EMBRYONIC STEM CELLS/INDUCED PLURIPOTENT STEM CELLS
Complete Meiosis from Human Induced Pluripotent Stem Cells
C. EGUIZABAL,a N. MONTSERRAT,a R. VASSENA,a M. BARRAGAN,a E. GARRETA,a L. GARCIA-QUEVEDO,b
F. VIDAL,b A. GIORGETTI,a A. VEIGA,a,c,d J. C. IZPISUA BELMONTEa,e
Center for Regenerative Medicine in Barcelona, Barcelona, Spain; bUnitat de Biologia Celular, Facultat de
Biociències, Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain; cServei de Medicina de la
Reproduccio, Institut Universitari Dexeus, Barcelona, Spain; dDepartment de Ciéncies Experimentals
i de la Salut, Universitat Pompeu Fabra, Barcelona, Spain; eGene Expression Laboratory,
The Salk Institute for Biological Studies, La Jolla, California, USA
a
Key Words. Human induced pluripotent stem cells • Germ cells • Meiosis • Haploid cells • In vitro differentiation
ABSTRACT
Gamete failure-derived infertility affects millions of people
worldwide; for many patients, gamete donation by unrelated
donors is the only available treatment. Embryonic stem cells
(ESCs) can differentiate in vitro into germ-like cells, but
they are genetically unrelated to the patient. Using an in
vitro protocol that aims at recapitulating development, we
have achieved, for the first time, complete differentiation of
human induced pluripotent stem cells (hiPSCs) to postmeiotic cells. Unlike previous reports using human ESCs,
postmeiotic cells arose without the over-expression of germline related transcription factors. Moreover, we consistently
obtained haploid cells from hiPSCs of different origin
(keratinocytes and cord blood), produced with a different
number of transcription factors, and of both genetic sexes,
suggesting the independence of our approach from the epigenetic memory of the reprogrammed somatic cells. Our work
brings us closer to the production of personalized human
gametes in vitro. STEM CELLS 2011; 29:1186–1195
Disclosure of potential conflicts of interest is found at the end of this article.
INTRODUCTION
Human germ cells transmit genetic information across generations. During development, primordial germ cells (PGCs) migrate
from the base of the allantoids, where they originate, to the genital
ridges [1, 2]. After colonizing the primordial gonad, PGCs
undergo global DNA demethylation, chromatin modifications,
erasure of parental imprinting [3–5], and, in the female, initiation
of meiosis. Functional gamete development to sperm and oocytes
ends with completion of meiosis years later: after puberty in the
spermatozoa of male and after fertilization of the oocyte in the
female. As developmental errors can result in infertility and other
pathological conditions [6], creation of in vitro models may not
only advance clinical applications of infertility treatments, but
also aid in in vitro production of donor-derived gametes.
Recent studies indicate that mouse [7–11] and human
[12–18] embryonic stem cells (ESCs) can differentiate in vitro
into oocyte- or sperm-like cells. However, gametes derived
from ESC lines would be unrelated to the patient in need of
fertility treatment. Moreover, the issue of identity of a new
individual, and that of the biological parent, would likely discourage this route of treatment. On the other hand, induced
pluripotent stem cells (iPSCs) overcome both issues, as they
are genetically related to the donor individual. Thus, we
sought to develop methodologies that would lead to meiosis
progression and gamete formation from iPSC.
MATERIALS
AND
METHODS
Culture of hESCs and hiPSC
Human ESCs (hESCs) and human iPSCs (hiPSCs) were
cultured on mitotically inactivated primary human foreskin
fibroblasts (iHFFs) in hESC medium with knockout (KO)–Dulbecco’s modified Eagle’s medium containing 20% KO serum
replacement, nonessential amino acids (1X), penicillin/streptomycin (50 U/ml), GlutaMAX (1X), and basic fibroblast growth
factor (bFGF; 8 ng/ml), all from Gibco (Invitrogen, Carlsbad,
CA, www.invitrogen.com). Medium was changed every 24
hours. ESCs and iPSCs were passaged mechanically onto new
iHFF layers every 7 days (Table 1).
Differentiation to Haploid Gamete-like Cells
Cells were maintained in a six-well plate for 3 weeks with
hESC media in the absence of bFGF. Subsequently, 1 lM retinoic acid (RA) (Sigma, St. Louis, MO, www.sigmaaldrich.com)
Author contributions: C.E.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript
writing, final approval of manuscript; N.M.: conception and design, collection and/or assembly of data, data analysis and interpretation;
R.V.: collection and/or assembly of data, data analysis and interpretation, manuscript writing; M.B.: collection and/or assembly of data,
data analysis and interpretation; E.G., L.G-Q., F.V., and A.G.: collection and/or assembly of data; A.V.: conception and design, final
approval of manuscript, J.C.I.B.: conception and design, final approval of manuscript, financial support.
Correspondence: J. C. Izpisua Belmonte, Ph.D., Gene Expression Laboratory, The Salk Institute for Biological Studies, 10010 North Torrey
Pines Road, La Jolla, CA 93027, USA. Telephone: 858-453-4100 (ext. 1130); Fax: 858-453-2573; e-mail: [email protected] (or) izpisua@
cmrb.eu Received February 28, 2011; accepted for publication May 10, 2011; first published online in STEM CELLS EXPRESS June 16,
C AlphaMed Press 1066-5099/2009/$30.00/0 doi: 10.1002/stem.672
2011. V
STEM CELLS 2011;29:1186–1195 www.StemCells.com
Eguizabal, Montserrat, Vassena et al.
