Spider silk for xeno-free long-term self-renewal and differentiation of human pluripotent stem cells

lable at ScienceDirect

Biomaterials 35 (2014) 8496e8502

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomateria ls

Spider silk for xeno-free long-term self-renewal and differentiationof human pluripotent stem cells

Siqin Wu a, Jan Johansson a, b, c, Pauliina Damdimopoulou d, Mansoureh Shahsavani e,Anna Falk e, Outi Hovatta d, Anna Rising a, b, *

a Department of Neurobiology, Care Sciences and Society (NVS), Karolinska Institutet, Novum 5th floor, 141 86 Stockholm, Swedenb Department of Anatomy, Physiology and Biochemistry, Swedish University of Agricultural Sciences, The Biomedical Centre, Box 575,751 23 Uppsala, Swedenc Institute of Mathematics and Natural Sciences, Tallinn University, Narva mnt 25, 101 20 Tallinn, Estoniad Department of Clinical Sciences, Intervention and Technology, Division of Obstetrics and Gynecology, Karolinska Institutet andKarolinska University Hospital, Huddinge, 141 86 Stockholm, Swedene Department of Neuroscience, Karolinska Institutet, Retzius v. 8, 171 77 Stockholm, Sweden

a r t i c l e i n f o

Article history:Received 24 April 2014Accepted 20 June 2014Available online 17 July 2014

Keywords:BiomaterialChemically definedHuman embryonic stem cellsHuman induced pluripotent stem cellsFunctionalized materialsScaffold

* Corresponding author. Department of Neurobiolo(NVS), Karolinska Institutet, Novum 5th floor, 141 86 S8 58583610.

E-mail addresses: [email protected], anna.rising@s

http://dx.doi.org/10.1016/j.biomaterials.2014.06.0390142-9612/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

Human pluripotent stem cells (hPSCs) can undergo unlimited self-renewal and have the capacity todifferentiate into all somatic cell types, and are therefore an ideal source for the generation of cells andtissues for research and therapy. To realize this potential, defined cell culture systems that allowexpansion of hPSCs and subsequent controlled differentiation, ideally in an implantable three-dimensional (3D) matrix, are required. Here we mimic spider silk e Nature's high performance mate-rial e for the design of chemically defined 2D and 3D matrices for cell culture. The silk matrices do notonly allow xeno-free long-term expansion of hPSCs but also differentiation in both 2D and 3D. Theseresults show that biomimetic spider silk matrices enable hPSC culture in a manner that can be applied forexperimental and clinical purposes.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Human pluripotent stem cells (hPSCs), including human em-bryonic stem cells (hESCs) and human induced pluripotent stemcells (hiPSCs), have the unique ability to form any specialized tissuein the human body. These cells therefore represent powerful re-sources for applications in regenerative medicine and pharma-ceutical development. However, several technical challenges mustbe addressed before hPSCs can be used routinely for clinical ther-apeutic applications and generation of tissues or organs [1]. First, inorder to generate sufficient number of cells, culture systems thatare cheap, easy-to handle and chemically defined are needed.Second, mechanically robust 3D matrices that are adaptable, welltolerated by the host and able to regulate stem cell fate commit-ments have to be developed. Recently, several groups have devel-opedmatrices for long-term xeno-free expansion of hPSCs [2e7]. In

gy, Care Sciences and Societytockholm, Sweden. Fax: þ46

lu.se (A. Rising).

these reports, hPSCs are maintained on recombinant extracellularmatrix (ECM) proteins or synthetic peptides derived from ECMproteins but none of these substrates have been reported togenerate 3D scaffolds that support proliferation and differentiationof hPSCs, cf below under 3.4. for further details.

Spider silk is an ideal biomaterial, since it is strong, extendibleand is well tolerated and degradedwhen implanted in living tissues[8,9]. However, spiders are difficult to farm and therefore nativespider silk is practically impossible to obtain at large scale. Pro-duction in heterologous hosts may be an alternative route to in-dustrial production of spider silk, but this strategy is associatedwith problems since the spider silk proteins are large and prone toaggregate. Spider silk proteins are composed of an extensive re-petitive region flanked by small folded terminal domains thatregulate silk assembly [10,11]. The low complexity of the about3000 amino acid residue long repetitive segment likely contributesboth to the impressive mechanical properties and presumed lowimmunogenicity of spider silk. Despite the technical problems,recent progress has resulted in cost-efficient methods to produceartificial spider silk in heterologous hosts [12]. We have found thata miniature spider silk protein, referred to as 4RepCT, is easilyproduced in Escherichia coli, can be purified to homogeneity and

S. Wu et al. / Biomaterials 35 (2014) 8496e8502 8497

assembled into mechanically robust films, foams or up to meter-long fibers under non-denaturing and sterile conditions [13,14].The 4RepCT matrices are potentially interesting to apply in tissueregeneration since they support self-renewal and differentiation ofrat neural stem cells, can be provided with bioactive information byfusion to protein domains and/or peptides and are well toleratedwhen implanted [15,16]. In the present study, we aimed to developa biomimetic spider silk cell culture system based on 4RepCT, thatenables hPSC long-term self-renewal and subsequent differentia-tion in a defined xeno-free environment.

