Cell lines and animals

G4 mouse Embryonic stem cells (mESC) were pruchased from the Lunenfeld-Tanenbaum Research Institute [28], mouse fibroblast induced pluripotent stem cell line (FiPSC) was obtained from System Biosciences [29] and mouse granulosa cell derived-iPSCs were generated in our laboratory. All female C57BL6/J mice were purchased from Charles River Laboratories (Wilmington, MA).

Collection of granulosa cells

Generation of retrovirus vectors and iPSCs

Mouse and human retroviral reprogramming vectors for the iPSC genes Oct4, Sox2, c-Myc or Klf4 were generated using standard retroviral production protocols using pMXs retroviral plasmids [30] (Addgene, Cambridge, MA). 293T cells were cultured in DMEM and 10% HI FBS until 40–50% confluent, and then transfected with the reprogramming vectors and VSV-G using FuGENE (Roche, Indianapolis, IN) for 48 hours. Virion particles were then harvested. The mouse and human retroviral vectors were generated using ecotropic and amphotropic virus, respectively. Fresh viral vectors are then employed to infect target GCs primary cultures along with the 6 μg/ml polybrene (Millipore, Billerica, MA). Mouse GCs tend to not proliferate well in culture, so primary cells were infected. Using the Yamanaka retroviral technique, which is replication-dependent, resulted in low infection efficiency. Efficiency was comparable to other published attempts at mouse iPSC generation, at 0.001–0.01% [2, 31, 32]. This variability in derivation efficiency was largely related to the poor in vitro proliferation of GC cells prior to reprogramming. After 24 hours, the viral media was rinsed and cultures observed for 2 weeks for emergence of stem cell-like colonies. Putative iPSC colonies were manually picked based on morphology and subcultured on feeder layers of mouse embryonic fibroblasts [MEFs] cells [Global Stem, Rockville, MD] that are mitotically-inactivated with mitomycin C. All stem cells cultures were grown in standard mouse or human stem cell media. Mouse ES media was comprised of DMEM, 10% ES-grade HI FBS, 1000 U/ml ESCRO LIF (Millipore), 2 mM L-Glutamine [GIBCO], 0.2 mM 2-mercaptoethanol. Human ES media consisted of DMEM-F12, 20% Knock Out Serum Replacement (KOSR, GIBCO), 1 mM L-Glutamine, 0.1 mM 2-mercaptoethanol, 8 ng/ml bFGF (Invitrogen, Grand Island, NY).

mGriPSC and hGriPSC colony expansion

Primitive colonies of iPSCs were grown on MEFs for several days and manually picked based on stem cell colony morphology and passaged onto fresh feeder plates. Further subculture was based on live-immunostaining screening [33], in which live colonies were immunostained while in culture for an external antigen that is characteristic of undifferentiated stem cells. Mouse embryonic stem cell marker stage-specific embryonic antigen-1 (SSEA-1, Millipore) was used for mGriPSC colonies (Fig. 1E-H) and human embryonic stem cell markers SSEA-4 and TRA-1–60 (Millipore) were used for hGriPSC colonies (S2 Fig.). Primary antibody was used at a 1:100 dilution in stem cell media at 37°C overnight followed by incubation in a 1:1000 dilution of secondary rhodamine- or fluoroscein-tagged antibodies in fresh stem cell media for a minimum of 2 hours at 37°C. Positively immunolabeled live colonies were then rinsed of antibody and expanded for further characterization.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. mGriPSCs are pluripotent.

Using G4-mESCs as a standard (A-D, I), mGriPSCs are alkaline phosphatase reactive (E) and express stem cell antigens Oct4 (F), SSEA1 (G) and Nanog (H, I). mGriPSCS express additional stem cell markers by RT-PCR (I) and are karyotypically normal (J). mGriPSC express endogenous copies of the introduced reprogramming genes and retroviral [trans] copies are present to varying degrees (K). mGriPSCs form EBs (L) that differentiate into three germ layers—ectodermal neurofilament (M), and mesodermal SMA (N), and endoderm alpha-fetoprotein (O, AFP). Teratomas (P) also show differentiation into three germ layers (Q-S). Scale bars: A-H 20 μm; L 200 μm; M-O, Q-S 10 μm.

