Embed Size (px)
Citation preview
432
CLONING AND STEM CELLSVolume 9, Number 3, 2007© Mary Ann Liebert, Inc.DOI: 10.1089/clo.2007.0033
Patient-Specific Stem Cell Lines Derived from HumanParthenogenetic Blastocysts
E.S. REVAZOVA,1 N.A. TUROVETS,1,2 O.D. KOCHETKOVA,1 L.B. KINDAROVA,2L.N. KUZMICHEV,2 J.D. JANUS,1 and M.V. PRYZHKOVA1
ABSTRACT
Parthenogenetic activation of human oocytes may be one way to produce histocompatiblecells for cell-based therapy. We report the successful derivation of six pluripotent human em-bryonic stem cell (hESC) lines from blastocysts of parthenogenetic origin. The parthenogenetichuman embryonic stem cells (phESC) demonstrate typical hESC morphology, express appro-priate markers, and possess high levels of alkaline phosphatase and telomerase activity. ThephESC lines have a normal 46, XX karyotype, except one cell line, and have been culturedfrom between 21 to 35 passages. The phESC lines form embryoid bodies in suspension cul-ture and teratomas after injection to immunodeficient animals and give differentiated deriv-atives of all three embryonic germ layers. DNA profiling of all six phESC lines demonstratesthat they are MHC matched with the oocyte donors. The study of imprinted genes demon-strated further evidence of the parthenogenetic origin of the phESC lines. Our research hasresulted in a protocol for the production of human parthenogenetic embryos and the deriva-tion of stem cell lines from them, which minimizes the presence of animal-derived compo-nents, making the derived phESC lines more suitable for potential clinical use.
INTRODUCTION
HUMAN EMBRYONIC STEM CELLS have the poten-tial to give significant therapeutic benefit to
patients, provided that the problem of immunerejection can be solved (Conley et al., 2004; Reubi-noff et al., 2000; Thomson et al., 1998). Embryonicstem cells that are genetically related to the re-cipient may overcome such rejection problems. Inthis regard, parthenogenetic stem cells may be analternative to embryonic stem cells derived by so-matic cell nuclear transfer (SCNT) technology forwomen of reproductive age.
Parthenogenetic activation of oocytes is a rela-tively simple method to create histocompatible
stem cells in comparison to SCNT, because it doesnot require the complex equipment necessary tomicromanipulate an oocyte (Marshall et al., 1998).Parthenogenetic stem cells are produced from un-fertilized oocytes and contain genetic material ex-clusively from the oocyte donor (the potential pa-tient). It is proposed that following directed celldifferentiation, autologous cells could be trans-planted without the threat of immune rejection.Before now, all attempts to produce humanparthenogenetic embryonic stem cell lines havefailed (Cibelli et al., 2001; Lin et al., 2003).Parthenogenetic mouse major histocompatibilitycomplex (MHC)-homozygous stem cell lines andone parthenogenetic primate heterozygous em-
1Lifeline Cell Technology, Walkersville, Maryland (a subsidiary of International Stem Cell Corporation, Oceanside,California).
2Scientific Center for Obstetrics, Gynecology, and Perinatology RAMS, Moscow, Russia.
PATIENT-SPECIFIC STEM CELL LINES 433
bryonic stem cell line (Cyno-1) have already beenderived, and cell pluripotency has been demon-strated in these lines (Lin et al., 2003; Vrana et al.,2003).
Possible clinical use of hESC lines will requirethat their derivation and culture be performed us-ing conditions that are free from components ofanimal origin. Attempts have been made in thisdirection (Kim et al., 2005); unfortunately, satis-factory methods for the derivation and culture ofhESCs were not developed.
In the present report we describe the deriva-tion of genetically MHC-matched pluripotent phESC lines from human parthenogenetic blas-tocysts using a protocol with minimal animal-derived components.
METHODS
Donor consent
Donors voluntarily donated oocytes, blood (forDNA analysis) with no financial payment.Donors signed comprehensive informed consentdocuments and were informed that all donatedmaterials were to be used for research and not forreproductive purposes, namely the developmentof methods to derive human embryonic stem (ES)cells and their differentiated progeny.
In our research we used only shared oocyte do-nation. The quantity of oocytes used in our ex-periments did not affect donor pregnancy rates.Four donors out of five became pregnant. Onedonor was not successful in achieving embryoimplantation, even though her in vitro fertiliza-tion (IVF) cycle utilized 11 mature oocytes.
Before ovarian stimulation, oocyte donors un-derwent medical examination for suitability ac-cording to Food and Drug Administration (FDA)eligibility determination guidelines for donors ofhuman cells, tissues, and cellular and tissue-based products (FDA, 2004) and Order N 67(02.26.2003) of the Russian Public Health Min-istry. It included chest X-ray, blood and urineanalysis, and liver function test. Donors were alsoscreened for chlamydia trachomatis, neisseriagonorrhoeae, syphilis, HIV, hepatitis B virus(HBV), and hepatitis C virus (HCV).
Donor superovulation
Each donor underwent ovarian stimulation byFSH (Gonal-F, Lab. Serono, Switzerland) from the
3rd to the 13th day of their menstrual cycle. A to-tal of 1500 IU were given. From the 10th to the14th day of the donor’s menstrual cycle go-nadoliberin antagonist Orgalutran (Organon,Holland) was injected at 0.25 mg/day. From the12th to the 14th day of the donor’s menstrual cy-cle a daily injection of 75 IU FSH � 75 IU LH(Menopur, Ferring GmbH, Germany) was given.If an ultrasound examination displayed folliculesbetween 18 and 20 mm in diameter, a single 8000IU dose of hCG (Choragon, Ferring GmbH, Ger-many) was administered on the 14th day of thedonor’s menstrual cycle. Transvaginal punctionwas performed 35 h after hCG injection on ap-proximately the 16th day. Follicular fluid was col-lected from the antral follicles of anesthetizeddonors by ultrasound-guided needle aspirationinto sterile test tubes.
Oocyte activation and culture of parthenogenetic embryos
Cumulus oocyte complexes (COCs) werepicked from the follicular fluid, washed inFlushing Medium (MediCult) and then incu-bated in Universal IVF medium (MediCult) withLiquid Paraffin (MediCult, Jyllinge, Denmark)overlay for 2 h in a 20% O2, 5% CO2, 37°C hu-midified atmosphere. Before activation, COCswere treated with SynVitro Hyadase (MediCult)to remove cumulus cells, followed by incubationin Universal IVF medium with Paraffin overlayfor 30 min. Further culture of oocytes and em-bryos was performed in a humidified atmo-sphere at 37°C with an O2 reduced gas mixture(90% N2 � 5% O2 � 5% CO2) with the exceptionof the ionomycin treatment, which was per-formed at conditions described for culture ofCOCs. Activation was performed in UniversalIVF medium with Paraffin overlay by consecu-tive exposure of oocytes to 5 �M calcium iono-mycin (Sigma, St. Louis, MO) for 5 min and 1mM 6-DMAP (Sigma) for 4 h with careful wash-ing of oocytes from each reagent in UniversalIVF medium. Oocytes were then placed in freshIVF medium with Paraffin overlay followingculture. The next day (day 1) further cultivationof the parthenogenetic embryos was performedusing sequential BlastAssist System media(MediCult) according to the manufacturer’s rec-ommendations. On days 5 through 6 of culture,derived blastocysts were used for the isolationof inner cell mass (ICM).