2. Materials and methods

2.1. Recombinant spider silk matrices

The recombinantminiature spider silk protein 4RepCTand variants thereof wereproduced in E. coli and purified as described previously [13], including depletion oflipopolysaccharides (LPS) [17].

2.2. hESC cell lines and derivation of human iPS cell lines

hESC lines HS181 [18] and HS360 [19] have been previously described. The hESCcell line H9 [20] was provided by WiCell Research Institute (Madison, Wisconsin).hiPSC lines C5 and C3 were derived using Sendai virus (Invitrogen), a non-integrating method with MOI of 3. Human skin fibroblasts from one female andone male donor were cultured in DMEM supplemented with 10% FBS, 1� Non-essential amino acids (NEAA), and 1� PenicillineStreptavidin (PEST) (all fromGibco). To generate hiPSCs, approximately 100,000 fibroblasts were transduced andone week later re-plated to irradiated human foreskin fibroblasts (hff, ATCC, CRL-2429). ES-like colonies are visible after 10 days, picked manually and transferredonto irradiated human fibroblasts approximately 3 weeks post-transduction. BothhiPSC lines were cultured in standard hES medium with 8 ng/ml FGF2 duringderivation. Mediumwas changed every day and hiPSC colonies were split manuallyevery five to seven days and extra colonies were frozen in STEM-CELLBANKER so-lution (ZENOAQ). hiPSCs were split until passage 10 before usage in this study.

2.3. Cell colony forming assay

Around 30,000 hES cells were manually transferred as aggregates to each cellculture plate coated with film made of different functionalized biomimetic spidersilk variants or matrigel. After 24 and 48 h incubation the number of colonies wascounted under microscope. The fraction of cell aggregates that formed colonies onthe coated surface was determined. The mean value and standard deviation of threeindependent experiments were calculated.

2.4. Long term hPSC cultures

The hESCs and hiPSC lines were cultured on irradiated human foreskin feedercell layer in Knockout DMEM medium supplemented with 20% knockout serumreplacement (KOSR), 1�NEAA,1� Glutamax, 1� PEST, 0.1 mM b-mecaptoethanol (allfrom Gibco), and 10 ng/ml FGF2 (R&D), manually dissociated and transferred onto4RepCT (and variants thereof) films. The cells were then cultured in xeno-freeNutristem medium (Biological Industries) and routinely passaged as aggregatesonce every 4e5 days using a sterile knife.

For passage as single cells, the hESCs and hiPSCs were treated with StemproAccutase or 0.05% Trypsin-EDTA (both from Gibco) for 3e5 min at 37 �C, washed inNutristemmedium, and seeded onto freshly prepared VN-4RepCT film orMatrigel inNutristem medium. 5 mM ROCK inhibitor Y-27632 was added to the medium duringthe first day after single cell plating.

2.5. RGD competing assay

For the RGD competing assay a cyclic RGD peptide, (cyclo-ArgeGlyeAsp-D-PheeLys from Peptides International, Inc. Louisville, KY, USA) was added to the cellculture at a final concentration of 25 mg/ml.

2.6. Cell proliferation assay

For cell proliferation assay, hESCs and hiPSCs cultured on VN-4RepCT film orMatrigel were passaged using Accutase. The cells were briefly washed in D-PBSwithout Ca and Mg, and then treated with Stempro Accutase solution (Gibco) for1e2 min at 37 �C. After the Accutase treatment, warm Nutristemmediumwas added.The cells were gently pipetted 5 times to dissociate the cells, centrifuged at 150� g for3 min, and then resuspended in warm Nutristemmedium. One tenth of the total cellswere seeded onto freshly prepared VN-4RepCT film or Matrigel in Nutristemmedium.To obtain a single cell suspension for counting using an ORFLO Moxi Z automated cellcounter, the rest of cells were treated with Accutase for 5 min at 37 �C.