1187
Table 1. List of human embryonic stem cell and induced
pluripotent stem cell lines used in this study
Name of the cell line
Cell type
Culture conditions
Sex
Factors
HS306
ES[6]
KiPS1
KiPS2
KiPS3
KiPS4
CBiPS1
CBiPS2
CBiPS3
CBiPS4
CBiPS5
hESC
hESC
KiPSC
KiPSC
KiPSC
KiPSC
CBiPSC
CBiPSC
CBiPSC
CBiPSC
CBiPSC
Feeders
Feeders
Feeders
Feeders
Feeders
Feeders
Feeders
Feeders
Feeders
Feeders
Feeders
F
M
M
M
M
F
M
M
F
M
F
4
4
3
4
4
2
3
2
2
Meiotic Spread Analysis
Abbreviations: CBiPSC, Cord blood induced pluripotent stem
cell; ESC, embryonic stem cell; F, females; hESC, human
embryonic stem cell; KiPSC, keratinocyte derived induced
pluripotent stem cell; M, male.
was added to the medium and the culture extended for 3 more
weeks. At the end of this period, cells were sorted and the desired
fraction was seeded at a concentration of 3.5 105 cells per well
in a six-well plate in the presence of 10 lM Forskolin (FRSK)
(Sigma), human recombinant leukemia inhibiting factor (rLIF;
1,000 U/ml) (Sigma), and bFGF (8 ng/ml) and the CYP26 inhibitor R115866 (1 lM) (generously supplied by Johnson & Johnson
Pharmaceutical Research and Development) for 2, 3, and 4 more
weeks, depending on the timing of the experiment. The medium
was changed every 2 days.
Immunofluorescence
Cells were grown on cover slides or seeded by Cytospin preparation, and fixed in 4% paraformaldehyde (PFA). The following
antibodies were used: TRA-1-81 (MAB4381, 1:100) and NESTIN
(AB5922, 1:200), both from Chemicon (Temecula, CA,
www.chemicon.com); stage-specific embryonic antigen 3 (SSEA3) (MC-631, 1:2) and SSEA-1 (MC-480, 1:1), both from the Developmental Studies Hybridoma Bank at the University of Iowa;
OCT-3/4 (C-10, SantaCruz [Santa Cruz, CA, www.scbt.com], sc5279, 1:100), NANOG (Everest Biotech [Oxfordshire, UK,
www.everestbiotech.com] EB06860, 1:100), VASA (ab13840,
Abcam [Cambridge, UK, www.abcam.com], 1:200; AF2030,
R&D Systems [Minneapolis, MN, rndsystems.com], 1:20),
VIMENTIN (V5255, Sigma, 1:200), 3b-HSD (37-2, Santa Cruz,
sc-100466, 1:50), a-fetoprotein (A0008, Covance [Princeton, NJ,
www.covance.com], 1:1000), forkhead box protein A2 (AF2400,
R&D system, 1:50), neuron-specific class III b-tubulin 1 (MMS435P, Covance, 1:500), glial fibrillary acidic protein (Z0334,
Dako [Glostrup, Denmark, www.dako.com], 1:400), a–smooth
muscle actin (A5228, Sigma, 1:400), PAX6 (PRB-278P, Covance,
1:100), and GATA4 (sc-9053, SantaCruz, 1:50). The Alexa Fluor
Series from Invitrogen (Carlsbad, CA, www.invitrogen.com) were
used as secondary antibodies (all 1:500). Images were taken using
a Leica SP5 confocal microscope.
Purification of Total RNA, Polymerase Chain
Reaction (PCR), and Retrotranscription
Qualitative PCR (RT-qPCR)
Isolation of total RNA from hESCs, keratinocyte-derived
iPSC (KiPSCs), and Cord blood-derived iPSCs (CBiPSCs)
was performed using Trizol reagent. Samples were treated
with DNase from DNA-free kit (Ambion, Austin, TX,
www.ambion.com) and 200 ng to 1 lg of total RNA was
used to synthesize cDNA using the Cloned AMV First strand
cDNA synthesis kit (Invitrogen). Approximately 25 ng of
www.StemCells.com
cDNA were used to quantify gene expression by qPCR
(7900HT, Applied Biosystems, Foster City, CA, www.
appliedbiosystems.com) and standard PCR using primers as
described (Table 2 and Supporting Information Table 1).
Samples were disaggregated to single cell by TrypLE Express
stable trypsin digestion and a hypotonic solution (pH 8.2) was
added for 10 minutes at room temperature (RT). Cells were
resuspended in 100 mM sucrose and placed onto slides previously dipped in 1% PFA-0,15% Triton-X-100, and kept in a
humid chamber at RT until the cells had settled. VASA, synaptonemal complex protein 3 (SCP3) (ab15093, Abcam,
1:100), and cH2AX (ab18311, Abcam, 1:200) were detected
by immunofluorescence.
Karyotyping
Karyotyping of the cell lines was carried out using standard
G-banding techniques every 10 passages and up to passage
70 in every cell line. Cells were incubated with colcemid KaryoMAX (0.1 lg/ml; Invitrogen) for up to 40 minutes, then dissociated with TrypLE Express (Invitrogen), and incubated for 10
minutes in prewarmed 0.075 M KCl hypotonic solution. Cells
were then resuspended in Carnoy fixative (3:1 v/v methanol/acetic acid). Metaphase spreads were prepared on glass microscope
slides and stained with 3:1 Wright stain (Sigma-Aldrich, St.
Louis, MO, www.sigmaaldrich.com). A minimum of 20 metaphase spreads per sample were analyzed.
Flow Cytometry Immunophenotyping
For surface phenotype and cell sorting, the following fluorochromes were used: fluorescein isothiocyanate (FITC), AlexaFluor-488, phycoerythrin (PE), or allophycocyanin (APC)-labeled monoclonal antibodies (mAbs), all from Becton
Dickinson Biosciences (Franklin Lakes, NJ, www.bd.com):
anti-CD9 FITC (M-L13), anti-CD49f PE (GoH3), anti-SSEA4 AF488 (MC-813-70), anti-CD90 APC (5E10), and antiSSEA-1 AF647 (MC-480). The specificity of the staining was
verified by matched isotype control mAbs. A total of 10,000
events were collected. Hoechst 33,528 (H258) was included at 1
lg/ml in the final wash to detect live/dead cells. All analyses
were performed on a MoFlo cell sorter (DakoCytomation,
Glostrup, Denmark, www.dako.com) applying Summit software.
