica molding * *** *** *** **

Chemical etching * ** * ** *

RIE ** *** * *** *

Electrospinning * ** ** * ***

Phase separation * ** ** ** **

Anodization * ** * ** *

Sintering * ** * ** **

* Low.** Medium.

*** High.

PDltmpssteroitcoaopwn5oaphelfings[

cell nanotopography sensitivity for engineering stem cellfate and function.

Fabrication of nanotopographical surfaces

Various nanoengineering tools and synthesis methods havebeen successfully developed and utilized to generate nan-otopographical surfaces or scaffolds for in vitro stemcell research. Based on their fabrication principles, thesetechniques can be classified into four different groups:lithographic patterning, pattern transfer, surface roughen-ing, and material synthesis (Fig. 1 and Tables 1 and 2).Lithographic patterning and pattern transfer are two top-down approaches that utilize predefined patterns to createnanotopographical features on two-dimensional planar sur-faces. Surface roughening and material synthesis, on thecontrary, directly generate nanostructures on material sur-faces from the bottom up using chemical or physicalmeans. Together, these methods present a wide spectrumof fabrication tools capable of generating nanotopographi-cal features of a wide range of sizes and geometries, andeven hierarchical (micro-)nanotopographical surfaces. Tosuccessfully utilize stem cell—nanotopography interactionsfor stem cell applications, it is important to understand andappreciate advantages and limitations of each of availablenanoengineering tools and synthesis methods for generatingextracellular nanotopography in terms of fabrication cost,throughput, materials, controllability of feature shape, sizeand accuracy (Table 1).

Lithographic patterning

A variety of lithographic patterning methods, includingphotolithography [34], electron beam lithography [35—38],
and colloidal lithography [39—44], have been successfullyapplied to generate extracellular nanotopography of dif-ferent size ranges and spatial organizations on planar 2Dsurfaces following pre-defined patterns (Table 2).
ETU

hotolithographyeveloped from semiconductor microfabrication, photo-

ithography, or optical lithography, is the most popularechnique for surface patterning at micron and sub-icron scales. In photolithography, defined geometricatterns are transferred from a photomask to a light-ensitive organic material (photoresist) coated on a planarubstrate via ultraviolet (UV) light exposure. After pho-olithographic patterning of positive/negative photoresist,xposed/protected regions of the photoresist can beemoved with the protected/exposed regions remainingn the substrate, serving as a lithographic mask faithfullynheriting pre-defined patterns from the photomask andransferring it to the substrate with subsequent etching pro-esses [45,46]. The finest resolvable dimension (resolution)f photolithography is limited by UV light wavelength as wells the ability of reduction lens to capture enough diffractionrders from illuminated photomask [45]. State of the arthotolithography using deep UV light from excimer lasersith wavelengths of 248 and 193 nm allows fabrications ofanoscale structures with a minimum feature size down to0 nm [45—48]. Due to the expense and limited accessibilityf photolithography instruments for sub-100 nm fabrication,pplication of photolithography for fabrication of nanoto-ographical features has been limited to a length scale ofundreds of nanometers (Table 1). It should be noted, how-ver, that the capability of photolithography in fabricatingarge-area, arbitrarily designed sub-micron topographicaleatures has rendered it the most popular surface pattern-ng technique for biomedical applications. As an example,anoscale gratings have been fabricated using photolitho-raphy on silicon substrates for studying nanotopographicensing by human ESCs and endothelial progenitor cells34,49].

lectron beam lithographyo overcome resolution of photolithography limited byV wavelength, other sources of illumination have been

762 W. Chen et al.

Figure 1 Fabrication of nanotopographic surfaces. Lithographic patterning. (a) A nanogrooved silicon substrates with 70 nm wideridge and 400 nm pitch fabricated by EBL [36]. Reproduced with permission [36]. Copyright 2003, Biologists Ltd. (b) Regular (left)and random (right) arrays of 120-nm-diameter, 100-nm-deep nanopits on silicon substrates fabricated by EBL [37]. Reproducedwith permission [37]. Copyright 2007, Nature Publishing Group. (c) A nanotopographic substrate fabricated by the self-assembly of110-nm-diameter nanoparticles [54]. Reproduced with permission [54]. Copyright 2003, IEEE. Pattern transfer. (d) Nanostructuredpolyurethane acrylate (PUA) surface with a patterned array of nanopillars fabricated by nanoimprinting from a silicon master. Thediameter of the pillars was 300 nm and the gap between the pillars was 900 nm [68]. Reproduced with permission [68]. Copyright2013, American Chemical Society. (e) PDMS nanograting produced by replica molding from a PMMA master [69]. Reproduced withpermission [69]. Copyright 2005, Elsevier. (f) AFM scan of a PCL surface with nanopits produced by replica molding from a pillaredquartz master [38]. Reproduced with permission [38]. Copyright 2002, IEEE. Surface roughening. (g) Nanostructured PCL with featuredimensions of 50—100 nm fabricated by NaOH etching [77]. Reproduced with permission [77]. Copyright 2003, Wiley Periodicals,Inc. (h) Nanoroughened Ti surface (surface roughness Ra = 0.87 ± 0.03 �m) fabricated by acid etching in hydrochloric acid/sulfuricacid [78]. Reproduced with permission [78]. Copyright 1998, John Wiley & Sons, Inc. (i) Glass surfaces with surface roughness of100 nm fabricated by RIE [80]. Reproduced with permission [80]. Copyright 2013, American Chemical Society. Material synthesis. (j)Aligned nanofibrous hydroxybutyl chitosan (HBC) scaffolds fabricated by electrospinning [104]. Reproduced with permission [104].Copyright 2007, WILEY-VCH Verlag GmbH & Co. (k) Nanofibrous PLLA matrix with an average fiber diameter of 148 ± 21 nm and aporosity of 92.9% fabricated by phase separation [114]. Reproduced with permission [114]. Copyright 2009, Mary Ann Liebert, Inc. (l)Self-aligned TiO2 nanotubes with a diameter of 100 nm generated by anodizing Ti sheets under a potential of 20 V [139]. Reproducedwith permission [139]. Copyright 2009, National Academy of Sciences of the USA. (m) Nanostructured alumina substrates with 24-nmgrain-like structures fabricated by sintering [146]. Reproduced with permission [146]. Copyright 2008, Wiley Periodicals, Inc.

Nanotopography

forstem

cellfatecontrol

763

Table 2 Summary of various nanotopographic methods for stem cell studies.

Fabrication technique Topography Feature size Material Cell type Major result Reference

Lithography &replica-molding

Grating 600 nm PDMS Endothelialprogenitor cells

Supports in vitro Capillary TubeFormation

[34]

Lithography &replica-molding

Grating 600 nm PDMS Human ESC Alters ESC morphology andproliferation

[49]

E-beam lithography Nanopit Array 120 nm Silicon Human MSC Control of human MSC differentiationusing nanoscale symmetry anddisorder

[37,276]

Colloid lithography Island 120—600 nm Silica Mouse ESC Maintains mouse ESC self-renewal [57]RIE Random roughness 1—150 nm Glass Human ESC Regulates ESC adhesion,

proliferation, self-renewal anddifferentiation

[80]

Electrospinning Random nanofibers 200—400 nm PCL Human & rat MSC Promoted osteogenesis of MSCs [95,96]Electrospinning Random nanofibers 500—700 nm PCL Human MSC Support in vitro chondrogenesis of

MSCs[97]

Electrospinning Random & alignednanofibers

250—930 nm PCL Rat NSC Aligned nanofibers promote neuronaldifferentiation

[98]

Electrospinning Random nanofibers 280 nm PCL Human ESC Support ESC colony formation andmaintain stemness

[99]

Electrospinning Random & alignednanofibers

250 nm PCL Mouse ESC Aligned nanofibers promote neuronaldifferentiation and neurite outgrowth

[100]

Electrospinning Random nanofiber 283—1452 nm PES Rat NSC Fiber diameter regulatedifferentiation and proliferation ofrat NSCs

[101]

Electrospinning Nanofiber 529 ± 114 nm PES Human HSC Nanofibrous surface supports HSCexpansion

[102]

Electrospinning Nanofiber 180 nm Polyamide Mouse ESC Promote the proliferation andself-renewal of mouse ESCs

[94,157]

Electrospinning Random and alignednanofibers

150—1150 nm PLLA Mouse NSC Aligned nanofibers enhance NSCdifferentiation

[87,88]


500 nm CNT/PLLA Mouse ESC Aligned nanofibers enhance NSCdifferentiation

[89]


196—253 nm P(LLA-CL) Mouse NSC Promote NSC proliferation anddifferentiation

[90]

Electrospinning Random nanofibers 230—620 nm P(LLA-CL) Mouse MSC Promote neuronal differentiation [91]Electrospinning Random nanofibers 250—300 nm PLGA Mouse ESC Affects ESC fate [92]Electrospinning Random nanofibers 50—300 nm/2—4 �m Fibrin/PLGA Human MSC Promotes cardiomyogenesis [148]Electrospinning Random nanofiber Not available PMGI Mouse ESC Maintain mouse ESC culture under

feeder-free conditions[103]

764W

.Chen

etal.

Table 2 (Continued)

Fabrication technique Topography Feature size Material Cell type Major result Reference

Electrospinning Aligned nanofiber 436 nm HBC Human MSC Induces cell alignment andmyogenesis

[104]


369—554 nm PHBV/HA Rat MSC Random nanofibers promote in vitroproliferation and early osteogenesis

[105]

Electrospinning Random nanofibers Not available HA/CTS Mouse MSC Enhances osteogenic differentiation [106]Phase separation Random nanofiber 50—500 nm PLLA Mouse ESC Promotes osteogenic differentiation [114]Phase separation Worm-like and

dot-like patterns160 nm Polystyrene-

blockpoly(2-vinylpyridine)

Human MSC Influences MSC morphology,proliferation

[127]

Anodization Nanotube 30—100 nm TiO2 Human MSC Regulates human MSC differentiationtoward osteoblast lineage

[139]

Anodization Nanotube 15—100 nm TiO2 Rat MSC Influences adhesion, morphologyproliferation and osteogenesis

[14]

Anodization Nanopillars 15—100 nm Titania Human MSC Feature height affects human MSCadhesion, morphology anddifferentiation

[140]

Sintering Grain-like structure 24—1500 nm Alumina, titania, andhydroxyapatite

Human MSC Affects human MSC adhesion andproliferation

[146]

Graft Random nanofibers Not available Poly(acrylicacid)/CNT

Human ESC Direct and promote human ESCderived EBs differentiation intoneuron cells

[166]

Nano-imprinting Grooves and pillars 300—1500 nm PUA Human NSC Nanopatterns influence morphology,alignment, focal adhesion, andneuronal differentiation

[68]

Nano-imprinting Ridge/groove 350 nm PUA Human ESC Induces neuronal differentiation [168]Nano-imprinting and

soft lithographyGrating 350—1000 nm PDMS Human MSC Induce differentiation of human MSCs

into neuronal lineage[75]

Nano-imprinting Nanopit 50 nm Polycarbonate Human ESC Influences human ESC differentiation [162]

P

Ptpfitbf

NNapotgtttspt(bt

a[pe[bm[gt

RWmeubrptrasrp

pthmt

Nanotopography for stem cell fate control

exploited for lithography, such as electron beams and x-rays.As a powerful nanofabrication method, electron beamlithography (EBL) [50] has been developed for creating well-defined patterns with feature sizes <10 nm, owing to areduced wavelength of electron waves. Instead of using pho-tomasks and UV light, in EBL a focused electron beam isutilized to selectively expose an electron-sensitive organicresist (such as poly(methyl methacrylate), or PMMA) follow-ing a pre-defined scanning path. Fabrication of periodic linepatterns with a line width of 5—7 nm has been successfullydemonstrated using EBL [35]. Owing to its versatility in gen-erating nanoscale features with dimensions comparable tocell adhesion structures, EBL has been commonly utilizedfor generating nanotopographical surfaces to regulate cell-ECM adhesive interactions on both silicon and fused silicasurfaces (Fig. 1a and b).

Although EBL can achieve precise surface patterning witha sub-100 nm resolution, given its ‘‘direct writing’’ nature,its throughput is extremely low, limiting its application tolow-volume productions of nanotopographical surfaces witha limited surface area. The high cost of EBL machine hasalso limited its access for pilot stem cell studies where alarge number of nanotopographical surfaces with differentgeometrical patterns are desired for high-content screeningof nanotopography that is optimal for specific stem cell fateregulation.

Colloidal lithographyBoth photolithography and EBL are top-down lithography-based approaches. As a totally different bottom-upapproach, colloid lithography has been developed to gen-erate surface topological features down to a nanometerscale with high-throughput, relatively simple fabricationstrategies and reasonable cost [38,53—56]. In colloidallithography, self-assembled nanoparticle crystal structureswith a short-range intrinsic order are created on planarsurfaces to serve as masks for subsequent etching pro-cesses (Fig. 1c) [39—42]. Colloidal lithography does notpossess the accuracy and on-demand control over spatialpatterns it can generate, as compared to photolithographyand EBL. To generate nanotopography with different spatialpatterns, colloidal lithography can utilize different crystalstructures of self-assembled colloidal masks and alter inci-dence angle of plasma that etches the underlying surface[39—42].

With proper surface functionalization, self-assembledcolloidal monolayers can be directly used as nanoto-pographical substrates for cell assays [54,57,58]. Inaddition, colloid lithography has been applied to gener-ate nanoscale topographical features on curvilinear surfacesof microscale particles and scaffolds to create hierar-chical topographical biomaterials that are difficult toobtain using conventional methods [39—41,56]. Combiningnanoscale patterned surfaces generated by photolithogra-phy and EBL with colloid self-assembly, it is also possibleto achieve regular patterns of nanotopography with curvi-linear local geometries [59]. Instead of serving as an
etching mask, self-assembled colloidal nanoparticle layerscan also act as a deposition mask to generate nanoto-pography with a negative pattern of the colloidal mask[60—62].
wPni

765

attern transfer

attern transfer utilizes pre-existing nanotopographical fea-ures on a rigid mold to transfer such features using moldingrocesses to other materials with high-efficiency and high-delity. Here we discuss two types of pattern transferechniques, nanoimprinting and replica molding, that haveeen commonly used for generating nanotopographical sur-aces for stem cell studies.

anoimprintinganoimprinting utilizes hot embossing of thermoplastics tochieve nanoscale surface patterning [63—65]. In nanoim-rinting, a thin layer of thermoplastic polymer is spreadr spin-coated onto a planar, featureless substrate. A hardemplate (‘‘mold’’) containing pre-fabricated nanotopo-raphical features is then brought into a direct contact withhe polymer layer (‘‘imprinting’’), and they are pressedogether under pressure. When heated up above the glassransition temperature of the polymer, the polymer on theubstrate melts and conforms to the pattern of the tem-late. The polymer layer forms an inverse replica of theemplate after cooling and separation from the template‘‘demolding’’) (Fig. 1d). In nanoimprinting, adhesionetween the polymer and the template is carefully con-rolled to allow proper release of polymeric substrates.

A key advantage of nanoimprinting is its ability tochieve high resolution nanofabrication (down to 14 nm66]) over a large surface area using a relatively simplerocedure; however, nanoimprinting requires specializedquipments to orchestrate its different processing steps67]. To date, synthetic nanotopographical surfaces haveeen created by nanoimprinting using thermoplastic poly-ers such as PMMA [69] and polyurethane acrylate (PUA)

68] and more recently with platinum-based bulk metalliclass alloy (Pt-BMG), an inorganic material susceptible tohermoplastic forming [70].

eplica moldinghile nanoimprinting uses thermoplastic materials, replicaolding is a method of replicating structures into an

lastomeric polymer material that hardens after bakingnder elevated temperature. Conceptually, replica moldingelongs to a larger class of methods called soft lithog-aphy that was developed complementary to traditionalhotolithography [71,72]. Polydimethylsiloxane (PDMS) ishe most frequently used elastomeric polymer material foreplica molding owing to its excellent physical propertynd biocompatibility. The procedure of replica molding istraightforward and easy for implementation in a regularesearch laboratory, rendering this method ideal for rapidrototyping.

