Published: January 11th, 2019
Primordial germ cells (PGCs) are common precursors of both sperm and eggs. Human embryonic PGCs are specified from pluripotent epiblast cells through interactions of cytokines. Here, we describe a 13-day protocol of inducing human cells transcriptomally resembling PGCs at the surface of embryoid bodies from primed-pluripotency induced pluripotent stem cells.
Primordial germ cells (PGCs) are common precursors of all germline cells. In mouse embryos, a founding population of ~40 PGCs are induced from pluripotent epiblast cells by orchestrated exposures to cytokines, including bone morphogenetic protein 4 (Bmp4). In human embryos, the earliest PGCs have been identified on the endodermal wall of yolk sac around the end of the 3rd week of gestation, but little is known about the process of human PGC specification and their early development. To circumvent the technical and ethical barriers of studying human embryonic PGCs, surrogate cell culture models have been recently generated from pluripotent stem cells. Here, we describe a 13-day protocol for robust production of human PGC-Like Cells (hPGCLCs). Human induced pluripotent stem cells (hiPSCs) maintained in the primed pluripotency state are incubated in the 4i naïve reprogramming medium for 48 hours, dissociated to single cells, and packed into microwells. Prolonged maintenance of hiPSCs in the naïve pluripotency state causes significant chromosomal aberrations and should be avoided. hiPSCs in the microwells are maintained for an additional 24 hours in the 4i medium to form embryoid bodies (EBs), which are then cultured in low-adherence plasticware under a rocking condition in the hPGCLC induction medium containing a high concentration of recombinant human BMP4. EBs are further cultured for up to 8 days in the rocking, non-adherent condition to obtain maximum yields of hPGCLCs. By immunohistochemistry, hPGCLCs are readily detected as cells strongly expressing OCT4 in almost all EBs exclusively on their surface. When EBs are enzymatically dissociated and subjected to FACS enrichment, hPGCLCs can be collected as CD38+ cells with up to 40-45% yield.
Primordial germ cells (PGCs) are common precursors of all germline cells in both sexes. Most of our knowledge on development of PGCs in mammalian embryos has been obtained through studying laboratory mice1,2. At embryonic day 6.0-6.5 of mouse embryos, 6 or similar small numbers of PGC precursors are located in the epiblast, and a founding population of ~40 PGCs are induced from them in a manner dependent on bone morphogenetic proteins Bmp2 and Bmp4 secreted from adjacent cells. The earliest human PGCs so far identified in embryos were on the endodermal wall of yolk sac at around the end of the third week of gestation3. Because this is the same place as migrating PGCs are observed in mouse embryos, it is likely that the observed human PGCs were in the path of migration but not the founding population. However, studies tracing back earlier stages of PGCs or PGC precursors in human embryos have been missing.
Access to human embryonic PGCs is challenging due to both technical and ethical obstacles. To overcome these hurdles, PGC-like cell culture models have been recently generated from human pluripotent stem cells (PSCs). Pluripotency is the cellular capability to differentiate into the germline and three embryonic germ layers4. Whereas human PSCs maintained in the mTeSR1 medium (a ready-to-use, commercially available medium formulated for maintenance of human PSCs in the primed pluripotency state) on dishes coated with the extracellular matrix protein have primed-state pluripotency4, in 2013 Jacob Hanna's lab showed that the primed pluripotency cells can be converted into a naïve pluripotency state by exposing to the naïve human stem cell medium (NHSM) containing chemical inhibitors to protein kinases ERK1/2, GSK3, JNK, ROCK, PKC, and p38 MAPK as well as growth factors LIF, TGF, bFGF5. From the naïve-pluripotency human PSCs, in 2015 a research group led by Hanna and Azim Surani accomplished the first robust production of human PGC-Like Cells (hPGCLCs) from PSCs6. Later, several other laboratories, including ours, reported generation of hPGCLCs from PSCs using slightly different protocols7,8,9,10. Our study provided evidence that hPGCLCs generated using different protocols (which are summarized in Table S1 of our previously published study10) are transcriptomally similar to each other10. Available evidence supports the resemblance of human PGCLCs to early-stage human embryonic PGCs prior to the global epigenetic erasure7 and/or chemotactic migration10.
