SEARCH

SEARCH BY CITATION

Keywords:

  • electroporation;
  • GFP;
  • somites;
  • chick embryo

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

In ovo electroporation is a well-established method of gene transfer into neural and mesenchymal tissue in chick embryos. Electroporation of somites, however, has been hampered by low efficiency due to technical difficulties. Here, we present a powerful technique to electroporate avian somites and subpopulations of somitic cells at high efficiency in ovo. We demonstrate specific targeting of distinct somitic compartments and their derivatives using single or combinations of plasmid expression vectors. This technique opens new perspectives to investigate the morphologic and genetic basis of somite development. Developmental Dynamics 229:643–650, 2004. © 2004 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

For many years, electroporation has been an established method to introduce exogenous DNA into eucaryotic cells in culture (Neumann et al., 1982; Potter et al., 1984; Potter, 1988; reviewed in Andreason and Evans, 1988). Recently, optimization of the electroporation technology toward higher cell survival opened its application to in vivo gene transfer into living tissue and entire organisms, including blood vessels (Nishi et al., 1996), cardiac tissue (Harrison et al., 1998), solid tumours (Hofmann et al., 1999), as well as embryos of Drosophila (Kamdar et al., 1995), ascidians (Corbo et al., 1997), zebrafish (Muller et al., 1993), Xenopus (Eide et al., 2000; Sasagawa et al., 2002), and mouse (Itasaki et al., 1999; Osumi and Inoue, 2001).

Despite its long scientific history, the impact of chick embryology has been somewhat decreasing in past years due to its limited accessibility to genetic manipulation. Transfection of chick embryonic tissue by electroporation is now helping to overcome this drawback. The first successful electroporation of chick tissue was performed on retinal explants in vitro (Pu and Young, 1990). Muramatsu et al. (1996, 1997) succeeded in electroporating chick embryos in ovo, and Momose et al. (1999) optimized the technique to specifically target tissues of interest in ovo by the use of microelectrodes (reviewed in Atkins et al., 2000). Since then, many functional studies based on electroporation-mediated gene transfer in chick have been published. Most of this work has been done in neural tissue (e.g., Araki and Nakamura, 1999; Inoue et al., 2001; Megason and McMahon, 2002), but also in head ectoderm (Ogino and Yasuda, 1998), limb mesenchyme (Takeuchi et al., 1999; Swartz et al., 2001a, b; Oberg et al., 2002), the segmental plate (Dubrulle et al., 2001), and other tissues (reviewed in Itasaki et al., 1999; Swartz et al., 2001a; for technical overviews see Yasuda et al., 2000; Nakamura et al., 2000; Nakamura and Funahashi, 2001). Recent reports on gene silencing by electroporation of dsRNA in the neural tube (Pekarik et al., 2003) and the use of photoactivatable green fluorescent protein (GFP) for selective labelling of cells and subcellular structures (Patterson and Lippincott-Schwartz, 2002) are further landmarks indicating the immense potential of electroporation to decipher the cellular and molecular mechanisms regulating vertebrate development.

Somites are amongst the most intensely studied structures of the chick embryo, as they give rise to a variety of derivatives, including the axial skeleton, skeletal muscle, blood vessels, and dermis (see Christ and Ordahl, 1995; Brand-Saberi et al., 1996; Marcelle et al., 2002; for reviews). Momose et al. (1999) and Itasaki et al. (1999) reported successful electroporation of somites by using a lacZ reporter construct. Swartz et al. ( 2001a) succeeded in electroporating somites with GFP. However, these authors agree on the technical difficulty and low efficiency of somite electroporation, and until recently, no functional study based on somite electroporation has been published. In an effort to optimize electroporation of somites in ovo, we developed a technique yielding high efficiency electroporation of distinct subpopulations of somitic cells, allowing us to specifically target various somitic derivatives during development. Recently, we published the first functional data obtained by somite electroporation, namely the overexpression of FGF8 in the lateral border of the dermomyotome (Marics et al., 2002). Here, we give a detailed description of our method to electroporate somites. We discuss the potential of this technique to investigate the morphologic and genetic basis of somite development during avian embryogenesis.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Applications of Somite Electroporation

Cell migration/lineage studies.