In replica molding, PDMS precursor in a liquid state isoured directly onto a mold with surface topography inverseo that is desired. Upon heating, PDMS cures, resulting in aardened PDMS layer that can be readily released from theold. The final PDMS layer presents a replica that retains

he size and geometry of the original topography on the mold
ith high fidelity (Fig. 1e). As typical elastomers includingDMS can be readily deformed by surface tension, they areot suitable for molding to generate nanoscale topograph-cal features with high fidelity. Instead, ‘‘hard’’ thermally

7

cbiTpgP(esf

S

Epdtinasorpr

CArpmwmfu(icpoft[tecmpA([

RBaanbda

ssetotgttgrr[bproRtgc

M

Desar

EEnacrsotc

vasrpsefencrn

it

66

urable polymers such as h-PDMS [73] and PUA [74] haveeen utilized for replica molding when desired topograph-cal features have critical dimensions of less than 500 nm.o study cellular responses to extracellular nanotopogra-hy, replica molding has been applied to generate nanoscaleratings on PDMS using either silicon [34] or nanoimprintedMMA molds [69,75]. Other than PDMS, polycaprolactonePCL), a biodegradable polyester, has also been used as alastomeric polymer for replica molding to transfer nanopittructures with a diameter of 50 nm from a silicon moldabricated by EBL [38] (Fig. 1f).

urface roughening

ven without external templates or masks, a uniformhysical or chemical etching environment can generate ran-om nanotopographic features on certain material surfaceshrough ‘‘surface roughening’’ processes [76—80]. Althought is intrinsically difficult for surface roughening to achieveanotopographic features with accurate feature patternsnd sizes, they are effective for generating nanotopographicurfaces of large surface areas owing to its simple methodol-gy and compatibility with bulk processing. To date, surfaceoughening has been utilized for generating random nanoto-ographic features using either chemical etching or physicaleactive ion etching (RIE) processes.

hemical etchings its name suggests, chemical etching relies on chemicaleactions between etchants and substrate surfaces to berocessed to remove materials from the surfaces. Due toaterial inhomogeneities at the atomic scale or nanoscaleithin the substrate, etchants locally react and removeaterials from the surface in an anisotropic fashion at dif-

erent rates, resulting in an uneven landscape with surfacendulations (surface roughness) down to a nanometer scaleFig. 1g and h). Surface roughening during chemical etch-ng depends on experimental conditions such as etchantomposition and concentration and etching time and tem-erature [76,81]. It is known, for example, that etchingf silicon surfaces with KOH can result in nanoscale sur-ace roughness, the level of which increases with etchingime but decreases with increasing reaction temperature76]. Importantly, chemical etching can also be appliedo some biocompatible materials commonly used in tissuengineering. For example, etching using acids or bases canreate nanotopography on surfaces of biocompatible poly-er materials such as poly (lactide-co-glycolide) (PLGA),oly-ether-urethane (PU), and polycaprolactone (PCL) [77].cid etching can also generate nanotopography on titaniumTi) surfaces, a common metallic material used for implant78].

eactive ion etchingesides chemical etching, nanotopographical surfaces canlso be generated on silica-based glass surfaces with RIE,

well-established semiconductor microfabrication tech-
ique. During RIE etching on glass surfaces, bombardmenty reactive ion species generated using SF6 and C4F8 gasesisrupts un-reactive glass substrate and causes damage suchs dangling bonds and dislocations, resulting in the glass
tooi

W. Chen et al.

urface reactive toward etchant species. Importantly, sincemall concentrations of impurities such as Al, K and Naxist in silica glass, these impurities result in accumula-ions of less volatile species (such as AlF3, KF, NaF, etc.)n glass surface during the RIE process [82,83]. Some ofhese less-volatile compounds are then backscattered ontolass surface and form randomly distributed small clustershat can shield glass surface from bombardment and reac-ion with reactive ions. These compound clusters effectivelyenerate the so-called ‘‘micro masking’’ effect that canandomly shadow glass surface and thus result in nanoscaleoughening of glass surface during the RIE process (Fig. 1i)82,83]. In practice, nanotopography of glass surface cane controlled by adjusting the RIE process duration. Com-ared with chemical etching, RIE usually has a lower etchingate and thereby allows a more precise control over nan-topography generated on substrate surfaces. In addition,IE is compatible with other semiconductor microfabricationechniques such as photolithography, and as such one canenerate patterned nanorough islands on a flat glass surfaceombing RIE with photolithography [80].

aterial synthesis

ifferent material synthesis methods developed from tissuengineering and other research fields, such as electro-pinning, phase separation, anodization, and sintering, canlso be utilized for fabricating nanotopographical biomate-ials useful for stem cell studies.

lectrospinninglectrospinning is a long-existing polymer processing tech-ique which was rediscovered in the early 1990s for itspplications in the field of tissue engineering owing to itsapability of generating nanofibrous constructs from a broadange of polymers in a simple setting [84—86]. Since electro-pinning does not require the use of coagulation chemistryr high temperature to produce solid threads from solution,his process is suited to production of fibers with large andomplex molecules.

In electrospinning, a high voltage in the range of kilo-olts is applied to a pendent droplet of polymer solution,nd the body of the solution becomes charged. Electro-tatic repulsion in the droplet counteracts surface tension,esulting in stretching of the droplet. When the voltageasses a certain threshold, force balance between electro-tatic repulsion and surface tension breaks, and a liquid jetrupts from the droplet surface. The liquid jet undergoesurther stretching due to electrostatic force and solventvaporation, reducing fiber diameter to as small as a fewanometers. Arrangements of electrospun nanofibers can beonveniently controlled from completely random to unidi-ectionally aligned, producing a wide span of extracellularanotopographical textures (Fig. 1j) [87—92].

A significant advantage of electrospinning for study-ng stem cell—nanotopography interactions is its capabilityo generate three-dimensional nanofibrous architectures
hat can properly mimic the structure and organizationf in vivo ECM [93]. Its compatibility with many typesf polymers and feasibility for multiplexed functional-zation has rendered electrospinning a popular method

dcpu(gib

bltbade‘c[bfpchPrwdparabtd

AAfsipdasfrrt

nethfsf


for different tissue engineering applications. For exam-ple, both random and aligned nanofibrillar matrices havebeen fabricated using electrospinning of polymers suchas polyamide [94], PCL [95—100], polyethersulfone (PES)[101,102], poly(L-lactic acid) (PLLA) [87—91], PLGA [92],polymethylglutari-mide (PMGI) [103], hydroxybutyl chitosan(HBC) [104], poly(3-hydroxybutyrate-co-3-hydroxyvalerate)(PHBV) [105], and hydroxyapatite/chitosan (HA/CTS) [106].Fiber functionalization can be achieved by simply mixingadhesive ECM proteins such as collagen and gelatin into poly-mer solution before electrospinning [90,91]. Multiplexedfunctionalization of electrospun nanofibrous constructs canbe accomplished by tailored grafting of polymer chains.Patel et al., for example, have developed a multifunc-tional PLLA nanofiber biomaterial presenting both lamininand basic fibroblast growth factor (bFGF) in addition tothe intrinsic nanotopography, via grafting through heparinand di-amino-poly(ethylene glycol) (di-NH2-PEG) as linkers[107]. Electrospinning has also been applied to fabricatenanofibrous structures composed of silk fibroin [108], show-ing its compatibility with native proteins.

Phase separationThermally induced phase separation has long been utilizedfor fabrication of nanoporous scaffolds for tissue engineer-ing applications [109—112]. Phase separation occurs whenthe concentration of polymer or polymer blends exceeds itssolubility in solvent. It is convenient to control the mixturetemperature to initiate phase separation in polymer solu-tion and subsequent generations of micro- and nanoscalepolymer-rich and polymer-lean phase domains within themixture [112]. After removal of solvent, polymer-rich phasedomains solidify and form polymeric foam that appearsas nanotopographical scaffolds (Fig. 1k). By varying thetypes of polymers and solvents, polymer concentration, andphase separation temperature, different nanotopographicmorphologies and structures can be created from phaseseparation. For example, to recapitulate the fibrous archi-tecture of type I collagen, phase separation has beenutilized for fabricating synthetic biodegradable PLLA nanofi-brous matrices with fiber diameters ranging from 50 nm to500 nm [113,114]. Other than nanofibrous matrices, phaseseparation has also been used for generation of nanoto-pography such as nanoscale islands and pits using blends ofpoly(p-bromostyrene) and poly(deuteriostyrene) in conjunc-tion with a spin-coating method [115].

Phase separation can also be induced at room tempera-ture to fabricate nanoscale islands of polymers based on abulk demixing process for a binary polymer mixture, suchas the polystyrene (PS) and poly(4-bromostyrene) (PBrS)mixture in a spin-cast thin film [116—119]. Dimension anddensity of nanotopographical features generated by polymerdemixing depend on mixture composition and componentconcentration. Using demixing of a PS/PBrS mixture, Dalbyet al. have successfully fabricated arrays of circular islandswith a diameter of 10—100 nm to study cellular responses tonanotopography [117]. Nanotopography patterns generated
by polymer demixing can also be regulated by pre-definedinterfaces at the polymer mixture-substrate boundary. Inan early study, Sprenger et al. have developed a hierarchi-cal nanotopographical surface by applying ternary polymer
om

b

767

emixing onto a pre-patterned substrate prepared by micro-ontact printing [120]. Specifically, a pre-defined microscaleattern of octadecylthiol was generated on a gold surfacesing microcontact printing to exclude poly(2-vinylpyridine)PVP) phase of the PS/PMMA/PVP mixture onto polar, bareold regions, where they retain a non-polar PS phase, result-ng in a wall-like PMMA phase of 200—300 nm thickness inetween the microscale patterns.

Block copolymer lithography is another phase separation-ased nanofabrication method with unique advantages ofow cost and rapid casting [121—124]. Due to phase separa-ion driven by interactions between different segments oflock copolymers, block copolymer lithography can gener-te a rich set of domain-like patterns with a characteristicimension ranging from micro- to nanoscale [122]. Inter-stingly, block copolymer lithography was named after‘lithography’’ because the patterns obtained via blockopolymer self-assembly was first used as lithography masks121]. In recent years, block copolymer lithography haseen directly applied to generate nanotopographical sur-aces for cell assays. Maclaine et al., for example, have usedoly(styrene-block-poly-2-vinylpyridine) (PS-b-P2VP) blockopolymer micelles to generate nanoscale islands of 20 nm ineight and 150 nm in in-plane periodicity [125]. Using a novelS-PDMS diblock copolymer, Salaun et al. have achieved fab-ication of PDMS nanopillars of 10 nm in height and 20 nm inidth [126]. Interestingly, by using PS-b-P2VP and PS-b-P4VPiblock copolymers, Khor et al. have generated two distinctatterns of nanotopography, namely, ‘‘dot-like’’ (‘‘salt-nd-pepper’’) and ‘‘worm-like’’ (‘‘labrynth’’) patterns,espectively [127]. In addition, emergent patterns gener-ted in functionalized block copolymer thin films can alsoe used to spatially define nanoscale arrays of gold nanopar-icles [124,128,129] and proteins [130], thus expanding theiversity of nanotopographical cues for cell assays.

nodizationnodization is an electrochemical process commonly usedor increasing thickness of the oxide layer on metal or alloyurfaces [131—138]. During anodization, metal is immersedn an electrolytic solution and connected as the anode (theositive electrode) of an electrical circuit. A current passingirectly through the electrolytic solution releases hydrogent the cathode (the negative electrode) and oxygen at theurface of the metal anode, building up an oxide layer of aew to tens of nanometers thick. Due to heterogeneous localeaction rates at the metal surface, the electrochemicaleaction in anodization can generate nano- and microscaleextures on the metal surface.

Anodization has been successfully used for fabrication ofanotopographic surfaces for stem cell studies (Fig. 1l). Forxample, Park et al. [14] have applied controlled anodiza-ion of Ti in a fluoride-containing electrolyte to generateighly ordered layers of titania (TiO2) nanotube structuresor promoting osteogenic differentiation of mesenchymaltem cells (MSCs) [14,139]. Similarly, Sjöström et al. haveabricated TiO2 nanopillars with pillar heights of 15—100 nm
n Ti surfaces using anodization through a porous aluminaask [140].Other than additive nanofabrication, anodization has also
een applied in fabrication of nanoporous substrates via

7

ehrsigteRn[

SSmdmaopcsDto1

N

Isspbptnf[s

M

Mcopctdip

MosPt(5s

dMtafdm

E[mrtlodoiceo

tcslwsisrddtZchwfndotooofpfsio

csutf

68

lectrochemical etching. In a recent study, Clements et al.ave developed a gradient nanoporous silicon (pSi) biomate-ial for rat MSC culture [141]. Using an asymmetric electrodeetup, a gradient of anodic etching generated a correspond-ng gradient of nanoscale pores in a Si wafer, with pore sizesradually decreasing from 920 nm to 20 nm as the distanceo the electrode increased. Surface functionalization usinglectrografting further generated an orthogonal gradient ofGD peptides and thus enabled study of synergistic effect ofanotopography and biochemical gradients on MSC behaviors141].

interingintering is a widely used method for creating objects frometal and ceramic powders [142—145]. In sintering, pow-ered materials are heated to a temperature below theaterial’s melting point but high enough for diffusion of

toms, causing fusion of different powder particles intone solid piece. Although sintering is mostly used in bulkrocessing, it can also create granular surface patterns withritical dimensions in nanometer size scales (Fig. 1m) whenintering temperature and duration are carefully controlled.ulgar-Tulloch et al., for example, have fabricated alumina,itania, and hydroxyapatite surfaces with grain-like nan-topography of critical dimensions ranging from 24 nm to500 nm for MSC culture [146].

anotopography controls stem cell fate

n vivo, stem cells interact with and interrogate theirurroundings at the micrometer and/or nanometer lengthcale. Plenty of evidence exists to suggest that nanoto-ographic signals from the local stem cell niche instructehaviors of stem cells. Here we provide illustrative exam-les using bioengineering and nanofabrication approacheso control nanotopographic features of the local stem celliche and, where evidence suggests, regulate stem cellate through synergistic regulations of stem cell shape16,22,23], CSK tension [30,31], integrin-mediated adhesionignaling [29], and nuclear mechanics [32,33].

esenchymal stem cells

esenchymal stem cells (MSCs) are multipotent cells thatan give rise to cells of the stromal lineage, namely,steoblastic (bone) [14,18,37,95,96,105,106,139,140], adi-ogenic (fat) [23,147], myoblastic (muscle) [104,148],hondrogenic (cartilage) [97], and fibroblastic (connectiveissue) [37] lineages. In recent studies, abundant evi-ence has been obtained to support that MSC fate can benfluenced by the size, symmetry, and regularity of nanoto-ographical features.

It has been commonly observed that both rat and humanSCs on nanofiber scaffolds show significantly enhancedsteogenic differentiation compared to conventional tis-ue culture plates [14,95,96,105,106,139]. In addition,ark et al. [14] observed that 15 nm diameter TiO2 nano-
ubes promoted osteogenic differentiation of rat MSCsFig. 2a). However, when the tube diameter increased to0 nm or above, osteogenic differentiation of rat MSCs wasignificantly diminished. More recently, Oh et al. [139]
tIns

W. Chen et al.

emonstrated that TiO2 nanotube arrays directed humanSC differentiation toward the osteoblast lineage even in

he absence of soluble inductive factors. In this study, amongrrays of 30, 50, 70, and 100 nm diameter TiO2 nanotubesabricated on Ti substrates, the one composed of 100 nmiameter nanotubes showed the highest potential for pro-oting human MSC osteogenic differentiation.Notably, while recent studies have shown enhanced

CM proteins adsorption on nanotopographical surfaces149,150], direct effect of such ECM protein adsorptionay be secondary compared to the effect of nanotopog-

aphy sensitivity on stem cell fate regulation. Supportinghis view, Oh et al. [139] proposed that instead of regu-ating stem cell behaviors through adsorbing more proteinsnto surface, instead, nanotopographic surfaces resulted inistinct sizes and spacing of aggregated ECM protein on nan-topographic surfaces, which in turn induced changes inntegrin-mediated focal adhesion (FA) formation and stemell function. Similar results were also obtained by Chent al. [80] for hESCs on glass surfaces with varied nanor-ughness.