Studies of mouse embryonic PGCs, mouse PGCLCs, and human PGCLCs (but with only very limited access to human PGCs) have revealed that molecular mechanisms of PGC specification differ significantly between mouse and human1,6,7,8,9,10,11,12,13,14,15,16,17. For example, Prdm14 plays critical roles in PGC specification in mouse embryos, but its role in human PGC specification seems limited1,15. In contrast, induction of SOX17 by EOMESODERMIN is essential for PGC specification6,11,14, whereas these transcription factors seem dispensable for mouse PGC specification15. These initial achievements of studies using hPGCLCs strongly support the importance of this cell culture model as a surrogate of human embryonic PGCs.
Recently published studies, involving our lab's deep sequencing evaluation of genomic DNA copy number analysis, have shown that prolonged maintenance of PSCs in the naïve pluripotency state significantly increases the risk of chromosomal instability and structural anomalies. This phenomenon was observed with both mouse18 and human19 PSCs. The original hPGCLC production protocol reported by Hanna/Surani was developed for human PSCs maintained in the 4i naïve pluripotency medium for at least 2 weeks6. To preserve the normal diploid karyotype of human PSCs and PGCLCs, we developed a modified protocol in which human PSCs are exposed to the 4i medium for only 72 hours10, which is presented in this article. Human iPSCs (hiPSCs) are maintained under the primed pluripotency state. Immediately before EB formation, cells are incubated in the 4i naïve reprogramming medium (a modified NHSM medium) for 48 hours. Cells are then dissociated and packed into microwells to form EBs for an additional 24 hours in the 4i medium. EBs are maintained in hPGCLC induction medium containing a high concentration of recombinant human BMP4 under a rocking condition for no-attachment culture for up to 8 days to obtain the maximum yield of hPGCLCs. After 8-day EB culture, hPGCLCs can be isolated from dissociated EB cells by FACS as CD38+ cells with up to ~40% yield in FACS-sortable single cell suspension. Whereas other published methods7,8,9, including the original protocol before our modifications6, typically generate hPGCLCs in spontaneously formed cell aggregates without specific localization, hPGCLC produced by our protocol are observed at the surface of embryoid bodies (EBs).
1. Cell Culture of hiPSCs in the Primed Pluripotency State
2. Generation of hPGCLCs
Note: Initiate the following steps when hiPSC cells in a 10 cm dish (mTeSR1 medium, extracellular matrix protein-coated) reaches to ~80% confluency (approximately 107 cells).
3. Immunohistochemical staining of EBs (Day 10 – Day 13)
4. FACS enrichment of hPGCLCs from EBs (Day 10 – Day 13)
The microwell plate used here is in the 24-well format and has 8 wells holding the microwell sheets, each of which supports formation of up to 1,200 EBs. From approximately 24 million of 4i naïve pluripotency cells, this microwell plate typically generates ~8,000 EBs consisting of ~3,000 cells per EB. During non-adherent culture of EBs with constant rocking, the number of intact EBs gradually decreases due to spontaneous self-dismantling, and ~3,000 EBs survive until Day 13 of the protocol. Most of these surviving EBs have 50-200 hPGCLCs on their surface (estimated by immunohistochemical detection of OCT4+ cells in serial sections of EBs; see Figure 8), yielding ~100,000 OCT4+ hPGCLCs in total. Enzymatic digestion of EBs into single cells is a relatively inefficient process, reducing the yield of FACS-enriched CD38+ hPGCLCs to 9,000 - 47,000 cells. In our hands, an average of six independent but consecutive batches of experiments was 14,038 ± 5,731 (mean ± SEM). Because CD38-negative EB cells cells express OCT4 mRNA (qPCR) or protein (Immunohistochemistry) only very weakly, if not completely absent, all EB cells strongly expressing OCT4 protein are practically equivalent to the whole population of the CD38+ hPGCLCs.
