Determine the quality of human embryonic stem colonies with laser light scattering patterns.

1.4 MB·English

by  Chen, Chi-Shuo

Chenetal.BiologicalProceduresOnline2013,15:2
Biological Procedures
http://www.biologicalproceduresonline.com/content/15/1/2
Online
RESEARCH Open Access
Determine the quality of human embryonic stem
colonies with laser light scattering patterns
Chi-Shuo Chen1, Matthew Biasca2, Catherine Le3, Eric Y-T Chen1, E Daniel Hirleman2 and Wei-Chun Chin1*
Abstract
Background: With the prompt developments of regenerative medicine, the potential clinical applications ofhuman
embryonic stemcells have attracted intense attention. However, the labor-intensive and complexmanualcell
selection processes required during embryonic stemcell culturing have seriously limited large-scale production and
broad applications. Thus, availability ofa label-free, non-invasive platform to replace the current cumbersome
manual selection has become a critical need.
Results: A non-invasive, label-free, and time-efficient optical platformfor determining the quality of human
embryonic stemcell colonies was developed by analyzing the scattering signals from thosestem cell colonies.
Additionally, confocal microscopyrevealed that thecell colony morphologyand surface structureswere correlated
withtheresulting characteristic light scattering patterns.Standardimmunostaining assay (Oct-4) was also utilizedto
validate the quality-determinationfrom this light scattering protocol. The platform developed here can therefore
provide identification accuracy ofup to 87% for colony determination.
Conclusions: Our study here demonstratedthat light scattering patternscan serve as a feasiblealternative
approach to replace conventional manual selection for human embryonic stem cell cultures.
Keywords: Light-scattering, Human embryonicstem cell, Pluripotency, Label-free detection
Background shown to offer partial Parkinson’s disease recovery in
Therehavebeenincreasinginterestsintheapplicationsof animal models [6,7]. Moreover, functional islet-like cells
human embryonic stem cells (hESCs). These hESCs derived from ESCs have been demonstrated to respond to
present almost unlimited applications and opportunities glucose and produce insulin, which hold promising poten-
for future advances in biotechnology and regenerative tial for treatments of diabetes [8]. In addition to regene-
medicine. Through the research of regenerative medicine, rating tissue replacements, human stem cells and/or
specificfunctionalcelltypeshavebeendifferentiatedfrom differentiated cells can serve as promising platforms for
hESCs for stem cell-based therapies [1,2]. For instance, drug discovery and toxicity testing within the pharma-
advanced protocols have been developed to enrich the ceuticalindustry.Thesehumancell-basedplatformsnot
functional cardiomyocytes differentiated from hESCs for only provide human pathology models, but are also im-
the treatment of irreversible cardiac tissue damage [3]. In portantplatformsforevaluatingnewdrugcompoundsin
addition, ESCs-derived cardiomyocytes were implanted human physiological environments. For example, diffe-
into adult dystrophic mice which formed intracardiac rentiated hepatocytes have been used for drug metabo-
graftsafter7-weeks[4].Stemcellscanfurthermorebeused lism studies in preclinical drug discovery; additionally,
toprovidefunctionalneuronalcellsforpotentialtreatments cardiomyocytes differentiated from hESCs have been
ofneurodegenerativedisorderssuchasAlzheimer’sdisease, utilized for cardiac drug discovery and cardiac safety
Huntington’s’ disease, and spinal cord injury [1,5]. Specifi- assessments [9]. Likewise, undifferentiated hESCs can
cally, dopaminergic neurons differentiated from ESCs have provideamodelforembryotoxicitytesting[10].
To fully develop these promising industrial applica-
*Correspondence:[email protected] tions, one of the most crucial issue is the maintenance
1Bioengineering,SchoolofEngineering,UniversityofCalifornia,Merced,CA, and expansion of the self-renewing undifferentiated
USA
hESCs that are able to retain the capacity to differentiate
Fulllistofauthorinformationisavailableattheendofthearticle
©2013Chenetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative
CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedtheoriginalworkisproperlycited.
Chenetal.BiologicalProceduresOnline2013,15:2 Page2of6
http://www.biologicalproceduresonline.com/content/15/1/2
into desired cell types [10]. In order to maintain developed to analyze the cell colonies (STEMvision,
the pluripotency of hESCs in cultures, manual microdis- STEMCELL technologies, Vancouver, Canada). With
section is broadly used for cell passaging, which requires methylcellulose-labeled assay, images of hematopoietic
laborious protocols and quality-controlled manual selec- coloniescanbeacquiredandscoredautomatically.How-
tion [9]. Typically, according to the protocols established ever, excessive labeling assays and the requirement of
by National Stem Cell Bank (NSCB) and biotechnology specific cultureware may hinder the large-scale hESC
companies [11,12], undifferentiated hESC cell colonies technological developments and have an unknown influ-
needtobemanuallyselectedandtransferredtonewplates enceonhESCdifferentiationcapacities.
every 7~10 days. The quality of undifferentiated hESC In order to create a label-free, non-invasive plat-
colonies is determined with bright-field/phase contrast form to evaluate the quality of stem cell colonies, we
light microscopy based on human-experience/assessment. introduced optical forward-scattering technology in
Variousrating scales/criteriahavebeenestablishedmainly this study (Figure 1). Previously, scattering technology
based on the morphology of individual colonies [13]. has been applied to assess micron spheres on silicon
Chiefly, good quality (good) hESCs, without stacking nor wafers [16]; it can also be applied to identify collected
undesired differentiations, grow as uniform flat colonies microbial contaminants from different food sources
with clear colony edges; low quality (bad) hESC colonies [17]. Various characteristics of bacterial colonies, such
show various shapes with visible surface structures and as colony matrices or shapes of colonies, govern the
irregular edges. The quality of hESCs plays a critical role interaction between incident light and the colonies
in the downstream applications. Consequently, determin- resulting in different scattering patterns (signatures)
ingtheundifferentiatedhESCcellcoloniesbythismanual [18]. However, the feasibility of rapid cell selection
selection processserves as an essential step in subculture; with scattering patterns has not yet been developed
only high-quality undifferentiated colonies should be for human stem cells. In this study, one of the most
transferred; conversely, transferring poor-quality colonies used stem cell lines, H9 (WiCell), was selected as the
results in the lost of pluripotency [13]. Nonetheless, this model line [19]. We demonstrated the feasibility of
time-consumingandlabor–intensivemanualselectioncan optical forward-scattering in determining the quality
hinder the development of large-scale culture for prac- of hESC cell colonies. Furthermore, standard immu-
tical/clinical applications. Moreover, lack of standard cri- nohistochemistry assay (Oct-4) was performed on
teria for rapid determination of stem cell qualities may hESC colonies to confirm the results from the newly
lead to the quality control/assurance issues of industrial- developed light scattering identification protocol.
scalereproducibilityfortherapeuticapplications[14].
A critical requirement for hESC research and techno-
logical developments is the efficient assessment of cell dif-
ferentiationstatus.Toaddressthisissue,severalapproaches
have been developed [13,14]. For instance, laser flow
cytometers havebeen broadlyused to provide comprehen-
siveinformationofstemcell differentiationwiththeuseof
fluorescentcellsurfaceantigensorproteinmarkerssuchas
Oct-4 and Nanog, or SSEA-3 in hESCs [13]. To further
simplifyspecimenpreparationandlowertheimmunohisto-
chemistry costs, the microfluidic dielectrophoresis (DEP)
platform have been introduced. Based on different cellular
dielectric properties, without cell-type specific markers,
neurons and astrocytes differentiated from mouse neural
precursorcellscanbedetectedandisolatedinmicrofluidic
channels(500μminwidthand50μminheight)[15].How-
ever,thesecurrenttechnologiesmaynotbeappropriatefor
hESCcolonyqualityevaluation,sincehESCcolonieswould
be squeezed within sheath flow and broken into undesired
small pieces while they are transported in the cytometry
Figure1Sketchofopticalsystem.Inthisopticalsystem,laser
pipe line or microfluidic channels. As a result, the disper- light(wavelength=633nm)wasguidedtoincidentonthehESCs
sionofcell colonies mayfurther complicatethe subculture colonyperpendicularly.Forwardscatteringpatternofthecolonywas
process. collectedwithCMOSsensorinthebackofculturedish.Duringthe
Tomaintainthe colonies’integrityduringtheselection scanningstep,specimensweretrackedandtransported
automaticallybymotorstage.