https://doi.org/10.1371/journal.pone.0119275.g001

Verification of mGriPSC and hGriPSC stem cell status

Cultures of manually selected, presumptive mouse and human GriPSC colonies were fixed using 4% paraformaldehyde (4°C)/4% sucrose in 0.1 N PBS for 15-minutes at room temperature and rinsed twice with 0.1N PBS for 5-minutes. An alkaline phosphatase reaction kit [Millipore] was used to evaluate enzyme activity in the iPSC colonies. Stem cell status was also assessed by immunocytochemistry (ICC) using commercial stem cell antibodies for mouse Oct4, SSEA-1, and Nanog, and for human OCT4, SSEA-4, TRA-1–60 and TRA-1–81 (Millipore; S2 Fig.; S1 Table). Differentiation was analyzed by ICC using antibodies to neurofilament [NF, ectoderm], alpha fetoprotein (AFP, endoderm; Santa Cruz, Santa Cruz, CA), and alpha smooth muscle actin (SMA, mesoderm; S1 Table). Controls included immunostaining for stem cell antigens on primary GCs to assess for expression of these genes prior to infection (S3 Fig.). Additional controls included omission of primary or secondary antibodies.

RNA was extracted using commercially available kits (Qiagen, Germantown, MD). Reverse transcription of cDNA was performed using qScript cDNA Synthesis (Quanta Biosciences, Gaithersburg, MD) and PCR was performed to confirm expression of additional stem cell markers. In the mGriPSCs, expression of Oct4, Nanog, SSEA1, Gdf3, and Dnmt3b were investigated using GoTaq Flexi polymerase (Promega, Madison, WI). For the hGriPSC line, the expression of OCT4, NANOG, DNMT3B, and GDF9 (S2 Table) were tested.

Absence of chromosomal abnormalities was determined by the Dana Farber/Harvard Cancer Center Cytogenetics Core facility (Boston, MA) in ten metaphase cells from mGriPSC colonies (Fig. 1J).

This process generated a total of 12 positively identified and characterized mGriPSC clones.

PCR analysis of reprogrammed iPSC line

Endogenous (native cellular), transgene (reprogramming vector), and total (capturing both) primers for all four stem cell genes that were introduced to the cells were used to determine the origin of stem cell gene expression (Fig. 2; S2 Table). These primers were used to characterize our mGriPSC line, G4 mESCs and control MEF cells. RT-PCR was performed using the kits and protocols described above, with the addition of 5% DMSO to the PCR reactions.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Methylation analysis of GC-derived iPS cells shows stem cell reprogramming with retention of target tissue methylation.

The stem cell gene Oct4 in mGriPSC is less methylated than it is in GCs (mGriPSC vs. GC, P<0.001) and is therefore more similar to Oct4 in G4 ESCs (A). Methylation patterns of ovarian and germ cell genes Foxl2 and Figla are comparable in both mGriPSC and GCs (B, Foxl2, P = 0.84; Figla, P = 0.606).

https://doi.org/10.1371/journal.pone.0119275.g002

Teratoma formation

Low-passage mGriPSCs cultured on irradiated MEFs [Global Stem] were detached using with collagenase IV (GIBCO) (1 mg/ml collagenase IV dissolved in 37°C DMEM media, filtered through 0.22 micron syringe filter) until detachment of the edges of the mGriPS colonies was detected. These colonies were harvested and dissociated with Accutase (Invitrogen) for 2–3 minutes to produce a single-cell suspension. Approximately 5–9 million mGriPSC cells in 100 μl DMEM media (pre-warmed at 37°C) were injected into the rear leg muscles of five-week-old severe combined immunodeficient (SCID) mice. A total of six mice were injected. Mice were sacrificed 7–8 weeks after injection and teratoma tissue was excised, fixed in cold 4% paraformaldehyde/4% sucrose, and processed for paraffin embedding. Sectioned slides containing various regions of tumors were stained by hematoxylin and eosin (H&E). Complex structures with various cell types were analyzed. Teratoma sections were also immunostained for differentiation markers (listed above).

This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Harvard Medical Area Standing Committee on Animals (Protocol: 750). All efforts were made to minimize suffering.