Isolation of ICM of blastocysts and culture ofparthenogenetic hESCs
The zona pellucida was removed by 0.5%pronase (Sigma) treatment. For immunosurgery,the blastocyst was incubated with horse anti-serum to human spleen cells followed by expo-sure to guinea pig complement. Trophectodermcells were removed from the ICM by gently pipet-ting the treated blastocyst. For the derivation ofphESC from the whole blastocysts, the blastocystswere placed on a feeder layer of mitomycin C mi-totically inactivated human neonatal skin fibro-blasts (NSF) in medium designed for the cultureof phESC. When a blastocyst attached and tro-phoblast cells spread, the ICM became visible.Through 3 to 4 days of additional culture, theICM was isolated through mechanical slicing ofthe ICM from the trophectoderm outgrowth us-ing a finely drawn glass pipette. Further cultureof isolated ICMs was performed on mitomycin Cmitotically inactivated human NSF derived un-der informed consent from a genetically unre-lated donor. NSF were derived using mediumcontaining human umbilical cord blood serum in-stead of animal serum. Before medium prepara-tion, serum was screened for syphilis, HIV, HBV,and HCV.
The medium for the culture of NSF consistedof 90% DMEM (high glucose, with L-glutamine)(Invitrogen, San Diego, CA), 10% human umbil-ical cord blood serum, and penicillin–strepto-mycin (100 U/100 �g) (Invitrogen).
For the culture of ICM and phESC we used Vit-roHES (Vitrolife, Kangsbacka, Sweden) supple-mented with 4 ng/ml hrbFGF (Chemicon,Temecula, CA), 5 ng/mL hrLIF (Chemicon), and10% human umbilical cord blood serum. The ICMwas mechanically plated on a fresh feeder layerand cultured for 3 to 4 days. The first colony wasmechanically cut and replated after 5 days of cul-ture. All subsequent passages were made after 5to 6 days in culture. For early passages, colonieswere mechanically divided into clumps and re-plated. Further passing of phESC was performedwith collagenase IV treatment and mechanicaldissociation. The propagation of phESC was per-formed at 37°C, 5% CO2 in a humidified atmo-sphere.
The phESC characterization
For immunostaining, phESC colonies werefixed with 4% paraformaldehyde during 20 min
at room temperature for staining of SSEA-1, 3, 4markers, or with 100% methanol for 5 min at �20°Cfor other markers. Monoclonal antibodies usedwere: SSEA-1 (MAB4301), SSEA-3 (MAB4303),SSEA-4 (MAB4304), OCT-4 (MAB4305), TRA-1-60(MAB4360), and TRA-1-81 (MAB4381) fromChemicon. Secondary antibodies Alexa Fluor 546(orange-fluorescent) and 488 (green-fluorescent)were from Molecular Probes (Invitrogen). Nucleiwere stained with DAPI (Sigma). Alkaline phos-phatase and telomerase activity were detectedwith an AP kit and a TRAPEZE Kit (Chemicon).Karyotype was determined by a standard G-banding method.
Embryoid body formation and further differentiation
The phESC colonies were mechanically di-vided into clumps and placed in the wells of 24-well cluster plates precoated with 2% agarose(Sigma) in medium containing 85% KnockoutDMEM, 15% human umbilical cord blood serum,1�MEM NEAA, 1 mM Glutamax, 0.055 mM �-mercaptoethanol, penicillin–streptomycin (50U/50 �g) (all from Invitrogen, except serum).Embryoid bodies were cultured for 14 days insuspension and then placed onto a dish to de-velop outgrowths, or cultivated in suspension foran additional week.
Neural differentiation was induced by the cul-tivation of 2-week-old embryoid bodies attachedto a culture dish surface over the period of a weekin differentiation medium: DMEM/F12, B27, 2mM Glutamax, penicillin–streptomycin (100U/100 �g) and 20 ng/mL hrbFGF (all from In-vitrogen). Some embryoid bodies gave rise to dif-ferentiated cells with neural morphology; otherswere dissected and additionally cultured to pro-duce neurospheres.
Beating embryoid bodies appeared sponta-neously following 5 days of culture after platingon an adhesive surface in the same medium aswas used for embryoid body generation.
Immunocytochemistry of differentiated derivativesof phESC
Embryoid bodies, neurospheres, or contractileembryoid bodies were placed on poly-D-lisyne(Sigma) treated microcover glasses (VWR Scien-tific Inc., West Chester, PA) and cultured for ap-proximately 1 week in the appropriate differen-
REVAZOVA ET AL.434
tiation medium. For immunostaining, differenti-ated cells were fixed with 100% methanol for 5min at �20°C.
For the detection of ectodermal markers, weused monoclonal mouse antineurofilament 68 an-tibody (Sigma), anti-human CD56 (NCAM) anti-body (Chemicon), and anti-beta III tubulin anti-body (Chemicon) to highlight neuronal markers.We used antiglial fibrillary acidic protein (GFAP)antibody (Chemicon) to detect the glial cellmarker.
For the detection of the mesodermal markersin 3-week-old embryoid bodies or in contractileembryoid bodies, we used monoclonal mouse an-tidesmin antibody (Chemicon), antihuman alpha-actinin antibody (Chemicon) as the muscle spe-cific markers, and antihuman CD31/PECAM-1antibody (R&D Systems, Minneapolis, MN), anti-human VE-Cadherin (CD144) antibody (R&DSystems) as the endothelial markers.
For the detection of the endodermal markers inembryoid bodies, we used monoclonal mouse an-tihuman alpha-fetoprotein antibody (R&D Sys-tems). Secondary antibodies Alexa Fluor 546 (orange-fluorescent) and 488 (green-fluorescent)were from Molecular Probes (Invitrogen). Nucleiwere stained with DAPI (Sigma).
HLA typing
Genomic DNA was extracted from blood, cu-mulus cells, phESC, and NSF with DynabeadsDNA Direct Blood from Dynal (Invitrogen). HLAtyping was performed by PCR with allele-specificsequencing primers (PCR-SSP, Protrans, Indi-anapolis, IN). All manipulations were performedaccordingly manufacturer’s recommendations.
Affymetrix SNP microarray analysis
Genomic DNA was isolated from blood, cumu-lus cells, phESC, and NSF by phenol/chlorophormextraction method. The DNA samples obtainedfrom four Caucasian subjects were genotypedwith Affymetrix Mapping 50 K Hind 240 Array(part of Affymetrix GeneChip Mapping 100 Kkit). Initially, the dataset contained 57,244 binarySNP markers. Since the number of markers ismore than would be necessary to identify theequivalency of genomic samples and to studyheterozygosity, 15 (chromosomes 1–15) out of 22autosomal chromosomes were chosen. Theshorter seven chromosomes were removed to re-duce the chance that no marker, or only a single
marker for a given chromosome, is selected dur-ing random sampling. The 1459 markers testedwere analyzed with Relcheck (version 0.67, copy-right (©) 2000 Karl W. Broman, Johns HopkinsUniversity, Licensed under the GNU GeneralPublic License version 2, June, 1991) (Boehnkeand Cox, 1997; Broman and Weber, 1998). TheRelcheck program identifies five types of rela-tionships; monozygous (MZ) twins, parent/off-spring pair, full siblings (full sibs), half sibs, andunrelated. Full sibs share approximately 50% oftheir genome identical by descent. In the resultsfrom paired samples, the marker data were con-sistent with this proportion. An “unrelated” pairshares less of their genome than “half-sibs.”
Analysis of imprinted genes
Total RNA was extracted as described (Chom-cznski and Sacchi, 1987) and precipitated withisopropanol. Residual genomic DNA was re-moved using an RNAse free DNAse treatment(Promega, Madison, WI). cDNA was synthesizedfrom 1 �g total RNA using RevertAid M-MuLVreverse transcriptase (Fermentas) in 20 �L of thereaction volume. The PCR reactions were per-formed with 1 �L cDNA, using Taq DNA poly-merase (Fermentas, Hanover, MD). All reactionswere performed according to the manufacturer’sinstructions.
