We demonstrate a method for grafting cultured cells into defined sites of early mouse embryos to determine their in vivo potential. We also introduce an optimized electroporation method that uses glass capillaries of known diameter, allowing the precise delivery of exogenous DNA into a few cells in the embryos.
Manipulation and culture of early mouse embryos is a powerful yet largely under-utilized technology enhancing the value of this model system. Conversely, cell culture has been widely used in developmental biology studies. However, it is important to determine whether in vitro cultured cells truly represent in vivo cell types. Grafting cells into embryos, followed by an assessment of their contribution during development is a useful method to determine the potential of in vitro cultured cells. In this study, we describe a method for grafting cells into a defined site of early postimplantation mouse embryos, followed by ex vivo culture. We also introduce an optimized electroporation method that uses glass capillaries of known diameter, allowing precise localization and adjustment of the number of cells receiving exogenous DNA with both high transfection efficiency and low cell death. These techniques, which do not require any specialized equipment, render experimental manipulations of the gastrulation and early organogenesis-stage mouse embryo possible, allowing analysis of commitment in cultured cell subpopulations and the effect of genetic manipulations in situ on cell differentiation.
Cell culture has been widely used in developmental biology studies. Mouse embryonic stem cells (ESCs) and epiblast stem cells (EpiSCs) can differentiate into all three germ layers in vitro and are a useful model for cell differentiation in early mammalian embryogenesis. The derivation of these cell lines has opened up an opportunity for in vitro manipulation and detailed investigation of the localized signaling events and transcriptional networks operating during early embryonic patterning. However, it remains important to determine the in vivo relevance of any manipulations performed in culture. The in vivo potential of preimplantation embryo-derived mouse ESCs has been assessed by introducing them back into preimplantation embryos (morulae or blastocysts)1. However, EpiSCs that represent the epiblast cells in postimplantation embryos cannot integrate efficiently in preimplantation embryos2,3. Our previous findings have shown that EpiSCs can efficiently generate chimeras and contribute to all germ layers, when grafted into postimplantation embryos4. Thus, the best way to evaluate the in vitro cultured cells is to introduce them to their corresponding environment in vivo.
Electroporation is a widely used method to deliver exogenous molecules into targeted cells in both in vivo and in vitro experiments. Electrical energy can generate a large number of pores in the cell membrane, which allows exogenous deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) to enter the cells. One of the biggest challenges for this technique is to combine optimal cell viability with high electrotransfection efficiency 5,6. For electroporation of nucleic acids in embryonic tissues, gold plated electrodes have most commonly been used, allowing targeting of cells in a broad spatial range7-9. To achieve a more localized gene transfer, a needle-shaped electrode has been utilized to achieve a focal electric field10,11. Using this method, the authors showed that after the electroporation, around 30-60 cells had taken up the DNA construct11. Nevertheless, it seems that accurately adjusting the number of electroporated cells remains difficult with a fixed-width electrode. The capillary electroporation technique has been used to deliver plasmids to single cells12-14. However, this technique has not been applied for electroporating plasmids to embryos ex vivo. More recently, a microdevice has been reported to locally electroporate a few distal visceral endoderm cells (less than 4 cells) in early postimplantation mouse embryos15. However, it is still unknown whether this device can efficiently target ectoderm and mesoderm ex vivo.
In this study we describe two novel methods to assess cellular and gene function in early post-implantation embryos. We first demonstrate how to graft in vitro cultured cells into defined sites in early mouse embryos to assess their in vivo potential. The integration of the grafted cells and their descendants, all labelled by a genetic tag (e.g., a green fluorescent protein (GFP), can be further examined by immunostaining of tissue specific proteins4. Secondly, we describe an improved method to precisely deliver DNA to localized sites in the embryo via electroporation. Rather than using a needle-shaped electrode, we inserted a thin wire inside a fine-tipped glass capillary, and demonstrate that this modification can deliver DNA to a small number of cells with high efficiency and limited cell death. Moreover, we show that by using glass capillaries with different opening sizes, we can control the number of electroporated cells. Therefore, we believe this method can be of great use to study early embryonic patterning involving small numbers of cells.
