We describe a series of methods to inject dyes, DNA vectors, virus, and cells in order to monitor both cell fate and phenotype of endogenous and grafted cells derived from embryonic or pluripotent cells within mouse embryos at embryonic day (E)9.5 and later stages of development.
Testing the fate of embryonic or pluripotent stem cell-derivatives in in vitro protocols has led to controversial outcomes that do not necessarily reflect their in vivo potential. Preferably, these cells should be placed in a proper embryonic environment in order to acquire their definite phenotype. Furthermore, cell lineage tracing studies in the mouse after labeling cells with dyes or retroviral vectors has remained mostly limited to early stage mouse embryos with still poorly developed organs. To overcome these limitations, we designed standard and ultrasound-mediated microinjection protocols to inject various agents in targeted regions of the heart in mouse embryos at E9.5 and later stages of development. Embryonic explant or embryos are then cultured or left to further develop in utero. These agents include fluorescent dyes, virus, shRNAs, or stem cell-derived progenitor cells. Our approaches allow for preservation of the function of the organ while monitoring migration and fate of labeled and/or injected cells. These technologies can be extended to other organs and will be very helpful to address key biological questions in biology of development.
More than a decade ago, human embryonic stem cells (HuESCs) have been derived from human blastocysts1. Since then, these cells have become the subject of an important field of research which addresses unmet questions in human developmental biology. HuESCs have furthermore provided hopes in regenerative medicine. In recent years, human induced pluripotent stem cells (iPSCs) have been generated from patient-specific somatic cells, providing models of genetic disease2. Many in vitro protocols for differentiation of embryonic or induced pluripotent stem cells towards various cell lineages, including heart lineages3, have been reported. The differentiated cells are often phenotyped by analysis of RNA and protein expression, immunostaining, and/or in vitro functional tests. However, pluripotent stem cell derivatives have to be placed in a proper embryonic environment in order to test whether they fully acquire the cell fate of their embryonic counterpart and whether they recapitulate the genuine in vivo function in response to regional cues. While tissue engineering is promising, it does not yet provide all the known and unknown cues of the proper in vivo developing embryonic tissue4,5.
Cell labeling with dyes or retroviral vectors in embryos, including mouse embryos, have brought important information as to the embryonic origin of cell lineages during cardiac development6. For example, dye injection into the pericardial space of mouse embryos ex vivo, followed by in vitro culture of isolated hearts, was used to label epicardial cells and their descendants7. However, dye and retroviral cell labeling have been mostly applied to early mouse embryos with still poorly developed organs, or chicken embryos, which are more easily accessible8. An exception is the brain, which is easier to target in embryos9,10. Such an approach has not yet been applied to the beating embryonic mouse heart.
To complement direct labeling with dyes or virus and to perform lineage tracing in more advanced stage mouse embryos and adult mice, the cell labeling approach has been combined with analysis of transgenic mice using the Cre/Lox technology. The Cre/Lox approach11 however features some limitations due to the spatiotemporal specificity of the genomic regulatory regions used to drive expression of the recombinase, and the efficiency of the Cre/Lox recombination12. Furthermore, this approach does not fully address the specific questions of cell migration-driven acquisition of cell fate as it can only label a precursor after activation of the regulatory region used to drive Cre expression. It also cannot apply to human embryos for obvious ethical issues.
Given these limitations, we designed a series of new protocols to inject a variety of cell labeling agents such as fluorescent dyes, virus, gene expression modulators such as shRNAs and DNA-based cell labeling vectors, or cells in the mouse embryo at E9.5 and later stages of development in targeted regions of the heart.
The DNA/cell injections use a stereomicroscope and a simple microinjection device combined with ex vivo embryo culture up to 48 hr, or isolated heart or embryonic explant culture for 48-72 hr. We also report an ultrasound-mediated microinjection protocol in mouse embryonic hearts in utero. This technique allows monitoring the development of embryos13 and allows for long-term follow-up of the injectates and/or labeled cells.