The critical parameter of this protocol includes cell density. When 2.0 x 105 human iPSCs are inoculated in an extracellular matrix protein-coated well of 6-well cell culture plate (9.60 cm2 growth area per well) with 2 mL Y27632-supplemented medium (2.2.9), cells will reach to about 20-30% confluency at 24 hours after inoculation (Figure 1). After an additional 24-hour culture in the mTeSR1 medium in the absence of Y7632, cells aggregate and form colonies, occupying ~30% of the growth area (Figure 2). Cells are then cultured in the 4i reprogramming medium (2.3.2). After 24 hours of culture in the 4i medium, cells become confluent (Figure 3). An additional 24-hour culture in the 4i medium makes cells densely packed (Figure 4). The exact timing of medium change (every 24 hours +/- 4 hours) and cell densities are critical for successful formation of EBs and hPGCLCs.
After 48-hour culture in the 4i medium, cells are dissociated and inoculated to the microwell plate, which has wells 400 µm in size (2.4). Although this commercially available microwell plate is coated for low-adhesion surface by the manufacturer, fresh re-coating with detergent (2.4.3) is recommended to reduce the risk of unwanted cell adhesion. 800 µm microwells resulted in reduced yield of hPGCLCs, suggesting the importance of EB size for properly directed differentiation. Cells inoculated in the 4i medium will form EBs in microwells at 24-30 hours, which can be observed using a standard, inverted phase contrast microscope (Figure 5). The circular contour of EBs become clearly visible at and after 24 hours of incubation. Harvesting EBs at an earlier time (e.g., 16 hours after inoculation) before their circular contour is clearly visible is not recommended because such EBs are very vulnerable to mechanistic damages and easily dismantle. Note that significant amounts of naïve hiPSCs are NOT incorporated in EBs, which is normal. These unincorporated cells will be washed away before initiation of rocking culture of EBs on low-adherent surface (2.5.2). Also note that, in our protocol, EBs are formed in the 4i naïve pluripotency medium - not in the hPGCLC medium. Pre-formation of solid EBs in the 4i medium before exposure to the hPGCLC medium is important for distribution of hPGCLCs on the surface of EBs.
EBs maintained in the hPGCLC medium under a rocking, non-adherent culture condition will maintain their spherical shape without aggregation or fusion (Figure 6 and Figure 7). Too weak rocking condition will cause EB aggregation and fusion, but too harsh condition will dismantle EBs. Human PGCLCs emerge as OCT4-expressing cells on the surface of EBs after as early as the 5-day culture in the hPGCLC medium, and their number increases until the 8-day culture (Figure 8). Further incubation of EBs may cause dismantling of EBs and loss of hPGCLCs. Human PGCLCs can be enriched from enzymatically dissociated EB cells after 5-8 days of culture in the hPGCLC medium by FACS as CD38+ cells (Figure 9).
Figure 1: Human iPSC cell culture in Y27632-supplemented medium 24 hours after inoculation. Cell density is about 20-30% confluent. In the presence of ROCK inhibitor Y27632, cells tend to spread well with long, spike-like elongation. Scale bar = 100 μm. Please click here to view a larger version of this figure.
Figure 2: Human iPSC cell culture 48 hours after inoculation. Cells aggregate to form colonies, occupying ~30% of growth area. Scale bar = 100 μm. Please click here to view a larger version of this figure.
Figure 3: Human iPSC cell culture incubated in the 4i reprogramming medium for 24 hours. Cells reach to confluency. Scale bar = 100 μm. Please click here to view a larger version of this figure.
Figure 4: Human iPSC cell culture incubated in the 4i reprogramming medium for 48 hours. Confluent cells are densely packed. Scale bar = 100 μm. Please click here to view a larger version of this figure.
Figure 5: Human iPSC EBs formed in microwells after 24-hour incubation in the 4i reprogramming medium. The circular contour of EBs become visible at 24-30 hours after inoculation. When the contour is confirmed under a phase contrast microscope, EBs are ready for transfer to a rocking culture condition. Scale bar = 500 μm. Please click here to view a larger version of this figure.
Figure 6: Human iPSC EBs incubated for 24 hours in the hPGCLC medium. EBs largely maintain their spherical shape. Scale bar = 500 μm. Please click here to view a larger version of this figure.
Figure 7: Human iPSC EBs incubated for 192 hours in the hPGCLC medium. EBs are enlarged compared to their appearance at 24-hour culture, but they still largely maintain spherical shapes with no aggregation or fusion. Scale bar = 500 μm. Please click here to view a larger version of this figure.