process, advanced image analysis systems have been
Chenetal.BiologicalProceduresOnline2013,15:2 Page3of6
http://www.biologicalproceduresonline.com/content/15/1/2
Results and discussion with different qualities. Optical-sliced images of DAPI
Inordertoestablishthecomputer-basedrecognitiondata- stained cells were collected and reconstructed to reveal
base to identify good/bad colonies, hESC colonies were detailed 3-dimensional (3-D) hESC colonies. Results of
first manually categorized into good/badsets.Representa- 3-D z-stack measurements showed the average height
tive colonies images were presented with phase contrast was 53±4.2 μm of good colonies and was 44±13.9 μm
microscopy (Figure 2a and b). Scattering patterns of 290 of bad colonies. Student t-test was applied to compare
pre-categorized colonies were collected into the database the different heights and indicated the significant height
to establish good/bad colony criteria. Representative difference (p-value<0.01) between good/bad colonies.
forward scattering patterns from good/bad colonies are The larger variations of height measurements also indi-
displayed in Figure 2c and d. Compared to the scattering cate the non-uniform stacking microstructures within
patternsof badhESCcolonies,thelight scatteringpattern bad colonies, rather than the homogenous spatial distri-
of good hESC colonies displayed a more symmetrical bution of cells within good colonies. Our 3-D colony
intensity distribution. In contrast, irregular or patchy sca- image analysis provided consistent quantified assess-
ttering patterns were observed in most of the bad hESC ments that were supported by the outcomes from the
colonies. human experience-basedevaluation.
Conventionally, one of the important criterion for As in the diffraction patterns created when light propa-
evaluating the quality of hESCs has been the determin- gatesthroughopticalapertures,theamplitudesandphases
ation of colony morphology by manual evaluation with of light are modulated when light passes through the
light microscopy. Since colony morphology will likewise biological specimens [18]. In previous studies [17,20], it
influence the scattering patterns of hESC colonies, we has been shown that the central thickness and radius of
aim to utilize this connection to determine colony qua- bacterial colonies may dominate the scattering pattern
lity from scattering patterns. In order to evaluate these formations. Served as the superposition of various aper-
key attributes, we used laser scanning confocal micros- tures,theobservednon-evenspatial colonyvariations can
copy to investigate the morphology of hESCs colonies leadtothenon-uniformscatteringpatterns.
Figure2Forward-scatteringpatternsofhESCscolonies.Asinglecolonywasilluminatedwithalaserbeamandthescatteringpatternswere
projectedonthedetectors.Thelightintensityofscatteringpatternscreatedbygoodcolonies(a)showedmorehomogeneousdistribution
(c)comparedtothosefrombadcolonies(b,scatteringpatternsind).
Chenetal.BiologicalProceduresOnline2013,15:2 Page4of6
http://www.biologicalproceduresonline.com/content/15/1/2
Inordertoestablishthe colonycharacteristicsdatabase, subsequentlyassessedwithimageanalysissoftware(ImageJ,
advanced images analyses were performed to further NIH). The fluorescence of DAPI staining (DNA stain) was
analyzescatteringpatterns. Eachscatteringimage features used to quantify the cell number within a single colony.
were extracted utilizing 2-D Zernike moment invariants Representative fluorescent images of Oct-4 and DAPI are
[21,22]. In general, the center of the image was set as the shown in Figure 3. Here, we used the Oct-4 to DAPI
origin and pixels were mapped into a specific coordinate, expressionratiotoindicatethepluripotencyofcellcolonies
thentheimageboundaryandcharacteristiccontentswere (Figure 4). Our results showed a higher Oct-4/DAPI
quantifiedwithasetofcomplexZernikemomentpolyno- expression ratio which was observed in classified-good
mials [22]. The 161 feature vectors were selected with colonies,ratherthanclassified-badones(Oct-4/DAPIratios
Fisher’s criterion to represent the image characteristics in were significantly different, p-value<0.001). Noticeably,
ordertooptimize classification outcomes[17].The classi- due to the heterogeneous cell populations within the
fication was performed with Support Vector Machine colony, certain portions of hESCs within classified-bad
(SVM)-basedalgorithm;apatternrecognitionmethodhas colonies expressed Oct-4 that can reduce the difference of
been widely used in handwritten recognition, objective Oct-4/DAPI expression level between good/bad classes.