In addition to in-plane nanotopographical features, thehird dimension, i.e., the nanoscale height of topologi-al features, can also play a potent role in regulatingtem cell fate. Using 15, 55, and 100 nm high Ti nanopil-ars, Sjöström et al. [140] showed that Ti nanopillarsith the smallest (15 nm) height resulted in the most

ignificant bone matrix nodule formation, suggesting anncreased level of osteogenic differentiation of humankeletal stem cells. A different observation, however, waseported recently by Zouani et al., wherein the authorseveloped nanotopographical features with similar in-planeimensions yet different depths using UV-mediated pho-odegradation of polyethylene terephthalate (PET) [151].ouani et al. observed that a greater nanotopographi-al depth (100 nm) promoted osteogenic differentiation ofuman MSCs, whereas much less osteogenic differentiationas observed for surfaces with shallower (10 nm depth)

eatures. The inconsistent observations about differentanotopographical feature sizes favorable for osteogenicifferentiation support cell-type specific response to nan-topography. Interestingly, recent studies have also shownhat MSC fate can be influenced by symmetry and regularityf nanotopographical patterns. Specifically, nanoscale dis-rder of a nanopit array significantly promoted osteogenesisf human MSCs even in the absence of soluble inductiveactors, whereas highly ordered nanotopographical surfacesroduced much limited cellular adhesion or osteogenic dif-erentiation [37]. Using silica nanohelices grafted on glassurfaces, Das et al. [18] also demonstrated that the periodic-ty of chiral extracellular topography directed commitmentf human MSCs toward the osteoblast lineage.

In addition to osteogenesis, other specific cell lineageommitments, including cardiomyogenesis [148], myogene-is [104] and chondrogenesis [97], have also been exploredsing nanotopographical biomaterials. For example, syn-hetic biodegradable PCL nanofibrous scaffolds were utilizedor enhancing in vitro chondrogenesis of MSCs compared to
he established cell pellet culture method (Fig. 2c) [97].n addition to chondrogenesis, it was demonstrated thatanofibrous topography regulated the actin CSK and nuclearhape of MSCs and induced MSC differentiation into the


Figure 2 Nanotopography regulates MSC lineage. (a) TiO2 nanotube diameter directs MSC osteogenic differentiation [14]. The toprow presents SEM images of highly ordered TiO2 nanotubes of small (15 nm; left) and large (100 nm; right) pore sizes. The bottomrow presents immunofluorescence images of cells with osteocalcin staining in red and F-actin staining is in green on 15 and 100 nmnanotubes after 2 weeks in culture in osteoblast differentiation medium. Osteogenic differentiation occurred 15 nm nanotubes asseen by osteocalcin staining but rarely detectable on 100 nm nanotubes. Reproduced with permission [14]. Copyright 2007, AmericanChemical Society. (b) Using of fibrin nanofiber and PLGA microfiber composite scaffold for human MSC differentiation towardmyocardial lineage [148]. The top row presents SEM micrographs of PLGA—fibrin electrospun membrane at different magnification.The bottom row presents confocal microscopy images of MSCs grown on PLGA—fibrin composite fibers after 14 days of inductionof cardiac differentiation expressing of cardiac specific markers including cardiac troponin, �-sarcomeric actinin, tropomyosinas indicated. Reproduced with permission [148]. Copyright 2013, Mary Ann Liebert, Inc. (c) Nanofibrous scaffold for human MSCchondrogenesis [97]. The left column shows SEM images of nanofibrous surface (top) produced by the electrospinning process showingrandom orientation of ultra-fine fibers with an diameter ranging from 500 to 900 nm and MSCs (bottom) with round, ECM-embeddedchondrocyte-like cells on the surface after 21 days culture in the presence of TGF-�1. The middle and right columns present thehistological analysis of human MSC cultured on nanofibrous surface in a chondrogenic medium supplemented with TGF-�1 for 21days. Sections from the upper and lower portions of the three-dimensional constructs were stained with H&E (middle column) andAlcian blue (right column). H&E staining showed flat fibroblast-like cells on the top zone (bracket, *), round chondrocyte-like cellsembedded in lacunae (arrows) in the middle zone (bracket, **), and small, flat cells at the bottom zone (bracket, ***). Alcian bluestaining showed the presence of sulfated proteoglycan-rich ECM in the construct. Reproduced with permission [97]. Copyright 2005,Elsevier. (d) Nanograting surfaces induce neuroal transdifferetiation of MSCs [75]. SEM (top) and immunofluorescence (bottom)images of human MSCs cultured on nanograting and unpatterned PDMS as indicated. In the immunofluorescence images, cells were

were

Mdeˇ

ce

differentiated for 14 days in the presence of retinoic acid andReproduced with permission [75]. Copyright 2007, Elsevier.

myogenic/myocardial lineages even without soluble induc-tive factors [104,148].

Nanotopographic cues have also been demonstrated inrecent studies for inducing neural transdifferentiation ofMSCs. A recent report by Woodbury et al. [152], for example,has shown that MSCs seeded on nanotopographical surfaces
can transdifferentiate into neural cells when cultured ina neural induction medium. Interestingly, another recentstudy by Yim et al. [75] has shown that grating-like nan-otopography alone can induce neural transdifferentiation of
CpMt

stained with neuronal marker Tuj1in red and GFAP in green.

SCs (Fig. 2d). Gene expression and immunostaining assaysemonstrated significant up-regulation of neuronal mark-rs such as microtubule-associated protein 2 (MAP2) and-tubulin III (Tuj1) for MSCs cultured on nanograting surfacesompared to unpatterned or micropatterned controls. Yimt al. [75] further proposed that nanotopography-induced
SK rearrangement and nuclei elongation in MSCs mightlay an important role for neural transdifferentiation ofSCs. Interestingly, Yim et al. [75] also reported that even
hough nanograting surfaces in conjunction with soluble

7 W. Chen et al.

irniaeso

N

Nsdfdatw

crahmdmapd

meCbcgoocaeetobnamomWnaoc

P

Pt

Figure 3 Nanotopography regulates NSC lineage [101]. SEM(left) and immunofluorescence (right) images of rat NSCs cul-tured on various substrates (laminin-coated PES films (flat film),283-nm, 749-nm, 1452-nm nanofibers) in the presence of 1 mMretinoic acid and 1% fetal bovine serum for 5 days. For theimmunofluorescence images, the cells are stained with RIP(oligodendrocyte marker), Tuj-1 (neuronal marker) or nestin(neural progenitor marker) as indicted. Scale bars for SEMimages are 10 �m, and for inserts are 2 �m. The arrows in theSEM images indicate cell attachment to nanofibers. Scale bar forall immunofluorescence images with are 100 �m. Circled cellson 283-nm fiber mesh are cells stained double positive for RIPaSE

oicdYasf

70

nductive factors such as retinoic acid could enhance neu-onal marker expression in MSCs to the maximal extent,anotopography showed a stronger independent effect onnducing neural transdifferentiation compared to retinoiccid alone on unpatterned surfaces. Similarly, Prabhakarant al. [91] utilized nanofibrous surfaces in conjunction withuitable inductive factors for neuronal transdifferentiationf human MSCs in vitro.

eural stem cells

eural stem cells (NSCs) are multipotent stem cells intrin-ically capable of self-renewal and differentiation intoifferent neural lineages, offering promising applicationsor NSC-based cell therapies to treat neurodegenerativeiseases and traumatic injuries [153]. However, clinicalpplications of NSC-based cell therapies are hindered dueo a lack of efficient methods for large-scale expansion asell as controlled differentiation of NSCs.

In recent studies, nanotopographical substrates, espe-ially nanofibrous scaffolds, have been widely applied foregulating NSC differentiation for neural tissue engineeringpplications [68,87,88,90,98,101,108]. NSC differentiationas been shown dependent on both nanofiber size and align-ent. For example, Yang et al. [87] showed an enhanced NSCifferentiation on aligned PLLA nanofibers compared to PLLAicrofibers. In other studies, nanofiber scaffolds with well-

ligned structural features were demonstrated for enhancedroliferation and differentiation of NSCs compared to ran-omly aligned nanofiber scaffolds [90,98].

Importantly, nanotopographical surfaces not only pro-ote NSC neural differentiation, but also selectively

nhance other lineage specifications of NSCs. For example,hristopherson et al. [101] demonstrated that on nanofi-rous matrix composed of 283 nm diameter fibers, rat NSCsould spread unhindered, adopt a cell morphology sug-esting glial lineages, and preferentially differentiate intoligodendrocytes; whereas cells seeded on matrix composedf larger fibers with diameters of 749 nm and 1452 nm wereonfined by individual fibers and preferentially differenti-ted into neural lineages (Fig. 3). More recently, by coatinglectrospun PCL nanofibers with graphene oxide (GO), Shaht al. [154] developed a hybrid nanotopographical bioma-erial that preferentially guided NSC differentiation intoligodendrocytes. Together, the nanofibrous materials haveeen widely demonstrated for functional regulation of NSCeural differentiation including differentiation efficiencynd lineage specifications. Notably, the nanotopographicalaterials (especially those with nanofibrous forms) may not

nly improve NSC differentiation, but also offer favorableicroenvironment for neural tissue engineering [155,156].ith careful design and engineering of material features at

anoscale, nanofibrous materials can potentially serve asunique matrix platform for neural regeneration through-

ut the process of stem cell neural differentiation, neuralulture, and transplantation.

luripotent stem cells

luripotent stem cells including both ESCs and iPSCs, possesshe ability of differentiating into any specialized cell type

esoo

nd Tuj1.ource: Reproduced with permission [101]. Copyright 2009,lsevier.

f the human body. The pluripotent nature of ESCs andPSCs opens unprecedented opportunities for potential stemell-based regenerative therapies for various degenerativeiseases and the development of drug discovery platforms.et, the current PSC research for regenerative medicine,lthough has attracted much enthusiasm, is still in a nascenttate largely due to a lack of effective technologies andunctional biomaterials for large scale maintenance and
xpansion of PSCs as well as accurate control of lineagepecification of PSCs. Recent advances in nanotechnol-gy and materials science have raised a widespread hopef developing promising nanoengineered biomaterials for


Figure 4 Nanotopography regulates ESC self-renewal. (a) and (b) Polymethylglutarimide (PMGI) nanofibrous surfaces serve asa cellular scaffold for maintaining self-renewal of mouse ESCs without MEFs [103]. (a) Bright field (top) and fluorescence images(middle) of R1-Oct4-EGFP mouse ESCs cultured on nanofibers at three different densities (i.e., low, medium and high) and con-trols. Bottom panels show the density of PMGI nanofibers doped with Rhodamine 6G for visualization. (b) The number of colonieswithin 5 mm2 after culturing for 3 days on various substrates. Data are presented as mean ± SD of three independent experiments.(*p < 0.001, **p < 0.01, n > 3). Reproduced with permission [103]. Copyright 2012 Springer. (c) and (d) Nanotopographical surfaceformed from silica colloidal crystal microspheres (SCC) maintain the expression of mouse ESC self-renewal in comparison to flatglass [57]. (c) SEM images of SCC substrates with particles of 124 ± 4 nm, 430 ± 4 nm, and 602 ± 12 nm diameter, respectively, show-ing their ordered face-centered cubic packing. Quantitative PCR showing (d) Nanog, a stem cell marker was less down-regulatedon the SCC substrates than on glass cover slips; N = 3. Reproduced with permission [57]. (e) and (f) Smooth vitronectin-coated glasssurfaces supported cell adhesion, rapid cell proliferation, and long-term self-renewal of human ESCs without using mouse MEFfeeder cells [80]. (e) Representative SEM images of glass surfaces (top) and immunofluorescence images of human ESCs (bottom)plated on glass surfaces with their root-mean-square (rms) nanoroughness Rq indicated. In the immunofluorescence images, the

(f) P7 da.

tnm([ifInwo[

cells were co-stained for Oct3/4 (red) and nuclei (DAPI; blue).different levels of nanoroughness as indicated, after culture forns (>0.05) and **(p < 0.01) (Student’s t-test). Adapted from [80]

PSC-based regenerative medicine, among which nanotopo-graphical matrixes mimicking the in vivo stem cell nichehave shown exceptional potentials of regulation of stem cellfunction and fate.

Stemness maintenance and self-renewalOne of the major challenges in PSC-based cell engineeringand regenerative medicine is to achieve large-scale, long-term PSC expansion without losing pluripotency. Althoughthe standard protocol maintaining PSCs on feeder cells issuccessful and widely adopted, such feeder based cultureprotocol is labor-intensive, prone to contaminations fromfeeder cells, and difficult for high-throughput automation.
Excitingly, recent developments of nanotopographical bio-materials have brought a few novel, feeder-free methodsfor PSC culture and thus pose promising routes for futurelarge-scale PSC engineering (Fig. 4).
aic

ercentage of Oct3/4 + human ESCs on the glass substrates withys. Error bars represent (standard error of the mean (SE, n = 3).

Recently, Nur-E-Kamal et al. [157] reported an elec-rospun nanofibrillar substrate composed of polyamideanofibers to promote proliferation and self-renewal ofouse ESCs via the Rac and phosphoinositide 3-kinase

PI3K) signaling. Similar results were obtained by Liu et al.103], where high density electrospun polymethylglutarim-de (PMGI) nanofibers were used to achieve sustainedeeder-free maintenances of mouse ESCs (Fig. 4a and b).n another related study, feeder-free mouse ESC mainte-ance was achieved using a topographical surface formedith silica colloidal crystal (SCC) spheres to generate highlyrdered topographical features of 120— 600 nm in diameter57] (Fig. 4c and d).

In addition to mouse ESCs, it is also challenging tochieve long-term culture of human ESCs while maintain-ng pluripotency. It has recently been reported that theurrent feeder based human ESC culture method can be

7

scbbr[inohtn[d[t1Efodthbnpne

DNtcmathaoc

NNsIsidearafwESccnpg

Ertrsin[rrsaai[

OAfdberwbsttgDnEtmdcodesm

rwssaitE[casgwrigidity [24,161], nanotopography in the stem cell microen-

72

ignificantly improved by introducing collagen or gelatinoated electrospun PCL nanofibrous scaffolds [99]. It shoulde noted, however, that a significant difference existsetween human and mouse ESCs for their culture envi-onment requirements for self-renewal and clonal growth8]. For example, human ESCs do not require leukemianhibitory factor (LIF) for self-renewal, which is howevereeded for mouse ESCs [8]. Such difference has also beenbserved in the context of mechanosensitive behaviors ofuman and mouse ESCs. For instance, although it was foundhat both intrinsic (cell) and extrinsic (substrate) soft-ess facilitated maintenance of mouse ESC pluripotency158,159], soft cell culture surfaces promoted spontaneousifferentiation rather than self-renewal of human ESCs24,160,161]. In another recent study, Chen et al. reportedhat nanorough substrates with topographical features of00 nm high promoted spontaneous differentiation of humanSCs, whereas flat, featureless coverslips were supportiveor stemness maintenance [80] (Fig. 4e and f). A similarbservation was reported by the Dalby group, wherein disor-ered nanopit arrays made of polycarbonates were observedo suppress self-renewal and promote early osteogenesis ofuman ESCs [162]. However, one conflicting observation haseen reported by Kong et al., wherein fibronectin-coatedanotopographical surfaces, rather than flat controls, sup-orted stemness maintenance of human ESCs, and switchinganotopography features from a hexagonal pattern to a hon-ycomb one further improved such an effect [60].

irected differentiationanotopography cues have also been shown impor-ant for directed differentiation of ESCs toward neuralells [89,100,163—168], osteoblasts [92,114,162,169,170],uscle cells [171,172], hematopoietic cells [173,174],

dipocytes [175], and chondrocytes [176]. Here for illustra-ions, we only discuss how nanotopographical biomaterialsave been utilized for improving directed neurogenesisnd osteogenesis from ESCs. For ESC differentiation towardther cell lineages, the readers are referred to referencesited above.

eural lineageanotopographical materials have been widely demon-trated for promoting neural lineage commitments of PSCs.n a study by Xie et al. [100], uniaxially aligned nanofiberubstrates were used to induce mouse ESC differentiationnto different neural lineages, such as neurons, oligoden-rocytes, and astrocytes (Fig. 5a—c). Aligned nanofibersnhanced not only neural differentiation efficiency butlso neurite outgrowth along nanofibers, critical for neu-onal network development. Carbon nanotubes (CNT) werelso recently utilized for promoting ESC neurogenic dif-erentiation. For example, CNT/PLLA composite scaffoldsere applied to enhance neural differentiation of mouseSCs even in the absence of soluble inductive factors [89].imilarly, human ESCs directly seeded onto type I collagen-oated CNT matrix showed enhanced ectodermal lineage
ommitment as well as cell alignment along collagen/carbonanotube fibrils [167] (Fig. 5d). Thin film scaffolds com-osed of biocompatible polymer poly(acrylic acid) or silkrafted with CNTs were also shown to promote human
vctw

W. Chen et al.

SC-derived embryoid body (EB) differentiation into neu-al cells with heightened cell viability [23,166]. In additiono nanofibrous materials and CNT composite scaffolds withandom nanotopological features, well controlled nanoscaleurface patterns such as nanoscale ridge/groove arrays alsomproved significantly differentiation of human ESCs intoeural lineages without any soluble inductive factor (Fig. 5e)168]. Interestingly, human ESCs cultured on such nanoscaleidge/grooves arrays preferentially differentiated into neu-ons but not glial cells, such as astrocytes. Such selectiveuppression of glial cell formation is desirable for potentialpplications in therapies for spinal cord injury, as astrocytesre known to contribute to glial scar formation, creat-ng barriers to axons in the central nerve system (CNS)177].

steogenic lineagenother important direction is the osteogenesis of PSCor bone repair and regeneration. Enhanced osteogenicifferentiation of mouse ESCs has been observed on nanofi-rous matrices (Fig. 6a and b) as indicated by heightenedxpression of osteogenic markers [114,170]. In anotherelated study, Massumi et al. [92] showed that in synergyith other scaffold properties such as Matrigel coating,oth the nanoroughness and alignment of PLGA nanofibrouscaffolds played a significant role in directing differentia-ion of mouse ESCs toward mesoderm lineages. Followingheir work showing the effect of symmetry of nanotopo-raphical cues on osteogenic differentiation of MSCs [37],alby and colleagues recently extended their polycarbo-ate nanotopographical substrates for regulation of humanSC differentiation [162] (Fig. 6c and d). Human ESCs cul-ured on nanopit arrays exhibited enhanced expression ofesenchymal or stromal markers, resulting in improved pro-uction of osteogenic progenitors. Gene expression analysisonfirmed that nanopit arrays did not induce endodermalr ectodermal lineage commitment. Interestingly, EB-basedifferentiations of human ESCs toward mesodermal lin-ages was less responsive to nanotopographical cues [162],uggesting efficient cell-substrate interactions critical forediating stem cell nanotopography sensitivity.Seeding mouse ESCs on PLLA nanofibrous matrices

esulted in upregulated �1 integrin expression, consistentith the important role of �1 integrins in ESC osteogene-

is and mesodermal lineage commitment [178]. In anothertudy, Chen et al. showed that nanotopography significantlyffected the local molecular arrangement and formation ofntegrin-mediated FA that might in turn regulate the spa-ial organization of myosin II activity, CSK contractility and-cadherin mediated intercellular adhesion of human ESCs80]. It suggests that the nanoscale topography in the stemell niche influences the molecular organization of integrinsnd triggers integrin-mediated FA signaling to elicit down-tream biochemical signals important for the regulation ofene expression and stem cell fate. Although similar effectsere also observed for other biophysical factors like matrix

ironment might serve as a more direct cue to affect integrinonformation and clustering and thus organizations of adap-or and signaling proteins in FAs due to their similarity in sizeith integrins [13].