2.7. Immunofluorescence

hPSCs were fixed with 4% (wt/vol) paraformaldehyde for 20 min at room temper-ature and subsequently washed with PBS for three times. For OCT4 and NANOG im-munostaining, cells were blocked with 5% Normal donkey serum (JacksonImmunoResearch) in PBS containing 0.3% (vol/vol) Triton X-100 for 1 h, and then incu-batedwith primaryantibodies diluted in the blocking buffer overnight at 4 �C. Next day,the cells were incubated with secondary antibodies for 1 h at room temperature. Aftereach incubation step the cells were washed with PBS for three times. For SSEA-4 andTRA-1-81 immunostaining, 5%Normaldonkeyserum inPBSwasusedasblockingbuffer.The primary antibodies used were as follows: rabbit anti-OCT4 (�100, sc-9081, SantaCruz), goat anti-NANOG (�200, NL-1997G, R&D), mouse anti-SSEA-4 (�20) andmouseanti-TRA-1-81 (�10, both kindly provided by professor Peter Andrews, University ofSheffield, UK), goat anti-human SOX17 (�40, AF1924, R&D), rabbit anti-FOXA2 (�1000,ab40874)andmouseanti-NESTIN (�1000, ab22035, both fromabcam), rabbit anti-PAX6(�200,AB2237,Millipore),mouse anti-cardiac TroponinT (cTnT) (�200,ms-295-p1, LabVision), and rabbit anti-NKX2.5 (�100, sc-14033, Santa Cruz).

2.8. Microscopy analysis

Microscopy analysis was performed using a Nikon Eclipse E800 microscope or aZeiss Axiovert 200M inverted confocal microscope. The software for confocal mi-croscopy was Zeiss Zen 2009. The excitation wavelength and emission filter usedwere 405 nm and BP420-480IR for DAPI (blue), 488 nm and BP505-530 for greenfluorescence, and 543 nm and LP615 for red fluorescence, respectively.

2.9. FACS analysis

hPSCs cultured on VN-4RepCT film were dissociated into single cells by treat-ment with 0.05% Trypsin-EDTA for 5min at 37 �C. H9 cells were at passage 20, HS181and C3 cells at passage 22, HS360 cells at passage 21, and C5 at passage 17 on VN-4RepCT film, respectively. The cells were pelleted by centrifugation at 200 g andthen resuspended in FACS buffer (D-PBS without Ca or Mg containing 2% FBS).1�105 cells were incubatedwithmousemonoclonal antibody against SSEA-1, SSEA-4, TRA-1-60, or TRA-1-81 in 50 ml FACS buffer for 30 min at 4 �C. After subsequentwashing with D-PBS without Ca or Mg, the cells were incubated with donkey anti-mouse secondary antibody in 50 ml FACS buffer for 30 min at 4 �C. After two washesin D-PBS without Ca or Mg, the cells were resuspended in 500 ml FACS buffer andanalyzed using a BD FACSCalibur flow cytometer. The antibodies used were mouseanti-SSEA4, mouse anti-TRA-1-60, and mouse anti-TRA-1-81 (�10) and Alexa Fluor488 conjugated donkey anti-mouse IgG antibody (�200, Jackson ImmunoResearch).

2.10. Karyotyping

hPSCs cultured on VN-4RepCT film for 10 passages (H9, C5), 11 passages (HS181),and 13 passages (HS360, C3), respectively, were treated with colcemid KaryoMax(0.025 mg/ml, GIBCO Invitrogen) for 4e5 h or overnight at 37 �C in 5% CO2. The cellswere then treated with 0.05% Trypsin-EDTA for 5 min at 37 �C, pelleted by centri-fugation at 150 g for 10 min, resuspended in 0.05 M KCl hypotonic solution andincubated for 40 min at room temperature. An equal volume of fixative (3:1methanol:acetic acid) was added to the solution. The cells were pelleted bycentrifugation at 214� g, washed once and resuspended in fixative solution. The cellsolutionwas dropped onto glass slides and metaphase chromosomes were analyzedunder microscope. Totally 25 metaphase spreads were analyzed for each cell line atLabmedicin Skåne, Lund university hospital.

2.11. Teratoma formation

hPSCs cultured on VN-4RepCT film were manually dissociated into small ag-gregates. We injected H9 cells after 12 passages, HS181 cells after 14 passages,HS360 cells after 13 passages, C5 cells after 14 passages, and C3 after 17 passages onVN-4RepCT. 5 � 105 cells per cell line were mixed with 5 � 105 irradiated humanforeskin feeder cells in totally 150 ml Nutristemmedium as previously described [21].75 ml Matrigel (BD Biosciences) were added to the cell suspension. Cell suspensionwith only irradiated human foreskin feeder cells was used as control. The wholesuspension volumewas injected subcutaneously into young SCIDmice (4 weeks old,Taconic). Four animals were used per cell line. Teratoma growth was determined bypalpation every week and the mice were sacrified when teratoma size reached 1 cmin diameter. The tumor tissue was excised, weighed and fixated for 24 h in 1%formalin and then transferred to 70% ethanol. The tissue was dehydrated, embeddedin paraffin and processed into 4 mm thick sections. The sections were stained withhematoxylin and eosin (HE) following standard protocols, and mounted undercoverslips.