Embryoid body formation and endocrine assays

EBs were generated from mGriPSC (Fig. 3A) mouse G4 mESC [28], mouse fibroblast-derived iPSCs [mFiPSC] [29], and hGriPSC (S2 Fig.) colonies on MEFs using the following protocol. Cultures were rinsed using Ca²⁺- and mG²⁺-free PBS and then treated with pre-warmed collagenase IV and Accutase as described above. To initiate differentiation of EBs, the mES media was replaced with EB media, consisting of DMEM-F12, 15% KOSR, 15% HI FBS, 1 mM L-glutamine, 0.1 mM 2-mercaptoethanol, and 1% Non-essential amino acids (NEAA, Invitrogen), 1% antibiotic-antimycotic solution (Invitrogen). EB media was used to resuspend the dissociated mGriPSC into a single-cell suspension. To minimize the presence of MEFs during EB formation, the cell suspension was placed to a gelatin-coated 100 mm culture plate in a 37°C, 5% CO2 incubator for 15–20 minutes. MEF cells typically adhere to gelatin faster than ESCs and iPSCs, thus the supernatant that is collected after 20 min contains fewer MEFs. This process was performed 2 times. Cell concentration was adjusted to approximately 6–10x10⁵ cells/ml using EB media. The suspension was then plated into ultra low-attachment 6-well plates (Corning, Tewksbury, MA) or 100mm plates coated with 2% poly-HEMA in ethanol which also achieves low-attachment. EBs were grown in suspension for 14 days, and subsequently attached to gelatin-coated plates for up to 30 days of additional culture. Half of the media in each well was collected and replaced daily with fresh EB media. This experiment was performed three times. Collected media was assayed for estradiol and progesterone by radioimmunoassay (RIA) at the Wisconsin National Primate Research Center (WNPRC) at the University of Wisconsin-Madison.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. mGriPSC-EBs preferentially synthesize estradiol.

Mouse GCs (A) infected with pluripotency genes generate iPSC colonies. Arrow indicates primary granulosa cells. mGriPSC-EBs, when allowed to attach, synthesize estradiol (B) at concentrations 5 to 12 times those of G4-EBs and mFiPSC-EBs. Two-tailed t-test, *P<0.001, different from other two cell lines at same time point. G4-EBs primarily secreted progesterone (C) while mFiPSC-EBs produced negligible amounts of either. Two-tailed t-test, *P<0.005, different from other two cell lines at same time point. Scale bar: A 100 μm.

https://doi.org/10.1371/journal.pone.0119275.g003

Ovarian tissue markers

Markers of ovarian tissue were employed to characterize differentiation of mGriPSC. EBs from all three cell lines were also dissociated using pre-warmed 0.05% Trypsin-EDTA (GIBCO) and reattached to gelatin-coated plates as a cell monolayer for improved visualization of individual cells. After fixation, ICC for antigens characteristic of ovarian cell types, including anti-Müllerian hormone and receptor (AMH/R), follicle-stimulating hormone receptor (FSHR), aromatase (Cyp19a1), inhibin β-A (Inha), and estrogen receptor (ER), was performed both in situ and in vitro for ovarian granulosa cells and differentiated stem cells. In the adult female reproductive system, AMH, a member of the TGF-beta superfamily, localizes to the granulosa cells of the ovary and is a marker of ovarian follicle reserve. Furthermore, AMH is also believed to play a role in antral follicle recruitment. In this study we use this as a marker for ovarian tissue specificity. FSHR and ER are expressed by postnatal ovarian granulosa cells and are also involved in follicle recruitment and growth. Similarly, Inha is expressed by ovarian granulosa cells and is associated with secondary follicles. One of the primary ovarian functions is steroidogenesis regulated by aromatases. Cyp19a1 is a primary aromatase that converts 19-carbon androgens into 18-carbon estrogens. Additional ovarian markers used in this study include Foxl2, an ovarian tissue commitment marker, and Figla, a marker of germ cell lineage.

Differentiated stem cels that were positively-immunolabeled for these ovarian markers were counted in 4 fields of view per antibody per cell line, and the percentage of cells expressing each of the ovarian antigens was calculated for dissociated mGriPSC-EBs, G4 mESC-EBs, and mFiPSC-EBs. Undifferentiated cell colonies from all three cell lines were also stained for these ovarian antigens as a negative control (S4 Fig.).

Statistical analyses

All endocirne data was analysed using a two-tailed T-test p<0.05 for significance. The microarray data were processed using the lumi package [34] for background subtraction, log2 transformation and quantile normalization. The R statistical environment was used to perform data analysis [http://www.r-project.org]. To filter out unresponsive probes, we removed probes that had detection P-values <0.01 (as determined by Illumina’s BeadStudio software) in less than or equal to one sample across all samples.