Fluorescence In Situ Hybridization
Samples were processed and labeled as described previously
[19]. A triple-color fluorescent in situ hybridization (FISH)
was used to determine ploidies of the cells using centromeric
probes against chromosomes 18, X, and Y (Spectrum Aqua,
Spectrum Green, and Texas Red, respectively; Vysis, Abbott
Laboratories, Abbott Park, IL, www.abbottmolecular.com).
Analyses were carried out using an Olympus BX60 epifluorescence microscope equipped with a triple-band pass filter
and specific filters for Aqua, FITC, and Cy3.
Fluorescence-Activated Cell Sorting (FACS)
Analysis for DNA Content
Single cell suspensions were prepared as described above.
Cell were fixed in 4% PFA for 15 minutes at RT, permeabilized in saponin containing buffer for 30 minutes and stained
with 40 ,6-diamidino-2-phenylindole overnight. All samples
were subjected to FACS analysis and sorting.
Promoter Methylation Analysis
Genomic DNA from hESCs and hiPSCs at various stages of
differentiation was extracted from about 1,000,000 cells using
QIAamp DNA Mini Kit (Qiagen, Germantown, MD,
Complete Meiosis from hiPSC
1188
Table 2. Polymerase chain reaction primer pairs used in this study
Primers
VASA
Stra8
Lin28
18S
0
0
F: 5 GATGTTCTTGCATGGTTGGA 3
R: 50 CCATGACTCATCATCTACTGGA 30
F: 50 CTGTGGCAGAACCTCTCGG 30
R: 50 GAACCTCACTTTTGTCCAGGAA 30
F: 50 AAGCGCAGATCAAAAGGAGA 30
R: 50 CTGATGCTCTGGCAGAAGTG 30
F: 50 GTAACCCGTTGAACCCCATTC 30
R: 50 CCATCCAATCGGTAGTAGCG 30
Product size (bp)
Reference
174
[14]
238
[49]
113
NM_024674
151
[50]
Figure 1. Timeline of in vitro differentiation of human induced pluripotent stem cells as a model for derivation of postmeiotic germ cells.
Abbreviations: FGF2, fibroblast growth factor 2; FRSK, Forskolin; hESC, human embryonic stem cells; hiPSC, human induced pluripotent stem
cells; hLIF, human leukemia inhibiting factor; RA, retinoic acid.
www.qiagen.com). Purified DNA was mutagenized with Epigentek (Bionova) or Epitect (Qiagen), according to the manufacturers’ specifications. The promoter sequences for H19
were amplified by a nested PCR with two subsequent reactions using primers and conditions described previously [18].
The resulting amplified products were cloned into pGEM T
Easy plasmids, amplified in TOP10 chemically competent
cells, purified, and sequenced.
RESULTS
CBiPSC and KiPSC lines used in this study were generated and
characterized following previously reported protocols [20, 21].
All hiPSC lines expressed the pluripotency-associated transcription factors OCT4, NANOG, SOX2, KLF4, C-MYC, DPPA4,
DNMT3B, REX1, SALL2, and LIN28 and surface markers SSEA3 and TRA-1-81 (Supporting Information Figs. 1, 2, 3A). All
iPSC lines were negative for the germ cell marker VASA, both at
the RNA and protein level (Supporting Information Figs. 1, 3C).
Moreover, the transgenes used to induce the pluripotent state
were silenced in all lines (Supporting Information Fig. 3B). All
cell lines exhibited a normal karyotype throughout the experimental period (Supporting Information Fig. 4), indicating that the
ploidies recorded were a product of the differentiation strategy.
The differentiation protocol that we developed (Fig. 1)
consists of a first step in which hiPSCs and hESCs were
allowed to differentiate for 3 weeks in monolayer in absence
of any growth cytokines. A second step consisted in culturing
the cells for 3 further weeks in presence of RA. After these 6
weeks of differentiation, we sorted all cells for a combination
of surface markers and cultured the positive fraction (see
below) in presence of LIF, bFGF, FRSK, and CYP26 inhibitor for 4 more weeks, reaching a 10 week-long differentiation
protocol. During the course of the differentiation protocol, we
evaluated the correct recapitulation of the developmental
events leading to germline differentiation.
To confirm that our protocol does induce differentiation,
we compared by RT-qPCR the expression of pluripotency
related genes after 3 weeks of differentiation (Supporting Information Fig. 3D) with undifferentiated cells (Supporting Information Fig. 3A). As an example, OCT4 expression was
found to be 18 times higher in pluripotent lines than in differentiated cells. As indirect evidence that germ cell differentiation does not usually occur during spontaneous differentiation,
we produced intradermal teratomas with the cell lines used in
this study and then performed immunofluorescence for detecting markers of the three germ layers and for germ cells
(VASA). We observed expression of the endoderm, mesoderm, and ectoderm markers but not of VASA in those teratomas (Supporting Information Fig. 5C).
RA signaling has been shown to stimulate both PGCs division and entrance into meiosis through the CYP26/Stra8
pathways in genital ridges in female PGCs and at puberty in
male germ cells [22, 23]. After 3 weeks of differentiation, we
added RA to the culture and extended the differentiation
protocol for another 3 weeks. It is interesting to note that
when RA was added to the culture from the beginning of the
protocol, it did neither increase nor speed up the appearance
of PGC markers such as VASA, as assessed by immunofluorescence at 4 and 6 weeks of differentiation (data not shown),
indicating that a certain level of cell commitment has to be
attained before the RA signaling pathways become active.