2.12. In vitro differentiation of hES and iPS cells after long-term culture on VN-4RepCT films

Differentiation towards neuroectoderm was carried out using a previouslydescribed protocol [22]. The hPSC lines H9, HS181, HS360, C3, and C5, cultured forover 35 passages on VN-4RepCT film, were seeded as single cells at a density of18,000 cells/cm2 on freshly prepared VN-4RepCT film, foam or mesh and culturedfor four days. Neuroectoderm differentiation was initiated by treating the cells with

S. Wu et al. / Biomaterials 35 (2014) 8496e85028498

500 ng/ml Noggin (R&D) and 10 mM SB431542 (Tocris) in the standard hES mediumwithout FGF2. However, KOSR was replaced with equal amount of Knockout SRXeno-Free CTS (Gibco) to give a xeno-free culture condition. After four days of dif-ferentiation, SB431542 was withdrawn and increasing amounts (25%, 50% and 75%)of N2mediumwas added every second day, and the cells were induced with 500 ng/ml Noggin for 6 more days.

Differentiation towards definitive endoderm was carried out essentiallyaccording to D’Amour et al. [23]. Briefly, H9 cells, cultured for 65 passages onVN-4RepCT film, were seeded at a density of 50000 cells/cm2 on freshlyprepared VN-4RepCT film and cultured until near confluency. Definitive endo-derm differentiation was initiated by treatment with 100 ng/ml Activin A(R&D) and 25 ng/ml Wnt3a (R&D) in RPMI-1640 medium. The next dayWnt3a was withdrawn and 1xB27 without insulin (Gibco) (replacing 0.2% fetalbovine serum (FBS) in the original protocol) was added to the medium topromote cell survival. The cells were treated with 100 ng/ml Activin A for fourmore days.

Differentiation towards cardiomyocytes was carried out essentially according toLaflamme et al. [24]. Briefly, H9 cells, cultured for 53 passages on VN-4RepCT film,were seeded at a density of 50,000 cells/cm2 on freshly prepared VN-4RepCT foamand cultured for five days. To induce cardiac differentiation, the cells were treatedwith 100 ng/ml Activin A for 24 h, and thenwith 10 ng/ml BMP4 (R&D) for four days,in RPMI-1640 medium supplemented with 1xB27 without insulin. The mediumwas

Fig. 1. Development of spider silk matrices for xeno-free culture of hPSCs. (A) Upper panel:panel: a fiber mesh (to the left), a foam (middle) and fibers in the bottom of a 15 ml test tub4RepCT, RGD-4RepCT, IKVAV-4RepCT, YIGSR-4RepCT, VN-4RepCT and Matrigel. See Fig. S1attachment of the cells. Scale bar corresponds to 500 mm and applies to all figures. (C) Copassaged as aggregates to VN-4RepCT film and Matrigel. The fraction of cell aggregates thatThe data represent the mean value and the standard deviation (error bars) of three indepepassage one tenth of cells were seeded onto VN-4RepCT films or Matrigel. The number of da(E) Appearance of cells after long-term culture (>10 passages) of three hESC (H9, HS360 a4RepCT film could be inhibited by addition of a cyclic RGD peptide (HS360 þ cRGD). Scale

then changed to RPMI-1640 medium supplemented with 1xB27 (Gibco), and thecells were differentiated for three more weeks, with medium change every 2e3days.

3. Results and discussion

3.1. Functionalized spider silk variants for cell culture

In order to optimize the 4RepCT-matrices for hPSC culture, weproduced four customized variants in which known cell bindingpeptides frommatrix proteins were genetically fused to 4RepCT; 1)IKVAV and 2) YIGSR from laminin, 3) RGD found in eg fibronectin,laminin and vitronectin, and 4) PQVTRGDVFTM (herein referred toas VN) from vitronectin (Fig. S1). All five recombinant spider silkvariants (4RepCT, IKVAV-4RepCT, YIGSR-4RepCT, RGD-4RepCT, andVN-4RepCT) could be expressed in E. coli, purified to homogeneityand thereafter processed into 2D films or 3D foams and fibers(Fig. 1A).