The sequence of the primers and PCR condi-tions were as follows: TSSC5 (Lee et al., 1998) for-ward primer 5�-GCTCTTCATGGTCATGTTCT-CCA-3� and reverse primer 5�-GGAGCAGT-GGTTGTACAGAGG-3�, at conditions of 94°C for4 min for one cycle; 94°C for 1 min, 55°C for 1min, and 72°C for 1 min for 33 cycles. The prod-uct size was 364 bp. H19 (Hashimoto et al., 1995)forward primer 5�-TACAACCACTGCACTAC-CTG-3� and reverse primer 5�-TGGCCATGAA-GATGGAGTCG-3�, at conditions of 94°C for 4min for one cycle; 94°C for 1 min, 52°C for 1 min,and 72°C for 1 min for 38 cycles. The product sizewas 148 bp. PEG1_1 (Li et al., 2002) forwardprimer 5�-GAG TCC TGT AGG CAA GGT CTTACC T-3� and reverse primer 5�-CTT GCC TGAAGA CTT CCA TGA GTG A-3�, at conditions of94°C for 4 min for one cycle; 94°C for 1 min, 55°Cfor 1 min, and 72°C for 1 min for 35 cycles. Theproduct size was 155 bp. PEG1_2 (Li et al., 2002)forward primer 5�-GCT GCT GGC CAG CTCTGC ACG GCT G-3� and reverse primer 5�-CTTGCC TGA AGA CTT CCA TGA GTG A-3�, at con-
PATIENT-SPECIFIC STEM CELL LINES 435
ditions of 94°C for 4 min for one cycle; 94°C for1 min, 65°C for 1 min, and 72°C for 1 min for 39cycles. The product size was 230 bp. SNRPN(Glenn et al., 1993) forward primer 5�-CTTAGC-TGAGACACCAAGAGG-3� and reverse primer 5�-GCAGCATCTTGCTACTCTTGC-3�, at condi-tions of 94°C for 4 min for 1 cycle; 94°C for 1 min,55°C for 1 min, and 72°C for 1 min for 33 cycles.The product size was 246 bp. GAPDH (Adjaye etal., 1999) forward primer 5�-ACCACAGTCCAT-GCCATCAC-3� and reverse primer 5�-TCCAC-CACCCTGTTGCTGTA-3�, at conditions of 94°Cfor 4 min for 1 cycle; 94°C for 1 min, 55°C for 1min, and 72°C for 1 min for 21 cycles. The prod-uct size was 450 bp.
The PCR products were analyzed by 5% poly-acrylamide gel electrophoresis (5 �L/line), stainedwith ethidium bromide, and documented using theBioImaging system (UVP, Upland, CA). RT-PCRexperiments were performed repeatedly with re-producible results. GAPDH served as a ubiqui-tously expressed control. Genomic contaminationwas ruled out by including an RT-negative sample(without reverse transcriptase, at the reverse tran-scription step) in each PCR set as a control.
Teratoma formation and immunohistochemical analysis
All animal procedures were carried out by Biocon (Rockville, MD) in accordance with Ani-mal Welfare Act; the Guide for the Care and Useof Laboratory Animals; and the Principles for theUse of Experimental Animals. Biocon is accred-ited by the Association for the Assessment andAccreditation of Laboratory Animal Care Inter-national (AAALAC) since 1981, #000529, and hasan approved Animal Welfare Assurance State-ment on file with the Office of Laboratory Ani-mal Welfare (OLAW), #A3267 01, and is regis-tered with the U.S. Department of Agriculture(USDA), #51 R 0032.
For injection to immunodeficient severe com-bined immunodeficient (SCID) beige mice andHSDRHCOr rats (Harlan) the phESC were en-zymatically removed from the culture dishes us-ing collagenase type IV and resuspended intoclumps. Approximately 2 to 5 million phESCcells were injected into the upper hind limb sub-cutaneous space. After approximately 2 monthsof growth, formed teratomas were removed andfixed with 4% paraformaldehyde. Half of the
tissue was cryoprotected in sucrose and theother half was mounted in 5% agar–agar andsectioned in 60-�m slices using a vibratome.Sections were mounted on glass slides andstained with hematoxylin/eosin or used for im-munochistochemistry. Hematoxylin was fromFisher and eosin-Y from Sigma. Antibodies usedwere mouse monoclonal antibody against neu-ronal class III beta-tubulin (TUJ1) (Covance,Richmond, CA) for ectoderm; mouse antihumanmuscle actin monoclonal antibody (Dako,Carpinteria, CA) and polyclonal rabbit antihu-man fibronectin antibody (Dako) for mesoderm;and mouse monoclonal anti-alpha-fetoproteinantibody (Sigma) for endoderm. Secondary goatantimouse Alexa Fluor 546 (orange-fluorescent)and goat antirabbit Alexa Fluor 488 (green-flu-orescent) antibody were from Molecular Probes(Invitrogen). Nuclei were stained with DAPI(blue) (Sigma).
About 2 to 5 million mitomycin-C-treated hu-man fibroblasts used as feeder layers for the phEScells were injected as controls. No teratomagrowth was observed in the control animals.
RESULTS
Generation of parthenogenetic embryos
Five oocyte donors, all over 31 years of age,participated in this study. Oocytes were obtainedusing hormonal stimulation with the primary in-tent of IVF. A total of 46 COCs were taken fromfive donors and used for this study (Table 1). Be-fore oocyte activation, COCs were held at atmo-spheric oxygen tension. After removal of cumu-lus cells only normal metaphase II oocytes withdistinct first polar body were taken for activationprocedure. We activated oocytes with 5 �M iono-mycin for 5 min, followed by incubation with 1mM 6-DMAP for 4 h to prevent the extrusion ofthe second polar body and produce diploid em-bryos. Manipulation and culture of oocytes andembryos was performed in MediCult media in ac-cordance with manufacturer’s recommendationsusing standard IVF procedures and under re-duced oxygen (90% N2 � 5% O2 � 5% CO2). Only40 oocytes were capable of cleavage afterparthenogenetic activation. These procedurespermitted us to produce 23 blastocysts on day 5or 6 of embryo culture. Eleven of the blastocystshad visible ICMs (Table 1).
REVAZOVA ET AL.436
PATIENT-SPECIFIC STEM CELL LINES 437
Derivation of parthenogenetic hESC lines
It is very important to minimize, if not elimi-nate, components of animal origin in the deriva-tion and culture of hESCs destined for clinicaluse. To this end, some researchers did not use im-munosurgery for ICM isolation, but rather usedmechanical means to isolate the ICMs from wholeblastocysts (Kim et al., 2005). Culture media andfeeder cells may also contain animal pathogens.To eliminate possible contamination, researchershave used human cells as feeder instead of mousefibroblasts (Cheng et al., 2003; Hovatta et al., 2003;Stojkovic et al., 2005) or have cultured hESCs infeeder-free and serum-free conditions (Amit etal., 2004; Klimanskaya et al., 2005). Taking intoaccount these previous investigations, we modi-fied the culture conditions under which we iso-lated the ICM and cultured the phESC.
For the derivation and culture of phESC weused mitomycin C mitotically inactivated humanNSF as feeder cells. These cells originally werederived and propagated with medium contain-ing human umbilical cord blood serum instead ofanimal serum. The phESC culture medium con-sisted of VitroHES medium (Vitrolife) supple-mented with human serum derived from umbil-ical cord blood, hrbFGF, and hrLIF. The phESC
were propagated in a 37°C, 5% CO2, humidifiedatmosphere.