All animal experiments were carried out in accordance with UK Home Office Regulations as specified in the Animals (Scientific Procedures) Act (1986) under Project Licence number 60/4435. To collect embryos at specific developmental stages, timed matings were set up O/N. Noon on the day of finding a vaginal plug was designated embryonic day (E) 0.5.
1. Dissecting the E7.5 or E8.5 Postimplantation Embryos for Ex Vivo Culture
2. Prepare the Embryo Culture Medium
3. Ex Vivo Culture Conditions for Different Embryonic Stages
4. Grafting Cultured Cells into E7.5 or E8.5 Mouse Embryos
5. Handmade Electroporation Materials and Apparatus Setup (Prepare the Following in Advance of Electroporation Experiments):
6. Electroporation E7.5 or E8.5 Mouse Embryos
Grafting
EpiSCs that ubiquitously express EGFP (r04-GFP, derived from E6.5 epiblast, and C2, derived in vitro from mESCs)4 were manually scraped from the culture dish and grafted into different sites of E7.5 embryos (Figure 2A). The embryos were cultured ex vivo and analysed after 24 hr. The distribution of donor cells was assessed by fluorescence microscopy. If donor cells incorporated, they proliferated and their derivatives dispersed within the host embryos (Figure 2B). It has been observed that grafts containing 10-16 cells incorporated efficiently in the host embryos (Figure 2A, and 2B), however, grafting more cells does not result in better chimaerism. Instead, grafted cells produced unincorporated clumps (Figure 2C and 2D).
Electroporation
To assess the efficiency of our electroporation system, we delivered GFP-expressing plasmids (pCAG-GFP and pCAG-Cre:GFP) to specific sites in the embryo. Consistent with a previous study11, GFP+ cells were detected in embryos 1-2 hr after electroporation (Figure 2E and 2G). When distal epiblast cells at the late primitive streak stage embryo were electroporated, labelled cells contributed to the neural ectoderm after 24 hr in culture (Figure 2E and 2F). This result corresponds well to known fate maps of epiblast cells in gastrulation stage embryos17. Similarly, when the GFP expression plasmid was electroporated in the primitive streak at E8.5 (2-5 somites), GFP+ cells contributed to the paraxial mesoderm (Figure 2G and 2H), consistent with known fate maps of the late primitive streak18. Moreover, we observed contribution to all three germ layers from electroporated cells (Figure 2I-K), suggesting that the electroporation procedure does not compromise cell behaviour in vivo. However, we also noticed that whilst epiblast (E7.5) or primitive streak cells (E8.5) were targeted, some endoderm cells were also electroporated (Figure 3C and Table 1).
One of the major advantages of using a capillary electrode is that the number of electroporated cells can be controlled, simply by changing the diameter of its opening. To determine the number of electroporated cells, embryos were fixed 2 hr after electroporation and imaged in wholemount on a confocal microscope. The number of GFP+ cells was manually counted in the confocal z-stacks. Table 1 shows that, for a given stage, increasing the opening size of the glass capillary from 20 to 30 µm results in DNA uptake by more cells. When a single opening size was compared between stages (E7.5 versus E8.5), more cells were found to be electroporated at the latter stage. This effect may be due to a higher concentration of DNA present in the amniotic cavity at E8.5. Because the DNA solution was mixed with the green food dye, we can use the green color to assess the DNA concentration in the amniotic cavity. In the microscope, it is clear that, when compared with E8.5 embryos, the green color after DNA injection is much lighter in the cavity of E7.5 embryos. Although the same concentration of DNA solution was injected into E7.5 and E8.5 embryos, more DNA solution was in the amniotic cavity of E8.5 embryos in order to fill it completely because they are larger in the size. After withdrawing the injection needle, there is always some degree of leakage of DNA solution from the amniotic cavity, and since the puncture hole is larger in comparison to the size of the amniotic cavity in earlier embryos, it is likely that there was proportionately more leakage from E7.5 than E8.5 embryos, leading to a lower DNA concentration. The different number of transfected cells could also be due to the different diameters or induced transmembrane voltage (ITV) thresholds of cells at different stages.
