Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te / sc r

Stem Cell Research (2012) 9, 167–170

REGULAR ARTICLE

Suspended in culture — Human pluripotent cells forscalable technologiesCarmel O'Brien⁎, Andrew L. Laslett

CSIRO Materials Science and Engineering, Stem Cells, Clayton, Victoria, AustraliaDepartment of Anatomy and Developmental Biology, Monash University, Clayton, Victoria, Australia

Abstract Human embryonic stem cells (hESCs) and human induced pluripotent stem cells (hiPSCs), collectivelytermed human pluripotent stem cells (hPSCs), are typically derived and maintained in adherent and semi-defined cultureconditions. Recently a number of groups, including Chen et al., 2012, have demonstrated that hESCs can now beexpanded efficiently and maintain pluripotency over long-term passaging as aggregates in a serum-free definedsuspension culture system, permitting the preparation of scalable cGMP derived hPSC cultures for cell banking, highthroughput research programs and clinical applications. In this short commentary we describe the utility and potential futureuses of suspension culture systems for hPSCs.

© 2012 Published by Elsevier B.V.

Introduction

Human pluripotent stem cells are notable for their remarkablecapacity to self-renew indefinitely in vitro and to give rise totissues comprising the three embryonic germ layers in theformation of teratomas in vivo. hPSCs are an important cellsource for the study of human disease processes, screening forcandidate therapeutic agents and potentially for regenerativecell-based therapies and tissue engineering applications. Therealisation of this potential still requires improved in vitrodifferentiation assays for the efficient generation of targetcell types as well as effective strategies for purification of thesefrom non-target somatic cells and residual hPSCs. Developmentof defined culture systems that enable scalable long term ex-pansion of hPSCs, is an important first step towards the repro-ducible seeding of robust differentiation technologies.

The current convention is to derive, culture and bank hPSCsfrom static two dimensional culture systems in a dish or flask

1873-5061/$ - see front matter © 2012 Published by Elsevier B.V.doi:10.1016/j.scr.2012.06.001

⁎ Corresponding author at: CSIRO Materials Science and Engineer-ing, Stem Cells, Bag 10, Clayton South MDC, Victoria, 3169 Australia.

E-mail addresses: carmel.obrien@csiro.au (C. O'Brien),andrew.laslett@csiro.au (A.L. Laslett).

with the use of supporting embryonic fibroblast feeder cells orsemi-defined extracellular matrix components. Apart frominadequately resembling the in vivo developmental environ-ment, these systems do not lend themselves to pre-clinical orfuture clinical application due to their inherent variability,time consuming passaging and handling requirements and lowyields. The culture of single hPSCs in a suspension system islikely the simplest and most robust means by which to gen-erate cell products suitable for large scale drug and therapyapplications. Since the first report of preformed embryoidbody (EB) culture of hESCs in stirred-suspension bioreactors(SSBs) in 2004 (Gerecht-Nir et al., 2004) there have been anumber of reports detailing modifications and improvements.These groups evaluated scalable aggregate, microcarrier andencapsulation SSB systems in the quest to improve outcomesfor hESC viability, expansion yield, homogeneity, differentia-tion capacity and applicability to industry processes (reviewedin Serra et al., 2012; Kehoe et al., 2010; Krawetz et al., 2010).SSBs that enable the expansion and self-renewal of hPSCs inclinically compatible conditions and relevant quantities willrequire single cell seeding of hPSCs in working volumes upwardsof a few hundred millilitres, the formation of uniform suspen-sion aggregates and ease of continuous passaging under definedculture conditions.

168 C. O'Brien, A.L. Laslett

Strategies to dissociate hPSC cultures for single cell seedingof suspension bioreactors have in the past been hampered byan increased loss of cell viability and self-renewal. The de-monstration that exposure of hESCs to the Rho-associatedkinase (ROCK) inhibitor Y-27632 can greatly diminish theapoptosis associated with single cell dissociation methods andimprove self-renewal (Watanabe et al., 2007, Harb et al., 2008)has improved outcomes for continual single cell passaging andmaintenance of hPSC cultures using commercially availableenzyme mixes and in both serum and serum-free conditions(Bajpai et al., 2008).