Figure 5 Nanotopography regulates ESCs toward neural lineage. (a)—(c) The differentiation of mouse ESCs on electrospun nanofi-brous surfaces into neural lineages [100]. (a) SEM images of aligned (top) and randomly oriented (bottom) PCL nanofibers preparedby electrospinning. (b) Representative immunofluorescence images of differentiated mouse ESCs on aligned (left) and random (right)PCL nanofibers for 14 days. The mouse ESCs expressed GFP in green color (top) and stained in red color with neuron marker Tuj1(middle), the bottom images show and superimposed image of the top and middle images in the same region. (c) Cell pheno-type analysis of mouse EBs cultured on PCL nanofibrous scaffolds for 14 days. The markers examined were SSEA-1 (stage specificembryonic antigen, for undifferentiated mouse ES cells), nestin (for neural precursors), Tuj1 (�-tubulin III, for neurons), O4 (foroligodendrocytes), and GFAP (glia fibrillary acidic protein, for astrocytes). # p < 0.05 for markers compared with embryonic stemcells. + p < 0.05 for markers compared with embryoid bodies. * p < 0.05 for markers compared with random PCL fibers. Reproducedwith permission [100]. Copyright 2009, Elsevier. (d) Direct human ESC neural differentiation on collagen/CNT matrix [167]. AFMcharacterization of the surface structures, cell morphology, and nestin expression of human ESCs after cultured in the medium ofspontaneous differentiation for three days on gelatin (top), collagen (middle) and collagen/CNT (bottom) matrices. Inset image sizein the AFM images: inset 2 × 2 �m2. The yellow arrows in the staining images indicate the coarse alignments of the cells. Repro-duced with permission [167]. Copyright 2009, Elsevier. (e) Direct differentiation of human ESCs into selective neurons on nanoscaleridge/groove pattern arrays [168]. The left column are representative SEM images of a bird’s eyes view (left top) and a cross-section(left middle) of 350-nm ridge/groove pattern arrays (spacing of 350 nm, height of 500 nm) and a SEM image showing human ESCs onthe 350-nm nanoscale ridge/groove pattern arrays (left bottom). The right column are representative immunofluorescence imagesof human ESCs stained with nuclei, neural and glial markers as indicated after cultured for 10 days on the 350-nm ridge/groovepattern arrays. Differentiated hESCs stained positively for HuC/D (human neuronal protein: RNA-binding protein) and MAP2 (matureneuronal marker: microtubule-associated protein 2), but were not for GFAP (intermediate filament proteins of mature astrocytes:

atedt 201

m

glial fibrillary acidic protein). This suggests that hESCs differentisuch as astrocytes. Reproduced with permission [168]. Copyrigh

Molecular insights for stem cell
nanotopography sensitivity
Much effort has been directed to study behaviors ofstem cells in response to nanotopographic cues. However,

rrns

into mature neurons without differentiation into a glial lineage0, Elsevier.

olecular mechanisms underlying stem cell nanotopog-
aphy sensitivity remain elusive. Some major questionsemain unanswered including detailed molecular mecha-isms relaying extracellular nanotopography to intracellularignaling pathways and regulatory networks controlling

774

Figure 6 Nanotopography regulates ESCs toward osteogeniclineage. (a) and (b) Enhancing osteogenic differentiation ofmouse ESCs by nanofibers matrix [114]. (a) SEM micrographs(top) of mouse ESCs after 12 h under differentiation conditions,calcium staining (middle) and immunofluorescence staining(bottom) images of late bone differentiation (Osteocalcin)marker expression after 26 days under osteogenic differ-entiation conditions on nanofibrous matrix, solid films, andgelatin-coated tissue culture plastic (Control). QuantitativePCR of osteogenic markers (b) osteocalcin (top) and bone sialo-protein (BSP; bottom) RNAs isolated from cells on nanofibrousmatrix, solid films, and gelatin-coated tissue culture plastic(Control) after 26 days of culture under osteogenic differenti-ation conditions. Reproduced with permission [114]. Copyright2009, Mary Ann Liebert, Inc. (c) and (d) Nanopit surfacesaugment mesenchymal differentiation of Human ESCs [162].(c) SEM image of polycarbonate nanopit substrate. Nanopitsof 120 nm diameter and 100 nm depth are arranged in a nearsquare geometry with center-center spacing 300 nm ± 50 nm.(d) Expression of endodermal (SOX17), ectodermal (Nestin)and osteogenic progenitor (ALCAM) markers was assessed byqPCR for self-renewing human ESCs, EBs and human ESCsseeded onto planar or nanopit substrates in basal media for14 days (n = 6). Nanopit substrates do not direct endodermalor ectodermal differentiation, but significantly enhancedow

sliim

ctdt

At

Phae8o��ifm

si[spdittdtiaaS

mAdtcsgrascinstivlwll

steoblastic differentiation. * p < 0.05, ** p < 0.01. Reproducedith permission [162]. Copyright 2013, Wiley-VCH (Germany).

tem cell fate. Hereby, we review some prominent cellu-
ar mechanosensory machineries that have been implicatedn stem cell nanotopography sensitivity. We discuss howntracellular signaling pathways downstream of theseachineries relay extracellular nanotopography through
oos

W. Chen et al.

ytoplasmic transducers, cytoskeletal integrators, andranscriptional actuators. Such an integrated mechanotrans-uction network may provide a roadmap for future studieso understand stem cell nanotopography sensitivity.

dhesion-based mechanosensors and cytoplasmicransducers

laying a pivotal role in cell—ECM adhesions, integrins areeterodimeric transmembrane proteins composed of an �nd a � subunit. Currently, there are 24 known integrin het-rodimers constituted by different combinations of 18 � and� subunits [179—182]. Human ESCs express a broad range

f integrins including �1, �2, �3, �5, �6, �7, �11, �V, �E, and1, �2, �3, �5, �6 integrins. Human iPSCs express mainly �5,6, �V, �1, and �5 integrins, with variations among different

PSC lines [183—185]. There are abundant studies showingunctional regulation of adult stem cells through integrin-ediated adhesion signaling [14,37,139,186].The first step adherent stem cells take to probe

urrounding ECM is mediated by ‘‘inside-out’’ and ‘‘outside-n’’ bidirectional regulation of integrin—ligand binding187—189]. Specifically, initial binding of integrin to adhe-ive molecules initiates clustering of intracellular adaptorroteins including talin, which in turn bind the cytoplasmicomain of integrin, resulting in a conformational change ofntegrin’s extracellular domain and its heightened affinityo ECM ligands. Firm binding of integrin to ECM ligand inurn activates integrin clustering through their intracellularomains and further recruits adaptor and signaling proteinso adhesion sites. Key molecular players in the early stage ofntegrin-mediated cell—ECM adhesion (also known as ‘‘focaldhesion’’, or FA) include adaptor proteins talin and paxillinnd tyrosine kinases such as focal adhesion kinase (FAK) andrc family kinases [190,191].

Importantly, FA formation is sensitive to spatial arrange-ent and presentation of extracellular ECM ligands [192].s the head of an integrin heterodimer is about 20 nm iniameter [13], a nanometer scale variation in substrateopography can directly affect integrin conformation andlustering and thus dynamic organization of adaptor andignaling proteins in FA. Results reported from differentroups have suggested that successful integrin clusteringequires a spacing between individual ECM ligands less than

threshold value between 50 nm and 70 nm, and a tooparse presentation of individual ligands inhibits integrinrosslinking and thus FA formation [124,193—195]. Interest-ngly, simply clustering individual ECM ligands into oligomeranopatterns without changing the overall ligand densitytill promotes FA formation and cell spreading, supportinghe crucial role of nanoscale organization of integrins in FAn controlling cell behaviors [196]. In line with this obser-ation, Huang et al. reported that a local disorder of ECMigand arrangement could still promote integrin clusteringhen the average ligand spacing was greater than 70 nm,

ikely owing to the emergence of local sub-70 nm-thresholdigand spacing [197].

A few recent studies have directly highlighted the effectf nanotopography on ligand spacing/density and localrder/disorder and thus integrin clustering and downstreamtem cell behaviors. For example, Park et al. observed that


Figure 7 Sensitivity of integrin clustering and FA morphogenesis to nanotopography. (a) (Top panel) Immunofluorescence imagesshowing that TiO2 nanotube arrays of tube diameter 15 nm promoted the growth of large FAs and prominent F-actin stress fibersin MSCs; while larger tube diameter (100 nm) strongly inhibited FA maturation and resulted in diffusive actin staining. (Middle andbottom panels) SEM immunogold staining images showing that the density of FAs (middle) and �1 integrin clustering was muchhigher on nanotube arrays of 15 nm diameter TiO2 nanotubes. Reproduced with permission [14]. Copyright 2007, American ChemicalSociety. (b) Immunofluorescence images showing decreased FA sizes in human ESCs cultured on nanotopographical surfaces of 100 nmroughness features. Adapted with permission from [80]. Copyright 2012, American Society of Chemistry. (c) (Top panel) SEM images

m paom [1

ieistaosoFssssavwc

FPwdiPtaaa

of nanopit arrays of disorder and ordered arrangements. (Bottoelongated FAs formed on disordered nanopit arrays. Adapted fr

the larger TiO2 nanotube diameter was (e.g., >50 nm), theless occurrence of integrin clustering and FA formation wasobserved (Fig. 7a) [14]. In another study by Chen et al., itwas also observed that FA formation was promoted in humanESCs cultured on nanotopographical surfaces with smaller(e.g., 1 nm), rather than greater (e.g., 100 nm), nanoroughfeatures (Fig. 7b) [80]. In a series of works published byDalby and colleagues [186,198], disordered arrangements ofnanopits were observed to induce larger and more matureFAs with elongated morphology in human MSCs comparedto ordered nanopit arrays (Fig. 7c). This observation againsupports the mechanism proposed by Huang et al. [197]that local disorder of ECM presentation might promote theemergence of integrin activation and FA formation under aconstant global average of ECM ligand density.

An important signaling axis downstream of integrin-mediated FA is the FAK-Src pathway [190,199,200].Activated by integrin ligation, FAK is recruited to FAby binding integrin, which in turn activates FAK throughautophosphorylation on Y397. Further phosphorylation ofFAK is completed by binding of FAK to Src family proteins.Together, FAK and Src form a signaling complex that relayssignals from integrin to control downstream Rho GTPaseactivities through, for instance, the FAK/p190RhoGEF/RhoA,FAK/p190RhoGAP/RhoA, and Grb2/SOS/Rac pathways [199].Integrin-regulated FAK-Src signaling also mediates theMAPK pathways to regulate cell proliferation, apopto-sis, and differentiation through the FAK/Ras/Raf/MEK/ERK,Fyn/Shc/Grb2/SOS/Ras, and p38 MAPK signaling [201—204].

Various studies have reported that nanotopographicalsubstrates could promote FAK phosphorylation [205]. How-ever, some other studies have suggested that maximalFAK phosphorylation and activation might occur at an

Eros

nel) Immunofluorescence images showing that larger and more86]. Copyright 2014, Wiley.

ntermediate nanotopographic level. For example, Parkt al. reported that TiO2 nanotubes of 15 nm in diameternduced greater FAK phosphorylation compared to eithermooth surfaces or TiO2 nanotube arrays of greater diame-ers [14]. Through FAK inhibition using small molecule drugsnd siRNA, Teo et al. further showed that FAK was notnly responsive to but also required for nanotopographyensitivity on 250 nm gratings and subsequent neurogenesisf human MSCs [205]. Importantly, constitutively activatedAK could rescue neurogenesis of human MSCs on smoothurfaces to an extent comparable to that on nanogratingurfaces, further underscoring the critical role of FAK intem cell nanotopography sensitivity. Given the dual, oppo-ite effects of FAK on RhoA activity [199,206,207] (e.g.,ctivation through p190RhoGEF [208,209] or inactivationia p190RhoGAP [210,211]), it remains an open questionhether FAK phosphorylation induced by nanotopographyould activate or inhibit RhoA activity.

Other signaling events downstream of integrin-mediatedAK-Src signaling, such as the FAK/MEK/ERK [14,212],I3K/Akt [213] and BMP and TGF�/SMAD [214] pathways,hich are important for cell proliferation, self-renewal, andifferentiation [215—217], might also be involved in regulat-ng nanotopography sensitivity of stem cells. For example,ark et al. observed maximal phosphorylation and activa-ion of ERK in human MSCs on TiO2 nanotube arrays with

tube diameter of 15 nm [14]. Through whole proteomenalysis, Kantawong et al. observed that disordered nanopitrrays elicited changes of protein expression involved in the
RK pathway for human MSCs [212]. Interestingly, Kim et al.eported that ERK in human MSCs was maximally activatedn nanogroove substrates with an optimal groove width:pacing ratio (1:3 and 1:1), indicating a biphasic activating

7

espBaibtFpMtpcan

wsFoRrCoRtcrieRtisafabttc2CtsCsa

C

AoetlaSss

roscnudntFr[

sii[tIiorpasYtSn2tFif

prSmfttGd

ootdaSsinamcc

76

ffect of nanotopography on ERK [218]. In another recenttudy, Yang et al. reported that disordered nanopit arraysromoted osteogenesis of human MSCs through co-localizingMP receptors to integrins, upregulating expression of BMPsnd BMP receptors, and thus enhancing transcriptional activ-ty of SMAD1/5 and Runx2 [219]. Although it has not yeteen demonstrated in nanotopography sensitivity per se,he Wnt/GSK3� signaling, which functions downstream ofAK/PI3K/Akt, was also implied by recent studies to be aotential regulator of differentiation, especially in humanSCs [220]. In light of these observations, a critical question

o answer in future studies is how these different signalingathways initiated by FA signaling and relayed by kinase cas-ades are interconnected and whether they converge onn integrative downstream signaling mechanism governinganotopography sensitivity of stem cells.