A drawback of electroporation is the associated cell death. Similar to the traditional gold plated or needle-shaped electrodes, electroporation using a capillary electrode also causes the cell death. After electroporation the targeted region appeared darker in colour compared to neighbouring regions (Figure 3A and 3B), indicating that some degree of cell death must have occurred in this area. To further determine the number of dead cells caused by the electroporation procedure, embryos were stained with a fluorescent cell membrane-impermeable nuclear dye. The nuclei of dead cells were labelled with a membrane-impermeable far-red fluorescence dye. The staining confirmed that this capillary electroporation technique only results in a small number of dead cells near the electroporation site (Figure 3D and Table 1).
We noticed that although dead cells appear at the electroporation site, GFP+ cells and dead cells are also most exclusive from each other (Figure 3E and 3F). Moreover, when the caudal lateral epiblast at E8.5 was electroporated with pCAG-GFP and a glass capillary opening of 20µm, a large number of GFP+ cells was detected after 48 hr in culture (Figure 3G and 3H). Taken together, these results suggest most GFP+ cells detected 2 hr after electroporation are still viable during the further culture.
We scored the number of GFP+ cells after 24 hr ex vivo culture. Six embryos were electroporated with pCAG-GFP at E7.5, using a capillary opening of 20µm diameter. 107±31 (mean ± s.d.) GFP+ cells/embryo were detected. Since at the start of culture (2 hr), 9 cells were electroporated on average per embryo (Table 1), this suggests that electroporated cells underwent 3-4 divisions within 2 hr. The average cell doubling time from E7.5 to E8.5 embryos is around 6-7 hr in all cells apart from those in the ventral node19,20. This suggests that the electroporation procedure does not hamper normal cell growth.
Figure 1. Circuit diagram showing the electroporation setup. The embryo containing DNA solution in its amniotic cavity was positioned between the two electrodes. Current at the chosen parameters was provided by a square wave pulse generator (power supply). A multimeter was connected in series to detect the electric current passing the embryo. Please click here to view a larger version of this figure.
Figure 2. The distribution of grafted or electroporated cells in host embryos. (A-H) GFP fluorescence overlays (green) on brightfield images of wholemount embryos (grayscale) (A) 10-16 GFP+ EpiSCs were grafted into the distal region of a late-streak stage embryo. (C) A larger clump of GFP+ EpiSCs was grafted into the distal region of a mid-streak stage embryo. (B and D) The distribution of EpiSCs derived cells (green) in the host embryos (shown in A and C) after 24 hr in culture. (B) GFP+ cells dispersed in the host embryo, suggesting correct integration of the donor cells. (D) Grafting larger cell clumps resulted in unincorporated clump formation in the host embryo. (E-K) pCAG-Cre:GFP plasmid electroporated into specific areas of wildtype embryos. Electroporating the distal region of an early bud stage embryo (E) or the primitive streak of a 2-5 somite stage embryo (G) resulted in GFP+ cells in these regions 2 hr after the procedure. (F and H) The distribution of GFP+ cells in the host embryos after 24 hr in culture, showing that electroporated cells contribute to the neuroectoderm (black arrow) (F) and paraxial mesoderm (white arrow) (H). (I-K) DAB immunostaining for the GFP+ cells showing that the electroporated cells can give rise to the neuroectoderm (I), mesoderm (J) and endoderm (J and K) after 24 hr in culture. Scale bar (A-H) = 250 µm; scale bar (I-K) = 100 µm. Note: Figure 1A and 1B are reprinted from our previous publication4. Please click here to view a larger version of this figure.
Figure 3. Distribution of GFP+ cells and dead cells in the embryos after electroporation. (A-C) pCAG-Cre:GFP plasmid electroporated into the caudal lateral epiblast cells of an E8.5 (2-5 somite stage) embryo (capillary opening size: 20 µm). (A) 2 hr after the procedure, the targeted region showed a dark color (white arrow) compared to other parts of the embryo. Inset shows an enlargement of the electroporated region. (B) Brightfield image (grayscale) overlaid with the green fluorescent channel showing the electroporated cells (green). (C) A confocal z-slice showing that two endoderm cells (green) also took up the plasmid when the caudal lateral epiblast cells were targeted. The cell nuclei are shown in red. (D-F) pCAG-Cre:GFP plasmid was electroporated in the caudal aspect of the node of an E8.5 embryo (capillary opening size: 30 µm). The embryo was cultured for 2 hr. Electroporated cells are shown in green and dead cells in red. (D) The electroporated area contains both GFP+ cells as well as dead cells. The area in the white box was further analyzed in a confocal microscope. Manual counting of the z-stack showed that there were 33 GFP+ cells and 23 dead cells in this area. Only two cells were both positive for both fluorophores. (E and F) XYZ view of a confocal z-slice from the white boxed region in D showing GFP+ cells are separate from the dead cells. The nuclei are shown in blue. (G and H) pCAG-Cre:GFP plasmid was electroporated into a few cells in the caudal lateral epiblast of an E8.5 embryo (capillary opening size: 20 µm) and imaged after two (G) and 48 (H) hours ex vivo culture Note: (H) The embryo were severed in two after culture. The head and heart regions were removed. Scale bar (A, B, D, G and H) = 250 µm; scale bar (C, E and F) = 100 µm. Please click here to view a larger version of this figure.