All derived parthenogenetic blastocysts wereinitially treated with 0.5% pronase to remove thezona pellucida. We isolated well-formed ICMsfrom two blastocysts obtained from the first oocytedonor using trypsin treatment (Li et al., 2003) andtraditional immunosurgery (Solter and Knowles,1975). The ICMs were further placed on humanfeeder cells at described conditions to produce ph-ESC. The ICM from trypsin-treated blastocyst didnot give live cells. The ICM derived after im-munosurgery displayed cell outgrowth, resultingin the creation of the phESC-1 cell line. The other21 whole blastocysts were initially placed on thefeeder at described conditions. The ICMs were iso-lated by mechanical slicing from sprawled tro-phoblast cells and replaced onto fresh feeder cells.Five phESC lines (from phESC-3 to phESC-7) weregenerated in this manner (Table 1).
The characterization of phESC lines
The phESC lines display a morphology ex-pected in hESCs and form colonies with tightlypacked cells, prominent nucleoli, and a smallcytoplasm to nucleus ratio (Fig. 1). These cellsexpress traditional hES cell markers SSEA-3,
TABLE 1. GENERATION OF PARTHENOTES AND PARTHENOGENETIC EMBRYONIC STEM CELL LINES
Blastocysts derivedf
Oocytes Oocytes Oocytes Parthenotes with without Lines DonorDonor derived donated activated createde ICM visible ICM generated destiny
1 8 4 4 4 2 — phESC-1 pregnantimmunosurgery
2 15 8 8 8 3 3 phESC-3 pregnantphESC-4 twinsphESC-5
all from wholeblastocysts
3 27 14 14.a,b 11 3 2 phESC-6 pregnantfrom wholeblastocyst
4 22 11 11.c 10 2 3 phESC-7 pregnantfrom wholeblastocyst
5 20 9.d 7 7 1 4 no cell line notgenerated pregnant
aTwo oocytes were not activated.bOne oocyte degenerated after activation.cOne oocyte was not activated.dTwo oocytes were at metaphase I stage and were discarded.eTotal parthenogenetically activated oocytes � 40.fTotal blastocytst derived � 23.
REVAZOVA ET AL.438
FIG. 1. Characterization phESC lines for specific markers. Undifferentiated colonies of phESC on human feeder layercells (A–F), negative staining for SSEA-1 (G–L), expression of cell surface markers SSEA-3 (M–R), SSEA-4 (S–X). Orig-inal magnification (A–E) �100; (F) �200; (G–X) �400. Alkaline phosphatase-positive staining of phESC colonies onfeeder cells (A–F), OCT-4 (G–L), TRA-1-60 (K–R), and TRA-1-81 (S–X). Original magnification (A, B, O, R) �100;(C–F, M, S, X) �200; (G–L, N, P, Q, T–W) �400.
SSEA-4, (Fig. 1) TRA-1-60, TRA-1-81, and OCT-4, (Fig. 1) and do not express SSEA-1, a posi-tive marker for undifferentiated mouse embry-onic stem cells (Fig. 1). The cells derived fromall lines demonstrate high levels of alkaline
phosphatase (Fig. 1) and telomerase activity(Fig. 2).
G-banded karyotyping showed that phESClines have a normal human 46,XX karyotype,with the exception of the phESC-7 line (Fig. 3).
PATIENT-SPECIFIC STEM CELL LINES 439
Approximately 91% of cells from the phESC-7line have a 47,XXX karyotype and 9% of the cellshave a 48,XXX,+6 karyotype. We observed a dif-ferent degree of X chromosome heteromorphismby analysis of 100 metaphases in our cell lines:approximately 12% of cells for the phESC-1 andphESC-6 lines showed X chromosome hetero-morphism; 42% of cells for the phESC-5 line andin 70%, 80%, and 86% for the cell lines phESC-7,
phESC-3, and phESC-4, respectively, showed Xchromosome heteromorphism (Fig. 3).
Differentiation capacity of phESC
The phESC-1 line remained undifferentiatedduring 10 months of culture spanning 35 pas-sages. The other cell lines were successfully cul-tivated over at least 21 passages. The cells from
FIG. 1. (Continued).
REVAZOVA ET AL.440
all phESC lines formed cystic embryoid bodies insuspension culture and gave rise to derivatives ofall three germ layers, ectoderm, mesoderm, andendoderm, after differentiation in vitro (Fig. 4).Approximately 5% of embryoid bodies from thephESC-1 line gave rise to beating cells 5 days fol-lowing plating (Movie 1, supplemental material;www.liebertonline.com/doi/suppl/10.1089/clo.2007.0033/suppl_file/MovieS1.avi). The phESC-6 line produced pigmented epithelial-like cells(Fig. 4I and K). Ectoderm differentiation is pre-sented by positive immunocytochemical stainingfor neuron specific markers neurofilament 68(Fig. 4A), NCAM (Fig. 4B), beta III-tubulin (Fig.4C), and the glial cell marker GFAP (Fig. 4D andM). Differentiated cells were positive for meso-derm markers including alpha-actinin (Fig. 4G)and desmin (Fig. 4J), which are muscle-specific markers, and the endothelial markersPECAM-1 (Fig. 4E) and VE-Cadherin (Fig. 4F).Endoderm differentiation is presented by posi-tive staining of differentiated derivatives for al-pha-fetoprotein (Fig. 4H and L). The ability of phESC lines to form derivatives from all threegerm layers was investigated in vivo by subcuta-neous injection of phESC into immunodeficientmice and rats (Fig. 5). Cells from all phESC lineswere capable of forming teratomas approxi-mately two months after injection. Histologicalexamination demonstrated the presence of orga-nized structures, including epithelia, capsula,smooth muscle, adipose tissue, hematogenic tis-
sue, neural tubes, and glandular epithelia (datanot shown). Immunohistochemical analysis re-vealed positive staining for beta-tubulin (Fig.5C)—ectoderm marker, fibronectin (Fig. 5B), andmuscle actin (Fig. 5A)—mesoderm markers, andalpha-fetoprotein (Fig. 5D)—endoderm marker.These data demonstrate that phESC can be dif-ferentiated in vivo into the three germ layers thatlead to all cell types found in a human body.
DNA profiling of phESC lines
We performed comparative DNA profiling ofall the phESC lines, the donor somatic cells, andthe feeder cells. These studies used Affymetrixsingle-nucleotide polymorphism (SNP) microar-rays (Mapping 50 K Hind 240 Arrays) to confirmthe genetic similarity of the phESC to the donor’ssomatic cells. A total of 1459 SNP markers across15 autosomes (chromosomes 1–15) were chosenwith median intermarker distance of 1.12 Mbp.All paired genotype relationships between phESC lines and their associated donor somaticcells were identified as “full siblings” (geneticallymatched), and all the other combinations of pairswere identified as “unrelated.” Internal controlsidentified the paired genotype relationship be-tween split cultures derived from the same phESC line as “monozygous twins” (Table 2).
Comparative analysis of SNP markers re-vealed, that on the whole, donor cells do not seemto exhibit a clear pattern of heterozygosity trend
FIG. 2. The phESC demonstrate high level of telomerase activity by comparison with positive control cells: “�”—the extract from 500 cells; “�”—heat-treated cell extract with inactivated telomerase; “Control �”—telomerase-pos-itive cell extract (applied with TRAPEZE Kit); “B”—CHAPS lysis buffer, primer-dimer/PCR contamination control;TSR8—telomerase quantitative control template (0.1 and 0.2 amol/�L); “M”—marker, DNA ladder.