Mouse primordial oocyte differentiation analysis

mGriPSCs and G4 mESCs were used to generate EBs, as described above, in defined media formulated by our lab to enrich ovarian cell types during differentiation (oocyte derivation media: DMEM, 15% KOSR, 15% HI ES-grade FBS, 2mM L-Glutamine, 1% NNEA, 0.1 nM 2-Mercaptoethanol, 100 U/ml ESGRO LIF, 5 ng/ml insulin, 0.01 U/ml PMSG, 50 μg/ml sodium pyruvate, 5 μg/ml Transferrin, 10 μg/ml Selenium, 50 ng/ml BMP-4, 5 μM Retinoic Acid, 5 μM Forskolin). After 2 weeks in suspension, EBs were dissociated and reattached to gelatin-coated plates, as described above, for an additional 10 days of culture as a monolayer of cells in oocyte derivation media. After fixation, ICC for mouse oocyte markers was performed with antibodies to Gdf9, Dazl, Boule, Mvh, and Zp1 (Santa Cruz; S1 Table). Positively-immunolabeled, spherical-shaped cells were counted in 7 fields of view and the percentage of cells expressing each of the oocyte antigens in both the dissociated mGriPSC-EBs and G4-EBs was calculated.

Microarray mRNA expression analysis

Microarrays were run using the Illumina WG-6 v2.0 kit. cDNA was amplified from RNA extracted from cell lines or tissues as described above and run in triplicate. Our analyses compared suspended and attached mGriPSC- and mESC-EBs with mESC colonies, postnatal day 2 (P2) mouse ovaries, and isolated adult mouse GCs. The R statistical environment was used to perform data analysis (http://www.r-project.org). The microarray data were processed using the lumi package [34] for background subtraction, log2 transformation and quantile normalization. To filter out unresponsive probes, we removed probes that had detection P-values < 0.01 (as determined by Illumina’s BeadStudio software) in less than or equal to one sample across all samples (18 samples in total). The remaining 25,294 probes were used for analysis. If multiple probes represent a single gene, the probe with the highest median expression across all samples was chosen to represent the expression of that gene. The final dataset contains 18,029 genes. Heat maps were generated to illustrate hierarchal clustering of genome-wide expression as well as genes in the steroidogenic biosynthesis pathways. The set of core steroid biosynthesis enzymes was taken from a review paper [35]. Principal component analysis (PCA) was performed on the gene expression dataset using R’s prcomp function (S1 Materials). Differential expression analysis was performed using limma [36]. These were then evaluated to explore common differentially regulated genes (DRGs) and relevant gene regulatory networks (GRNs). Differential gene expression relevant to gonadogenesis and steroidogenesis was further analyzed using Ingenuity Pathways Analysis (IPA) software (Ingenuity Systems, Redwood City, CA).

Genome-wide profiling of microRNA using next generation sequencing

Bar-coded multiplexed cDNA library preparation was performed as previously described [37]. For each sample, barcoded 3’ adapters were ligated to 2 μg of total RNA. Size fractionation during adapter ligations using radiolabeled 19- and 35-nucleotide markers were used to gel purify the miRNA fraction. Sequencing of amplified cDNA libraries was performed using 50 cycles of single-end sequencing by synthesis (Illumina HiSeq 2000). Barcodes were extracted and reads were aligned to the genome (hg19) and annotated using an in-house pipeline [37, 38]. The sum of all reads with fewer than 2 annotation mismatches was used to determine miRNA abundance, as previously described [39].