Three weeks after RA supplementation, 6 weeks of total differentiation, we were able to identify VASAþ/SSEA-1 cells
by FACS analysis in some of the samples, indicating possible
exit from the PGC state of this population and progression
toward premeiotic cells [24]. Interestingly, VASAþ cells
were found surrounded by VIMENTINþ (marker of Sertoli
cells), NESTINþ, and 3b-HSDþ (markers of Leydig cells)
cells (Fig. 2A), suggesting the possibility of reconstitution of
an elementary in vitro testicular niche [25, 26].
Some of these markers are not strictly specific for Sertoli
and Leydig cells, but they also detect neural cells and
Eguizabal, Montserrat, Vassena et al.
1189
Figure 2. Differentiation towards human germ-like cells at 6 weeks of culture. (A): Immunofluorescence staining for VIMENTIN, VASA,
NESTIN, and 3b-HSD in differentiated human embryonic stem cells and human induced pluripotent stem cells before sorting. Human testis is
used as a positive control. Scale bar ¼ 50 lm. (B): Polymerase chain reaction for detecting VASA at 6 weeks differentiation in all cell lines
used. Abbreviations: DAPI, 40 ,6-diamidino-2-phenylindole.
cardiomyocytes. To clarify the identities of each cell type, we
assayed different combinations of markers by immunofluorescence (Supporting Information Fig. 5A, 5B).
When assessing markers of neural cells (NESTIN and
PAX6) in combination with VIMENTIN, we found in the
www.StemCells.com
same well a ‘‘neural niche’’ comprised by cells expressing
NESTIN and PAX 6 (but not VIMENTIN), as well as other
areas whose cells did not express PAX6, but they did express
NESTIN and VIMENTIN only, confirming the presence of a
differentiating population of non-neural origin, and presenting
1190
Complete Meiosis from hiPSC
Figure 3. Characterization of human germ-like cells after enrichment by FACS sorting at 6 weeks of culture. (A): Upper dot-plots: FACS analysis and
sorting of CD9þ/CD49fþþ/CD90/SSEA-4 population (orange) and CD9þþ/CD49fþ/CD90þ/SSEA-4þ population (red) at 6 weeks of differentiation
sample. Lower dot-plots: FACS analysis of different population in human testis as a control; CD9/CD49f/CD90þ population (red), CD9þ/CD49fþþ/
CD90 population (orange) and CD9þþ/CD49fþ/CD90 population (green). (B): Immunofluorescence of VASA in CD9þ/CD49þþ/CD90 and
CD9þþ/CD49þ/CD90þ fractions at 6 weeks of differentiation (see arrowhead representing VASA positive cell) Scale bar ¼ 25 lm. (C): Polymerase
chain reaction for the detection of VASA at 6 weeks of differentiation after enrichment by FACS sorting. (D): Representative morphologies of sorted fractions from KiPS1 and KiPS2 cultured at 8 weeks of differentiation (see arrowheads representing round cells). Scale bar ¼ 50 lm. Abbreviations: APC, allophycocyanin; DAPI, 40 ,6-diamidino-2-phenylindole; FITC, fluorescein isothiocyanate; PE, phycoerythrin; SSEA4, stage-specific embryonic antigen 4.
testicular markers (NESTIN and VIMENTIN) instead. Also,
assaying for a marker of cardiomyocytes (GATA4) in combination with NESTIN and VIMENTIN, we also found noncardiac origin and confirmed the presence of testicular markers.
To evaluate further differentiation, we analyzed the cells
after 6 weeks of differentiation. We detected the expression of
VASA in all cell lines tested (Fig. 2B), indicating the repeatability of the current differentiation protocol among cell lines.
Eguizabal, Montserrat, Vassena et al.
In primates, CD49f has been shown to be expressed in a subset of early spermatogonia [27, 28], whereas CD9 is a marker of
spermatogonia [29, 30] and CD90 expression extends to the
early spermatid in human [31]. A combination of CD9, CD49f,
CD90, and SSEA-4 staining (Fig. 3A) revealed that the CD9þ/
CD49fþþ/CD90/SSEA4-fraction comprised about 45% of
all cells (orange population) and, most importantly, contained a
small VASAþ population, roughly 2%, which is in line with the
1191
final proportion of haploid cells, as shown by immunofluorescence and RT-PCR (Fig. 3B, 3C). As a control, we characterized
samples of human testis with the same combination of markers
except for SSEA-4 (Fig. 3A, orange population). We identified
the same CD9þ/CD49fþþ/CD90 population in the testis,
about 29%, which comprised germ cells. Taken together, these
data indicate that our protocol is conducive to the differentiation
of spermatogonia-like cells [32].
Thus, we attempted to further advance differentiation toward meiosis of the CD9þ/CD49fþþ/CD90/SSEA-4 fraction by culturing the cells in the presence of human LIF
(hLIF), FRSK, bFGF, and R115866, a CYP26 inhibitor, for
either 2, 3, and 4 more weeks. The CD9þþ/CD49fþ/CD90þ/
SSEA-4þ cell fraction was also cultured in parallel. We
observed different morphologies between the two fractions
(Fig. 3D); specifically, where CD9þ/CD49fþþ/CD90/
SSEA-4 cells displayed a round shape and a high nucleus to
cytoplasm ratio (as seen in germline cells), CD9þþ/CD49fþ/
CD90þ/SSEA-4þ cells were smaller, of variable shape, and
tended to aggregate in colonies of ESC-like appearance. As
expected, this population expressed SSEA-4, an ESC marker.
Conversely, the cells in the CD9þ/CD49fþþ/CD90/SSEA4 fraction expressed the male germline marker ACROSIN
(Fig. 5D). This population was also negative for the female
markers ZP1, ZP3, and FIG 1a (data not shown), indicating a
predominant presence of male germ-like cells.