PATIENT-SPECIFIC STEM CELL LINES 441
FIG. 3. The G-banded karyotyping of phESC lines. The phESC-1 (A), phESC-3 (B), phESC-4 (C), phESC-5 (D), andphESC-6 (E) lines have a normal 46, XX karyotype. The phESC-7 line has 47,XXX karyotype (F).
across distances from centromeres, whereas stemcells display somewhat lower proportions of het-erozygosity near centromeres and telomeres incomparison to heterozygosity proportions in the
middle as a result of probable chromosome re-combination (Supplemental material; www.liebertonline.com/doi/suppl/10.1089/clo.2007.0033/suppl_file/SNP_Analysis.pdf).
Based on the HLA-typing results, stem cellsderived of all phESC lines appeared MHC-matched with the oocyte donors, making this apossible method to create cells for therapeuticuse (Table 3). HLA-analysis of the genetic ma-terial from the human fibroblasts used as feedercells revealed no contamination of the phESC
lines with material from the human fibroblasts(Table 3).
Analysis of imprinted genes
Alterations of genomic imprinting in humanembryos can contribute to the development of
PATIENT-SPECIFIC STEM CELL LINES 443
FIG. 5. The in vivo differentiation of phESC and teratoma formation in SCID mice. Immunofluorescence staining forthe markers of three germ layers. The muscle actin, a mesodermal cell marker, is organized surrounding other com-ponents and is clearly identifiable (A). The presence of fibronectin in higher quantities is specific for connective tis-sue of mesodermal origin (B) The areas of neural differentiated cells (ectodermal origin) are extensive and labeled in-tensively with antybodies for beta tubulin (C). Alpha-fetoprotein, an immature endodermal cell marker, can beretrieved in areas of glandular appearance (D). The nuclei were staned with DAPI (blue)—(A, D). Original magnifi-cation (A), (C), (D) �100; (B) �200.
FIG. 4. The in vitro differentiation of phESC into derivatives of all three germ layers. Ectoderm differentiation is pre-sented by positive immunocytochemical staining for neuron-specific markers neurofilament 68 (A), NCAM (B), betaIII-tubulin (C), and glial cell marker GFAP (D, M). Differentiated cells were positive for mesoderm markers: musclespecific alpha-actinin (G) and desmin (J), endothelial markers PECAM-1 (E), and VE-Cadherin (F). Endoderm dif-ferentiation is presented by positive staining for alpha-fetoprotein (H, L). The phESC produce pigmented epithelial-like cells (I, K). Original magnification (I) �100; (A–H, J–M) �400.
TA
BL
E2.
I DE
NT
IFY
ING
DN
A S
AM
PL
ES
FRO
MP
HE
SC A
ND
RE
LA
TE
DD
ON
OR
S
Gen
otyp
eG
enot
ype
Put
ativ
eIn
ferr
ed1
2re
lati
onsh
ipre
lati
onsh
ip0
12
n_ty
ped
MZ
twin
sP
ar/o
ffFu
llsib
sH
alfs
ibs
Unr
elat
ed
12
unre
late
dun
rela
ted
166
662
631
1459
�15
03.0
3�
300.
45�
23.1
5�
8.41
01
3un
rela
ted
unre
late
d24
161
660
214
59�
1560
.65
�43
4.85
�28
.04
�12
.22
01
4un
rela
ted
unre
late
d22
562
361
114
59�
1535
.94
�40
0.61
�31
.39
�14
.39
01
5un
rela
ted
unre
late
d22
562
361
114
59�
1535
.94
�40
0.61
�31
.39
�14
.39
01
6un
rela
ted
unre
late
d24
364
457
214
59�
1642
.35
�44
5.78
�31
.74
�14
.54
01
7un
rela
ted
unre
late
d25
263
856
914
59�
1641
.11
�45
3.5
�29
.25
�12
.86
01
8un
rela
ted
unre
late
d25
064
356
614
59�
1656
.02
�46
0.02
�32
.86
�15
.32
01
9un
rela
ted
unre
late
d21
965
758
314
59�
1605
.31
�38
2.39
�27
.37
�11
.58
01
10un
rela
ted
unre
late
d15
870
759
414
59�
1591
.43
�27
9.21
�26
.37
�10
.89
01
11un
rela
ted
unre
late
d19
366
859
814
59�
1584
.71
�35
4.76
�29
.65
�13
.31
01
12un
rela
ted
unre
late
d16
667
162
214
59�
1523
.1�
300.
5�
30.5
3�
13.9
20
23
unre
late
dfu
ll si
bs0
282
1177
1459
�44
0.02
�14
6.3
0�
167.
42�
363.
632
4un
rela
ted
unre
late
d23
362
759
914
59�
1569
.66
�42
3.24
�28
.24
�12
.91
02
5un
rela
ted
unre
late
d23
362
759
914
59�
1569
.66
�42
3.24
�28
.24
�12
.91
02
6un
rela
ted
unre
late
d21
765
059
214
59�
1584
.75
�38
8.44
�22
.62
�8.
530
27
unre
late
dun
rela
ted
243
650
566
1459
�16
45.9
4�
437.
91�
23.2
3�
8.72
02
8un
rela
ted
unre
late
d22
564
958
514
59�
1603
.18
�40
4.41
�27
.04
�11
.97
02
9un
rela
ted
unre
late
d21
063
961
014
59�
1532
.75
�36
0.46
�24
.72
�9.
890
210
unre
late
dun
rela
ted
144
683
632
1459
�14
91.1
8�
243.
56�
16.8
2�
4.51
02
11un
rela
ted
unre
late
d17
268
060
714
59�
1556
.46
�31
0.03
�23
.5�
9.7
02
12un
rela
ted
unre
late
d17
666
761
614
59�
1538
.57
�32
7.95
�27
.31
�12
.06
03
4un
rela
ted
unre
late
d33
645
766
514
59�
1391
.57
�59
9.92
�30
.6�
14.6
20
35
unre
late
dun
rela
ted
336
457
665
1459
�13
91.5
7�
599.
92�
30.6
�14
.62
03
6un
rela
ted
unre
late
d32
248
265
514
59�
1415
.98
�57
1.23
�26
.08
�11
.86
03
7un
rela
ted
unre
late
d36
944
264
814
59�
1432
.05
�66
4.95
�27
.39
�11
.93
03
8un
rela
ted
unre
late
d33
448
364
214
59�
1449
.86
�59
7.75
�31
.68
�15
.14
03
9un
rela
ted
unre
late
d30
749
365
914
59�
1395
.19
�53
0.45
�24
.56
�10
03
10un
rela
ted
unre
late
d21
562
362
114
59�
1503
.92
�36
4.97
�17
.26
�4.
430
311
unre
late
dun
rela
ted
264
582
613
1459
�15
31.9
1�
473.
48�
28.4
1�
12.8
10
312
unre
late
dun
rela
ted
254
595
610
1459
�15
44.7
3�
460.
57�
29.9
2�
13.8
80
45
unre
late
dM
Z t
win
s0
014
5914
590
�37
9.58
�45
.47
�40
1.67
�67
7.74
46
unre
late
dun
rela
ted
334
475
650
1459
�14
36.5
9�
599.
55�
342.
73�
15.1
90
47
unre
late
dun
rela
ted
365
439
655
1459
�14
18.3
4�
656.
01�
31.6
�14
.56
04
8un
rela
ted
unre
late
d32
948
664
414
59�
1450
.75
�58
6.4
�32
.06
�14
.88
04
9un
rela
ted
unre
late
d33
246
666
114
59�
1395
.18
�59
0.12
�28
.69
�12
.94
04
10un
rela
ted
unre
late
d24
560
660
814
59�
1542
.32
�43
8.93
�28
.75
�12
.74
04
11un
rela
ted
unre
late
d27
356
961
714
59�
1530
.97
�49
2.84
�29
.03
�12
.34
04
12un
rela
ted
full
sibs
022
412
3514
59�
326.