Figure 8: Human PGCLCs expressing OCT4 are localized on the surface of hiPSC EBs incubated for 192 hours in the hPGCLC medium. EBs were embedded in extracellular matrix protein and processed for FFPE slide immunohistochemical staining of OCT4 (DAB substrate). Scale bar = 1 mm. Please click here to view a larger version of this figure.
Figure 9: Human PGCLCs are enriched from enzymatically dissociated EB cells by FACS as CD38+ cells. After incubation in the hPGCLC medium for 5-8 days, EBs can be dissociated by enzymatic digestion to prepare single cell suspension. hPGCLCs can be enriched as CD38+ cells by FACS (red dots). EB cells that do not express CD38 (blue dots) should also be collected as negative control. FACS gates of CD38-positive and CD38-negative cells should be separated with a wide margin (green dots) to avoid contamination of each type of cells. The upper and lower panels show FACS profiles without or with anti-CD38 antibody staining, respectively. Please click here to view a larger version of this figure.
Robust production of hPGCLCs using the protocol described here was confirmed with three independent clones of human iPSCs with the normal diploid karyotype10. These iPSC clones were derived from the same human neonatal dermal skin fibroblast cell culture10. They will be provided by authors of this article to investigators upon request and under appropriate materials transfer agreement and shipping arrangement of frozen live human cells. It is presently unknown as to whether normal karyotype is required for robust hPGCLC production using our protocol or those reported by other laboratories.
Recent studies have shown that production of hPGCLCs from hiPSCs11 or ESCs14 using a protocol described by Saitou's group of Kyoto University8 is dependent on expression of EOMESODERMIN, a T-box transcription factor required for induction of SOX17. SOX17 seems to function as the master lineage determining transcription factor in germline differentiation of human pluripotent stem cells6. EOMESODERMIN is encoded by the EOMES gene, and CRISPR/Cas9 knockout of EOMES caused almost complete absence of SOX17 induction in the hPGCLC producing condition11, and expression of other genes followed the same pattern of the SOX17-null knockout cells. Overexpression of EOMESODERMIN from an inducible vector in the EOMES-null knockout cells during hPGCLC induction culture efficiently rescued the robust hPGCLC production as well as induction of germline genes, including SOX17. In contrast, induced overexpression SOX17 also rescued the robust PGCLC production but without inducing EOMES. Thus, EOMESODERMIN is a critical upstream inducer of SOX17, and this seems the single most important role of EOMESODERMIN in hPGCLC induction from human pluripotent stem cells. Our protocol induces SOX17 in hiPSCs10, but its dependency on EOMESODERMINE induction awaits to be determined.
This protocol converts primed-pluripotency human iPSCs maintained in the mTeSR1 medium to ERK-independent naïve pluripotency for 96 hours in the 4i reprogramming medium6, which is a modified naïve human stem cell medium (NHSM)5. Our attempts to generate hPGCLCs starting with the same human iPSC clones but maintained in other commercially available human iPSC growth media before culture in the 4i reprogramming medium resulted in varying degrees of lower hPGCLC yields. Although whether longer-term adaptation in other media improves hPGCLC production or not remains be determined in future studies, this observation suggests that the exact state of the primed pluripotency of human iPSCs before the 4i reprogramming significantly impacts EB formation in the 4i medium and EB differentiation in hPGCLC medium.
Production of hPGCLCs from hiPSCs following our protocol is robust and highly reproducible, partly owing to the use of the microwell plates that enable efficient production of a large number of EBs (~8,000 EBs per batch) with a uniform size (3,000 hiPSCs per EB). The number of EBs readily produced in a single batch of experiment using our protocol may be far greater than methods using the regular U-bottom cell culture wells. Production of a large number of equally sized EBs uniformly studded with hPGCLCs may provide unique opportunities of high-throughput chemical screenings to identify small-molecular-weight activators or inhibitors affecting PGC specification or their biological characteristics such as epigenetic reprogramming. Such EBs may also be useful for toxicological assessments of large numbers of germline cell toxicants, including not only environmental pollutants but also clinically prescribed medications such as chemotherapeutic agents.