The ability to culture hPSCs in suspension systems, oftenas embryoid bodies, for the purpose of producing differen-tiated cell types is well documented (Kurosawa, 2007). Morerecently, there have been several reports demonstrating somesuccess with the maintenance, expansion and serial passagingof undifferentiated hPSCs in stirred and static suspensionculture systems and with the use of ROCK inhibitors (Singh etal., 2010; Amit et al., 2011; Krawetz et al., 2010; Kehoe et al.,2010; Olmer et al., 2010, Steiner et al., 2010; Zweigerdt et al.,2011). These studies have reported very promising outcomesfor the successful expansion and maintenance for a limitednumber of passages of hESCs, hiPSCs and a nonhuman primateESC line in both static low attachment plates and stirredsuspension cultures in small scale petri dish, Erlenmeyer andspinner flask formats. hPSC expansion rates in these studieshave been variable however, with respect to initial seedingdensities and overall cell yields were not always comparable tothose achieved with adherent cell cultures (reviewed by Serraet al., 2012).

Long term maintenance of hPSCs in a scalableand defined suspension culture system

A report published recently in Stem Cell Research (Chen etal., 2012, 8:388–402) extends from these previous studies todemonstrate for the first time, the long term expansion ofhESCs retaining the characteristic traits of pluripotency forover 20 passages without loss of viability or growth rate in ascalable GMP compliant spinner flask culture system that ismatrix, serum-free and reagent defined. A calculated cumu-lative fold expansion over 1×1013 was achieved following 21passages of the H9 hESC line. Chen et al. demonstrated that atthis passage these cells were still undifferentiated by flowcytometric analysis of pluripotency markers, and karyotypi-cally normal.

Chen et al. went on to describe the development of astirred spinner flask culture system for the continuous singlecell passaging and aggregate expansion of 3 hESC lines (HES-2,H1 and H9). This system is equivalent or superior to adherentculture systems in terms of growth rate, expansion capacity,viability, maintenance of the pluripotent phenotype and finalyield. Suspension culture conditions were optimized by com-paring commercially available serum-free defined hPSCmediaand determining the best seeding density and cell passagingintervals. In this production system, prior to initiating sus-pension culture, the hESC cultures are first adapted fromconventional adherent cultures on MEF feeders to feeder-freeadherent culture on a defined xeno-free matrix (CELLstart,Invitrogen) and in a serum-free defined medium (StemPro hESCSFM, Invitrogen), andmaintained for more than 3 passages. The

three cell lines adapted to this defined culture system grew asmonolayer cultures of cells that were shown to express highlevels of the pluripotency markers Tra-1-60, Tra-181, SSEA-4and Oct-4 by flow cytometry analysis andwere reported to formteratomas containing lineages representative of the threeembryonic germ layers in immunocompromised mice.

A number of experiments were initially undertaken byChen et al. in small scale orbital rotation low attachmentculture wells to determine conditions for transit directlyfrom the CELLstart adherent hESC cultures to an optimiseddefined suspension culture system. Consistent with previousreports from other groups, the adherent hESC cultures weredemonstrated to form viable cell aggregates following pre-treatment with the Rho kinase inhibitor Y-27632 and singlecell dissociation using the commercially available enzyme mixAccutase (Millipore). Similarly, continued exposure to the ROCKinhibitor appeared to be a key requirement for the furtherexpansion of aggregates formed in suspension. Aggregates weretypically of 100–120 μm size at day 2 and 150–200 μm at day 3,retained characteristic pluripotency markers by confocal fluo-rescence analysis and displayed compact hPSC colony morphol-ogy and strong alkaline phosphatase staining when replatedback to MEF adherent culture conditions after 4 days in thedefined medium suspension culture. From three media condi-tions and two seeding densities trialled, the culture of adaptedhESC cells seeded at 2.5×105 cells/ml in StemPro hESC SFMsupplemented with 40 ng/ml human recombinant basic fibro-blast growth factor (bFGF) and 0.1 mM β-mercaptoethanol andin the presence of 10 μM Y-27632 was determined as beingoptimal for the generation of aggregates with respect tonumbers formed, smaller size and homogeneity (typicallyb200 μm), high cell viability, retained expression of markersfor pluripotency by flow cytometry (N90% at day 3) and overallcell yields. The authors did note some variations in results forthe above parameters for the different cell lines evaluated inthis study. While there is no direct comparison given for theparent MEF adherent cultures for adaptation to either the smallscale or spinner flask suspension cultures described, the widelydocumented variation in both self-renewal and differentiationoutcomes generally seen between different hPSC lines inculture would suggest variable amenity to this defined cultureadaptation for other hESC and hiPSC cell lines.