17�
162.
340
�18
3.44
�39
3.46
56
unre
late
dun
rela
ted
334
475
650
1459
�14
36.5
9�
599.
55�
32.7
3�
15.1
90
57
unre
late
dun
rela
ted
365
439
655
1459
�14
18.3
4�
656.
01�
31.6
�14
.56
05
8un
rela
ted
unre
late
d32
948
664
414
59�
1450
.75
�58
6.4
�32
.06
�14
.88
0
IBS
LOD
59
unre
late
dun
rela
ted
332
466
661
1459
�13
95.1
8�
590.
12�
28.6
9�
12.9
40
510
unre
late
dun
rela
ted
245
606
608
1459
�15
42.3
2�
438.
93�
28.7
5�
12.7
40
511
unre
late
dun
rela
ted
273
569
617
1459
�15
30.9
7�
492.
84�
29.0
3�
12.3
40
512
unre
late
dfu
ll si
bs0
224
1235
1459
�32
6.17
�16
2.34
0�
183.
44�
393.
466
7un
rela
ted
full
sibs
4517
612
3814
59�
277.
78�
217.
210
�16
5.72
�39
0.62
68
unre
late
dfu
ll si
bs44
187
1228
1459
�28
9.8
�20
1.32
0�
153.
75�
365.
516
9un
rela
ted
unre
late
d33
348
164
514
59�
1436
.5�
595.
4�
30.3
�13
.77
06
10un
rela
ted
unre
late
d24
060
161
814
59�
1518
.17
�42
5.03
�27
.11
�11
.53
06
11un
rela
ted
full
sibs
016
412
9514
59�
209.
27�
191.
660
�21
3.25
�44
0.56
612
unre
late
dun
rela
ted
234
615
610
1459
�15
47.1
5�
416.
14�
30.2
1�
13.6
40
78
unre
late
dfu
ll si
bs38
225
1196
1459
�32
6.62
�15
0.16
0�
121.
55�
334.
097
9un
rela
ted
unre
late
d35
947
362
714
59�
1479
.28
�64
2.41
�30
.61
�14
.47
07
10un
rela
ted
unre
late
d25
262
358
414
59�
1598
.35
�44
3.81
�28
.88
�13
.09
07
11un
rela
ted
full
sibs
023
012
2914
59�
318.
49�
137.
930
�15
9.55
�38
9.58
712
unre
late
dun
rela
ted
265
583
611
1459
�15
39.3
3�
472.
91�
30.5
5�
13.8
70
89
unre
late
dun
rela
ted
347
480
632
1459
�14
72.4
1�
625.
68�
30.9
3�
14.3
10
810
unre
late
dun
rela
ted
244
614
601
1459
�15
61.3
�43
4�
28.0
7�
12.3
70
811
unre
late
dfu
ll si
bs0
175
1284
1459
�22
3.73
�17
8.56
0�
200.
12�
428.
048
12un
rela
ted
unre
late
d23
661
061
314
59�
1539
.08
�17
.14
�29
.32
�13
.14
09
10un
rela
ted
full
sibs
022
812
3114
59�
315.
15�
152.
880
�17
4.27
�39
2.91
911
unre
late
dun
rela
ted
269
567
623
1459
�15
02.6
9�
479.
57�
28.4
7�
12.5
50
912
unre
late
dun
rela
ted
245
612
602
1459
�15
57.2
5�
438.
53�
26.0
7�
11.1
50
1011
unre
late
dun
rela
ted
187
635
637
1459
�14
78.7
�32
8.06
�25
.52
�10
.60
1012
unre
late
dun
rela
ted
181
662
616
1459
�15
34.3
6�
329
�25
.2�
10.6
011
12un
rela
ted
unre
late
d18
964
562
514
59�
1520
.01
�33
7.88
�27
.33
�11
.72
0
DN
A s
ampl
es w
ere
num
bere
d a
s fo
llow
s: 1
—hu
man
neo
nata
l ski
n fi
brob
last
s, 2
—ph
ESC
-7 li
ne d
onor
, 3—
phE
SC-7
line
, 4—
phE
SC-1
line
, 5—
phE
SC-1
line
, 6—
phE
SC-
3 lin
e, 7
—ph
ESC
-4 l
ine,
8—
phE
SC-5
lin
e, 9
—ph
ESC
-6 l
ine,
10—
phE
SC-6
lin
e d
onor
, 11—
phE
SC-3
to
phE
SC-5
lin
es d
onor
, 12—
phE
SC-1
lin
e d
onor
. The
IB
S co
lum
ns i
nth
e ou
tput
dis
play
the
num
ber
of m
arke
rs a
t w
hich
the
pai
r ar
e bo
th t
yped
and
sha
re 0
, 1, o
r 2
alle
les
iden
tica
l by
sta
te. (
For
MZ
tw
ins
und
er t
he i
dea
l co
ndit
ion
of n
oge
noty
ping
err
ors,
all
mar
kers
mus
t be
pla
ced
und
er I
BS
�2)
. The
out
put
doe
s no
t d
ispl
ay P
(ob
serv
ed m
arke
rs⏐g
iven
rel
atio
nshi
p) d
irec
tly,
but
it
dis
play
s L
OD
sco
re�
log 1
0[P
(ob
serv
ed m
arke
rs⏐p
utat
ive
rela
tion
ship
)/P(
obse
rved
mar
kers
⏐rel
atio
nshi
p fo
r w
hich
max
imum
lik
elih
ood
was
obt
aine
d a
nd t
hus
the
call
was
mad
e)]
as a
mea
sure
of
sim
ilari
ty. T
he s
mal
ler
this
LO
D s
core
is, t
he le
ss li
kely
is t
he p
utat
ive
rela
tion
ship
bet
wee
n tw
o sa
mpl
es.
disorders linked to maternally or paternally ex-pressed genes (Gabriel et al., 1998). Studies of im-printing in phESC require a detailed investiga-tion because of the possible influence uponphESC differentiation and functionality of theirderivatives. As a preliminary study, we per-formed expression analysis of the human im-printed genes (Morison et al., 2005) TSSC5, H19,PEG1, and SNRPN in undifferentiated phESC(Fig. 6). Two hESC lines derived from discardedIVF embryos were used as controls. The tran-scripts of maternally expressed genes TSSC5(Morison et al., 2005) and H19 (Morison et al.,2005) were observed in all phESC lines and alsoin control lines. The human PEG1 gene is tran-scribed from two alternative promoters (Li et al.,2002). The gene region from the first promoter isbiallelically expressed, and the gene region fromthe second promoter (isoform 1) is paternally ex-pressed (Li et al., 2002). Expression of the PEG1gene from the first promoter was not affected inour phESC lines. Analysis of the paternally ex-pressed region of the PEG1 gene and the pater-nally expressed SNRPN gene (Morison et al.,2005) demonstrated that expression of thesegenes was significantly downregulated in phESClines in comparison with control hESC lines (Fig.6). These results provide further evidence of theparthenogenetic origin of the described phESClines.