Beside multifaceted FAK signaling, the RhoA/ROCK path-ay originated from G-protein coupled receptor (GPCR)

ignaling [221—224] and downstream of integrin-mediatedA signaling has also been shown important for mechan-transduction in stem cells [223,225]. When functioning,hoA/ROCK signaling could feedback to mediate integrin-egulated adhesion signaling through its effect on the actinSK contractility. Specifically, FAK activation downstreamf integrin ligation activates Rho family GTPases, includinghoA, which in turn activates two effectors ROCK and mDiao regulate structural quality of filamentous actin CSK andontractile activity of non-muscle myosin II motor proteins,espectively [221]. Actin CSK contractility has been shownmportant for FA maturation and signaling, which in turnnhances RhoA/ROCK activation. Altogether, it implicateshoA, ROCK, and CSK contractility as an interconnected,ripartite module responsible for nanotopography sensitiv-ty. This hypothesis, in fact, is supported by a few recenttudies. For example, McMurray et al. reported that ROCKctivity and CSK contractility were required for sensing dif-erent arrangement orders of nanopit arrays by human MSCs,s their inhabitation by small molecule inhibitors Y27632 andlebbistatin, respectively, abolished nanotopography sensi-ivity of human MSCs [226]. Similarly, Teo et al. also showedhat inhibition of either ROCK or CSK contractility was suffi-ient to inhibit nanotopography sensitivity of human MSCs on50 nm nanogratings [227]. In addition, in a recent study byhen et al., the authors demonstrated decreased CSK con-ractility on nanorough polymeric substrates, implying thattem cells might sense nanotopography through changes ofSK contractility [79]. However, it is still unclear whetheruch nanotopography-regulated CSK contractility is medi-ted directly through the RhoA/ROCK pathway.

ytoskeletal integrators

lthough mounting evidence has suggested involvementsf integrin-mediated adhesion signaling and downstreamffectors, such as FAK, ERK, RhoA/ROCK, and CSK con-ractility, in stem cell nanotopography sensitivity, a criticalink between such diverse cytoplasmic signal transducers
nd downstream transcriptional regulators is stillmissing.ome important hints provided from mechanotransductiontudies have supported the actin CSK and its integrity toerve as an integrator of the multitude of upstream signals
veuf

W. Chen et al.

elayed from extracellular mechanical cues including nan-topography (Fig. 8) [228]. Supporting this view, recenttudies have demonstrated disorganized F-actin CSK andompromised CSK contractility in adherent cells cultured onanorough substrates [80]. Herein we briefly review currentnderstanding of the role of the actin CSK in mechanotrans-uction and speculate its possible involvement in stem cellanotopography sensitivity through controlling downstreamranscriptional activity of Hippo/YAP and MAL/SRF signaling.or detailed discussions about Hippo/YAP signaling, theeaders can refer to excellent reviews published elsewhere229—232].

The Hippo/YAP pathway, which is important for organize control and cancer [231,233], has recently beendentified as a downstream effector of the quality, dynam-cs as well as intrinsic contractility of the F-actin CSK234—240]. Hippo/YAP signaling has also been shown impor-ant in controlling stem cell differentiation [240,241].nterestingly, experimental evidence has been accumulatedmplicating an emerging pattern of mechanical regulationf Hippo/YAP activity: extracellular mechanical cues (e.g.,igid substrate, external stretch, large cell size) resulting inrominent F-actin CSK formation promote high nuclear YAPctivity, while those cues (e.g., soft substrate and small cellize) compromising the integrity of F-actin CSK lead to lowAP activity and dominant cytoplasmic YAP retention due toheir association with the scaffold protein 14-3-3 [234,235].uch a role of the F-actin CSK as a potent mechanical sig-al integrator has been supported by recent studies in bothD and 3D contexts on a myriad of cell behaviors and func-ions [235]. Furthermore, convergent signaling through the-actin CSK to suppress YAP activity has also been shownnvolved in regulating mechanosensitive motor neuronal dif-erentiation of human ESCs [24].

In addition to the Hippo/YAP pathway, the integrity andolymerization of actin CSK have also been implicated inegulating SRF (serum response factor) signaling [242,243].pecifically, enhanced polymerization of cytoplasmic actinonomers (G-actin) releases MAL, a transcription co-factor,

rom its association with G-actin, resulting in MAL nuclearranslocation and thus elevated SRF signaling. Followinghis mechanism, large cell size with elevated cytoplasmic-actin level has been found to inhibit SRF signaling andifferentiation of epidermal stem cells [242].

Combining these observations with the recent findingf diffusive F-actin CSK in human ESCs cultured on nan-topographical substrates [80], it is tempting to speculatehat nanotopography may either activate the actin CSK-ependent Hippo pathway and thus suppress nuclear YAPctivity or inhibit nuclear translocation of MAL and thusRF signaling (Fig. 8). Yet, how Hippo/YAP and/or MAL/SRFignaling are involved in stem cell nanotopography sensitiv-ty has not been specifically examined so far. It is worthoting that although cellular sensing of nanotopographynd other extracellular mechanical cues (such as substrateechanics, geometric confinements, and surface electric

harge, as reviewed by Higuchi et al. [244]) might shareommon signaling pathways, like those already been shown
ia RhoA/ROCK, FAK, and CSK contractility, a possibilityxists that nanotopography sensitivity might elicit certainnique intracellular mechanisms. It calls for future care-ul examinations of the actin CSK and its integrity and


Figure 8 Potential mechanotransduction mechanisms in cellular responsiveness to nanotopographical biomaterials. (Left) Onpro-integrin clustering surfaces, which could be either smooth or composed of certain nanotopographical features, integrins couldundergo free lateral recruitment and ligation with ECM proteins, and thus cluster and form mature, stable FAs together with theunhindered recruitment of FA adaptor proteins. This process initiates signals from plasma membrane, enhancing FAK/ERK signaling,as well as RhoA activity, which further promoting CSK contractility (through RhoA/ROCK) as well as stress fiber formation. CSKcontractility, as an important mediator of mechanotransduction, could provide positive feedback to the FAs. In addition, CSK forcecould also induce phosphorylation of emerin, a nuclear envelope-located protein, and thus initiate nuclear mechanotransductionthrough enhanced nuclear actin polymerization and subsequent nuclear translocation of MAL, a transcription co-factor of SRF. Inthe meantime, prominent stress fibers formed in cells on smooth surfaces could promote the nuclear translocation of YAP/TAZ, atranscription co-factor of TEAD. Nuclear shuttling of SMAD 2/3 (R-SMAD), the transcription factor downstream of TGF� signaling, isalso controlled by YAP/TAZ translocation. (Right) On surfaces containing anti-integrin clustering nanotopographical cues, althoughintegrins could still freely diffuse laterally, the nanoscale surface features restrict the ligation of additional integrins to the ECMproteins, which further restrict successful recruitment and clustering of integrins and other FA proteins, resulting in smaller, lessstable FAs. Such process disrupts the activation of RhoA and therefore limits the formation of stress fibers and might induce highcytoplasmic G-actin level, which inhibits the nuclear translocation of MAL and thus SRF signaling. Unlike nanotopographical surfacesthat promote integrin clustering, anti-integrin clustering surfaces containing different types of nanoscale structures could notsustain FAK activation and downstream signaling. Last but not the least, compromised stress fibers and FAs on nanotopographical

y enf YAP

psctFEe

substrates could potentially intersect with Hippo/YAP pathway b-independent mechanisms, resulting in cytoplasmic retention o

their functional relationship with Hippo/YAP and MAL/SRFsignaling in stem cell nanotopography sensitivity.

Nuclear mechanosensors and transcriptionalactuators

Although deeply embedded within a cell, the nucleusis an emerging, active mechanosensor [33,245]. As themost important mechanical component of the nucleus, thenuclear lamina is composed mainly of intermediate filament

a

So

hancing YAP/TAZ phosphorylation via either Lats-dependent or/TAZ.

roteins lamin A and C, endowing a significant mechanicaltiffness to the nucleus. Lamin A/C has been demonstratedritical for mechanotransduction and transcription regula-ion through signaling molecules such as MAL, emerin, and-/G-actin [12,33,246—248]. Importantly, the nucleus ofSCs, which lacks lamin A/C, is extremely deformable, andmergence of lamin A/C in the nucleus has been proposed
s an ESC differentiation marker [249].
Through its connection to the actin CSK via KASH andUN domain proteins (together known as the LINC — linkerf nucleoskeleton and cytoskeleton — complex) across the

7

nsta[sfomffir

iicbsnatfmru

btcdFfiiIcrc

mmlnftupm

tAithccaci[gn

rct

ctttSmmtcpwmpacir[inApest

C

Siaamcmtoifm

nbDbomiftemc

78

uclear envelope as well as perinuclear adaptor proteinsuch as nesprin and plectin, the nuclear lamina is struc-urally connected to the mechanoregulatory network ofctin filaments, microtubules, and intermediate filaments33,250,251]. Thus, it is not surprising that the nucleus itselferves an important role in mechanoresponsive stem cellate regulation. Herein, we briefly review nuclear mechan-transduction mechanisms involved in cellular sensing ofechanical cues such as substrate mechanics and external

orces, in the hope of providing prospective mechanisms foruture studies on the role of nuclear mechanotransductionn stem cell nanotopography sensitivity, which has so faremained a largely unexplored field.

Using a cysteine shot gun technique in combination withn vitro shear force application [252], Discher and colleaguesdentified a group of proteins, including lamin A, thatould change conformation and expose cryptic hydropho-ic domains under unfolding forces [253]. Recently, theame authors observed a positive scaling relation betweenuclear lamin expression and in vivo ECM rigidity, suggestingfunctional link between rigidity-dependent CSK contrac-

ility and force-dependent nuclear lamin expression andunction [253]. Nuclear mechanical stiffness, as endowedainly by lamin A/C expression, thus may scale with ECM

igidity, consistent with a recent study by Liu and colleaguessing nanoindentation [254].

Recently, Poh et al. observed that an intact linkageetween the actin CSK and nucleoskeleton was required forransmission of integrin-mediated external forces from theytoplasm membrane to the Cajal bodies in the nucleus toirectly dissociate coilin and SMN protein complexes [255].urthermore, transmission of CSK tension within actin stressbers and microtubules has been found critical for mediat-

ng long-range force-sensing involving the nucleus [251,256].n addition, it has been reported that confining adherentell shapes to high aspect ratio morphology results in a cor-esponding high aspect ratio of the nucleus and prominenthromatin condensation [257].

In another very recent study, the inner nucleusembrane-located protein, emerin, is demonstrated to beechanosensitive and undergo force-dependent phosphory-

ation [258]. Using isolated nuclei, this study demonstrateduclear stiffening in response to external cyclic pullingorces and its regulation by recruitment of lamin A/Chrough Src-mediated tyrosine-phosphorylated emerinsnder tension. Interestingly, such force-mediated emerinhosphorylation depended on both substrate rigidity andechanical forces applied through integrin ligation.So far, there are few studies of nuclear mechanotransduc-

ion in the context of stem cell nanotopography sensitivity.lthough it has been reported that nanotopography can

nduce nucleus deformation [104,259], detailed examina-ions of such nanotopography-dependent nuclear eventsave not yet been demonstrated until very recently. Specifi-ally, Dalby and colleagues have reported mechanosensitivehromosome positioning in response to disordered nanopitrrays, wherein nanotopographic cue poses chromosome 1loser to the nuclear membrane and increases its mean
nter-territory distance during human MSC osteogenesis260]. The authors have also observed that positions ofene regulation along the chromosome are sensitive toanotopography [261], with nanotopography-enhanced gene
cnea

W. Chen et al.

egulation concentrated toward the telomeric end of thehromosome, where osteogenesis-related genes are clus-ered [260].

A few mechanosensitive nuclear actuators relaying extra-ellular mechanical signals to transcription activities inhe nucleus have been identified recently. As a nuclearranscription co-factor, YAP/TAZ can bind transcription fac-ors TEAD as well as co-translocate SMAD 2/3 (regulatoryMAD, R-SMAD) to the nucleus, thus relaying cytoplas-ic mechanotransductive signals to transcriptional controlachineries [24,231,262,263]. Importantly, the nucleus pro-

ein emerin and nuclear localization of lamin A/C are bothritical for MAL/SRF signaling in the nucleus [248,266]. Inarticular, emerin promotes nuclear actin polymerization,hich in turn prohibits MAL nuclear export and thus pro-otes transcriptional activity of MAL. In addition, emerinhosphorylation is correlated with both substrate rigiditynd nuclear YAP localization and activity [258]. Mechani-al tension through nesprin-1 on isolated nuclei can alsonduce RhoA activation in the nucleus, although the exactole of RhoA in transcriptional control remains unclear258]. Together, it is foreseeable that mechanistic stud-es of nanotopography sensitivity of stem cells involvinguclear deformation, nuclear envelope mechanics, lamin/C expression, nuclear lamina protein expression and phos-horylation, and nuclear transcription activation will helplucidate nuclear mechanosensitive players completing theignaling axis through which substrate nanotopography con-rols stem cell fate (Fig. 7).

onclusion remarks and outlook

tem cells are promising cell sources for tissue engineer-ng, regenerative medicine, drug screening, and toxicityssays. The major challenge faced by stem cell biologistsnd bioengineers is the large-scale, long-term stem cellaintenance and high-specificity, high-yield, directed stem

ell differentiation toward clinically relevant lineages withature tissue-specific functions. Rapid advances in nano-

echnology and materials science have leveraged regulationf stem cell fate and function via microenvironmental phys-cal factors such as nanotopographical cues to enhanceunctional performances of synthetic nanoengineered bio-aterials in stem cell engineering.In this review, we have summarized state of the art

anofabrication methods for generating nanotopographicaliomaterials and surfaces for regulating stem cell fate.ifferent synthetic nanotopographical biomaterials haveeen successfully developed to control behaviors and phen-types of stem cells, such as adhesion, migration, andorphology, and eventually determine transcriptional activ-

ties and stem cell fate. By designing nanotopographicaleatures mimicking in vivo stem cell niches, a diverseoolbox of nanotopographical biomaterials have enabledither long-term maintenance or directed lineage commit-ent of stem cells, producing promising solutions for stem

ell-based applications. Although successful as a proof of
oncept, static nanotopographical biomaterials alone mayot sufficiently dictate stem cell fate desired for differ-nt applications. Soluble biochemical cues, dynamic controlnd regulation of topographical features, as well as cell

aecamasip

A

WFIApCRstaa

R


co-culture systems, have all been demonstrated to playin synergy with physical cues in regulating stem cell fate[11,12,89,166,167,267,268]. In the future, design and fab-rication of nanotopographical biomaterials will need tobe integrated with multifunctional biochemical modifica-tions and even spatiotemporal patterning for simultaneouslycontrol of multiple aspects of in vitro stem cell microen-vironment and fine tuning of specific terminal lineagecommitment at a large scale.

Besides controlled stem cell differentiation, new strate-gies to support long-term stem cell self-renewal andlarge-scale expansion (maintaining an undifferentiated stemcell state without spontaneous differentiation) will be cru-cial to obtain sufficient cell source for stem cell-basedregenerative therapies. While it is currently still challengingto design well-defined biomaterials and protocols promotingstem cell culture, nanotopographical materials have beenshown great potential to form a favorable microenvironmentfor maintenance of stem cell potencies. Such approaches,alongside the potential of nanotopography to modulate stemcell functions and to mimic the stem cell niche, remain tobe investigated in depth.

It is also worth noting that possible variance in nan-otopography sensitivity might occur as a consequence ofstudies using different cell lines with distinct culture con-ditions and genetic origins and from different species. Forexample, rate and human MSCs have been shown distinctpreferences for different nanotopographic feature sizes foroptimal osteogenic differentiation [14,139]. There are alsoa growing number of studies indicating differences betweeniPSCs and ESCs [269,270]. As such, it is important thatmultiple cell lines from the same species are examinedand compared in the same set of nanotopography studies.Especially for PSCs, iPSCs are always recommended to beinvestigated alongside ESCs.

Another important direction in stem cell research isinduction of pluripotency or transdifferentiation of one celltype to another [271,272]. Abundant evidence has shownpotent regulation of adult stem cell function and signalingby nanotopography sensitivity, yet questions of whetherand how nanotopography at the cell—substrate interfacecan be leveraged for improving cell reprogramming andtransdifferentiation remain unexplored. Such studies willpotentially provide novel approaches to improve efficiencyof reprogramming and transdifferentiation processes andyield significant mechanistic insights related to mechanoreg-ulation of development and diseases [9,273—275].