As cell aggregates increase in size in culture, so too willthe difficulty for nutrients and cytokines required for themaintenance of self-renewal to penetrate the 3D cultures,leading to decreased expansion of hPSCs and increasing dif-ferentiation. For adapted hESCs seeded in the optimisedmediaconditions to scaled culture in 125 ml spinner flasks withmagnetic stirrers (Thermo Scientific) set at a pre-determined70 rpm, Chen et al. demonstrated by flow cytometric analyses asignificant decline in the expression of pluripotency markersTra-1-60, Tra-1-81 and Oct-4 after 4 days of suspension culture.Similarly, growth rates were calculated as starting to decreaseafter day 4 although, cell viability remained high at ~90% at day6 in the same cultures.

Apart from the elimination of feeder and matrices, scal-ability is the key advantage offered by suspension cultureformats. Adaptation of scalable defined suspension systems tobioreactor formats where process parameters can be auto-matically controlled will enable the commercial scale gener-ation of cGMP compliant cells in the folds that will be arequirement for pre-clinical and clinical applications. In the

169Suspended in culture — Human pluripotent cells for scalable technologies

study by Chen et al. an optimal hPSC expansion rate and long-termmaintenance of pluripotency were demonstrated for thedefined media suspension system scaled to 125 ml or 500 mlspinner flask cultures with continual single cell passagingevery 3–4 days using Accutase and ROCK inhibitor. Following 9passages for the three hESC lines and 21 passages (64 days) fora further study of one line (H9, 64 days), all three hESC linesconsistently displayed a 3–4 fold average expansion perpassage (~1×106 cells/ml each 3–4 days) under these condi-tions, retained a normal karyotype, and consistently displayedover 90% expression of pluripotency markers by flow cytom-etry analyses. hESC cultures had maintained consistently highviabilities of N90% prior to each passage event in this system,and an overall expansion of 1.5×1013 fold was calculatedfollowing 21 passages. Compared with MEF adherent culturesmaintained over a 3–4 week period, the suspension culturesand adherent CELLstart cultures for three hESC lines displayedhigher expansion rates and cell viability. Population doublingrates were similar for suspension and adherent cultures,indicating that normal growth characteristics were retained.

All cell lines demonstrated contribution to tissues represen-tative of the 3 germ layers in teratoma assays at the end ofsuspension culture. The H9 line was also examined for in vitrodifferentiation capacity in EBs formed from day 3 spinner flaskcultures with the addition of serum. RT-PCR analyses of EBsplated to matrigel demonstrated undetectable Oct4 and Nanogexpression by day 6 and the onset of T, Sox17, Sox1 and Tubb33expression during 16 days of adherent culture. The directeddifferentiation of H9 suspension cultures in a continuous serum-free spinner flask EB culture to a cardiomyocyte fate was alsoachieved with a reasonable efficiency, as confirmed by flowcytometry, immunohistochemistry and gene expression analy-ses for a range of cardiac specific markers. The authors' objec-tive in this manuscript is to demonstrate a scalable and definedhPSC maintenance suspension culture system, however, avaluable extension would be to compare the directed differen-tiation outcome with parent MEF cultures and in particular toevaluate the differentiation outcomes for additional hPSC linesgrown to scale in the described suspension culture system. Thatthe suspension system described by Chen et al. is scalable aswell as compliant with both cGLP and cGMP manufacturing andcell banking requirements, is an important outcome of thisstudy. The authors also demonstrated the successful cryopres-ervation of hESC expanded in this system using a serum-freedefined cyroprotective medium with the inclusion of ROCKinhibitor, and it is noteworthy that post-thaw cultures demon-strate at least 85% viability and a high capacity to self-renewboth directly in suspension and conventional adherent cultureconditions. To demonstrate the utility of the described hESCcell production system as a whole, three hESC lines were ex-panded from 125 ml to 500 ml scale spinner flasks as previouslydescribed and 200–400 cryovials each with 4×106 cells/vialwere banked under cGLP or cGMP conditions for all cell lines.

Conclusions and future directions

The studies reported by Chen et al. demonstrate a cell pro-duction system for the reagent-defined adaptation and longterm expansion of hPSC suspension cultures in scalablespinner flasks for over 20 passages, with no loss of viability orgrowth rate. The described GMP compliant suspension system

is equivalent or superior to conventional adherent MEF ormatrix systems with respect to maintenance of pluripotency,growth characteristics, expansion capability and overall yield,and the hPSCs generated are amenable to cryopreservation inreagent-defined conditions. As cell densities equivalent toadherent culture for long term hPSC suspension passaging arenow demonstrated in a spinner flask system, it is reasonable topropose that this system when introduced to a more sophis-ticated bioreactor format with controllable processes will beable to achieve even greater cell yields.