Mouse ovarian GCs are readily reprogrammed into ovarian tissue-derived iPSCs

G4 mESCs were used as a standard for characteristic alkaline phosphatase reactivity and stem cell antibody ICC (Fig. 1A-D). The stem cell identity of the mGriPSC line as was verified by confirming positive alkaline phosphatase reactivity (Alk Phos, Fig. 1E) and expression of stem cell markers Oct4, SSEA-1, and Nanog (Fig. 1F-H). Colony morphology was comparable to mESCs. RT-PCR of ICC-confirmed Oct4, Nanog, and SSEA-1, as well as additional stem cell markers Gdf3 and Dnmt3b, was demonstrated in G4 mESCs and further verified pluripotency of the mGriPSCs (Fig. 1I). We confirmed that the mGriPSC line was consistently karyotypically normal by chromosome analysis, with 10 metaphase cells possessing 40 normal chromosomes (Fig. 1J). Additional culture of mGriPSC-EBs (Fig. 1L) in differentiation conditions resulted in generation of cell types from all three developmental germ layers as demonstrated by neurofilament (NF, ectoderm, Fig. 1M), smooth muscle actin (SMA, mesoderm, Fig. 1N) and alpha fetoprotein (αFP, endoderm, Fig. 1O) immunoreactivity. 100% of the six injected SCID mice formed a teratoma (Fig. 1P), which displayed differentiation into cells types of all germ layers [Fig. 1Q-S]. Collectively, these cytological characterization and functional differentiation studies confirm the successful generation of a novel ovarian tissue iPSC line.

Characterization of endogenous stem cell gene expression in mGriPSCs

We interrogated activation of stem cell loci for Oct4, Sox2, cMyc and Klf4 in retrovirally reprogrammed mGriPSC (Fig. 1K). Our results demonstrate endogenous activation of all four genes reflecting successful reprogramming of our derived iPSC cells. PCR analysis with primers for the total transcripts [endogenous and transgene] confirmed presence of these genes in our iPS line, while transgene expression demonstrated various levels of silencing.

Human GCs are also reprogrammable to generate hGriPSCs

Employing the same retroviral infection technique, we derived an analogous human granulosa cell-derived iPSC line and generated hGriPSC-EBs (S2A-D Fig.). The hGriPSC cultures morphologically resemble hESC colonies in that colonies are larger in diameter and flatter than mGriPSC colonies. Furthermore, they demonstrate expression of known human stem cell antigens OCT4, SSEA4, TRA-1–81, and NANOG by ICC analysis (S2E-H Fig.) and alkaline phosphatase reactivity (S2I Fig.). RT-PCR re-confirmed the expression of OCT4 and NANOG, as well as other stem cell markers DNMT3B and GDF3 (S2J Fig.). These results reliably verify the stem cell status of our human GC-derived iPSC line.

Methylation analysis supports reprogramming and retention of GC signature

The methylation status of CpG islands #25–29 in the Oct4 gene was analyzed in mGriPSCs, G4 mESCs, and GCs (Fig. 2A). In mGriPSC, methylation of the CpG islands ranged from 32.7–48.8%, in mESCS 19.4–31.5%, and in GCs 61.4–85.7%. Average methylation was different in all three cells lines in the Oct4 gene (P≤0.001).

Methylation of the Foxl2 ovarian gene and Figla germ cell gene were not different in the pluripotent mGriPSCs and GC starting tissue (Fig. 2B-C; Foxl2 P = 0.84; Figla P = 0.606). Methylation of CpG islands #49–55 ranged from 0–5.2% in mGriPSC and 0–4.0% in GCs. Methylation of CpG islands #2–12 and #20–21 in the Figla gene ranged from 14.3–91.9% in mGripSC and 14.0–93.5% in GCs.

mGriPSC EBs demonstrate preferential differentiation into estrogenic cell types

We compared steroid hormone production by EBs formed from mGriPSCs, mouse fibroblast-derived iPSCs (mFiPSCs), and G4 mESCs under identical culture conditions (mGriPSC: Fig. 3A). After 11 days, we detected estradiol (E2) at 344.1 pg/ml in the mGriPSC-EB conditioned media compared to 19.9 pg/ml and 28.1 pg/ml in the media of G4-EBs and mFiPSC-EBs, respectively. At the three assayed time points during the experiments, differentiating mGriPSC-EBs synthesized E2 at concentrations ranging from 5.3 to 17.3 times higher than those of the G4- or FiPSC-EBs (Fig. 3B). By contrast, G4-EBs produced primarily progesterone (P4), at a concentration of 1.95 ng/ml by day 11 in culture. mFiPSC-EBs made negligible E2 and P4 (Fig. 3C). Baseline RIA assay detected that “control,” unconditioned media contained 20.1 pg/ml of E2 and 0.04 ng/ml of P4. Our results not only demonstrate a significant disparity in the concentrations of the hormones produced, but also in the type of steroidogenic cells present in these cultures (estrogenic versus progestogenic).