Diameter of opening of the capillary tube | Embryo stage | Electroporation efficiency: no. embryos containing GFP+ cells after 2h / total no. of electroporated embryos (no. GFP+ embryos that developed normally after 24 or 48h culture) | Average number of GFP+ cells per embryo ±s.d. (n= no. of examined embryos) | No. of GFP+ endodermal cells per each embryo ±s.d. (n= no. of examined embryos) |
20μm | E7.5 (LS-LB) | 7 / 9 (7) | 9±3 (n=4) | 4±2 (n=4) |
30μm | E7.5 (LS-LB) | 13 / 15 (12) | 17±2(n=4) | 6±1 (n=4) |
20μm | E8.5 (2-5 somites) | 12 / 13 (10) | 21±4 (n=4) | 11±4 (n=4) |
30μm | E8.5 (2-5 somites) | 2 / 2 (2) | 33 and 26 (n=2) | 14 and 16 (n=2) |
Table 1. Electroporation efficiency of pCAG-Cre:GFP plasmid in mouse embryos.
Abbreviation: LS, late primitive streak stage; LB: late bud stage. Embryos are staged according to Downs and Davies12
Grafting
The critical step for cell grafting experiments is the insertion of a coherent string of cells ideally in a single action, to avoid breakup of the clump. This technique requires some practice in controlling the mouth pipette. If donor cells incorporate well in the host, their derivatives will disperse in the embryo. To further determine whether the dispersed donor derived-cells differentiate appropriately in the host, immunostaining can be performed on the embryo sections. If donor cells are not compatible with the host environment, they either cannot be detected (as they are expelled from the embryo) or form unincorporated clumps in the embryos after culture. If both dispersed cells and cell clumps were observed, this may indicate that too many cells were grafted and excessive donor cells which cannot interact with surrounding host cells resulted in clump formation. In this case, additional grafts containing a smaller number of cells can be performed.
The major limitation of the cell grafting technique is that it is not possible to determine the full in vivo potential of cells since mouse ex vivo culture over periods longer than 48 hr has not been achieved. However, if combined with ultrasound-guided cell injection, it may be possible to transfer cultured cells to the embryos in utero. To summarize, cell grafting experiments have been widely used in our group and have given us valuable clues about the in vivo potential of various cell types4,21,22. It is a technique of general utility to assess the in vivo potential of in vitro cultured cells in early postimplantation embryos.
Electroporation
Although in this study we have only shown that it is efficient to use the capillary electroporation technique to target the epiblast, it is also possible to intentionally target other germ layers such as endoderm cells. The critical step for the capillary electroporation technique is to minimize the time taken to electroporate each embryo (< 5 min per embryo) since PBS is very suboptimal for early mouse embryos. Our data above has shown that, in most areas in the embryos, electroporation does not affect the embryo growth. However, electroporation in the node caused abnormal development and led to the premature death of the embryo. This is likely due to damage or death of the cells that form important signalling centres23. Hence, this region would have to be avoided with this technique. A further caveat is that, as mentioned in the results section, whilst epiblast or primitive streak cells were targeted, some endoderm cells were also electroporated. This may be because DNA reaches to the endoderm through gaps under the epiblast epithelium. Endoderm is composed of epithelial cells and in our experience these cells have a higher propensity to take up DNA. Therefore, when applying this technique for fate mapping, it is important to assess which cells initially take up DNA.