Next, we evaluated the expression of meiosis related
markers; we detected SCP3 and cH2AX expression in a fraction of CD9þ/CD49fþþ/CD90/SSEA-4 cells, indicating
meiotic competence (Fig. 4A). This was further confirmed by
detecting meiotic prophase I in chromosomal spreads of this
fractions (Fig. 4B) and by the expression of premeiotic gene,
STRA8 (Fig. 4C). We concluded that cells in the CD9þ/
CD49fþþ/CD90/SSEA-4 fraction enters meiosis by 9
weeks of differentiation. We found that both hESC and hiPSC
can initiate meiosis, as assessed by SCP3 and cH2AX expression (Fig. 4A, 4B). However, as detailed below, only hiPSCs
were able to complete meiosis and generate haploid cells
(Fig. 5A–5D and Supporting Information Fig. 7).
Furthermore, we assessed the ability of our differentiation
protocol to epigenetically reprogram the differentiating cells.
We selected six imprinted genes, two of which are normally
expressed from the maternally inherited allele: pleckstrin
homology-like domain, family A, member 2 (PHLDA2) [33]
and cyclin-dependent kinase inhibitor 1C (CDKN1C) [33],
and four that are expressed from the paternally inherited one:
mesoderm specific transcript homolog (MEST) [34], insuline
like growth factor 2 (IGF2) [35], neuronatin (NNAT) [36],
and small nuclear ribonucleoprotein polypeptide N (SNRPN)
[37]; (Supporting Information Fig. 6). The expression level of
all genes after 10 weeks of differentiation was similar to that
Figure 4. Meiotic progression of germ-like cells at 9 weeks of culture.
(A): Immunofluorescence of meiotic spreads at 9 weeks of differentiation. Detection of VASA, SCP3, and cH2AX proteins in human embryonic stem cell (hESC) and human induced pluripotent stem cell (hiPSC)
lines. (B): Upper pictures: Profase I morphology after Leishmann staining
of CD9þ/CD49fþþ (arrowheads); lower pictures: representative morphology of the nucleus in the CD9þþ/CD49fþ fraction. Scale bar ¼ 10
lm. (C): Polymerase chain reaction detecting the expression of VASA,
Stra8, and S18 in some cell lines used in this study. (D): CpGs in the H19
promoter DMR were analyzed by bisulphite sequencing in undifferentiated hiPSC and hESC and after 6 and 10 weeks of differentiation. Methylated CpGs are represented as filled circles; unmethylated CpGs are
represented as open circles. Abbreviations: DAPI, 40 ,6-diamidino-2-phenylindole; SCP3, synaptonemal complex protein 3.
www.StemCells.com
1192
Complete Meiosis from hiPSC
Figure 5. Haploid formation of germ-like cells at 10 weeks culture. (A): DNA content analysis of differentiation process of KiPS2: (i) undifferentiated, (ii) 10 weeks, (iii) 10 weeks þ R115866, (iv) control of human sperm. (B): Fluorescent in situ hybridization against chromosomes X
and Y and 18 in a female (KiPS4) and male (KiPS2) samples of differentiated human induced pluripotent stem cells over 10 weeks in presence
of inhibitor. Scale bar ¼ 10 lm. (C): Percentages of 1n, 2n, 4n, and aneuploid cells in all samples analyzed at 10 weeks of differentiation. (D):
Immunofluorescence of haploid marker, Acrosin, and Vasa in 10 weeks of differentiation sample (unsorted and sorted haploid population) in presence of inhibitor (see arrowheads showing a positive cell). Human testis was used as a positive control. Scale bar ¼ 50 lm. Abbreviations:
DAPI, 40 ,6-diamidino-2-phenylindole.
of mature spermatozoa, whereas the expression at 6 weeks
was consistently different. We also assayed the expression
level of the stem cell specific gene telomerase reverse transcriptase (TERT) and the gene controlling X inactivation
(XIST). In both cases, the expression level of hiPSCs after 10
weeks of differentiation was comparable with spermatozoa,
while it remained higher at 6 weeks of differentiation (Supporting Information Fig. 6). Furthermore, we evaluated the
DNA methylation status of the promoter of the maternally
expressed, paternally imprinted gene H19 (Fig. 4D). Four of six
cell lines tested (one hESC and five hiPSC) increased the methylation status of H19 from an average 48.5% (range 43–51) to
Eguizabal, Montserrat, Vassena et al.
an average 75.8% (range 53–100). One hiPSC line decreased in
H19 methylation level from 72% to 53%. Moreover, the culture
of the cells in presence of the CYP26 inhibitor, which acts at a
stage of differentiation posterior to imprinting re-establishment,
did not affect imprinting re-establishment.
Finally, we asked whether the CD9þ/CD49þþ/CD90/
SSEA-4 fraction contained haploid cells. DNA content analysis by FACS showed putative presence of haploid cells, indicative of meiosis completion, by 10 weeks (Fig. 5A and
Supporting Information Fig. 7); FACS positive samples were
further evaluated by FISH with centromeric probes on
chromosomes 18, X, and Y, and the haploidy was confirmed
(Fig. 5B). Using FISH, which we considered the most
stringent assay of ploidies, after performing several controls
(Supporting Information Table 3), we consistently observed
between 0.4% and 2.3% of haploid cells per sample derived
from hiPSC (Fig. 5C). The heterogeneity of results could be
due to either the variable differentiation ability of each individual hiPSC line, the effect of uncontrolled variables in the
protocol, or differences of sensibility among techniques (DNA
content vs. FISH). Also, we performed a RT-qPCR for detecting the silencing of the transgenes, and we found that most of
the lines had silenced them at 6 weeks of differentiation;
however, three of eight lines had slightly reactivated the Oct4
transgene (Supporting Information Fig. 3E). Also, we
observed that haploid cells appeared only when the inhibitor
R115866 and the rest of the growth factors (FRSK, FGF2,
and hLIF) were added to the differentiation medium (Supporting Information Table 3). Finally, we verified that the haploid
cells were masculine by detecting ACROSIN by immunofluorescence at 10 weeks of differentiation (Fig. 5D).