Differentiated mGriPSCS express ovarian and primitive germ cell antigens

Distinct subpopulations of differentiated cells within the mGriPSC-EBs cultures expressed the ovarian markers AMHR, FSHR, Cyp19a1, ER, and Inhibin β-A (Inha; Fig. 4A-E). The average percent of positively-immunolabeled cells was calculated for EBs generated from the same three cell lines (mGriPSCs, FiPSCs, and G4 mESCs) cultured under identical conditions. AMHR, Cyp19a1, ER, and Inha were expressed in a larger proportion of cells in the mGriPSC-EB monolayered cultures compared to the other cell lines (Fig. 4G).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. mGriPSC-EBs differentiate to express ovarian and primordial oocyte markers.

Differentiating mGriPSC-EBs express ovarian markers AMHR (A), FSHR (B), Cyp19a (C), Inha (D), and ER (E) (phase contrast; F) at higher frequencies than G4- and mFiPSC-EB (G). Expression of germ cell markers Mvh (H), Dazl (I), Gdf9 (J), Boule (K), and Zp1 (L) (phase contrast; M) are also increased in mGriPSC-EBs (N). Two-tailed t-test, *P < 0.05. Scale bars: A-F 10 μm; H-M 20 μm.

https://doi.org/10.1371/journal.pone.0119275.g004

An increased cell population of cells within the differentiating mGriPSC-EBs also differentiated into cell types that expressed the primordial oocyte and germ cell antigens Mvh, Dazl, Gdf9, Boule, and Zp1 (Fig. 4H-L) compared to G4 mESC-EBs when cultured in identical, oocyte-enriching conditions. Compared to G4 mESC-EBs, the efficiency of primitive germ cell generation from mGriPSC was markedly improved, as evidenced by the 2.8-, 1.8-, and 1.9-fold fold expression of Dazl, Gdf9, and Boule, respectively (Fig. 4N). These results, along with the steroidogenic functionality of cultured mGriPSC-EBs, provide further support for the preferential homotypic differentiation of mGriPSCs into ovarian tissue.

Genome-wide profiling of mRNA and microRNA supports the stem cell status of the mGriPSCs and hGriPSCs

Microarray analysis of markers of ovarian tissue and gametogenesis.

We investigated the gene expression patterns that coincide with the initiation of steroidogenesis, a primary function of the mammalian ovary, in our estrogen-producing mGriPSC-EBs and in the adult mouse ovary compared to suspended mGriPSC-EBs, suspended mESC-EBs, and the immature, postnatal day 2 (P2) mouse ovary (Fig. 5A-C, S5F-K Fig.). We computed genome-wide Pearson correlation coefficient of every pair of samples in the microarray dataset (Fig. 5A). We found two main clusters: [i] P2 and Adult ovary, and [ii] mES and EBs. This indicates that regardless of the source of EB (mESC or from iPSC); suspended or attached, their genome-wide expression profiles more closely resemble that of mESC colonies than that of postnatal primary tissue. We further analyzed the expression patterns of key genes that encode enzymes in the steroid hormone biosynthesis pathways. There is a clear increase in expression of these genes in adult ovary compared to P2, but the expression in other EB and ES cells are generally lower (Fig. 5B-C, S5H Fig.).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Genome-wide analysis of mRNA and microRNA confirms efficient pluripotent reprogramming of mGriPSCs and hGriPSCs.

mGriPSC-EBs and mESC-EBs have comparable microarray expression profiles (A). All ESC or EB samples show minimal steroidogenic mRNA compared to adult ovary samples (B). Principal component analysis (PCA) indicates attached and unattached mGriPSC-EBs are more disparate relative to the analogous mESC-EBs (C). microRNA heatmaps support efficient stem cell reprogramming of mGriPSCs (D).

https://doi.org/10.1371/journal.pone.0119275.g005

Principal Component Analysis (PCA) was performed to identify important axis of variability among the samples. PC1 and PC2 collectively account for 61% of the variability among the samples. PC1 separates primary tissues (including P2 and Adult ovary) from EB and mES cells; PC2 separates iPSC-derived EBs from mESC-derived EBs (Fig. 5C). The difference between suspended and attached iPSC-derived EBs is much larger than the analogous difference between suspended and attached ESC-derived EBs.

Additionally, microarray analysis confirms expression of ovarian genes as well as markers of gametogenesis (S5L-P Fig.) supporting our ICC results.