It should also be noted that although pCAG-GFP and pCAG-Cre:GFP plasmids can be efficiently delivered using the electroporation parameters shown in this study, the efficiency of other DNA constructs may vary and need individual optimisation. Alterations in DNA concentration, electroporation voltage or the number of pulses can be made if plasmids are found to be difficult to transfect.
To summarize, our optimized capillary electroporation system can efficiently and reproducibly deliver GFP or Cre:GFP plasmids into a very few cells in the embryo with limited cell death. Since this method does not require expensive or highly specialized equipment, it can be of great use for cell tracking studies or in testing the effect of ectopic expression or conditional deletion of genes in early embryos, if electroporation is performed in embryos carrying floxed conditional mutant alleles. Therefore, this electroporation technique provides a useful functional tool for understanding on a cell-by-cell basis the roles of cell-intrinsic factors in the context of localized wildtype embryonic environments.
The authors have nothing to disclose.
We thank Filip Wymeersch and Anestis Tsakiridis for comments on the manuscript, staff in the SCRM animal unit for help with animal maintenance and Prof. Stuart Forbes for immunohistochemistry reagents. This work was supported by MRC grant Mr/K011200/1 and the China Scholarship Council
Forceps | Dumostar | T5390 | |
Dissecting stereomicroscope | Zeiss | Stemi 2000-C | |
Stereomicroscope system with fluorescence | Nikon | AZ100 | |
Inverted microscope with a digital camera | Olympus | Olympus BX61 | |
Inverted confocal microscope | Leica Microsystems | Leica TCS SP8 | |
Low melting point agarose | Life Technologies | 16520-050 | |
Pasteur pipettes | Fisher Scientific | 11397863 | |
30mm Petri dishes | Fisher Scientific | 121V | |
4-well plates | Thermo scientific | 179820 | |
M2 medium | Sigma-Aldrich | M7167 | |
Phosphate Buffered Saline (PBS) | Life Technologies | 10010015 | |
Paraformaldehyde (PFA) | Sigma-Aldrich | P6148 | |
Pipettes | NICHIRYO | Nichipet | |
tips | Greiner Bio One | 685280 | |
Cell culture incubator | SANYO | MCO-17AIC | |
Roller culture apparatus | BTC Engineering | ||
Syringe filters 0.45µm, sterile | Sigma-Aldrich | 10462100 | |
Glasgow Minimum Essential Medium (GMEM) | Sigma-Aldrich | G5154 | |
non-essential amino acids (NEAA) | Life Technologies | 11140050 | |
L-glutamine | Fisher Scientific | SH30549.01 | |
Sodium pyruvate solution | Fisher Scientific | SH30239.01 | |
Penicillin and Streptomycin 10.000UI/ml | Lonza | DE17-602E | |
Gas Cartridge for Portable Meker Burner | COLEMAN | COLEMAN 250 | |
Thin Wall Borosilicate Capillary Glass with Fillament, OD 1.0 mm, ID 0.78 mm | Harvard Apparatus | 640798 | |
Aspirator tube assemblies for calibrated microcapillary pipettes | Sigma-Aldrich | A5177-5EA | |
Flaming/Brown micropipette puller | Sutter Instrument Company | P-97 | |
Microforge | De Fonbrune | BS030301 | |
Pneumatic pico pump | World Precision Instruments | PV830 | |
Microloader tips | Eppendorf | 5242956.003 | |
ECM 830 square wave pulse generator | BTX | 45-0002 | |
Green food coloring dye | Sigma-Aldrich | C.I. 42053 | |
A far-red cell membrane-impermeable nuclear dye | Biotium | 40060-T | |
pCAG-Cre:GFP | Addgene | #13776 | |
pCAG-GFP | Addgene | #16664 | |
Multimeter | Excel | XL830L | |
Micromanipulators | Leitz | ||
0.2mm diameter platinum wire | Agar Scientific | E404-2 | |
Anti-GFP antibody | Abcam | ab13970 | |
Goat anti-Chicken IgY, HRP | Santa Cruz | sc-2428 | |
Liquid DAB+ Substrate Chromogen System | Dako | K3467 | |
4',6-diamidino-2-phenylindole (DAPI) | Life Technologies | D21490 | |
A far-red fluorescence nuclear counterstain | Life Technologies | T3605 |