DISCUSSION
Gamete differentiation in the human species is a complex process that starts in the early embryo when PGCs arise and ends
as much as 40 years later with the completion of meiosis after
fertilization of the female oocyte. Along the way, the developing gamete undergoes a finely orchestrated series of events,
which renders them the most specialized cells in the body and
the only pluripotent one. Perhaps for this reason, in vitro differentiation of ESCs and iPSCs to germ cells has been the
most challenging and the least successful.
We reasoned that hiPSCs could be induced to reach the
meiotic stage of differentiation by recapitulating known steps
of gamete development. Thus, we set out to transition the
cells from a pluripotent state to a committed one. This was
initially achieved by culturing the cells without bFGF for 3
weeks, releasing them from the pluripotency pool and allowing unguided differentiation. Direct differentiation onto monolayers of human fibroblasts rather than by embryoid bodies
ensured more consistent differentiation results, as shown by a
more widespread decrease of pluripotency markers under this
condition [14]. At the end of this period, the expression of the
pluripotency associated markers SSEA-4 and OCT4 were
downregulated [14, 18] (Supporting Information Figs. 3, 8).
SSEA-4 is one of the last surface markers to be lost during
hiPSC and hESC differentiation [38] and is, therefore, a good
indicator of a permanent exit from the pluripotent program.
We speculate that during the initial 3 weeks of unguided
differentiation, some of the cells in the pluripotent population
will stochastically start down the germline differentiation
path, but this initial ‘‘inclination’’ would be lost without the
RA signaling cascade being activated. Our hypothesis gains
strength by the observation that adding RA to the differentiawww.StemCells.com
1193
tion protocol from the very beginning does neither speed up
nor increase the appearance of this initial population.
LIF is produced in the male gonad by the peritubular cells
and promotes survival and proliferation of rat gonocytes [39].
FRSK induces proliferation in germ cells by activation of
cyclic adenosine monophosphate [40] and is involved in
meiosis induction [41]. bFGF is produced by Sertoli cells,
where its role is to help balancing self-renewal and differentiation of spermatogonial stem cells [42]. In various animal
models, the expression of bFGF is involved in the autocrine
and paracrine regulation of proliferation and differentiation of
spermatogonia and spermatocytes via its receptors [43]. The
CYP26 inhibitor R115866 acts by suppressing the inhibitory
effect of CYP26 on the meiosis regulator gene STRA8 [24].
Specifically, CYP26 is part of a regulatory pathway that regulates the timing of meiotic initiation by increasing RA degradation and consequent STRA8 inactivation in the developing
gonad. This results in blocking germ cells from entering
meiosis. As the effect of R115866 on this pathway has been
shown in different biological contexts [44, 45], we speculated
that it could improve meiotic entry in our system.
It is interesting to note that omitting R115866 from the
differentiation cocktail inhibited meiotic progression, as evidenced by the consistent lack of haploid cells in control
experiments, as detected by both chromosomal spreads and
FACS analysis (Supporting Information Table 3). During
human development, PGCs migrate along the dorsal mesentery and into the genital ridges, and their differentiation lasts
about 6–8 weeks [2], while a premeiotic to postmeiotic seminiferous epithelium cycle (the time needed for a dividing
spermatogonia to reach the haploid sperm stage) is about 7
weeks long [46]. Our differentiation protocol is broadly consistent with this timing and faithfully recapitulates sequential
steps of development. We were able to direct the differentiation of hiPSC to germ-like cells without the overexpression of
any developmentally related genes. However, we could not
detect haploid cells even after 10 weeks of differentiation
when using hESCs in any of the experiments. As germline
differentiation has been achieved in hESC and hiPSCs by subsequent forced overexpression of DAZL, DAZ, and BOULE [17,
47], one can speculate that hiPSCs are somehow epigenetically
predisposed to differentiation along the germ line path as Panula et al. [47] suggested in their recent publication. One reason
might be the epigenetic state of hiPSCs, which has been shown
to be different from that of hESCs in some imprinted loci,
which might confer hiPSCs an advantage in differentiating
along the germline [48]. Another possible reason might be
some predisposing epigenetic memory from the original somatic cell population, although we tested hiPSCs from different
somatic sources and all were able to differentiate to haploid
cells in vitro. Finally, the reactivation of transgene OCT4, a
gene involved in early PGCs biology, might play a role in this
phenomenon, but in our study OCT4 was slightly reactivated in
three of eight cell lines, thus dispensable and not the main cause
of haploid cells formation. Further studies using nonintegrative
or excisable reprogramming strategies will be needed to dissect
the reason for the apparent facility of hiPSCs to differentiate
into germ-like cells.
One of the most crucial points on the development of a germline cell is the establishment of its imprinting mark to a monoparental state. The imprinted genes are mostly involved in fetal and
placental development. To assess the state of imprinted genes
reprogramming to a monoparental state, we selected six imprinted
genes, two of which are normally expressed from the maternally
inherited allele and four are expressed from the paternally inherited one. We reasoned that, if our differentiation protocol was
directing the cells toward a sperm-like phenotype, the expression
Complete Meiosis from hiPSC
1194
level of imprinted genes after 10 weeks of differentiation should
have been similar to that of mature spermatozoa. We indeed found
a strong tendency in all six transcripts to sperm-like level of
expression at 10 weeks of differentiation, while the expression at
6 weeks was consistently different. We also assayed the expression level of the gene TERT and the gene controlling XIST. In
both cases, the gene expression level in hiPSCs after 10 weeks of
differentiation was comparable with spermatozoa, while it
remained higher at 6 weeks of differentiation (Supporting Information Fig. 7). We then moved on to assess the epigenetic reprogramming of differentiated cells by evaluating the DNA methylation status of the promoter of the maternally expressed, paternally
imprinted gene H19. Of six cell lines that we tested (one ESC and
five hiPSC), four increased the methylation status of H19 from an
average 48.5% (range 43–51), indicative of biparental imprinting,
to an average 75.8% (range 53–100), displaying a clear tendency
to monoparental (paternal) imprinting. One hiPSC line decreased
in H19 methylation level from 72% to 53%, thus displaying faulty
imprinting re-establishment. Finally, another hiPSC line displayed
a strong methylation mark (94%) since the beginning of culture
and did not change significantly. This last instance is analogous to
what was reported by Park et al. [18].