We found that these approaches preserve the function of the organ and provide a more representative environment than in vitro testing of stem cell potential. It also provides the opportunity to follow migration of labeled and/or injected cells to monitor their fate. Ultimately, this should yield a better understanding of regional tissue patterning and key biological processes.
1. Preparation
Animal procedures
Obtain approval from an animal ethical committee and follow institutional guidelines for work with virus, HuESC and/or iPSC (when applicable), as well as mouse handling, obtaining mouse embryos, and performing mouse surgery. For timed matings, the day of the plug is considered embryonic day (E)0.5 / 0.5 days post-coitum.
2. Collection of E9.5 and E10.5 Embryos for Ex vivo Injection
3. DNA or Cell Injection Under a Stereomicroscope
4. Embryo, Isolated Heart, and Explant Culture
5. Ultrasound-guided Injection in the Heart In utero
Using the injection protocols described above, cells can be labeled and/or injected into the embryonic mouse heart. As proof of concept, several examples are shown in which the injection protocol and the ex vivo AVC explant, isolated heart, or whole embryo culture were combined (Figure 1).
Figure 1 shows the preparation of the embryo before cell injection. The E9.5 embryo is removed from its decidua while maintaining integrity of the yolk sac (Figure 1A). The yolk sac is opened just above the heart (Figure 1B). The pipette is approached (Figure 1C) and the cells injected. The heart retains its shape (Figure 1D) and remains beating.
To address a more specific biological question in developmental biology such as the epithelial to mesenchyme transition (EMT), an E9.5 AVC explant was injected with a shRNA that down-regulates a protein required for endothelial-mesenchymal transition of endocardial cells (Figure 2A). The control embryo was injected with an empty backbone vector. Likewise, HuES cell–derived endothelial pre-valvular cells expressing GFP under the control of the Sox9 promoter were injected in the AVC at E10.5, followed by 48 hr explanted heart culture. These cells acquire markers of EMT such as periostin similar to the endogenous endocardial cells (Figures 2B and 2C).
These ex vivo approaches are relatively easy, but limited to short-term follow-up (maximally 48-72 hr). The ultrasound-guided injection procedure provides a technically more challenging approach, but with the option of long-term, even post-natal, in utero follow-up. The in utero injection of dye or viral vectors to label cells in specific cardiac regions is shown in Figures 3A-3C. With this method, specific labeling of epicardial cells can be achieved with fluorescent dyes (Figures 3D and 3E) or virus (Figure 3F), and the method can be expanded upon with cells or alternative injectates.
Figure 1. Cell injection in the heart. A: E9.5 embryo in the yolk sac. B: visualization of the heart after opening the yolk sac just above the heart. The inset below shows a magnification of the heart and points the AV canal. C: approach of the pipette towards the AVC. D: injected embryo after 48 hr culture (H = heart).
Figure 2. Injection of a shRNA in the AVC of an E9.5 embryo that prevents ex vivo EMT of endocardial cells in a cardiac explant. A: control backbone vector or a shRNA was injected together with lipotectamine in the AVC of an E9.5 mouse embryo; 3 hr later, the AVC was dissected out and grown on a collagen gel in order to trigger EMT of endocardial cells. The shRNA downregulated a protein that is required for EMT. B-C: HuES cell-derived valvular cells engineered to express GFP under the control of the Sox9 promoter were injected in the AVC of E10.5 embryos. The heart was cultured for 2 days (B), and subsequently fixed and stained with an anti-periostin antibody (C). The inset shows a magnification of the AVC region.