recognition, and face detection [17,23,24]. In order to in- These small expression differences not only demonstrate
vestigate the precision of our computerized identification the complexity in evaluating the heterogeneous colonies
protocolforgood/badcolonyidentification,additional100 qualitywithimmunostaining,butalsohighlighttheunique
manually-classified colonies were assessed with the BAR- advantage of light scattering assay having high sensiti-
DOT (Bacterial Rapid Detection using Optical scattering vity. Moreover, it has been shown that extracellular matrix
Technology) system under an optimized system setting, properties can alter, resulting in a change to a cell’s
which was the same setting for database training. Scatte- organization during the progression of cell differentiation
ring patterns were analyzed with the same setting para- [27,28]; accordingly, the refractive index and the micro-
metersforthesetestingsamples.AsshowninTable1,87% structure of colonies can contribute to the variations of
of manually-classified-good colonies could be determined scattering patterns [17,29]. The accuracy of this scattering
into good-colony classification, and 83% of manually- system could be further improved by taking these para-
classified-bad ones could be resolved. Factors causing false metersintoconsideration.
determination may come from either the manual selection
processorthecurrentsystemset-up.Asdiscussed,manual
selection is based on researcher’s experience and this indi- Conclusions
vidual human bias may lead to certain false determination. In this study, we demonstrate the feasibility in determin-
Largerimagevariationcausedbytheheterogeneouscolony ing the quality of hESC cell colonies with optical
shapes, can also lead to the false classification outcome. In forward-scattering technology. Integrating with SVM
spite of this, an improved image analysis algorithm may machine learning, this technology provides a critical
helptodecreasethefalsedeterminationrateaswell. module for automatic hESC cell colony selection. With-
In addition to testing the accuracy of BARDOT deter- out any biochemistry labeling processes or manual labor,
mination with manually classified specimens, standard the determination results from our protocol showed
immunological staining was also used to quantify the highcorrelationwiththeresultsofstandardimmunohis-
classification outcomes. Oct-4 is an important transcrip- tochemistry assay. In addition, this new label-free light
tion factor during cellular development that serves to scattering process significantly reduces the cost and time
regulate the pluripotency and self-renewal of embryonic for specimen preparation. Though only two fundamental
stem cells. The decrease of Oct-4 expression levels has categories of the hESC database were assessed here, our
beenwidelyusedasanindicatorforhESCdifferentiation results demonstrated that the light scattering patterns
(loss of pluripotency) [25,26]. To assess pluripotency of could provide unique signature standards for hESC
the classified stem cell colonies, colonies were labeled categorization. Additionally, this non-invasive optical
with anti-Oct-4 (stem cell pluripotency biomarker) and protocol demonstrates its potential capacity for applica-
then imaged with epi-fluorescent microscopy (N=18, tions on other various stem cell colony types with
from each classification). The Oct-4 expressions were specific differentiation lineages, such as iPS (induced
pluripotent stem) cells, neuronal stem cells, cells grown
Table1Distinguishrate(%)ofgood/badcolonies on Matrigel, etc. Due to the unique simplicity of this
non-destructive technology, the hESC cell colonies
BARDOTdetermined GoodhESCs BadhESCs
colonies colonies scoring protocol developed in this study is expected to
Humancategorized
provide a general calibrator to facilitate industrial scale
GoodhESCscolonies 87.3% 12.7%
productivity of consistent quality stem cells for applica-
BadhESCscolonies 16.4% 83.6%
tionsofregenerativemedicine.
Chenetal.BiologicalProceduresOnline2013,15:2 Page5of6
http://www.biologicalproceduresonline.com/content/15/1/2
Figure3FluorescentimagesofhESCscolonies.ImmunohistologicalstainingassaywasappliedtodeterminethequalitiesofhESCscolonies.