To improve performance of nanotopography-driven stemcell fate regulation, it is imperative to understand themechano-sensing and -transductive mechanisms underly-ing cellular responses to extracellular nanotopography. Inthis review, we have discussed a few key mechano-sensingand -transductive mechanisms implicated in regulation ofstem cell fate via nanotopographical biomaterials. Basedon current understanding of stem cell mechanobiology, wehave also proposed a few speculations regarding how dif-ferent mechanosensitive signaling events may eventually beintegrated and converge onto several core mechanotrans-
duction axes, which constitute the ‘‘chain of commands’’from extracellular nanotopography to nuclear gene expres-sion. Future studies of nanotopography-responsive stem cellbehaviors may have to involve spatiotemporal dynamics
779

nd regulations of molecular mechanosensors, transduc-rs, integrators, and actuators residing in different cellularompartments and organelles. Elucidating interconnectionsnd cross-regulations between each component involved inechanotransduction remain the central goal for the field to

ddress in the future. Multidisciplinary approaches mergingtem cell biology, materials science, biomedical engineer-ng, and nanotechnology will be the most promising andowerful route toward such goal.

cknowledgments

e acknowledge financial support from the National Scienceoundation (CMMI 1129611 and CBET 1149401), the Nationalnstitutes of Health (R21 HL114011 and R21 EB017078), themerican Heart Association (12SDG12180025), the UM Com-rehensive Cancer Center Prostate SPORE Pilot Project (P50A069568), and the Michigan Institute for Clinical & Healthesearch (MICHR) Pilot Program (UL1 RR024986). W. Chen isupported in part by the American Heart Association Predoc-oral Fellowship (13PRE16510018). Finally, we extend ourpologies to all our colleagues in the field whose work were unable to discuss formally because of space constraints.

eferences

[1] G.Q. Daley, D.T. Scadden, Cell 132 (2008) 544—548.[2] S.I. Nishikawa, L.M. Jakt, T. Era, Nat. Rev. Mol. Cell Biol. 8

(2007) 502—507.[3] C. Jopling, S. Boue, J.C.I. Belmonte, Nat. Rev. Mol. Cell Biol.

12 (2011) 79—89.[4] I. Singec, R. Jandial, A. Crain, G. Nikkhah, E.Y. Snyder, Annu.

Rev. Cell Dev. Biol. 58 (2007) 313—328.[5] L.E. O’Brien, D. Bilder, Annu. Rev. Cell Dev. Biol. 29 (2013)

107—136.[6] R.R. Nadig, J. Conserv. Dent. 12 (2009) 131—138.[7] H.E. Young, A.C. Black, Anat. Rec. A 276A (2004) 75—102.[8] J.A. Thomson, J. Itskovitz-Eldor, S.S. Shapiro, M.A. Waknitz,

J.J. Swiergiel, V.S. Marshall, J.M. Jones, Science 282 (1998)1145—1147.

[9] K. Takahashi, K. Tanabe, M. Ohnuki, M. Narita, T. Ichisaka, K.Tomoda, S. Yamanaka, Cell 131 (2007) 861—872.

[10] C.E. Murry, G. Keller, Cell 132 (2008) 661—680.[11] R.M. Tenney, D.E. Discher, Curr. Opin. Cell Biol. 21 (2009)

630—635.[12] T. Harada, J. Swift, J. Irianto, J.W. Shin, K.R. Spinler, A.

Athirasala, R. Diegmiller, P.C.D.P. Dingal, I.L. Ivanovska, D.E.Discher, J. Cell Biol. 204 (2014) 669—682.

[13] J. Takagi, B.M. Petre, T. Walz, T.A. Springer, Cell 110 (2002)599—611.

[14] J. Park, S. Bauer, K. von der Mark, P. Schmuki, Nano Lett. 7(2007) 1686—1691.

[15] M. Crisan, S. Yap, L. Casteilla, C.W. Chen, M. Corselli, T.S.Park, G. Andriolo, B. Sun, B. Zheng, L. Zhang, C. Norotte,P.N. Teng, J. Traas, R. Schugar, B.M. Deasy, S. Badylak, H.J.Buhring, J.P. Giacobino, L. Lazzari, J. Huard, B. Peault, CellStem Cell 3 (2008) 301—313.

[16] E. Kingham, R.O. Oreffo, ACS Nano 7 (2013) 1867—1881.[17] H.N. Kim, A. Jiao, N.S. Hwang, M.S. Kim, D.H. Kang, D.H. Kim,

K.Y. Suh, Adv. Drug Deliv. Rev. 65 (2013) 536—558.
[18] R.K. Das, O.F. Zouani, C. Labrugere, R. Oda, M.C. Durrieu,
ACS Nano 7 (2013) 3351—3361.[19] M. Dalby, N. Gadegaard, R. Oreffo, Nat. Mater. 13 (2014)

558—569.

7
80
[20] L.E. McNamara, T. Sjostrom, K. Seunarine, R.D. Meek, B. Su,M.J. Dalby, J. Tissue Eng. 5 (2014), 2041731414536177.

[21] I.R. Lemischka, K.A. Moore, Nature 425 (2003) 778—779.[22] A. Spradling, D. Drummond-Barbosa, T. Kai, Nature 414 (2001)

98—104.[23] Y. Sun, C.S. Chen, J. Fu, Annu. Rev. Biophys. 41 (2012)

519—542.[24] Y. Sun, K.M.A. Yong, L.G. Villa-Diaz, X. Zhang, W. Chen, R.

Philson, S. Weng, H. Xu, P.H. Krebsbach, J. Fu, Nat. Mater. 13(2014) 599—604.

[25] D.H. Kim, P.P. Provenzano, C.L. Smith, A. Levchenko, J. CellBiol. 197 (2012) 351—360.

[26] F. Guilak, D.M. Cohen, B.T. Estes, J.M. Gimble, W. Liedtke,C.S. Chen, Cell Stem Cell 5 (2009) 17—26.

[27] K.C. Wu, C.L. Tseng, C.C. Wu, F.C. Kao, Y.K. Tu, E.C. So, Y.K.Wang, Sci. Technol. Adv. Mater. 14 (2013) 054401.

[28] V. Vogel, M. Sheetz, Nat. Rev. Mol. Cell Biol. 7 (2006)265—275.

[29] B. Geiger, J.P. Spatz, A.D. Bershadsky, Nat. Rev. Mol. Cell Biol.10 (2009) 21—33.

[30] C.C. DuFort, M.J. Paszek, V.M. Weaver, Nat. Rev. Mol. CellBiol. 12 (2011) 308—319.

[31] N. Wang, J.D. Tytell, D.E. Ingber, Nat. Rev. Mol. Cell Biol. 10(2009) 75—82.

[32] A. Buxboim, I. Ivanovska, D. Discher, J. Cell Sci. 123 (2010)297—308.

[33] P. Isermann, J. Lammerding, Curr. Biol. 23 (2013)R1113—R1121.

[34] C.J. Bettinger, Z.T. Zhang, S. Gerecht, J.T. Borenstein, R.Langer, Adv. Mater. 20 (2008) 99—103.

[35] W. Chen, H. Ahmed, Appl. Phys. Lett. 62 (1993) 1499—1501.[36] A.I. Teixeira, G.A. Abrams, P.J. Bertics, C.J. Murphy, P.F.

Nealey, J. Cell Sci. 116 (2003) 1881—1892.[37] M.J. Dalby, N. Gadegaard, R. Tare, A. Andar, M.O. Riehle, P.

Herzyk, C.D.W. Wilkinson, R.O.C. Oreffo, Nat. Mater. 6 (2007)997—1003.

[38] J. Gallagher, K. McGhee, C. Wilkinson, M. Riehle, IEEE Trans.Nanobiosci. 1 (2002) 24—28.

[39] S.M. Yang, S.G. Jang, D.G. Choi, S. Kim, H.K. Yu, Small 2(2006) 458—475.

[40] G. Zhang, D.Y. Wang, Chem. Asian J. 4 (2009) 236—245.[41] S.P. Bhawalkar, J. Qian, M.C. Heiber, L. Jia, Langmuir 26

(2010) 16662—16666.[42] A. Iovan, M. Fischer, R. Lo Conte, V. Korenivski, Beilstein J.

Nanotechnol. 3 (2012) 884—892.[43] M.J. Dalby, C.C. Berry, M.O. Riehle, D.S. Sutherland, H.

Agheli, A.S.G. Curtis, Exp. Cell Res. 295 (2004) 387—394.[44] M.J. Dalby, M.O. Riehle, D.S. Sutherland, H. Agheli, A.S. Cur-

tis, Biomaterials 25 (2004) 5415—5422.[45] R.C. Jaeger, Introduction to Microelectronic Fabrication, 2nd

ed., Prentice Hall, Upper Saddle River, NJ, 2002.[46] K. Jain, C.G. Willson, B.J. Lin, Electron Device Lett. 3 (1982)

53—55.[47] K. Jain, R.T. Kerth, Appl. Opt. 23 (1984) 648—650.[48] A.K.-K. Wong, Resolution Enhancement Techniques in Optical

Lithography, SPIE Press, Bellingham, WA, 2001.[49] S. Gerecht, C.J. Bettinger, Z. Zhang, J.T. Borenstein,

G. Vunjak-Novakovic, R. Langer, Biomaterials 28 (2007)4068—4077.

[50] C. Vieu, F. Carcenac, A. Pepin, Y. Chen, M. Mejias, A. Lebib, L.Manin-Ferlazzo, L. Couraud, H. Launois, Appl. Surf. Sci. 164(2000) 111—117.

[53] M. Wood, D. Meredith, G. Owen, IEEE Trans. Nanobiosci. 1(2002) 133—140.

[54] A.S. Andersson, J. Brink, U. Lidberg, D.S. Sutherland, IEEE
Trans. Nanobiosci. 2 (2003) 49—57.
[55] M. Wood, C. Wilkinson, A. Curtis, IEEE Trans. Nanobiosci. 5(2006) 20—31.

W. Chen et al.

[56] M.A. Wood, J. R. Soc. Interface 4 (2007) 1—17.[57] L. Ji, V.L. LaPointe, N.D. Evans, M.M. Stevens, Eur. Cell Mater.

23 (2012) 135—146.[58] A. Solanki, S. Shah, P. Yin, K.-B. Lee, Sci. Rep. 3 (2013) 1553.[59] S. Grego, T.W. Jarvis, B.R. Stoner, J.S. Lewis, Langmuir 21

(2005) 4971—4975.[60] Y. Kong, C. Tu, P. Donovan, A. Yee, Acta Biomater. 9 (2013)

6369—6380.[61] J. Malmström, J. Lovmand, S. Kristensen, M. Sundh, M. Duch,

D. Sutherland, Nano Lett. 11 (2011) 2264—2271.[62] V. Ng, Y.V. Lee, B.T. Chen, A.O. Adeyeye, Nanotechnology 13

(2002) 554—558.[63] S.Y. Chou, P.R. Krauss, P.J. Renstrom, Science 272 (1996)

85—87.[64] L.J. Guo, J. Phys. D: Appl. Phys. 37 (2004) R123—R141.[65] H. Schift, J. Vac. Sci. Technol. B 26 (2008) 458—480.[66] M.D. Austin, H.X. Ge, W. Wu, M.T. Li, Z.N. Yu, D. Wasser-

man, S.A. Lyon, S.Y. Chou, Appl. Phys. Lett. 84 (2004)5299—5301.

[67] H. Tan, L. Kong, M.T. Li, C. Steere, L. Koecher, Emerging Litho-graphic Technologies VIII, vol. 5374, SPIE, 2004, pp. 213—221.

[68] K. Yang, K. Jung, E. Ko, J. Kim, K.I. Park, J. Kim, S.-W. Cho,ACS Appl. Mater. Interfaces 5 (2013) 10529—10540.

[69] E.K. Yim, R.M. Reano, S.W. Pang, A.F. Yee, C.S. Chen, K.W.Leong, Biomaterials 26 (2005) 5405—5413.

[70] J. Padmanabhan, E. Kinser, M. Stalter, C. Duncan-Lewis, J.Balestrini, A. Sawyer, J. Schroers, T. Kyriakides, ACS Nano 8(2014) 4366—4375.

[71] Y. Xia, G. M Whitesides, Annu. Rev. Mater. Sci. 28 (1998)153—184.

[72] D. Qin, Y. Xia, G. Whitesides, Nat. Protoc. 5 (2010) 491—502.[73] T.W. Odom, C.J. Love, D.B. Wolfe, K.E. Paul, G.M. Whitesides,

Langmuir 18 (2002) 5314—5320.[74] D.H. Kim, K. Han, K. Gupta, K.W. Kwon, K.Y. Suh, A.

Levchenko, Biomaterials 30 (2009) 5433—5444.[75] E.K.F. Yim, S.W. Pang, K.W. Leong, Exp. Cell Res. 313 (2007)

1820—1829.[76] E.D. Palik, O.J. Glembocki, I. Heard, P.S. Burno, L. Tenerz, J.

Appl. Phys. 70 (1991) 3291—3300.[77] A. Thapa, T.J. Webster, K.M. Haberstroh, J. Biomed. Mater.

Res. A 67A (2003) 1374—1383.[78] B.D. Boyan, R. Batzer, K. Kieswetter, Y. Liu, D.L. Cochran, S.

Szmuckler-Moncler, D.D. Dean, Z. Schwartz, J. Biomed. Mater.Res. 39 (1998) 77—85.

[79] W. Chen, Y. Sun, J. Fu, Small 9 (2013) 81—89.[80] W. Chen, L. Villa-Diaz, Y. Sun, S. Weng, J. Kim, R. Lam, L. Han,

R. Fan, P. Krebsbach, J. Fu, ACS Nano 6 (2012) 4094—4103.[81] I. Zubel, M. Kramkowska, Sens. Actuator A: Phys. 93 (2001)

138—147.[82] E. Metwalli, C.G. Pantano, Nucl. Instrum. Methods Phys. Res.

B: Beam Interact. Mater. At. 207 (2003) 21—27.[83] P.W. Leech, Vacuum 55 (1999) 191—196.[84] N. Bhardwaj, S.C. Kundu, Biotechnol. Adv. 28 (2010) 325—347.[85] D. Li, Y.N. Xia, Adv. Mater. 16 (2004) 1151—1170.[86] A. Greiner, J.H. Wendorff, Angew. Chem. Int. Ed. 46 (2007)

5670—5703.[87] F. Yang, R. Murugan, S. Wang, S. Ramakrishna, Biomaterials

26 (2005) 2603—2610.[88] L. He, S. Liao, D. Quan, K. Ma, C. Chan, S. Ramakrishna, J.

Lu, Acta Biomater. 6 (2010) 2960—2969.[89] M. Kabiri, M. Soleimani, I. Shabani, K. Futrega, N. Ghaemi,

H.H. Ahvaz, E. Elahi, M.R. Doran, Biotechnol. Lett. 34 (2012)1357—1365.

[90] E. Kijenska, M.P. Prabhakaran, W. Swieszkowski, K.J. Kurzyd-lowski, S. Ramakrishna, J. Biomed. Mater. Res. B: Appl.
Biomater. 100 (2012) 1093—1102.
[91] M.P. Prabhakaran, J.R. Venugopal, S. Ramakrishna, Biomate-rials 30 (2009) 4996—5003.

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

[

[

[

[

[

[[

[

[

[


[92] M. Massumi, M. Abasi, H. Babaloo, P. Terraf, M. Safi, M. Saeed,J. Barzin, M. Zandi, M. Soleimani, Tissue Eng. A 18 (2012)609—620.

[93] G.A. Abrams, S.L. Goodman, P.F. Nealey, M. Franco, C.J. Mur-phy, Cell Tissue Res. 299 (2000) 39—46.

[94] M. Schindler, I. Ahmed, J. Kamal, A. Nur-E-Kamal, T.H. Grafe,H.Y. Chung, S. Meiners, Biomaterials 26 (2005) 5624—5631.

[95] C. Gandhimathi, J. Venugopal, R. Ravichandran, S. Sundarra-jan, S. Suganya, S. Ramakrishna, Macromol. Biosci. 13 (2013)696—706.

[96] H. Yoshimoto, Y.M. Shin, H. Terai, J.P. Vacanti, Biomaterials24 (2003) 2077—2082.

[97] W.J. Li, R. Tuli, C. Okafor, A. Derfoul, K.G. Danielson, D.J.Hall, R.S. Tuan, Biomaterials 26 (2005) 599—609.

[98] S.H. Lim, X.Y. Liu, H. Song, K.J. Yarema, H.Q. Mao, Biomate-rials 31 (2010) 9031—9039.

[99] K. Gauthaman, J.R. Venugopal, F.C. Yee, G.S. Peh, S. Rama-krishna, A. Bongso, J. Cell. Mol. Med. 13 (2009) 3475—3484.

[100] J.W. Xie, S.M. Willerth, X.R. Li, M.R. Macewan, A. Rader, S.E.Sakiyama-Elbert, Y.N. Xia, Biomaterials 30 (2009) 354—362.