Overall, there is a strong tendency to reimprint in a
monoparental, paternal way for most of the assayed hESC
and hiPSC lines, but the process is not complete. Furthermore,
the presence of the CYP26 inhibitor, which should only push
meiosis completion, thus act at a stage of differentiation
which posterior to imprinting re-establishment, did not affect
imprinting re-establishment, in accordance to prediction.
males and females, there is an increasing need for patient specific gametes able to differentiate properly into a new individual. We show here, for the first time, complete and robust
meiotic competence of hiPSC-derived cells, and although further assays will be needed to assess the developmental ability
of meiotic cells, our work paves the way for the production
of in vitro gametes in the human species.
ACKNOWLEDGMENTS
We thank G. Tiscornia, A. Herrerias, and members of the laboratory for comments on the manuscript; J. M. Andres-Vaquero for
assistance with flow cytometry; B. Aran and M. Carrio for assistance with cell culture techniques; L. Mulero, C. Morera, C.
Pardo, and M. Marti for bioimaging assistance; T. Ventura and
L. Batlle for assistance with animal samples; and T. Pepe, B.
Sese, F. Gonzalez, E. Sleep, M. Salerno, C. Cortina, E. Miquel,
C. Gomez, and A. Lopez for technical assistance. This work was
supported by the Ministerio de Ciencia e Innovacion (MICINN),
Terapia Celular (TERCEL), Networking Center of Biomedical
Research in Bioengineering, Biomaterials and Nanomedicine
(CIBER), the Cellex and G. Harold and Leila Y. Mathers Charitable Foundations and Sanofi-Aventis. N.M. and E.G. were partially supported by the Juan de la Cierva and Sara Borrell
program, respectively.
DISCLOSURE
CONCLUSION
Infertility affects millions of individuals each year, and with
the current tendency to postpone childbearing age in both
REFERENCES
1
2
3
4
5
6
7
8
9
10
11
12
13
McLaren A. Primordial germ cells in the mouse. Dev Bio 2003;262:
1–15.
De Felici M, Scaldaferri ML, Lobascio M et al. Experimental
approaches to the study of primordial germ cells lineage and proliferation. Hum Reprod Update 2004;10:197–206.
Monk M. Genomic imprinting. Memories of the mother and father.
Nature 1987;328:203–204.
Hajkova P, Erhardt S, Lane N et al. Epigenetic reprogramming in
mouse primordial germ cells. Mech Dev 2002;117:15–23.
Lee J, Inoue K, Ono R et al. Erasing genomic imprinting memory in
mouse clone embryos produced from day 11.5 primordial germ cells.
Development 2002;129:1807–1817.
Hull MG, Glazener CM, Kelly MJ et al. Population study of
causes, treatment, and outcome of infertility. Br Med J 1985;291:1693–1697.
Hübner K, Fuhrmann G, Christenson LK et al. Derivation of oocytes
from mouse embryonic stem cells. Science 2003;300:1251–1256.
Toyooka Y, Tsunekawa N, Akasu R et al. Embryonic stem cells can
form germ cells in vitro. Proc Natl Acad Sci USA 2003;100:
11457–11462.
Geijsen N, Horoschak M, Kim K et al. Derivation of embryonic germ
cells and male gametes from embryonic stem cells. Nature 2004;427:
148–154.
Qing T, Shi Y, Qin H et al. Induction of oocyte-like cells from mouse
embryonic stem cells by co-culture with ovarian granulosa cells. Differentiation 2007;75:902–911.
Nagano MC. In vitro gamete derivation from pluripotent stem cells:
Progress and perspective. Biol Reprod 2007;76:546–551.
Clark AT, Bodnar MS, Fox M et al. Spontaneous differentiation of
germ cells from human embryonic stem cells in vitro. Hum Mol
Genet 2004;13:727–739.
West FD, Machacek DW, Boyd NL et al. Enrichment and differentiation of human germ-like cells mediated by feeder cells and
OF
OF
POTENTIAL CONFLICTS
INTEREST
The authors indicate no potential conflicts of interest.
14
15
16
17
18
19
20
21
22
23
24
basic fibroblast growth factor signalling. Stem Cells 2008;26:
2768–2776.
Tilgner K, Atkinson SP, Golebiewska A et al. Isolation of primordial
germ cells from differentiating human embryonic stem cells. Stem
Cells 2008;26:3075–3085.
Bucay N, Yebra N, Cirulli V et al. A novel approach for the derivation of putative primordial germ cells and Sertoli cells from human
embryonic stem cells. Stem Cells 2009;27:68–67.
Tilgner K, Atkinson SP, Yung S et al. Expression of GFP under the control of the RNA helicase VASA permits fluorescence-activated cell sorting isolation of human primordial germ cells. Stem Cells 2010;28:84–92.
Kee K, Angeles VT, Flores M et al. Human DAZL, Daz and Boule
genes modulate primordial germ-cell and haploid gamete formation.
Nature 2009;462:222–225.
Park TS, Galic Z, Conway AE et al. Derivation of primordial germ
cells from human embryonic and induced pluripotent stem cells is significantly improved by coculture with human fetal gonadal cells. Stem
Cells 2009;27:783–795.
Blanco J, Egozcue J, Vidal F. Meiotic behaviour of the sex chromosomes in three patients with sex chromosome anomalies (47,XXY,
mosaic 46,XY/47, XXY and 47/XYY) assessed by fluorescence in-situ
hybridization. Hum Reprod 2001;16:887–892.
Giorgetti A, Montserrat N, Aasen T et al. Generation of induced pluripotent stem cells from human cord blood using OCT4 and SOX2.
Cell Stem Cell 2009;5:353–357.