Examples of biomimetic spider silk matrices in cell culture wells (96 well plate). Lowere (to the right). Scale bar corresponds to 1 mm. (B) hESC (HS181) adhesion after 24 h tofor explanations of the abbreviations. Only VN-4RepCT and Matrigel allowed efficientlony formation on VN-4RepCT film and Matrigel. 30,000 HS181 cells were manuallyformed colonies on the coated surface one and two days after plating were calculated.ndent experiments. (D) Cell growth curves on VN-4RepCT film and Matrigel. For eachys in culture is shown on the X-axis and the fold increase in cell number on the Y-axis.nd HS181) and two hiPSC lines (C3 and C5) on VN-4RepCT film. Cell adhesion to VN-bar is 500 mm and applies to all figures.

S. Wu et al. / Biomaterials 35 (2014) 8496e8502 8499

3.2. Cell adhesion and proliferation on VN-4RepCT film

The hESC line HS181 was used to compare the cell bindingcapacity of films made from each of the five 4RepCT-based pro-teins, and Matrigel, a non-defined matrix derived from a tumorcell line was used as positive control. After 24 h, cell colonieswere observed on VN-4RepCT and on Matrigel, but not on any ofthe other films (Fig. 1B). The cell colony forming efficiency andproliferation rate on VN-4RepCT was practically identical to thaton Matrigel (Fig. 1C, D). After attachment to VN-4RepCT film, thecell aggregates flattened out and formed a monolayer, indicatingthat they bind to the VN-4RepCT film at least as strongly as theybind to each other. hPSCs express a variety of integrins that an-chors the cells to the surrounding matrix and to other cells [2,25].Attachment of hPSCs via integrins is not a prerequisite for stem

Fig. 2. Long-term self-renewal of hPSCs on VN-4RepCT film. (A) Karyotype analysis of fivenormal karyotype. (B) The expression of pluripotency markers OCT4 (red) and NANOG (greenvisualized by DAPI-staining (blue). The expression of nuclear marker OCT4 and NANOG was epassages. To the right, the expression of cell surface markers SSEA-4 (red), TRA-1-60 (greepassage 20 (H9), 22 (HS181 and C3), 21 (HS360), and 17 (C5), respectively. Scale bar correspothis figure legend, the reader is referred to the web version of this article.)

cell maintenance [26], as attachment may also be mediated bye.g. cell surface glucosaminoglycans [27]. We found that adhesionto VN-4RepCT film is critically dependent on integrin avß3 and/oravß5 since adhesion was blocked by a cyclic RGD peptide (aspecific inhibitor of binding to the above integrins) (Fig. 1E,middle picture of lower panel). Somewhat surprisingly, RGD-4RepCT, containing the integrin binding motif RGD, did notsupport hESC attachment (Fig. 1B), which suggest that the seg-ments flanking RGD in the VN peptide are important for thecorrect presentation of the RGD motif on the matrix surface and/or for proper attachment of hESCs, as has been previously sug-gested [3]. The cyclic RGD peptide did not show any effect on celladhesion to Matrigel, thus ruling out the possibility that its in-hibition of cell adhesion to VN-4RepCT film is due to cytotoxity orother unspecific mechanisms.

hPSC lines cultured on VN-4RepCT film for >10 passages. All cell lines maintained a) in hESCs and hiPSCs after long-term culture on VN-4RepCT film are shown. Nuclei arexamined by immunostaining after 11 (H9), 13 (HS181), 11 (HS360), 16 (C3), and 14 (C5)n) and TRA-1-81 (blue) are shown. The cells were analyzed using flow cytometry innds to 100 mm and applies to all figures. (For interpretation of the references to color in

S. Wu et al. / Biomaterials 35 (2014) 8496e85028500

3.3. Pluripotency of hPSCs after long-term culture on VN-4RepCTfilm

To determine if VN-4RepCT supports long-term propagation ofhPSCs, we cultured three different hESC cell lines (HS181 [18],HS360 [19] and H9 [20]) and two iPSC lines (C5 and C3; cf materialsand methods) on VN-4RepCT film in xeno-free Nutristem medium(Fig. 1E). All five hPSC lines showed the same proliferation rate ason Matrigel, and unaltered morphology for more than 10 passages(Fig. 1D, E), and their karyotypes remained intact after 10 passagesor more (Fig. 2A). Immunofluorescence analysis after >10 passageson VN-4RepCT film revealed that the pluripotency markers OCT4and NANOG were expressed at high levels throughout the colonies(Fig. 2B). Moreover, in all five hPSC lines at least 97% of the cellswere positive for the pluripotency cell surface markers SSEA-4,TRA-1-60 and TRA-1-81 (Fig. 2B) as detected by fluorescence-activated cell sorting (FACS). Remarkably, all hPSC lines could becultured on VN-4RepCT film for over 30 passages without signifi-cant changes in growth rate and/or morphology.