An important application for developments to date forthe expansion of human PSCs in reagent defined and feeder/matrix-free suspension culture systems will be the transfer ofcell reprogramming technologies from static culture in dishes toderive, maintain and expand hiPS cells in suspension. Elimina-tion of TGFβ cytokine signalling from both fetal bovine serumand feeders will improve the reprogramming environment withrespect to the mesenchymal-to-epithelial transition that isrequired for the initiation of reprogramming to pluripotency(Li et al., 2010). Very recently, two groups independentlydemonstrated success with the derivation of iPSCs fromsomatic mouse cells in suspension culture conditions withoutthe supporting feeder or matrix substratum (Fluri et al., 2012Shafa et al., 2012). Fluri et al. further demonstrated thedirected differentiation of resulting suspension iPSCs in acontinuous suspension culture to contractile cardiac cells. Itwill be of much interest going forward to see if a suspensionculture approach can be achieved for the reprogramming ofhuman cells, which if additional challenges with generating aclonal hiPSC line can be overcome, would facilitate scale-up ofpatient-matched hiPSCs for seeding differentiation assaysgeared to clinical trial applications.

At the time of writing, there are new reports describingimprovements to the mechanical strategies and culture envi-ronment tomaximise the formation, maintenance and yield ofhPSC aggregates while minimising shear stress and loss ofviability (for e.g. Bhatia et al., 2012; Abbasalizadeh et al.,2012). Other important requirements for translation of basicresearch to industry/clinical scale systems are the develop-ment of robust and sensitive characterisation tools usinggenomic, proteomic, metabolomic and/or functional assaysthat enable accurate monitoring of cell types in bioreactorformats. Ultimately, automated cell production systems thatcircumvent the need for repeated dissociation and reforma-tion of aggregates will greatly facilitate future applications.No doubt there are further developments underway as thisexciting field of work advances towards clinical stem cellproduction.

References

Abbasalizadeh, S., Larijani, M.R., Samadian, A., Baharvand, H., 2012.Bioprocess development for mass production of size-controlledhuman pluripotent stem cell aggregates in stirred suspensionbioreactor. Tissue Eng. C Methods. http://dx.doi.org/10.1089/ten.TEC.2012.0161.

Amit, M., Laevsky, I., Miropolsky, Y., Shariki, K., Peri, M., Itskovitz-Eldor, J., 2011. Dynamic suspension culture for scalableexpansion of undifferentiated human pluripotent stem cells.Nat. Protoc. 6 (5), 572–579.

Bajpai, R., Lesperance, J., Kim, M., Terskikh, A.V., 2008. Efficientpropagation of single cells accutase-dissociated human embryonic

170 C. O'Brien, A.L. Laslett

stem cells. Mol. Reprod. Dev. 75 (5), 818–827. http://dx.doi.org/10.1002/mrd.20809.

Bhatia, H., Sharma, R., Dawes, J., Khillan, J.S., 2012. Maintenanceof feeder free anchorage independent cultures of ES and iPS cells byretinol/vitamin A. J. Cell. Biochem. http://dx.doi.org/10.1002/jcb.24177

Chen, V.C., Couture, S.M., Ye, J., Lin, Z., Hua, G., Huang, H.I.P.,Wu, J., Hsu, D., Carpenter, M.K., Couture, L.A., 2012. ScalableGMP compliant suspension culture system for human ES cells. StemCell Res. 8 (3), 388–402.

Fluri, D.A., Tonge, P.D., Song, H., Baptista, R.P., Shakiba, N.,Shukla, S., Clarke, G., Nagy, A., Zandstra, P.W., 2012.Derivation, expansion and differentiation of induced pluripotentstem cells in continuous suspension cultures. Nat. Methods 9 (5),509–516 (http://www.nature.com/nmeth/journal/v9/n5/abs/nmeth.1939.html#supplementary-information).

Gerecht-Nir, S., Cohen, S., Itskovitz-Eldor, J., 2004. Bioreactorcultivation enhances the efficiency of human embryoid body (hEB)formation and differentiation. Biotechnol. Bioeng. 86 (5),493–502. http://dx.doi.org/10.1002/bit.20045.