Anti-Oct-4antibodywasusedtolabeltheself-renewingpluripotentstemcellsincolonies.ResultsindicatedtheexistenceofmoreOct-4marked
cellsincategorizedgoodhESCscolonies(a)versusthoseofbadhESCscolonies(b).
Methods Laserforward-scatteringsystemandimageanalysis
Humanstemcellculture In this study, an automated BARDOTsystem was modi-
(Madison, WI, passage 32–60) were cultured following fiedforhESCcolonyscanning.Thissystemwascomposed
the NSCB protocols. Briefly, undifferentiated hESC of a laser diode (0.95 mW, 632 nm), monochromatic
weremaintainedonfeedercell layers,whichweremito- CMOSimagesensor,andanx-yscanningstage.Thelaser
mycin C treated mouse embryonic fibroblasts (MEF). beam was illuminated on a single colony with computer-
Serum-free medium was used in this experiment and aided positioning. Theresulting forward-scatteringsignals
the medium was composed of DMEM-F-12, Knockout werecollectedwithaCMOSimagesensor.Toanalyzethe
Serum Replacement, basic fibroblast growth factor, features from colony scattering patterns, Pseudo-Zernike
L-glutamine, and MEM non-essential amino acid [30]. moments (PZMs) were applied in our current models.
Medium was changed daily and hESCs were manually Support vector machine (SVM)-based algorithm was
selected to pass every 7 days. All hESC colonies were employed for recognition and classification of the images
analyzed on day 7 to acquire forward light scattering [17,29]. In order to establish the colony characteristic
patterns in order to eliminate the variations from cul- database,290colonieswerescannedwithBARDOT.Their
tureconditions. scatteringpatternswerethencollectedandanalyzed.
Immunohistologicalstaining
In order to quantify pluripotency, after culturing in
medium for 7 days, hESCs were fixed with 4% wt para-
formaldehyde for 20 min at room temperature. Parafor-
maldehyde was removed with 3 subsequent PBS rinses.
Triton-X(0.1%,Sigma)wasusedtopenetratecellmem-
branes. The sample was then incubated in 2% BSA for
40 min to block non-specific binding. Primary antibody
anti-Oct-4 (Millipore, 1:100) was used to identify pluri-
potent stem cells within colonies [13]. Alexafluor-488
conjugated antibody was used for secondary antibody
0
staining. All samples were stained with 4, 6-diamdino-
2-phenylindole (DAPI, Invitrogen, 1:3000) for quantify-
ingcellnumbers.
FluorescentimagingofhESCcolonies
Figure4ExpressionlevelsofOct-4/DAPIinhESCscolonies.
Coloniescategorizedintogood/badcoloniesshoweddifferent ColonyimagesforOct-4quantitiesanalysiswerecollected
Oct-4/DAPIexpressionratios.Theexpressionwasstatistically with fluorescent microscopy. Fluorescence of Alexafluor-
differentingood/badclassifications(p-value<0.001).Higher 488wasexcitedat488nmandtheemissionwascollected
Oct-4/DAPIexpressionratiosindicatedahigherpercentageof
between 500 nm and 560 nm. DAPI staining signals were
pluripotentstemcellsintheclassified-goodcolonies.
collectedtoidentifythecellnumbers.Inthisstudy,images
Chenetal.BiologicalProceduresOnline2013,15:2 Page6of6
http://www.biologicalproceduresonline.com/content/15/1/2
ofcolonieswerecategorizedintotwoseparatedgroupsby 11. WiCell[http://www.wicell.org/]
BARDOTcollection (N=18 in each classification). Laser 12. Viacyte[http://www.viacyte.com/]
13. LoringJF,RaoMS:Establishingstandardsforthecharacterizationof
scanning confocal microscopy was utilized to investigate humanembryonicstemcelllines.STEMCELLS2006,24(1):145–150.
thesurfacemorphologyandcolonyhomogeneityofgood/ 14. TersteggeS,LaufenbergI,PochertJ,SchenkS,Itskovitz-EldorJ,EndlE,
badcolonies. BrustleO:Automatedmaintenanceofembryonicstemcellcultures.