[101] G.T. Christopherson, H. Song, H.Q. Mao, Biomaterials 30(2009) 556—564.

[102] K.N. Chua, C. Chai, P.C. Lee, Y.N. Tang, S. Ramakrishna, K.W.Leong, H.Q. Mao, Biomaterials 27 (2006) 6043—6051.

[103] L. Liu, Q. Yuan, J. Shi, X. Li, D. Jung, L. Wang, K. Yamauchi,N. Nakatsuji, K. Kamei, Y. Chen, Biotechnol. Lett. 34 (2012)1951—1957.

[104] J.M. Dang, K.W. Leong, Adv. Mater. 19 (2007) 2775—2779.[105] L.X. Lu, X.F. Zhang, Y.Y. Wang, L. Ortiz, X. Mao, Z.L. Jiang,

Z.D. Xiao, N.P. Huang, ACS Appl. Mater. Interfaces 5 (2013)319—330.

[106] H. Peng, Z. Yin, H. Liu, X. Chen, B. Feng, H. Yuan, B. Su, H.Ouyang, Y. Zhang, Nanotechnology 23 (2012) 485102.

[107] S. Patel, K. Kurpinski, R. Quigley, H. Gao, B. Hsiao, M.-M. Poo,S. Li, Nano Lett. 7 (2007) 2122—2128.

[108] J. Qu, D. Wang, H. Wang, Y. Dong, F. Zhang, B. Zuo, H. Zhang,J. Biomed. Mater. Res. A 101 (2013) 2667—2678.

[109] Y.S. Nam, T.G. Park, J. Biomed. Mater. Res. 47 (1999) 8—17.[110] R.Y. Zhang, P.X. Ma, J. Biomed. Mater. Res. 44 (1999) 446—455.[111] S.H. Lee, B.S. Kim, S.H. Kim, S.W. Kang, Y.H. Kim, Macromol.

Biosci. 4 (2004) 802—810.[112] L.A. Smith, X.H. Liu, P.X. Ma, Soft Matter 4 (2008) 2144—2149.[113] P.X. Ma, R. Zhang, J. Biomed. Mater. Res. 46 (1999) 60—72.[114] L.A. Smith, X. Liu, J. Hu, P. Wang, P.X. Ma, Tissue Eng. A 15

(2009) 1855—1864.[115] S. Affrossman, G. Henn, S.A. ONeill, R.A. Pethrick, M. Stamm,

Macromolecules 29 (1996) 5010—5016.[116] S. Walheim, M. Boltau, J. Mlynek, G. Krausch, U. Steiner,

Macromolecules 30 (1997) 4995—5003.[117] M.J. Dalby, M.O. Riehle, H.J.H. Johnstone, S. Affrossman,

A.S.G. Curtis, Tissue Eng. 8 (2002) 1099—1108.[118] M.J. Dalby, D. Giannaras, M.O. Riehle, N. Gadegaard, S.

Affrossman, A.S. Curtis, Biomaterials 25 (2004) 77—83.[119] M.J. Dalby, D. Pasqui, S. Affrossman, IEEE. Proc. Nanobiotech-

nol. 151 (2004) 53—61.[120] M. Sprenger, S. Walheim, C. Schafle, U. Steiner, Adv. Mater.

15 (2003) 703—706.[121] M. Park, C. Harrison, P.M. Chaikin, R.A. Register, D.H. Adam-

son, Science 276 (1997) 1401—1404.[122] C.J. Hawker, T.P. Russell, MRS Bull. 30 (2005) 952—966.[123] I.W. Hamley, Nanotechnology 14 (2003) R39—R54.[124] R. Glass, M. Möller, J. Spatz, Nanotechnology 14 (2003) 1153.[125] S.E. Maclaine, N. Gadhari, R. Pugin, R.M.D. Meek, M. Liley,

M.J. Dalby, J. Orthop. Res. 30 (2012) 1190—1197.
[126] M. Salaun, M. Zelsmann, S. Archambault, D. Borah, N. Keha-
gias, C. Simao, O. Lorret, M.T. Shaw, C.M.S. Torres, M.A.Morris, J. Mater. Chem. C 1 (2013) 3544—3550.

[

781

127] H.L. Khor, Y. Kuan, H. Kukula, K. Tamada, W. Knoll,M. Moeller, D.W. Hutmacher, Biomacromolecules 8 (2007)1530—1540.

128] M. Haupt, S. Miller, R. Glass, M. Arnold, R. Sauer, K. Thonke,M. Moller, J.P. Spatz, Adv. Mater. 15 (2003) 829—831.

129] S.J. Ku, S.A. Woo, D. Seo, H. Song, J.B. Kim, Eur. Polym. J.47 (2011) 305—310.

130] I.C. Reynhout, G. Delaittre, H.C. Kim, R.J.M. Nolte, J.J.L.M.Cornelissen, J. Mater. Chem. B 1 (2013) 3026—3030.

131] K. Lee, A. Mazare, P. Schmuki, Chem. Rev. 114 (2014)9385—9454.

132] X.M. Zhou, N.T. Nguyen, S. Ozkan, P. Schmuki, Electrochem.Commun. 46 (2014) 157—162.

133] C. Yao, T.J. Webster, J. Nanosci. Nanotechnol. 6 (2006)2682—2692.

134] C. Yao, V. Perla, J.L. McKenzie, E.B. Slamovich, T.J. Webster,J. Biomed. Nanotechnol. 1 (2005) 68—73.

135] S.H. Oh, R.R. Finones, C. Daraio, L.H. Chen, S.H. Jin, Bioma-terials 26 (2005) 4938—4943.

136] T. Ohgai, X. Hoffer, L. Gravier, J.P. Ansermet, J. Appl. Elec-trochem. 34 (2004) 1007—1012.

137] K. Watanabe, M. Sakairi, H. Takahashi, S. Hirai, S. Yamaguchi,J. Electroanal. Chem. 473 (1999) 250—255.

138] H. Tsuchiya, J.M. Macak, A. Ghicov, Y.C. Tang, S. Fujimoto,M. Niinomi, T. Noda, P. Schmuki, Electrochim. Acta 52 (2006)94—101.

139] S. Oh, K.S. Brammer, Y.S. Li, D. Teng, A.J. Engler, S. Chien, S.Jin, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 2130—2135.

140] T. Sjostrom, M.J. Dalby, A. Hart, R. Tare, R.O.C. Oreffo, B.Su, Acta Biomater. 5 (2009) 1433—1441.

141] L.R. Clements, P.Y. Wang, W.B. Tsai, H. Thissen, N.H. Voelcker,Lab Chip 12 (2012) 1480—1486.

142] M. Jarcho, C.H. Bolen, M.B. Thomas, J. Bobick, J.F. Kay, R.H.Doremus, J. Mater. Sci. 11 (1976) 2027—2035.

143] T. Nishimura, M. Mitomo, H. Hirotsuru, M. Kawahara, J. Mater.Sci. Lett. 14 (1995) 1046—1047.

144] M. Nygren, Z.J. Shen, Solid State Sci. 5 (2003) 125—131.145] J. Rankin, B.W. Sheldon, Mater. Sci. Eng. A: Struct. Mater.

Prop. Microstruct. Process. 204 (1995) 48—53.146] A.J. Dulgar-Tulloch, R. Bizios, R.W. Siegel, J. Biomed. Mater.

Res. A 90A (2009) 586—594.147] J. Fu, Y.K. Wang, M.T. Yang, R.A. Desai, X. Yu, Z. Liu, C.S.

Chen, Nat. Methods 7 (2010) 733—736.148] P.R. Sreerekha, D. Menon, S.V. Nair, K.P. Chennazhi, Tissue

Eng. A 19 (2013) 849—859.149] S. Namgung, T. Kim, K.Y. Baik, M. Lee, J.M. Nam, S. Hong,

Small 7 (2011) 56—61.150] D. Khang, S.Y. Kim, P. Liu-Snyder, G.T.R. Palmore, S.M. Durbin,

T.J. Webster, Biomaterials 28 (2007) 4756—4768.151] O.F. Zouani, C. Chanseau, B. Brouillaud, R. Bareille, F.

Deliane, M.P. Foulc, A. Mehdi, M.C. Durrieu, J. Cell Sci. 125(2012) 1217—1224.

152] D. Woodbury, E.J. Schwarz, D.J. Prockop, I.B. Black, J. Neu-rosci. Res. 61 (2000) 364—370.

153] F.H. Gage, S. Temple, Neuron 80 (2013) 588—601.154] S. Shreyas, T.Y. Perry, M.U. Thiers, C. Sy-Tsong Dean, Y. Letao,

L. Ki-Bum, Adv. Mater. 26 (2014) 3673—3680.155] S. Lee, M.K. Leach, S.A. Redmond, S.Y.C. Chong, S.H. Mellon,

S.J. Tuck, Z.Q. Feng, J.M. Corey, J.R. Chan, Nat. Methods 9(2012) 917.

156] J. Gerardo-Nava, T. Fuhrmann, K. Klinkhammer, N. Seiler,J. Mey, D. Klee, M. Moller, P.D. Dalton, G.A. Brook,Nanomedicine (London) 4 (2009) 11—30.

157] E.K.A. Nur, I. Ahmed, J. Kamal, M. Schindler, S. Meiners, Stem
Cells 24 (2006) 426—433.
158] F. Chowdhury, S. Na, D. Li, Y.C. Poh, T.S. Tanaka, F. Wang, N.Wang, Nat. Mater. 9 (2010) 82—88.

7

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

[

[

[

[

[

[

[

[

[

[[

82

159] F. Chowdhury, Y.Z. Li, Y.C. Poh, T. Yokohama-Tamaki, N. Wang,T.S. Tanaka, PLoS ONE 5 (2010).

160] A.J. Keung, P. Asuri, S. Kumar, D.V. Schaffer, Integr. Biol. 4(2012) 1049—1058.

161] Y.B. Sun, L.G. Villa-Diaz, R.H.W. Lam, W.Q. Chen, P.H. Kreb-sbach, J.P. Fu, PLoS ONE 7 (2012).

162] E. Kingham, K. White, N. Gadegaard, M.J. Dalby, R.O.C.Oreffo, Small 9 (2013) 2140—2151.

163] S.H. Lee, N. Lumelsky, L. Studer, J.M. Auerbach, R.D. McKay,Nat. Biotechnol. 18 (2000) 675—679.

164] H. Wichterle, I. Lieberam, J.A. Porter, T.M. Jessell, Cell 110(2002) 385—397.

165] O. Brustle, K.N. Jones, R.D. Learish, K. Karram, K. Choudhary,O.D. Wiestler, I.D. Duncan, R.D. McKay, Science 285 (1999)754—756.

166] T.I. Chao, S. Xiang, C.S. Chen, W.C. Chin, A.J. Nelson, C.Wang, J. Lu, Biochem. Biophys. Res. Commun. 384 (2009)426—430.

167] I. Sridharan, T. Kim, R. Wang, Biochem. Biophys. Res. Com-mun. 381 (2009) 508—512.

168] M.R. Lee, K.W. Kwon, H. Jung, H.N. Kim, K.Y. Suh, K. Kim,K.S. Kim, Biomaterials 31 (2010) 4360—4366.

169] N.I. zur Nieden, G. Kempka, H.J. Ahr, Differentiation 71(2003) 18—27.

170] E. Garreta, E. Genove, S. Borros, C.E. Semino, Tissue Eng. 12(2006) 2215—2227.

171] J. Yamashita, H. Itoh, M. Hirashima, M. Ogawa, S. Nishikawa,T. Yurugi, M. Naito, K. Nakao, S. Nishikawa, Nature 408 (2000)92—96.

172] M. Drab, H. Haller, R. Bychkov, B. Erdmann, C. Lindschau, H.Haase, I. Morano, F.C. Luft, A.M. Wobus, FASEB J. 11 (1997)905—915.

173] R.M. Schmitt, E. Bruyns, H.R. Snodgrass, Gene Dev. 5 (1991)728—740.

174] M.V. Wiles, G. Keller, Development 111 (1991) 259—267.175] C. Dani, A.G. Smith, S. Dessolin, P. Leroy, L. Staccini, P.

Villageois, C. Darimont, G. Ailhaud, J. Cell Sci. 110 (1997)1279—1285.

176] J. Kramer, C. Hegert, K.M. Guan, A.M. Wobus, P.K. Muller, J.Rohwedel, Mech. Dev. 92 (2000) 193—205.

177] K.N. Chua, C. Chai, P.C. Lee, S. Ramakrishna, K.W. Leong,H.Q. Mao, Exp. Hematol. 35 (2007) 771—781.

178] M.C. Siebers, P.J. ter Brugge, X.F. Walboomers, J.A. Jansen,Biomaterials 26 (2005) 137—146.

179] M.A. Schwartz, M.H. Ginsberg, Nat. Cell Biol. 4 (2002)E65—E68.

180] M. Arnaout, S. Goodman, J.-P. Xiong, Curr. Opin. Cell Biol. 19(2007) 495—507.

181] O. Rossier, V. Octeau, J.-B. Sibarita, C. Leduc, B. Tessier, D.Nair, V. Gatterdam, O. Destaing, C. Albigès-Rizo, R. Tampé,L. Cognet, D. Choquet, B. Lounis, G. Giannone, Nat. Cell Biol.14 (2012) 1057—1067.

182] H. Schiller, M.-R. Hermann, J. Polleux, T. Vignaud, S. Zanivan,C. Friedel, Z. Sun, A. Raducanu, K.-E. Gottschalk, M. Théry,M. Mann, R. Fässler, Nat. Cell Biol. 15 (2013) 625—636.

183] L. Li, S.A.L. Bennett, L.S. Wang, Cell Adhes. Migr. 6 (2012)59—70.

184] Y. Meng, S. Eshghi, Y.J. Li, R. Schmidt, D.V. Schaffer, K.E.Healy, FASEB J. 24 (2010) 1056—1065.

185] T.J. Rowland, L.M. Miller, A.J. Blaschke, E.L. Doss, A.J. Bon-ham, S.T. Hikita, L.V. Johnson, D.O. Clegg, Stem Cells Dev. 19(2010) 1231—1240.

186] P. Tsimbouri, N. Gadegaard, K. Burgess, K. White, P. Reynolds,P. Herzyk, R. Oreffo, M.J. Dalby, J. Cell. Biochem. 115 (2014)
380—390.
187] R. Hynes, Cell (2002).188] J. Qin, O. Vinogradova, E. Plow, PLoS Biol. 2 (2004).

W. Chen et al.

[189] J.A. Askari, P.A. Buckley, A.P. Mould, M.J. Humphries, J. CellSci. 122 (2009) 165—170.

[190] B. Geiger, J. Spatz, A. Bershadsky, Nat. Rev. Mol. Cell Biol. 10(2009) 21—33.

[191] R. Zaidel-Bar, B. Geiger, J. Cell Sci. 123 (2010) 1385—1388.[192] L.Y. Koo, D.J. Irvine, A.M. Mayes, D.A. Lauffenburger, L.G.

Griffith, J. Cell Sci. 115 (2002) 1423—1433.[193] M. Arnold, E. Cavalcanti-Adam, R. Glass, J. Blümmel, W. Eck,

M. Kantlehner, H. Kessler, J. Spatz, ChemPhysChem 5 (2004)383—388.

[194] E. Cavalcanti-Adam, T. Volberg, A. Micoulet, H. Kessler, B.Geiger, J. Spatz, Biophys. J. 92 (2007) 2964—2974.

[195] M. Schvartzman, M. Palma, J. Sable, J. Abramson, X. Hu, M.Sheetz, S. Wind, Nano Lett. 11 (2011) 1306—1312.

[196] G. Maheshwari, G. Brown, D.A. Lauffenburger, A. Wells, L.G.Griffith, J. Cell Sci. 113 (2000) 1677—1686.

[197] J. Huang, S. Grater, F. Corbellini, S. Rinck, E. Bock, R.Kemkemer, H. Kessler, J. Ding, J. Spatz, Nano Lett. 9 (2009)1111—1116.

[198] M. Dalby, N. Gadegaard, R. Tare, A. Andar, M. Riehle,P. Herzyk, C. Wilkinson, R. Oreffo, Nat. Mater. 6 (2007)997—1003.

[199] S. Mitra, D. Hanson, D. Schlaepfer, Nat. Rev. Mol. Cell Biol. 6(2005) 56—68.

[200] J. Seong, A. Tajik, J. Sun, J.L. Guan, M.J. Humphries,S.E. Craig, A. Shekaran, A.J. Garcia, S.Y. Lu, M.Z. Lin, N.Wang, Y.X. Wang, Proc. Natl. Acad. Sci. U. S. A. 110 (2013)19372—19377.