The critical factors of the robust and reproducible production of hPGCLCs using the presented protocol include (i) the use of healthy hiPSCs maintained in mTeSR1 on extracellular matrix protein, (ii) to inoculate exact number of cells as specified and strictly follow the timings of medium change and subculture, (iii) to select a good lot of human recombinant BMP4, and (iv) to minimize physical damages of EBs during the rocking culture. It is our experience that the best lot of BMP4 reagent worked at 100 ng/mL concentration whereas other lot of BMP4 reagents required 2X or greater doses. On the other hand, yield of hPGCLCs production using the best lot of BMP4 rather decreased at higher doses of BMP4 (e.g., 200 ng/mL). We recommend to test several different lots of recombinant human BMP4 reagents obtained from multiple vendors for their performance in supporting hPGCLC generation and to secure a large amount of the best lot.
A unique feature of our hPGCLC protocol is that hPGCLCs are localized on the outermost surface layer of EBs10 (Figure 8), whereas other protocols can generate hPGCLCs in the middle of cell aggregates6,8. Embryoid bodies tend to form multiple distinct layers such as surface, outer shell, inner shell, and core, and the central core regions are often necrotic due to limited supply of nutrients, oxygen, as well as pro-surviving growth factors provided form the culture medium by diffusion20. Localization of hPGCLCs on the surface of EBs without possible restrictions due to limited diffusion towards the center of EBs may be beneficial for direct, time- and dose-controlled exposure of hPGCLCs to drugs or toxic substances for pharmacological or toxicological studies.
Whereas mouse PGCLCs show robust genome-wide DNA demethylation involving the imprinting control regions at least partly7,19,20, the degree of global gDNA demethylation in hPGCLCs seems weaker than mouse PGCLCS or PGCs6,7. Transcriptomal profiles suggest that hPGCLCs may resemble an earlier stage of embryonic PGCs than mouse PGCLCs10. It has been reported that prolonged culture of EBs under the condition of hPGCLC production caused an increased degree of gDNA demethylation7; however, whether an extended period of culture of hPGCLCs in EBs or as isolated cells can achieve more advanced stages of germline differentiation needs to be determined by future studies.
The authors have nothing to disclose.
We acknowledge Shiomi Yawata and Chie Owa for technical assistance during initial studies. This study was supported by NIEHS/NIH grants R01 ES023316 and R21ES024861 to TS, and by Flight Attendant Medical Research Institute (FAMRI) grant to JHH.
|Primed pluripotency hiPSC culture
|Aliquot Matrigel at the volume indicated by the manufacturer (about 200 μL). Use cold tubes. Store at -80 °C
|Thermo Fisher Scientific
|Store at 4 °C.
|Reconstitute by adding 5X Supplement to Basal Medium. The reconstituted medium can be stored at 4 °C for up to
4 weeks without affecting cell culture performance.
|Dilute 5 mg Y27632 (MW 320.26) with 312.2 μL water to prepare 50 mM Y27632 stock solution. Sterilize by filtration (0.22 μm). Aliquot 20 μL/tube x 15 tubes and store at -20 °C.
|Store at 4 °C.
|Innovative cell technologies
|Cell dissociation enzyme; aliquot 40 mL/tube x 12 tubes and store at -20 °C.
|4i hiPSC culture
|Thermo Fisher Scientific
|KnockOut Serum Replacement
|Thermo Fisher Scientific
|Aliquot 40 mL/tube x 12 tubes and store at -20 °C.
|Thermo Fisher Scientific
|MEM Non-Essential Amino Acids Solution (100x)
|Thermo Fisher Scientific
|Insulin from bovine pancreas
|Add 0.1 mL glacial acetic acid to 10 mL water. Sterilize by filtration (0.22 um). Dilute 100 mg insulin lyophilized powder with cold 10 mL the acidified water to make 10 mg/mL stock solution. Aliquot 650 μL/tube x 15 tubes and store at -80 °C.
|Reconstitute 250 μg human LIF with 250 μL water. Add 750 μL of 0.1% bovine serum albumin in PBS(-) to make 250 μg/mL human LIF stock solution. Specific activity of this product is >10,000 units/μg. Aliquot 40 ul/tube x 25 tubes and store at -80 °C.
|Reconstitute 1 mg human FGF2 with 5 mL PBS(-) to make 200 μg/mL human FGF2 stock solution. Aliquot 100 μL/tube x 50 tubes and store at -80 °C.
|Reconstitite 50 μg with 50 μL of 10 mM Citric Acid (pH 3.0). Add 150 μL of 0.1% bovine serum albumin in PBS(-) to make 250 μg/mL human TGF-β1 stock solution. Aliquot 4 μL/tube x 50 tubes and store at -80 °C.
|4i basal medium
|Mix the following reagents to make 4i basal medium.