Harb, N., Archer, T.K., Sato, N., 2008. The Rho–Rock–Myosinsignaling axis determines cell–cell integrity of self-renewingpluripotent stem cells. PLoS One 3 (8), e3001.

Kehoe, D.E., Jing, D., Lock, L.T., Tzanakakis, E.S., 2010. Scalablestirred-suspension bioreactor culture of human pluripotent stemcells. Tissue Eng. A 16 (2), 405–421. http://dx.doi.org/10.1089/ten.TEA.2009.0454.

Krawetz, R., Taiani, J.T., Liu, S., Meng, G., Li, X., Kallos, M.S.,Rancourt, D.E., 2010. Large-scale expansion of pluripotent humanembryonic stem cells in stirred-suspension bioreactors. TissueEng. C Methods 16 (4), 573–582. http://dx.doi.org/10.1089/ten.TEC.2009.0228.

Kurosawa, H., 2007. Methods for inducing embryoid body forma-tion: in vitro differentiation system of embryonic stem cells.J. Biosci. Bioeng. 103 (5), 389–398. http://dx.doi.org/10.1263/jbb.103.389 (doi: S1389-1723(07)70078-6 [pii]).

Li, R., Liang, J., Ni, S., Zhou, T., Qing, X., Li, H., He, W., Chen, J.,Li, F., Zhuang, Q., Qin, B., Xu, J., Li, W., Yang, J., Gan, Y., Qin,

D., Feng, S., Song, H., Yang, D., Zhang, B., Zeng, L., Lai, L.,Esteban, M.A., Pei, D., 2010. A mesenchymal-to-epithelial transi-tion initiates and is required for the nuclear reprogramming ofmouse fibroblasts. Cell Stem Cell 7 (1), 51–63.

Olmer, R., Haase, A., Merkert, S., Cui, W., Paleček, J., Ran, C.,Kirschning, A., Scheper, T., Glage, S., Miller, K., Curnow, E.C.,Hayes, E.S., Martin, U., 2010. Long term expansion of undiffer-entiated human iPS and ES cells in suspension culture using adefined medium. Stem Cell Res. 5 (1), 51–64.

Serra, M., Brito, C., Correia, C., Alves, P.M., 2012. Process engineeringof human pluripotent stem cells for clinical application. TrendsBiotechnol. 30 (6), 350–359. http://dx.doi.org/10.1016/j.tibtech.2012.03.003 (doi: S0167-7799(12)00035-2 [pii]).

Shafa, M., Day, B., Yamashita, A., Meng, G., Liu, S., Krawetz, R.,Rancourt, D.E., 2012. Derivation of iPSCs in stirred suspensionbioreactors. Nat. Methods 9 (5), 465–466 (http://www.nature.com/nmeth/journal/v9/n5/abs/nmeth.1973.html#supplementary-information).

Singh, H., Mok, P., Balakrishnan, T., Rahmat, S.N., Zweigerdt, R.,2010. Up-scaling single cell-inoculated suspension culture ofhuman embryonic stem cells. Stem Cell Res 4 (3), 165–179.http://dx.doi.org/10.1016/j.scr.2010.03.001 (S1873-5061(10)00017-6 [pii]).

Steiner, D., Khaner, H., Cohen, M., Even-Ram, S., Gil, Y., Itsykson,P., Turestsky, T., Idelson, M., Aizenman, E., Ram, R., Berman-Zaken, Y., Reubinoff, B., 2010. Derivation, propagation andcontrolled differentiation of human embryonic stem cells insuspension. Nat. Biotechnol. 28 (4), 361–364. http://dx.doi.org/10.1038/nbt.1616.

Watanabe, K., Ueno, M., Kamiya, D., Nishiyama, A., Matsumura, M.,Wataya, T., Takahashi, J.B., Nishikawa, S., Muguruma, K., Sasai,Y., 2007. A ROCK inhibitor permits survival of dissociatedhuman embryonic stem cells. Nat. Biotechnol. 25 (6), 681–686(http://www.nature.com/nbt/journal/v25/n6/suppinfo/nbt1310_S1.html).

Zweigerdt, R., Olmer, R., Singh, H., Haverich, A., Martin, U., 2011.Scalable expansion of human pluripotent stem cells in suspensionculture. Nat. Protoc. 6 (5), 689–700.