To further investigate if the hPSC lines remained pluripotentafter long-term culture (>12 passages) on VN-4RepCT film weinjected them subcutaneously into severe combined immunodefi-ciency (SCID) mice. H9, HS181, HS360, C3, and C5 cells formedteratoma after 4e8 weeks, and generated cells derived from allthree germ layers (Fig. 3).

3.4. In vitro differentiation on VN-4RepCT matrices after long-termculture

Next, H9, HS181, HS360, C3, and C5 cells that had beenexpanded for over 35 passages on VN-4RepCT film were differ-entiated towards neuroectoderm on 2D VN-4RepCT film, as well ason 3D mesh and foam. The expression of PAX6 and NESTIN wasdetected in all five cell lines differentiated on either mesh or foamand the level of expression equaled that on VN-4RepCT film(Fig. 4).

C5 cells that had been cultured over 20 passages on VN-4RepCTfilm were in vitro differentiated, and the expression of markers

Fig. 3. Pluripotency of hPSCs after long-term culture on VN-4RepCT film. Pluripotency of fiformation. The hPSC lines were injected subcutaneously into SCID mice after >12 passages otissues of the three germ layers. Sections were stained with hematoxylin and eosin (HE). ResS3. Scale bar corresponds to 100 mm and applies to all figures.

(SOX17/FOXA2) confirmed the successful differentiation intoendoderm (Fig. S2). Moreover, H9 cells that had been culturedover 50 passages on VN-4RepCT film were differentiated towardsdefinitive endoderm on 2D VN-4RepCT film and towards car-diomyocytes on 3D foam. The expression of markers for definitiveendoderm (SOX17/FOXA2) and cardiac lineages (NKX2.5/cTnT),respectively, were detected (Fig. S3). In summary, both teratomaformation and in vitro differentiation studies showed that VN-4RepCT film supports xeno-free long-term self-renewal ofpluripotent hESC and hiPSC, and support differentiation towardslineages of the three germ layers under xeno-free definedconditions.

VN-4RepCT foam and mesh are the first described defined 3Dmatrices that support xeno-free differentiation of hPSC. Efficientdifferentiation of hPSCs into functional tissues requires 3D scaf-folds, as they better mimic the cellular microenvironment in vivothan 2D surfaces do. Recently, four defined surfaces based onpeptides or proteins have been shown to support long-termexpansion of both hESCs and hiPSCs in xeno-free media; recom-binant laminin-511 [2] and laminin-521 [7], a synthetic peptidederived from vitronectin (KGGPQVTRGDFTMP) [3,6], and recom-binant C-terminal regions of the a-, b- and g-chains of laminin (E8fragments) [5]. Additionally, long-term xeno-free expansion ofeither hiPSCs or hESCs has been accomplished on recombinantvitronectin [4] and on a peptide derived from bone sialoprotein(KGGNGEPRGDTYRAY) [3], respectively. None of these proteins orpeptides, however, has been shown to formmechanically robust 3Dscaffolds.

Stem cells respond to biophysical cues from the surroundingmatrix and, if the matrix is degraded, to degradation-mediatedcellular traction [28,29]. Optimal scaffolds should therefore givethe cells the proper mechanical support and guidance, and bedegraded at the implant site. Three biomaterials have been previ-ously employed to generate 2D and/or 3D scaffolds for long-termpropagation and differentiation of hPSCs; Collagen type I extrac-ted from animal sources supports both self-renewal and hepatocytedifferentiation of hESCs in 2D and 3D culture systems [30,31]; hy-aluronic acid hydrogel, a fully synthetic 3D matrix, supports short

ve hPSC lines after long-term culture on VN-4RepCT film as determined by teratoman VN-4RepCT. All five cell lines (H9, HS181, HS360, C3 and C5) formed teratomas withults from in vitro endodermal and mesodermal differentiation are shown in Fig. S2 and

Fig. 4. In vitro differentiation of hPSCs on VN-4RepCT matrices in 2D and 3D. H9, HS181, HS360, C3, and C5 cells were induced towards neuroectoderm using Noggin and SB431542for 10 days. The expression of PAX6 (red) and NESTIN (green) was examined by immunostaining and confocal microscopy. Nuclei are shown in blue (DAPI-staining). Dashed linesmark the location of the fibers. Scale bar corresponds to 50 mm and applies to all figures. (For interpretation of the references to color in this figure legend, the reader is referred tothe web version of this article.)