BiotechnolBioeng2007,96(1):195–201.
15. FlanaganLA,LuJ,WangL,MarchenkoSA,JeonNL,LeeAP,MonukiES:
Abbreviations
UniqueDielectricPropertiesDistinguishStemCellsandTheir
BARDOT:BacterialRapidDetectionusingOpticalscatteringTechnology;
BSA:Bovineserumalbumin;DAPI:40,6-diamdino-2-phenylindole; DifferentiatedProgeny.STEMCELLS2008,26(3):656–665.
16. WeberDC,HirlemanED:Light-ScatteringSignaturesofIndividualSpheres
hESCs:Humanembryonicstemcells;MEF:Mouseembryonicfibroblasts;
onOpticallySmoothConductingSurfaces.ApplOptics1988,
PBS:Phosphatebufferedsaline;PZMs:Pseudo-Zernikemoments;
27(19):4019–4026.
SVM:Supportvectormachine.
17. BaeE,BaiN,AroonnualA,BhuniaAK,HirlemanED:Label-free
identificationofbacterialmicrocoloniesviaelasticscattering.Biotechnol
Competinginterest
Bioeng2011,108(3):637–644.
Drs.Chen,ChinandHirlemanarelistedasco-inventorsofapatent
18. BaeE,BaiN,AroonnualA,RobinsonJP,BhuniaAK,HirlemanED:Modeling
applicationfiledbyUCMercedbasedonthisstudy.
lightpropagationthroughbacterialcoloniesanditscorrelationwith
forwardscatteringpatterns.JBiomedOpt2010,15(4):.
Authors’contributions
19. ScottCT,McCormickJB,Owen-SmithJ:Andthenthereweretwo:useof
CSC,WCC,EYTCandEDHconceivedandcoordinatedthestudyanddrafted hESClines.NatBiotechnol2009,27(8):696–697.
themanuscript.MBandCLperformedsomeofthestemcellcultureand
20. BanadaPP,GuoS,BayraktarB,BaeE,RajwaB,RobinsonJP,HirlemanED,
lightscatteringexperiments.CSCperformedmostoftheresearchand
BhuniaAK:Opticalforward-scatteringfordetectionofListeria
analyzedthedata.Allauthorsreadandapprovedthefinalversionofthis
monocytogenesandotherListeriaspecies.BiosensBioelectron2007,
manuscript. 22(8):1664–1671.
21. KhotanzadA,HongYH:INVARIANTIMAGERECOGNITIONBYZERNIKE
Acknowledgements MOMENTS.IEEETransPatternAnalMachIntell1990,12(5):489–497.
ThisstudywassupportedinpartbyagrantfromtheMuscularDystrophy
22. BayraktarB,BanadaPP,HirlemanED,BhuniaAK,RobinsonJP,RajwaB:
Association(MDA).CSCwassupportedbytheUCMercedGRCsummer
Featureextractionfromlight-scatterpatternsofListeriacoloniesfor
fellowshipandtheJaneVilasFellowship.EYCwasinpartsupportedbyNIH
identificationandclassification.JBiomedOpt2006,11(3):.
(1R15HL095039).
23. RajwaB,VenkatapathiM,RaghebK,BanadaPP,HirlemanED,LaryT,
RobinsonJP:Automatedclassificationofbacterialparticlesinflowby
Authordetails
multianglescattermeasurementandsupportvectormachineclassifier.
1Bioengineering,SchoolofEngineering,UniversityofCalifornia,Merced,CA, CytometryPartA2008,73A(4):369–379.
USA.2SchoolofEngineering,UniversityofCalifornia,Merced,CA,USA.
24. BurgesCJC:AtutorialonSupportVectorMachinesforpattern
3SchoolofNaturalSciences,UniversityofCalifornia,Merced,CA,USA. recognition.DataMiningandKnowledgeDiscovery1998,2(2):121–167.
25. ReubinoffBE,PeraMF,FongCY,TrounsonA,BongsoA:Embryonicstem
Received:23August2012Accepted:5January2013
celllinesfromhumanblastocysts:somaticdifferentiationinvitro.Nat
Published:14January2013 Biotechnol2000,18(4):399–404.