[201] D.D. Schlaepfer, S.K. Hanks, T. Hunter, P. Vandergeer, Nature372 (1994) 786—791.

[202] S.Y. Hong, Y.M. Jeon, H.J. Lee, J.G. Kim, J.A. Baek, J.C. Lee,Mol. Cell. Biochem. 335 (2010) 263—272.

[203] K.K. Wary, A. Mariotti, C. Zurzolo, F.G. Giancotti, Cell 94(1998) 625—634.

[204] R. Aikawa, T. Nagai, S. Kudoh, Y.Z. Zou, M. Tanaka, M. Tamura,H. Akazawa, H. Takano, R. Nagai, I. Komuro, Hypertension 39(2002) 233—238.

[205] B.K.K. Teo, S.T. Wong, C.K. Lim, T.Y.S. Kung, C.H. Yap, Y. Ram-agopal, L.H. Romer, E.K.F. Yim, ACS Nano 7 (2013) 4785—4798.

[206] D.A. Hsia, S.K. Mitra, C.R. Hauck, D. Streblow, J.A. Nelson,D. Ilic, S. Huang, E.G. Li, G.R. Nemerow, J. Leng, K.S.R.Spencer, D.A. Cheresh, D.D. Schlaepfer, J. Cell Biol. 160(2003) 753—767.

[207] D.D. Schlaepfer, S.K. Mitra, D. Ilic, BBA Mol. Cell Res. 1692(2004) 77—102.

[208] J.B. Zhai, H. Lin, Z.Y. Nie, J.H. Wu, R. Canete-Soler,W.W. Schlaepfer, D.D. Schalepfer, J. Biol. Chem. 278 (2003)24865—24873.

[209] B. Rico, H.E. Beggs, D. Schahin-Reed, N. Kimes, A. Schmidt,L.F. Reichardt, Nat. Neurosci. 7 (2004) 1059—1069.

[210] J.D. Hildebrand, J.M. Taylor, J.T. Parsons, Mol. Cell. Biol. 16(1996) 3169—3178.

[211] X.D. Ren, W.B. Kiosses, D.J. Sieg, C.A. Otey, D.D. Schlaepfer,M.A. Schwartz, J. Cell Sci. 113 (2000) 3673—3678.

[212] F. Kantawong, K.E.V. Burgess, K. Jayawardena, A. Hart, R.J.Burchmore, N. Gadegaard, R.O.C. Oreffo, M.J. Dalby, Bioma-terials 30 (2009) 4723—4731.

[213] H. Xia, R.S. Nho, J. Kahm, J. Kleidon, C.A. Henke, J. Biol.Chem. 279 (2004) 33024—33034.

[214] J. Du, X. Chen, X. Liang, G. Zhang, J. Xu, L. He, Q. Zhan,X.-Q. Feng, S. Chien, C. Yang, Proc. Natl. Acad. Sci. U. S. A.108 (2011) 9466—9471.

[215] L. Armstrong, O. Hughes, S. Yung, L. Hyslop, R. Stewart, I.Wappler, H. Peters, T. Walter, P. Stojkovic, J. Evans, M. Sto-
jkovic, M. Lako, Hum. Mol. Genet. 15 (2006) 1894—1913.
[216] S. Chambers, C. Fasano, E. Papapetrou, M. Tomishima, M.Sadelain, L. Studer, Nat. Biotechnol. 27 (2009) 275—280.

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[

[[

[


[217] A.M. Singh, D. Reynolds, T. Cliff, S. Ohtsuka, A.L. Mattheyses,Y.H. Sun, L. Menendez, M. Kulik, S. Dalton, Cell Stem Cell 10(2012) 312—326.

[218] J. Kim, H.N. Kim, K.T. Lim, Y. Kim, H. Seonwoo, S.H. Park,H.J. Lim, D.H. Kim, K.Y. Suh, P.H. Choung, Y.H. Choung, J.H.Chung, Sci. Rep. 3 (2013).

[219] J. Yang, L. McNamara, N. Gadegaard, E. Alakpa, K. Burgess,R. Meek, M. Dalby, ACS Nano 8 (2014) 9941—9953.

[220] L. Azzolin, F. Zanconato, S. Bresolin, M. Forcato, G. Basso,S. Bicciato, M. Cordenonsi, S. Piccolo, Cell 151 (2012)1443—1456.

[221] K. Burridge, K. Wennerberg, Cell 116 (2004) 167—179.[222] A. Mammoto, S. Huang, K. Moore, P. Oh, D.E. Ingber, J. Biol.

Chem. 279 (2004) 26323—26330.[223] K. Bhadriraju, M. Yang, S. Alom Ruiz, D. Pirone, J. Tan, C.

Chen, Exp. Cell Res. 313 (2007) 3616—3623.[224] C. Guilluy, V. Swaminathan, R. Garcia-Mata, E.T. O’Brien, R.

Superfine, K. Burridge, Nat. Cell Biol. 13 (2011) 722—727.[225] R. McBeath, D.M. Pirone, C.M. Nelson, K. Bhadriraju, C.S.

Chen, Dev. Cell 6 (2004) 483—495.[226] R. McMurray, N. Gadegaard, P. Tsimbouri, K. Burgess, L. McNa-

mara, R. Tare, K. Murawski, E. Kingham, R. Oreffo, M. Dalby,Nat. Mater. 10 (2011) 637—644.

[227] B.K. Teo, S. Wong, C. Lim, T. Kung, C. Yap, Y. Ramgopal, L.Romer, E. Yim, ACS Nano 7 (2013) 4785—4798.

[228] C. DuFort, M. Paszek, V. Weaver, Nat. Rev. Mol. Cell Biol. 12(2011) 308—319.

[229] G. Halder, S. Dupont, S. Piccolo, Nat. Rev. Mol. Cell Biol. 13(2012) 591—600.

[230] S. Piccolo, S. Dupont, M. Cordenonsi, Physiol. Rev. 94 (2014)1287—1312.

[231] J.S. Mo, H.W. Park, K.L. Guan, EMBO Rep. 15 (2014)642—656.

[232] G. Halder, F. Camargo, Cancer Res. (2013).[233] J.X. Dong, G. Feldmann, J.B. Huang, S. Wu, N.L. Zhang, S.A.

Comerford, M.F. Gayyed, R.A. Anders, A. Maitra, D.J. Pan,Cell 130 (2007) 1120—1133.

[234] S. Dupont, L. Morsut, M. Aragona, E. Enzo, S. Giulitti, M. Cor-denonsi, F. Zanconato, J. Le Digabel, M. Forcato, S. Bicciato,N. Elvassore, S. Piccolo, Nature 474 (2011) 179—183.

[235] M. Aragona, T. Panciera, A. Manfrin, S. Giulitti, F. Michielin, N.Elvassore, S. Dupont, S. Piccolo, Cell 154 (2013) 1047—1059.

[236] K.I. Wada, K. Itoga, T. Okano, S. Yonemura, H. Sasaki, Devel-opment 138 (2011) 3907—3914.

[237] B.G. Fernandez, P. Gaspar, C. Bras-Pereira, B. Jezowska, S.R.Rebelo, F. Janody, Development 138 (2011) 2337—2346.

[239] L. Sansores-Garcia, W. Bossuyt, K.I. Wada, S. Yonemura, C.Y.Tao, H. Sasaki, G. Halder, EMBO J. 30 (2011) 2325—2335.

[240] C. Yang, M.W. Tibbitt, L. Basta, K.S. Anseth, Nat. Mater. 13(2014) 645—652.

[241] Y.B. Sun, K.M. Aw Yong, L.G. Villa-Diaz, X.L. Zhang, W.Q.Chen, R. Philson, S.N. Weng, H.X. Xu, P.H. Krebsbach, J.P.Fu, Nat. Mater. 13 (2014) 599—604.

[242] J.T. Connelly, J.E. Gautrot, B. Trappmann, D.W.M. Tan, G.Donati, W.T.S. Huck, F.M. Watt, Nat. Cell Biol. 12 (2010)711—718.

[243] G. Schratt, U. Philippar, J. Berger, H. Schwarz, O. Heidenre-ich, A. Nordheim, J. Cell Biol. 156 (2002) 737—750.

[244] A. Higuchi, Q.-D. Ling, Y. Chang, S.-T. Hsu, A. Umezawa,Chem. Rev. 113 (2013) 3297—3328.

[245] G.V. Shivashankar, Annu. Rev. Biophys. 40 (2011) 361—378.[246] J. Lammerding, C. Schulze, T. Takahashi, S. Kozlov, T. Sullivan,

R.D. Kamm, C.L. Stewart, R.T. Lee, Circulation 108 (2003)289.

[247] J. Lammerding, P.C. Schulze, T. Takahashi, S. Kozlov, T. Sulli-van, R.D. Kamm, C.L. Stewart, R.T. Lee, J. Clin. Invest. 113(2004) 370—378.

[

783

248] C.Y. Ho, D.E. Jaalouk, M.K. Vartiainen, J. Lammerding, Nature497 (2013) 507—511.

249] J.D. Pajerowski, K.N. Dahl, F.L. Zhong, P.J. Sammak, D.E. Dis-cher, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 15619—15624.

250] M.L. Lombardi, D.E. Jaalouk, C.M. Shanahan, B. Burke,K.J. Roux, J. Lammerding, J. Biol. Chem. 286 (2011)26743—26753.

251] N. Wang, J. Tytell, D. Ingber, Nat. Rev. Mol. Cell Biol. 10 (2009)75—82.

252] C.P. Johnson, H.Y. Tang, C. Carag, D.W. Speicher, D.E. Discher,Science 317 (2007) 663—666.

253] J. Swift, I. Ivanovska, A. Buxboim, T. Harada, P. Dingal, J.Pinter, J. Pajerowski, K. Spinler, J.-W. Shin, M. Tewari, F.Rehfeldt, D. Speicher, D. Discher, Science 341 (2013) 1240104.

254] H. Liu, J. Wen, Y. Xiao, J. Liu, S. Hopyan, M. Radisic, C.Simmons, Y. Sun, ACS Nano 8 (2014) 3821—3828.

255] Y.C. Poh, S.P. Shevtsov, F. Chowdhury, D.C. Wu, S. Na, M.Dundr, N. Wang, Nat. Commun. 3 (2012).

256] S. Na, O. Collin, F. Chowdhury, B. Tay, M.X. Ouyang, Y.X. Wang,N. Wang, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 6626—6631.

257] M. Versaevel, T. Grevesse, S. Gabriele, Nat. Commun. 3 (2012)671.

258] C. Guilluy, L.D. Osborne, L. Van Landeghem, L. Sharek, R.Superfine, R. Garcia-Mata, K. Burridge, Nat. Cell Biol. 16(2014) 376—381.

259] K.J. Chalut, K. Kulangara, M.G. Giacomelli, A. Wax, K.W.Leong, Soft Matter 6 (2010) 1675—1681.

260] P.M. Tsimbouri, K. Murawski, G. Hamilton, P. Herzyk, R.O.C.Oreffo, N. Gadegaard, M.J. Dalby, Biomaterials 34 (2013)2177—2184.

261] M.J. Dalby, N. Gadegaard, P. Herzyk, D. Sutherland, H. Agheli,C.D.W. Wilkinson, A.S.G. Curtis, J. Cell. Biochem. 102 (2007)1234—1244.

262] T. Mammoto, A. Mammoto, D.E. Ingber, Annu. Rev. Cell Dev.Biol. 29 (2013) 27—61.

263] X. Varelas, R. Sakuma, P. Samavarchi-Tehrani, R. Peerani, B.M.Rao, J. Dembowy, M.B. Yaffe, P.W. Zandstra, J.L. Wrana, Nat.Cell Biol. 10 (2008) 837—848.

266] C. Baarlink, H.C. Wang, R. Grosse, Science 340 (2013)864—867.

267] J. Kim, H.N. Kim, K.T. Lim, Y. Kim, S. Pandey, P. Garg, Y.H.Choung, P.H. Choung, K.Y. Suh, J.H. Chung, Biomaterials 34(2013) 7257—7268.

268] M. Guvendiren, J.A. Burdick, Adv. Healthc. Mater. 2 (2013)155—164.

269] J. Xi, M. Khalil, N. Shishechian, T. Hannes, K. Pfannkuche, H.Liang, A. Fatima, M. Haustein, F. Suhr, W. Bloch, M. Reppel,T. Saric, M. Wernig, R. Janisch, K. Brockmeier, J. Hescheler,F. Pillekamp, FASEB J. 24 (2010) 2739—2751.

270] A. Doi, I.H. Park, B. Wen, P. Murakami, M.J. Aryee, R. Irizarry,B. Herb, C. Ladd-Acosta, J. Rho, S. Loewer, J. Miller, T.Schlaeger, G.Q. Daley, A.P. Feinberg, Nat. Genet. 41 (2009)1350—1353.

271] M.L. Condic, M. Rao, Stem Cells 26 (2008) 2753—2758.272] S. Corti, M. Nizzardo, C. Simone, M. Falcone, M. Nardini,

D. Ronchi, C. Donadoni, S. Salani, G. Riboldi, F. Magri, G.Menozzi, C. Bonaglia, F. Rizzo, N. Bresolin, G.P. Comi, Sci.Transl. Med. 4 (2012), 165ra162.

273] G. Amabile, A. Meissner, Trends Mol. Med. 15 (2009) 59—68.274] H. Lee, J. Park, B.G. Forget, P. Gaines, Regen. Med. 4 (2009)

759—769.275] M. Wernig, A. Meissner, R. Foreman, T. Brambrink, M.C. Ku, K.

Hochedlinger, B.E. Bernstein, R. Jaenisch, Nature 448 (2007)318—324.

276] R.J. McMurray, N. Gadegaard, P.M. Tsimbouri, K.V. Burgess,L.E. McNamara, R. Tare, K. Murawski, E. Kingham, R.O.Oreffo, M.J. Dalby, Nat. Mater. 10 (2011) 637—644.

7

v

cC

technology to illuminate biological systems atboth the molecular and cellular levels. Dr. Fu is the recipient of theAmerican Heart Association Scientist Development Grant (2012) andthe National Science Foundation CAREER Award (2012).

84

Weiqiang Chen has been an assistant profes-sor of Mechanical and Aerospace Engineeringat New York University since 2014. Dr. Chenreceived his PhD in Mechanical Engineeringfrom the University of Michigan in 2014. Hewas an American Heart Association Predoc-toral Fellow at the University of Michiganfrom 2013 to 2014. Dr. Chen’s research inter-ests focus on Bio-MicroelectromechanicalSystems (BioMEMS), Lab-on-a-Chip, biomate-rials, mechanobiology and stem cell biology.

Yue Shao is a PhD candidate in the Depart-ment of Mechanical Engineering at theUniversity of Michigan, Ann Arbor. He receivedhis BE and ME degrees on EngineeringMechanics and Solid Mechanics from TsinghuaUniversity, China, in 2008 and 2011, respec-tively. His current research interests includebiomechanics of cellular mechanosensing andmechanotransduction of extracellular phys-ical cues, stem cell mechanobiology, andmicro/nanoengineering ex vivo cell microen-

ironment for regenerative medicine.

Xiang Li is a PhD candidate in the Departmentof Mechanical Engineering at the Universityof Michigan, Ann Arbor. He received his BSdegree in Theoretical and Applied Mechanics
from Peking University in 2012. His currentresearch interest is to develop novel micro-engineering tools for single cell studies andtissue engineering.
W. Chen et al.

Gang Zhao has been a professor at Biomedi-cal Engineering Research Center of Universityof Science and Technology of China (USTC)since April 2011. Dr. Zhao received his PhDdegree in Power Engineering and Engineer-ing Thermophysics from USTC in June 2004.He was a post-doctor at USTC from July2004 to September 2006, and a JSPS (JapanSociety for Promotion of Science) researchfellow from July 2008 to October 2010. Hisresearch interests include cryobiology, biomi-

ro/nanofluidics and biomedical microsystems. He received theAREER Award for Young Teachers (CAYT) of USTC in 2013.

Jianping Fu has been an assistant professorof Mechanical and Biomedical Engineering atthe University of Michigan, Ann Arbor since2009. Dr. Fu received his PhD degree fromthe Massachusetts Institute of Technology in2007. He was an American Heart Associa-tion Postdoctoral Fellow at the Universityof Pennsylvania from 2007 to 2009. Dr. Fu’sresearch focuses on mechanobiology, stemcell biology, and applying microfabrication

nanotopographical surfaces for stem cell fate control: engineering...

Documents