Figure 3. In utero injection. A: scheme depicting the experimental setup. The pregnant mouse is positioned supine and a Petri dish with silicone membrane is stabilized above the incision site. 1 = microinjector, 2 = transducer, 3 = Petri dish with silicone membrane, 4 = ultrasound gel, 5 = mouse with uterine horns (red), 6 = Play-Doh clay, 7 = mouse handling table. B: still frame of an ultrasound movie (M mode) showing the position of the needle and embryonic heart prior to injection. The female ECG is shown at the bottom (green), and heart and respiration rate in the lower right (yellow). H = heart. C: ultrasound image showing the position of the needle and embryonic heart during injection. Inset: the tip has passed the pericardium, but does not touch the myocardium, so that the injectate can be injected into the pericardial space of the atrio-ventricular groove. Black dotted lines mark the borders of the atrium (A) and ventricle (V). D: whole mount image of an E11.5 mouse embryo showing strong fluorescence of the dye in the pericardial space. E: representative section of an E11.5 embryonic mouse heart showing pericardial and epicardial cell labeling with a fluorescent dye (CDCFDA-SE, green) 4 hr after injection. Nuclei are labeled with DAPI (blue). LA = left atrium, LV = left ventricle, RA = right atrium, RV = right ventricle. F: representative section of an E14.5 embryonic mouse heart showing mosaic epicardial cell labeling with GFP-expressing lentivirus 3 days after injection. To confirm GFP specificity, the GFP protein was also stained with a GFP antibody (red). Nuclei are labeled with DAPI (blue).
The intra-cardiac ex vivo injection protocols described above are designed to preserve myocardial function for at least 48 hr in mid-stage (E9.5- E11.5) mouse embryos. These injection approaches allow for spatially targeted injection of DNA or cells. The few examples shown in Figures 1-3 provide proof of concept for delineating ex vivo and in vivo molecular mechanisms of developmental processes that take place in restricted cardiac regions, such as EMT of endocardial or epicardial cells. Most importantly, these protocols provide an opportunity to follow migration and fate of injected cells such as human embryonic cells or huESC derivatives in a proper embryonic environment, a so far unmet challenge. In vivo labeling (like the epicardial cells in Figure 3) with saturating or limiting dilutions of dye or virus allows for lineage tracing of cell (sub)populations on a short- as well as long-term basis, without the need for Cre/lox technology (although combining these techniques may be useful as well).
Several critical steps while applying these protocols were observed. First, the embryos are very sensitive to light. It is therefore recommended to practice the injection protocol under a stereomicroscope so that the procedure can be carried out with high reproducibility and efficiency within 15 min per embryo. In the same line, the entire ultrasound-mediated injection procedure should be performed within 30 min to preserve a good viability of embryos and to not compromise their in utero development. Manipulation of the uterus should be kept to a minimum. Additionally, it should be noted that surgery on pregnant mice tends to result in premature delivery (1-2 days earlier than normal). However, neonates should be generally healthy and develop normally. Further, the penetration of the pipette through the uterine wall, the yolk sac and the embryo to reach the pericardium and finally the myocardium is a delicate procedure (step 5.3.13). This step should be done carefully and gently. For optimization, it is recommended to perform several short-term dye injections to determine the reproducibility and exclude injection in non-targeted areas. Finally, we have not found developmental cardiac defects related to the injection procedures for the examples given above. However, a more detailed, stage-by-stage, morphological and functional analysis should be performed to completely ascertain normal development and function during the stages of interest.
The limitations of these technologies are several. (i) The injection under the stereomicroscope is limited by the light and temperature sensitivity of the mouse embryos. This can be overcome by the practice of the microinjection procedure in order to be as fast as possible, regularly replacing or warming the medium, and working in a dark room. (ii) The ex vivo culture of embryos or explants is limited in time, in contrast to the in utero ultrasound-guided injection, which allows for long-term follow-up. Vital dyes tend to label cells almost immediately and remain visible up to three to four days post-injection. However, the intensity of the dye will be diluted by successive cell divisions. Labeling cells with lentivirus may overcome this problem. However, uptake of the virus takes several hours and expression is optimal after 24-48 hr, which prevents short-term analysis. (iii) The ultrasound-mediated injection is limited by the cost of the equipment. The resolution of the 40 MHz transducer is sufficient for mid- and late-stage embryos, but more limiting for stages prior to E11.5.