Aasen T, Raya A, Barrero MJ et al. Efficient and rapid generation of
induced pluripotent stem cells from human keratinocytes. Nat Biotechnol 2008;26:1276–1284.
Trautmann E, Guerquin MJ, Duquenne C et al. Retinoic acid prevents
germ cell mitotic arrest in mouse fetal testes. Cell Cycle 2008;7:656–664.
Pellegrini M, Filiponni D, Gori M et al. ATRA and KL promote differentiation toward the meiotic program of male germ cells. Cell
Cycle 2008;7:3878–3888.
Koubova J, Menke DB, Zhou Q et al. Retinoic acid regulates sex-specific timing of meiotic initiation in mice. Proc Natl Acad Sci USA
2006;103:2474–2479.
Eguizabal, Montserrat, Vassena et al.
25 Vogl AW, Vaid KS, Guttman JA. The sertoli cell cytoskeleton. Adv
Exp Med Biol 2008;636:186–211.
26 Lobo MV, Arenas MI, Alonso FJ et al. Nestin, a neuroectodermal
stem cell marker molecule, is expressed in Leydig cells of the human
testis and in some specific cell types from human testicular tumors.
Cell Tissue Res 2004;316:369–376.
27 Shinohara T, Avarbock MR, Brinster RL. b1- and a6-integrin are surface markers on mouse spermatogonial stem cells. Proc Natl Acad Sci
USA 1999;96:5504–5509.
28 Maki CB, Pacchiarotti J, Ramos T et al. Phenotypic and molecular
characterization of spermatogonial stem cells in adult primate testes.
Hum Reprod 2009;24:1480–1491.
29 Kanatsu-Shinohara M, Toyokuni S, Shinohara T. CD9 is a surface marker
on mouse and rat male germline stem cells. Biol Reprod 2004;70:70–75.
30 Lee JH, Engel W, Nayernia K. Stem cell protein Piwil 2 modulates
expression of murine spermatogonial stem cell expressed genes. Mol
Reprod Dev 2006;73:173–179.
31 Conrad S, Renninger M, Hennenlotter J et al. Generation of pluripotent
stem cells from adult human testis. Nature 2008;456:344–349.
32 He Z, Kokkinaki M, Jiang J et al. Isolation, characterization, and culture of human spermatogonia. Biol Reprod 2010;82:363–372.
33 Fisher RA, Nucci MR, Thaker HM, Complete hydatidiform mole
retaining a chromosome 11 of maternal origin: Molecular genetic
analysis of a case. Mod Pathol 2004;17:1155–1160.
34 Kobayashi S, Kohda T, Miyoshi N et al. Human PEG1/MEST,
an imprinted gene on chromosome 7. Hum Mol Genet 1997;6:781–786.
35 Giannoukakis N, Deal C, Paquette J et al. Parental genomic imprinting
of the human IGF2 gene. Nat Genet 1993;4:98–101.
36 John RM, Aparicio SA, Ainscough JF et al. Imprinted expression of
neuronatin from modified BAC transgenes reveals regulation by distinct and distant enhancers. Dev Biol 2001;236:387–399.
37 Leff SE, Brannan CI, Reed ML et al. Maternal imprinting of the
mouse Snrpn gene and conserved linkage homology with the human
Prader-Willi syndrome region. Nat Genet 1992;2:259–264.
38 Draper JS, Pigott C, Thomson JA et al. Surface antigens of human
embryonic stem cells: Changes upon differentiation in culture. J Anat
2002;200:249–258.
1195
39 Kanatsu-Shinohara M, Inoue K, Ogonuki N et al. Leukemia inhibitor
factor enhances formation of germ cell colonies in neonatal mouse
testis culture. Biol Reprod 2007;76:55–62.
40 Eguizabal C, Boyano MD, Diez-Torre A et al. Interleukin-2 induces
the proliferation of mouse primordial germ cells in vitro. Int J Dev
Biol 2007;51:731–738.
41 Byskov AG, Fenger M, Westergaard L et al. Forskolin and the meiosis inducing substance synergistically initiate meiosis in fetal male
germ cells. Mol Reprod Dev 1993;34:47–52.
42 de Rooij DG. Regulation of spermatogonial stem cell behavior in vivo
and in vitro. Anim Reprod 2006;2:130–134.
43 Steger K, Tetens F, Jürgen S et al. Localization of fibroblast
growth factor 2 (FGF-2) protein and the receptors FGFR1–4 in normal human seminiferous epithelium. Histochem Cell Biol 1998;110:
57–62.
44 Ocaya P, Gidlöf AT, Olofsson PS et al. CYP26 inhibitor R115866
increases retinoid signalling in intimal smooth muscle cells. Arterioscler Thromb Vasc Biol 2007;27:1542–1548.
45 Pavez-Loriè E, Li H, Vahlquist A et al. The involvement of
cytochrome p450 (CYP) 26 in the retinoic acid metabolism of
human epidermal keratinocytes. Biochim Biophys Acta 2009;1791:
220–228.
46 Amann RP. The cycle of the seminiferous epithelium in humans: A
need to revisit? J Androl 2008;29:469–487.
47 Panula S, Medrano JV, Kee K et al. Human germ cell differentiation
from fetal- and adult-derived induced pluripotent stem cells. Hum Mol
Gen 2011;20:752–762.
48 Stadtfeld M, Apostolou E, Akutsu H et al. Aberrant silencing of
imprinted genes on chromosome 12qF1 in mouse induced pluripotent
stem cells. Nature 2010;465:175–181.
49 Chen B, Wang YB, Zhang ZL et al. Xeno-free culture of
human spermatogonial stem cells supported by human embryonic
stem cell-derived fibroblast-like cells. Asian J Androl 2009;11:
557–565.
50 Li J, Quirt J, Do HQ et al. Expression of c-kit receptor tyrosine kinase
and effect on beta-cell development in the human fetal pancreas. Am
J Physiol Endocrinol Metab 2007;293:475–483.
See www.StemCells.com for supporting information available online.