500 mL of KnockOut DMEM
100 mL of KnockOut Serum Replacement
6 mL of Penicillin-Streptomycin-Glutamine (100x)
6 mL of MEM Non-Essential Amino Acids Solution (100x)
650 µL of 10 mg/mL Insulin from bovine pancreas
40 µL of 250 µg/mL human LIF
20 µl of 200 µg/ml human FGF2
4 µl of 250 µg/ml human TGF-β1
Aliquots 40 ml/tube and store at -80 °C.
|Reconstitute 25 mg CHIR99021 (MW 465.34) with 1791 μL DMSO to make 30 mM CHIR99021 stock solution. Aliquot 50 μL/tube x 35 tubes and store at -20 °C.
|Reconstitute 5 mg PD0325901 (MW 482.19) with 1037 μL DMSO to make 10 mM PD0325901 stock solution. Aliquot 50 μL/tube x 20 tubes and store at -20 °C.
|Reconstitute 10 mg BIRB796 (MW 527.66) with 948 μL DMSO to make 20 mM BIRB796 stock solution. Aliquot 50 μL/tube x 18 tubes and store at -20 °C.
|Reconstitute 10 mg SP600125 (MW220.23) with 908 μL DMSO to make 50 mM SP600125 stock solution. Aliquot 50 μL/tube x 18 tubes and store at -20 °C.
|Reconstitute 5 g Pluronic F-127 in 100 mL water to make 5%(w/v) Pluronic F127 stock solution. Sterilize by filtration (0.22 um). Aliqout 50 mL/tube x 2 tubes and store at r.t.
|Thermo Fisher Scientific
|Sodium pyruvate (100 mM)
|Thermo Fisher Scientific
|hPGCLC basal medium
|Mix the following reagents to make hPGCLC basal medium.
500 mL of Glasgow’s MEM
75 mL of KnockOut Serum Replacement
6 mL of MEM Non-Essential Amino Acids Solution (100x)
6 mL of 100 mM Sodium pyruvate
6 mL of Penicillin-Streptomycin (x100)
Aliquots 40 mL/tube and store at -20 °C.
|Dilute 350 μL 2-Mercaptoethanol (14.3 M) in 9.65 mL PBS(-) to make 500 mM 2-Mercaptoethanol stock solution . Store at 4 °C.
|Reconstitute 400 mg L-Ascorbic acid with 20 mL water to make 20 mg/mL L-Ascorbic acid stock solution. Sterilize by filtration (0.22 um). Aliquot 500 μL/tube x 40 tubes and store at -20 °C.
|Recombinant human BMP4
|Reconstitute 1 mg recombinant human BMP4 with 10 mL of 4mM HCl aq. to make 100 μg/mL recombinant human BMP4 stock solution. Aliquot 250 μL/tube x 40 tubes and store at -80 °C.
|Reconstitute 250 μg human SCF with 500 μL of 0.1% bovine serum albumin in PBS(-) to make 500 μg/mL human SCF stock solution. Aliquot 35 μL/tube x 14 tubes and store at -80 °C.
|Reconstitute 1 mg human EGF with 1 mL of 0.1% bovine serum albumin in PBS(-) to make 1 mg/mL human EGF stock solution. Aliqot 100 μL/tube x 10 tubes. Dilute 100 μL of 1 mg/ml human EGF stock solution with 300 μL of 0.1% bovine serum albumin in PBS(-) to make 250 μg/mL humanEGF stock solution. Aliquot 20 μL/tube x 20 tubes. Store at -80 °C.
|Pore size = 40 μm.
|50 ml polypropylene conical tube
|Low attachment plate
|Vari-Mix Platform Rocker
|BLOXALL Blocking Solution
|Normal Horse serum
ImmPRESS Reagent Anti-Goat IgG
|Embryoid Body Dissociation Kit
|Anti-CD38 antibody conjugated to APC
Copyright © 2024 MyJoVE Corporation. All rights reserved