In summary, we propose that the cardiac injection protocols described above should be used in mid-stage mouse embryos. The ex vivo technique is compatible with the culture of embryos, isolated hearts, or explants. The in vivo procedure allows for long-term follow-up, including post-natal development. These techniques will aid significantly in testing the differentiation potential of human stem cell derivatives in a proper environment, as well as in addressing specific questions of developmental biology such as cell migration and regional cues, which are key events in the road toward shaping a functional heart. Additionally, our methods and those of others10 may in the future be used for investigation of other organs than the heart, dependent on available ex vivo culture protocols and/or developmental stage in vivo.
The authors have nothing to disclose.
The authors acknowledge the Foundation Leducq (MITRAL) and the Agence Nationale pour la Recherche (grant ANR Specistem) for funding this research.
Setups / Hardware | |||
40 MHz Transducer | VisualSonics | MS550S | |
Microinjector | VisualSonics | ||
Microinjector | Eppendorf | 5242 | |
Micromanipulator | Eppendorf | 5171 | |
Nitrogen | required to pressurize the injector | ||
Rail system | VisualSonics | ||
rotator | to rotate glass tube with embryos inside the incubator | ||
Standard incubator | 5% CO2, 37 C | ||
stereomicroscope | Zeiss | Discovery. V8 | |
Vevo 2100 | VisualSonics | ||
Microinjection | |||
borosilicate capillary tubes | World Precision Instrument | KTW-120-6 | 1.2 mm external diameter |
pipette puller | Sutter | Model P87 | |
microinjection needles | Origio-Humagen | C060609 | OD/ID 1.14mm/53mm, with 50/35 um OD/ID tip |
Hamilton syringes | |||
Petridishes | 10 cm diameter | ||
Mineral oil | Sigma | M8410 | |
Silicon membrane | Visualsonics | 4.3×4.3 cm | |
Play-Doh | |||
Isoflurane | Vet One | ||
hair removal agent | Nair | ||
eye lubricant | Optixcare | 31779 | |
Electrode gel (Signa) | Parker | ||
Suture | Sofsilk 5-0 | S1173 | |
Ultrasound gel | Aquasonic | ||
Buprenex Buprenex (buprenorphine hydrochloride) | Reckitt Benckiser Pharmaceuticals Inc. | NDC 12496-0757-1 | 0.05-0.1 mg/kg in saline |
Other | |||
Silicone Elastomer | Dow Corning | Sylgard 184 | |
Glass petridishes | Fine Science Tools | 60mm diameter | |
insect pins | Fine Science Tools | 26002-20 | |
Media and culture reagents | |||
Optimem medium | Life Technologies | 51985026 | |
M2 medium | Sigma | M7167 | |
Dulbecco’s Eagle Medium | Lonza | BE12-640F | high glucose and 50% rat serum |
M16 medium | Sigma | M7292 | |
rat serum | Janvier | ODI 7158 | |
pennicilin/streptomycin | Life Technologies | 15140-12 | |
oxygen 40% | Air liquid | required to oxygenate the embryo culture medium | |
fetal calf serum | Fisher | RVJ35882 | |
matrigel | BD | 356230 | |
collagen type I | BD | 354236 | to coat culture dishes for explant culture |
culture dishes | Dutcher /Orange | 131020 | |
Injectates | |||
CDCFDA-SE | Invitrogen/Molecular Probes | C1165 | 25mg/ml DMSO. Store at -20 C. Dilute 1:100-200 in saline before use. |
PGK-GFP-expressing lentivirus | ~8E9 transducing units/ml DMEM | ||
lipofectamine 2000 | Life Technologies | 11668019 |