S. Wu et al. / Biomaterials 35 (2014) 8496e8502 8501

term self-renewal and vascular differentiation of encapsulatedhESCs in conditioned media [32]; and chitin-alginate microfibrousscaffolds support self-renewal and neural differentiation of hESCs[33]. However, none of these studies used xeno-free defined cultureconditions.

4. Conclusions

Herein, we present a customized recombinant spider silk matrixthat provides a chemically defined and xeno-free 2D long-termculture system for both hiPSCs and hESCs. Subsequent differentia-tion of the cells under xeno-free conditions on 3D biomimeticspider silk scaffolds was successful. Biomimetic spider silk scaffoldsthus have great potential for use in regenerative medicine.

Acknowledgments

We are grateful to Dr. Thomas Sakmar for valuable input on themanuscript. The cyclic RGD peptide was kindly provided by Dr.Staffan Johansson at Uppsala University. The work was funded bythe Swedish Research Council and Spiber Technologies AB. Photosof biomimetic spider silk matrices in Fig. 1A were provided bySpiber Technologies AB and Dr. Lena Holm.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.biomaterials.2014.06.039.

References

[1] Chen KG, Mallon BS, McKay RD, Robey PG. Human pluripotent stem cellculture: considerations for maintenance, expansion, and therapeutics. CellStem Cell 2014;14:13e26.

[2] Rodin S, Domogatskaya A, Strom S, Hansson EM, Chien KR, Inzunza J, et al.Long-term self-renewal of human pluripotent stem cells on human recom-binant laminin-511. Nat Biotechnol 2010;28:611e5.

[3] Melkoumian Z, Weber JL, Weber DM, Fadeev AG, Zhou Y, Dolley-Sonneville P,et al. Synthetic peptide-acrylate surfaces for long-term self-renewal andcardiomyocyte differentiation of human embryonic stem cells. Nat Biotechnol2010;28:606e10.

[4] Chen G, Gulbranson DR, Hou Z, Bolin JM, Ruotti V, Probasco MD, et al.Chemically defined conditions for human iPSC derivation and culture. NatMethods 2011;8:424e9.

[5] Miyazaki T, Futaki S, Suemori H, Taniguchi Y, Yamada M, Kawasaki M, et al.Laminin E8 fragments support efficient adhesion and expansion of dissociatedhuman pluripotent stem cells. Nat Commun 2012;3:1236.

[6] Jin S, Yao H, Weber JL, Melkoumian ZK, Ye K. A synthetic, xeno-free peptidesurface for expansion and directed differentiation of human induced plurip-otent stem cells. PLoS One 2012;7:e50880.

[7] Rodin S, Antonsson L, Niaudet C, Simonson OE, Salmela E, Hansson EM, et al.Clonal culturing of human embryonic stem cells on laminin-521/E-cadherinmatrix in defined and xeno-free environment. Nat Commun 2014;5:3195.

[8] Radtke C, Allmeling C, Waldmann KH, Reimers K, Thies K, Schenk HC, et al.Spider silk constructs enhance axonal regeneration and remyelination in longnerve defects in sheep. PLoS One 2011;6:e16990.

[9] Simmons AH, Michal CA, Jelinski LW. Molecular orientation and two-component nature of the crystalline fraction of spider dragline silk. Science1996;271:84e7.

[10] Askarieh G, Hedhammar M, Nordling K, Saenz A, Casals C, Rising A, et al. Self-assembly of spider silk proteins is controlled by a pH-sensitive relay. Nature2010;465:236e8.

[11] Hagn F, Eisoldt L, Hardy JG, Vendrely C, Coles M, Scheibel T, et al. A conservedspider silk domain acts as a molecular switch that controls fibre assembly.Nature 2010;465:239e42.

[12] Xia XX, Qian ZG, Ki CS, Park YH, Kaplan DL, Lee SY. Native-sized recombinantspider silk protein produced in metabolically engineered Escherichia coli re-sults in a strong fiber. Proc Natl Acad Sci U S A 2010;107:14059e63.

S. Wu et al. / Biomaterials 35 (2014) 8496e85028502

[13] Widhe M, Bysell H, Nystedt S, Schenning I, Malmsten M, Johansson J, et al. Re-combinant spider silk asmatrices for cell culture. Biomaterials 2010;31:9575e85.

[14] Stark M, Grip S, Rising A, Hedhammar M, Engstrom W, Hjalm G, et al.Macroscopic fibers self-assembled from recombinant miniature spider silkproteins. Biomacromolecules 2007;8:1695e701.