DISCUSSION
The present paper describes the protocol for invitro human parthenogenetic embryo developmentand subsequent embryonic stem cell derivation. Forthe generation of parthenogenetic human embryos,we used MediCult media developed for humanembryo culture and procedures applied for stan-dard IVF embryos, and cultured them at reducedoxygen. The use of low oxygen concentration in thegas mixture for the development of humanparthenogenetic embryos to the blastocyst stage invitro was critical. In general, the oxygen tension ina mammal oviduct and uterus is much less thanhalf of that found in the normal atmosphere (Fis-cher and Bavister, 1993; Kaufman and Mitchell,1994). For successful culture of human embryos af-ter IVF, oxygen concentrations of 20% as well as 5%have been used. However, increased oxygen cangenerate reactive oxygen species that can induceapoptosis (Van Soom et al., 2002). It has been re-ported that low oxygen concentration increases theviability of preimplantation embryos, assists theirnormal development, and gives higher incidence ofthe formation of healthy blastocysts as indicated bygreater cell number and a well formed ICM (Du-moulin et al., 1999). In previous investigations, hu-man parthenogenetic embryos were developed invitro using gas mixtures with high (20%) oxygencontent (Cibelli et al., 2001; Lin et al., 2003). To pro-
REVAZOVA ET AL.446
TABLE 3. HLA TYPING RESULTS
MHC I MHC II
HLA-A HLA-B HLA-C DRB1 DQB1 DQA1
phESC-1 A*01 B*15 (63) Cw*04 DRB1*12 DQB1*06 DQA1*01A*02 B*35 Cw*0708 DRB1*13 DQB1*03 DQA1*0505
phESC-1 A*01 B*15 (63) Cw*04 DRB1*12 DQB1*06 DQA1*01donor A*02 B*35 Cw*0708 DRB1*13 DQB1*03 DQA1*0505
phESC-3, A*02 B*52 Cw*03 DRB1*01 DQB1*05 DQA1*01014, 5 A*03 B*22 Cw*04 DRB1*03 DQB1*02 DQA1*05
phESC-3, A*02 B*52 Cw*03 DRB1*01 DQB1*05 DQA1*01014, 5 donor A*03 B*22 Cw*04 DRB1*03 DQB1*02 DQA1*05
phESC-6 A*02 B*07 Cw*04 DRB1*04 DQB1*06 DQA1*01A*03 B*27 Cw*07 DRB1*15 DQB1*03 DQA1*03
phESC-6 A*02 B*07 Cw*04 DRB1*04 DQB1*06 DQA1*01donor A*03 B*27 Cw*07 DRB1*15 DQB1*03 DQA1*03
phESC-7 A*01 B*38 Cw*06 DRB1*13 DQB1*06 DQA1*0106A*02 B*57 Cw*12 DRB1*14 DQB1*06 DQA1*0103
phESC-7 A*01 B*38 Cw*06 DRB1*13 DQB1*06 DQA1*0106donor A*02 B*57 Cw*12 DRB1*14 DQB1*06 DQA1*0103
NSF A*25 B*15(62) Cw*12 DRB1*04 DQB1*06 DQA1*01A*32 B*18 Cw*12 DRB1*15 DQB1*03 DQA1*03
PATIENT-SPECIFIC STEM CELL LINES 447
duce human parthenogenetic embryos, we used agas mixture containing 5% oxygen and achieved of57.5% success rate in the formation of blastocysts.
The next step in the study was to select opti-mal methods for ICM isolation and phESC cul-ture giving the best possible chance for clinicaluse of the phESC lines. To accomplish this goal,we derived human skin fibroblasts, propagatedthem with human umbilical cord blood serum(HUCBS) instead of animal serum and used themas feeder cells. Derivation and culture of phESClines was performed in VitroHES medium (Vit-rolife) designed for hESC culture with the addi-tion of HUCBS. The use of HUCBS in productionof hESCs has not been reported in earlier work,and we observed it had positive effects on ICMoutgrowth and phESC propagation. The growthof phESC in VitroHES medium was better withthe addition of HUCBS than without it. The iso-lation ICMs from whole blastocysts by mechani-cal slicing from the trophectoderm outgrowth ap-peared to be a more gentle and preferable methodversus immunosurgery and trypsin treatment.Moreover, this method permitted the exclusion ofinteraction with animal-derived reagents.
The six phESC lines were derived from oocytesgiven by four donors. IVF procedures resulted inpregnancies in all four donors. A contributing fac-tor to this high success rate for donor IVF preg-nancy was the selection of donors with a goodprognosis for IVF pregnancy. The IVF procedurefor donor 5 did not result in a pregnancy. Incontrast to the other four, the quality of par-thenogenetically activated oocytes from donornumber 5 was poor, and we were unable to gen-erate phESC lines. So the development of humanparthenogenetic embryos may depend on thequality of donor oocytes and may be conditionedby donor peculiarities.
In prior research, parthenogenetic activation ofmouse oocytes resulted in MHC-homozygousembryonic stem cell lines (Lin et al., 2003). In ourstudy the suppression of the second meiotic di-vision after parthenogenetic activation of humanmetaphase II oocytes and the generation ofdiploid embryos led to the derivation of MHC-heterozygous phESC.
Although the phESC lines we generated pres-ent typical characteristics displayed by hESClines, they show unique characteristics, includinggenotypes that are practically identical to thoseof the oocyte donors, as seen in the partheno-genetically derived monkey ES cell line Cyno-1
(Vrana et al., 2003). The phESC lines provide aunique model for clinical and scientific research,and the creation of hESC lines from partheno-genetic embryos may be a superior way to gen-erate MHC-matched and possibly histocompati-ble ES cells in comparison to SCNT.
Further investigations of the characteristics ofphESC lines and their immune matching are nec-essary to determine their suitability for use in celltherapy. As an example, the altered karyotype ofphESC-7 may be a reason to exclude it from clin-ical use. Doubts have been raised regarding thecapability of parthenogenetic stem cells to be dif-ferentiated into functional derivatives. In this re-gard, the additional analysis of the influence ofimprinted genes on functionality of phESC dif-ferentiated derivatives is needed. However, asprevious studies of mouse and monkey partheno-genetic stem cells have shown, these cells can formteratomas with derivatives from all three embry-onic germ layers (Lin et al., 2003; Vrana et al.,2003). Monkey parthenogenetic ES cells under se-lective culture conditions have been differentiatedinto neural cells and functional dopaminergic andserotonergic neurons (Vrana et al., 2003). In ourstudy we have shown that phESC can also be dif-ferentiated into derivatives of all three germ lay-ers in vitro and in vivo and are pluripotent. More-over, embryoid bodies from phESC were capableof giving rise to beating cardiomyocyte-like cells(Movie 1, supplement material).
We have demonstrated a method of creatingparthenogenetic human embryonic stem cells andhave generated experimental data showing thatphESC can be differentiated into functional cellsthat may be of great value in future treatment ofhuman degenerative diseases and for use in stemcell research.
ACNOWLEDGMENTS
We would like to thank H.S. Keirstead (Uni-versity of California at Irvine, Irvine, CA) and J.Hammond (Lifeline Cell Technology, Walk-ersville, MD) for helpful discussions and teratomainvestigation; V. Yu. Abramov (National ResearchInstitute of Transplantalogy and Artificial Organs,Moscow, Russia) for HLA-typing; O. I. Sokova(Russian N. N. Blokhin Memorial Cancer Re-search Center, RAMS, Moscow, Russia) for karyo-typing of phESC lines; E. S. Zakharova (Instituteof Gene Biology, RAS, Moscow, Russia) for the de-
tection of the telomerase activity; A. V. Kibardin(Institute of Gene Biology, RAS, Moscow, Russia)for gene imprinting analysis; P. H. Schwartz (Or-ange County Research Institute, Orange, CA) forhis helpful discussions and immunocytochemicalassistance; and L. S. Agapova (Russian N. N.Blokhin Memorial Cancer Research Center,RAMS, Moscow, Russia) for her help in immuno-cytochemical staining. Microarray experimentsand statistical analyses were performed at The
Centre for Applied Genomics at The Hospital forSick Children, Toronto, Canada. We are gratefulto Drs. Celia Greenwood and Sooyeol Lim for sta-tistical analysis, Dr. Chao Lu and Quyen Tran formicroarray experiments, and Drs. Christian Mar-shall and Richard Wintle for helpful discussionsof the manuscript. Funding for this work was pro-vided by Lifeline Cell Technology, LLC, Walk-ersville, MD, a division of International Stem CellCorporation, Oceanside CA.