26. LohYH,WuQ,ChewJL,VegaVB,ZhangWW,ChenX,BourqueG,GeorgeJ,
References
LeongB,LiuJ,etal:TheOct4andNanogtranscriptionnetworkregulates
1. VolarevicV,LjujicB,StojkovicP,LukicA,ArsenijevicN,StojkovicM:Human
pluripotencyinmouseembryonicstemcells.NatGenet2006,
stemcellresearchandregenerativemedicine—presentandfuture. 38(4):431–440.
BritishMedicalBulletin2011,99(1):155–168.
27. HalvorsenYDC,FranklinD,BondAL,HittDC,AuchterC,BoskeyAL,
2. WobusAM,BohelerKR:EmbryonicStemCells:Prospectsfor
PaschalisEP,WilkisonWO,GimbleJM:Extracellularmatrixmineralization
DevelopmentalBiologyandCellTherapy.PhysiolRev2005,85(2):635–678.
andosteoblastgeneexpressionbyhumanadiposetissue-derived
3. XuC,PoliceS,RaoN,CarpenterMK:CharacterizationandEnrichmentof stromalcells.TissueEng2001,7(6):729–741.
CardiomyocytesDerivedFromHumanEmbryonicStemCells.CircRes
28. JaiswalN,HaynesworthSE,CaplanAI,BruderSP:Osteogenicdifferentiation
2002,91(6):501–508.
ofpurified,culture-expandedhumanmesenchymalstemcellsinvitro.
4. KlugMG,SoonpaaMH,KohGY,FieldLJ:Geneticallyselected JCellBiochem1997,64(2):295–312.
cardiomyocytesfromdifferentiatingembronicstemcellsformstable
29. BaeEW,AroonnualA,BhuniaAK,RobinsonJP,HirlemanED:System
intracardiacgrafts.JClinInvest1996,98(1):216–224.
automationforabacterialcolonydetectionandidentification
5. KellerG:Embryonicstemcelldifferentiation:emergenceofanewerain
instrumentviaforwardscattering.MeasSciTechnol2009,20(1).
biologyandmedicine.GenesDev2005,19(10):1129–1155.
30. ChenC-S,SoniS,LeC,BiascaM,FarrE,ChenE,ChinW-C:Humanstem
6. BjörklundLM,Sánchez-PernauteR,ChungS,AnderssonT,ChenIYC,
cellneuronaldifferentiationonsilk-carbonnanotubecomposite.
McNaughtKSP,BrownellA-L,JenkinsBG,WahlestedtC,KimK-S,etal:
NanoscaleResearchLetters2012,7(1):126.
Embryonicstemcellsdevelopintofunctionaldopaminergicneurons
aftertransplantationinaParkinsonratmodel.ProcNatlAcadSci2002,
99(4):2344–2349. doi:10.1186/1480-9222-15-2
Citethisarticleas:Chenetal.:Determinethequalityofhuman
7. KimJ-H,AuerbachJM,Rodriguez-GomezJA,VelascoI,GavinD,LumelskyN,
embryonicstemcolonieswithlaserlightscatteringpatterns.Biological
LeeS-H,NguyenJ,Sanchez-PernauteR,BankiewiczK,etal:Dopamine ProceduresOnline201315:2.
neuronsderivedfromembryonicstemcellsfunctioninananimalmodel
ofParkinson'sdisease.Nature2002,418(6893):50–56.
8. JiangJ,AuM,LuK,EshpeterA,KorbuttG,FiskG,MajumdarAS:Generation
ofInsulin-ProducingIslet-LikeClustersfromHumanEmbryonicStem
Cells.STEMCELLS2007,25(8):1940–1953.
9. SartipyP,BjörquistP,StrehlR,HyllnerJ:Theapplicationofhuman
embryonicstemcelltechnologiestodrugdiscovery.DrugDiscovery
Today2007,12(17–18):688–699.
10. AméenC,StrehlR,BjörquistP,LindahlA,HyllnerJ,SartipyP:Human
embryonicstemcells:Currenttechnologiesandemergingindustrial
applications.CriticalReviewsinOncology/Hematology2008,65(1):54–80.