[15] Lewicka M, Hermanson O, Rising AU. Recombinant spider silk matrices forneural stem cell cultures. Biomaterials 2012;33:7712e7.

[16] Fredriksson C, Hedhammar M, Feinstein R, Nordling K, Kratz G, Johansson J,et al. Tissue response to subcutaneously implanted recombinant spider silk:an in vivo study. Materials 2009;2:1908e22.

[17] Hedhammar M, Bramfeldt H, Baris T, Widhe M, Askarieh G, Nordling K, et al.Sterilized recombinant spider silk fibers of low pyrogenicity. Bio-macromolecules 2010;11:953e9.

[18] Hovatta O, Mikkola M, Gertow K, Stromberg AM, Inzunza J, Hreinsson J, et al.A culture system using human foreskin fibroblasts as feeder cells allowsproduction of human embryonic stem cells. Hum Reprod 2003;18:1404e9.

[19] Lappalainen RS, Salomaki M, Yla-Outinen L, Heikkila TJ, Hyttinen JA,Pihlajamaki H, et al. Similarly derived and cultured hESC lines show variationin their developmental potential towards neuronal cells in long-term culture.Regen Med 2010;5:749e62.

[20] Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA, Swiergiel JJ,Marshall VS, et al. Embryonic stem cell lines derived from human blastocysts.Science 1998;282:1145e7.

[21] Gropp M, Shilo V, Vainer G, Gov M, Gil Y, Khaner H, et al. Standardization ofthe teratoma assay for analysis of pluripotency of human ES cells andbiosafety of their differentiated progeny. PLoS One 2012;7:e45532.

[22] Chambers SM, Fasano CA, Papapetrou EP, Tomishima M, Sadelain M, Studer L.Highly efficient neural conversion of human ES and iPS cells by dual inhibitionof SMAD signaling. Nat Biotechnol 2009;27:275e80.

[23] D'Amour KA, Bang AG, Eliazer S, Kelly OG, Agulnick AD, Smart NG, et al.Production of pancreatic hormone-expressing endocrine cells from humanembryonic stem cells. Nat Biotechnol 2006;24:1392e401.

[24] Laflamme MA, Chen KY, Naumova AV, Muskheli V, Fugate JA, Dupras SK, et al.Cardiomyocytes derived from human embryonic stem cells in pro-survivalfactors enhance function of infarcted rat hearts. Nat Biotechnol 2007;25:1015e24.

[25] Meng Y, Eshghi S, Li YJ, Schmidt R, Schaffer DV, Healy KE. Characterization ofintegrin engagement during defined human embryonic stem cell culture.FASEB J 2010;24:1056e65.

[26] Nagaoka M, Si-Tayeb K, Akaike T, Duncan SA. Culture of human pluripotentstem cells using completely defined conditions on a recombinant E-cadherinsubstratum. BMC Dev Biol 2010;10:60.

[27] Klim JR, Li L, Wrighton PJ, Piekarczyk MS, Kiessling LL. A definedglycosaminoglycan-binding substratum for human pluripotent stem cells. NatMethods 2010;7:989e94.

[28] Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix elasticity directs stem celllineage specification. Cell 2006;126:677e89.

[29] Khetan S, Guvendiren M, Legant WR, Cohen DM, Chen CS, Burdick JA.Degradation-mediated cellular traction directs stem cell fate in covalentlycrosslinked three-dimensional hydrogels. Nat Mater 2013;12:458e65.

[30] Furue MK, Na J, Jackson JP, Okamoto T, Jones M, Baker D, et al. Heparin pro-motes the growth of human embryonic stem cells in a defined serum-freemedium. Proc Natl Acad Sci U S A 2008;105:13409e14.

[31] Baharvand H, Hashemi SM, Kazemi Ashtiani S, Farrokhi A. Differentiation ofhuman embryonic stem cells into hepatocytes in 2D and 3D culture systemsin vitro. Int J Dev Biol 2006;50:645e52.

[32] Gerecht S, Burdick JA, Ferreira LS, Townsend SA, Langer R, Vunjak-Novakovic G. Hyaluronic acid hydrogel for controlled self-renewal and dif-ferentiation of human embryonic stem cells. Proc Natl Acad Sci U S A2007;104:11298e303.

[33] Lu HF, Narayanan K, Lim SX, Gao S, Leong MF, Wan AC. A 3D microfibrousscaffold for long-term human pluripotent stem cell self-renewal underchemically defined conditions. Biomaterials 2012;33:2419e30.