REVAZOVA ET AL.448
FIG. 6. RT-PCR analysis of imprinted gene expression. Two hESC lines hES1 and hES2 from discarded IVF embryoswere used as positive controls for the expression analysis of the paternally expressed genes SNRPN and PEG1_2 andthe maternally expressed genes TSSC5 and H19 in phESC lines: phESC-1, phESC-3, phESC-4, phESC-5, phESC-6, andphESC-7 (1, 3, 4, 5, 6, and 7, respectively.) The PEG1_1 gene is biallelically expressed and was used as an additionalcontrol. GAPDH was included as mRNA quantitative control. RT data demonstrate no genomic contamination of RTsamples.
REFERENCES
Adjaye, J., Bolton, V., and Monk, M. (1999). Develop-mental expression of specific genes detected in high-quality cDNA libraries from single human preimplan-tation embryos. Gene 237, 373–383.
Amit, M., Shariki, C., Margulets, V., et al. (2004). Feederlayer- and serum-free culture of human embryonicstem cells. Biol. Reprod. 70, 837–845.
Boehnke, M., and Cox, N.J. (1997). Accurate inference ofrelationships in Sib-pair linkage studies. Am. J. Hum.Genet. 61, 423–429.
Broman, K.W., and Weber, J.L. (1998). Estimation of pair-wise relationships in the presence of genotyping errors.Am. J. Hum. Genet. 63, 1563–1564.
Cheng, L., Hammond, H., Ye, Z., et al. (2003). Humanadult marrow cells support prolonged expansion of hu-man embryonic stem cells in culture. Stem Cells 21,131–142.
Chomcznski, P., and Sacchi, N. (1987). Single-stepmethod of RNA isolation by acid guanidinium thio-cyanate-phenol-chloroform extraction. Anal. Biochem.162, 156–159.
Cibelli, J.B., Kiessling, A.A., Cunniff, et al. (2001). Rapidcommunication: somatic cell nuclear transfer in hu-mans: pronuclear and early embryonic development. e-biomed: J. Reg. Med. 2, 25–31.
Conley, B.J., Young, J.C., Trounson, et al. (2004). Deriva-tion, propagation and differentiation of human embry-onic stem cells. Int. J. Biochem. Cell Biol. 36, 555–567.
Dumoulin, J.C., Meijers, C.J., Bras, M., et al. (1999). Effectof oxygen concentration on human in vitro fertilizationand embryo culture. Hum. Reprod. 14, 465–469.
Fischer, B., and Bavister, B.D. (1993). Oxygen tension inthe oviduct and uterus of rhesus monkeys, hamstersand rabbits. J. Reprod. Fertil. 99, 673–679.
Food and Drug Administration. (2004). Guidance for In-dustry: Eligibility Determination for Donors of HumanCells, Tissues, and Cellular and Tissue-Based Products(HCT/Ps).
Gabriel, J.M., Higgins, M.J., Gebuhr, T.C., et al. (1998). Amodel system to study genomic imprinting of humangenes. Proc. Natl. Acad. Sci. USA 95, 14857–14862.
Glenn, C.C., Porter, K.A., Jong, M.T., et al. (1993). Func-tional imprinting and epigenetic modification of the hu-man SNRPN gene. Hum. Mol. Genet. 2, 2001–2005.
Hashimoto, K., Azuma, C., Koyama, M., et al. (1995). Lossof imprinting in choriocarcinoma. Nat. Genet. 9,109–110.
Hovatta, O., Mikkola, M., Gertow, K., et al. (2003). A cul-ture system using human foreskin fibroblasts as feedercells allows production of human embryonic stem cells.Hum. Reprod. 18, 1404–1409.
Kaufman, D.I., and Mitchell, J.A. (1994). Intrauterine oxy-gen tension during estrous cycle in the hamster: pat-terns of change. Comp. Biochem. Physiol. Comp. Phys-iol. 107, 673–678.
Kim, H.S., Oh, S.K., Park, Y.B., et al. (2005). Methods for
derivation of human embryonic stem cells. Stem Cells23, 1228–1233.
Klimanskaya, I., Chung, Y., Meisner, L., et al. (2005). Hu-man embryonic stem cells derived without feeder cells.Lancet 365, 1636–1641.
Lee, M.P., Reeves, C., Schmitt, A., et al. (1998). Somaticmutation of TSSC5, a novel imprinted gene from hu-man chromosome 11p15.5. Cancer Res. 58, 4155–4159.
Li, M., Zhang, D., Hou, Y., et al. (2003). Isolation and cul-ture of embryonic stem cells from porcine blastocysts.Mol. Reprod. Dev. 65, 429–434.
Li, T., Vu, T.H., Lee, K.O., et al. (2002). An imprintedPEG1/MEST antisense expressed predominantly in hu-man testis and in mature spermatozoa. J Biol. Chem.277, 13518–13527.
Lin, H., Lei, J., Wininger, D., et al. (2003). Multilineage po-tential of homozygous stem cells derived frommetaphase II oocytes. Stem Cells 21, 152–161.
Marshall, V.S., Wilton, L.J., and Moore, H.D. (1998).Parthenogenetic activation of marmoset (Callithrix jac-chus) oocytes and the development of marmosetparthenogenones in vitro and in vivo. Biol. Reprod. 59,1491–1497.
Morison, I.M., Ramsay, J.P., and Spencer, H.G. (2005). Acensus of mammalian imprinting. Trends Genet. 21,457–465.
Reubinoff, B.E., Pera, M.F., Fong, C.Y., et al. (2000). Em-bryonic stem cell lines from human blastocysts: somaticdifferentiation in vitro. Nat. Biotechnol. 18, 399–404.
Solter, D., and Knowles, B.B. (1975). Immunosurgery ofmouse blastocyst. Proc. Natl. Acad. Sci. USA 72,5099–5102.
Stojkovic, P., Lako, M., Stewart, R., et al. (2005). An auto-geneic feeder cell system that efficiently supportsgrowth of undifferentiated human embryonic stemcells. Stem Cells 23, 306–314.
Thomson, J.A., Itskovitz-Eldor, J., Shapiro, S.S., et al.(1998). Embryonic stem cell lines derived from humanblastocysts. Science 282, 1145–1147.
Van Soom, A., Yuan, Y.Q., Peelman, L.J., et al. (2002).Prevalence of apoptosis and inner cell allocation inbovine embryos cultured under different oxygen ten-sion with or without cysteine addition. Theriogenology57, 1453–1465.
Vrana, K.E., Hipp, J.D., Goss, A.M., et al. (2003). Nonhu-man primate parthenogenetic stem cells. Proc NatlAcad Sci USA 100, 11911–11916.
Address reprint requests to:J.D. Janus
Lifeline Cell TechnologyWalkersville, MD (a division of International
Stem Cell Corporation, Oceanside, CA)
E-mail: [email protected]
PATIENT-SPECIFIC STEM CELL LINES 449
This article has been cited by:
1. Ian Wilmut . 2007. Embryo Stem Cells from Parthenotes and Embryos Produced by Nuclear Transfer: The Distinction BetweenThem and Their Potential Value in Cell Therapy. Cloning and Stem Cells 9:3, 291-292. [Citation] [PDF] [PDF Plus]