GFP fluorescence in the neural tube has been previously observed from 2 hr to 10 days after electroporation (Swartz et al., 2001a). In the somites, we detect GFP fluorescence as early as 3 hr and up to at least 4 days after electroporation in somitic muscle derivatives (Fig. 4f). This makes selective somite electroporation with fluorescent reporter constructs a powerful technique to follow cell migration processes and cell fate in vivo. Plasmid DNA is thought to remain episomal in transfected cells. Thus, the intensity of the GFP signal depends not only on the properties of the plasmid, but also on the number of cell divisions of transfected cells. In highly proliferative tissue, the signal will decrease earlier than in postmitotic cells. In lineage studies, it is often useful to compare multiple cell fates and/or movements simultaneously in the same embryo. By using the technique described here, it is possible to target two different somitic cell populations within the same somite by two differently oriented pulsings (Fig. 4d). Differential expression of different reporter constructs, e.g., GFP and dsRed or YFP, could be extremely useful to follow these subpopulations during further development.

thumbnail image

Figure 4. Whole-mount chick embryos electroporated in various somitic compartments. a,c,e,f: Photomicrographs taken on a dissection microscope equipped with fluorescence. b,d: Images were taken on a confocal microscope. In a–d, electroporated somites are shown in dorsolateral view, cranial to the right. These embryos have been electroporated at embryonic day (E) 2.5 and incubated overnight, except d, which was incubated 8 hr. In e and f, limb buds are shown in dorsolateral view, cranial to the top. These embryos have been electroporated at E2.5 and incubated overnight (e) or for 3 days (f). a: Electroporation of the dorsomedial quadrant of an epithelial somite with empty pCLGFPA. Cells in the medial lip of the dermomyotome (arrowheads) and myotomal fibres (arrows) are labeled. Note the difference in the number of labelled cells in neighbouring somites due to local variations in DNA injection and current flow. The electroporated somites and the neural tube (NT) are outlined. b: When less plasmid is injected, few fibers are labeled with green fluorescent protein (GFP). Arrows indicate two fibres expressing the GFP reporter, the remaining myotomal fibres are visualized by whole-mount immunostaining for the embryonic form of the myosin heavy chain (monoclonal antibody MF20, Developmental Studies Hybridoma Bank), in red. c: Same setup as in a, but in this case, pCLGFPA contained an insert (a cDNA coding for Xenopus dishevelled). The dishevelled mRNA expression is visualized by in situ hybridization, in red. Note the differences in expression levels in neighbouring somites due to variations in DNA injection and current flow. The electroporated somites and the neural tube (NT) are schematically outlined. d: Electroporation of the cranial and caudal borders of an epithelial somite. This was performed by two sets of pulses in respectively inverse orientation (see Experimental Procedures section). The electroporated somite and the neural tube are outlined. e: Electroporation of the lateral portion of epithelial somites at hindlimb level labels the migratory myogenic precursor cells, which then populate the leg bud mesenchyme (arrow). Note that a subpopulation of labeled cells is remaining in the lateral somites (arrowhead). f: Same setup as in e but incubated up to Hamburger and Hamilton stage 28 (3 days after electroporation) when limb muscles have formed. GFP labeling is found in distinct groups of muscle fibres in the leg (arrows). Scale bars = 50 μm in a–d, 100 μm in e,f.

Download figure to PowerPoint

Gene expression studies.

In principle, any gene or genetic construct cloned in an appropriate vector can be expressed in tissues after electroporation (Fig. 4c). Thus, gain-of-function studies with endogenous or foreign genes (Marics et al., 2002), as well as loss-of-function studies with dominant negative constructs or RNAi (e.g., Delise and Tuam, 2002; Pekarik et al., 2003) are possible in the chick embryo. A major advantage of gene transfer by electroporation is the possibility to transfect cells by more than one DNA construct simultaneously (Megason and McMahon, 2002; our unpublished results). This coelectroporation technique opens a variety of approaches in somite molecular embryology, including rescue experiments and simultaneous inhibition of redundant signals.

Limitations and Problems of Somite Electroporation

The electroporation technique described here is an efficient method to transfect somite cells with plasmid expression vectors. However, as already mentioned above, there are some critical aspects to the technique that should be taken into account by researchers.

Technical parameters.

There are no strict rules to the optimal physical parameters of the experimental set-up. Most labs will use a different combination of electroporation devices and will work in different biological conditions. A large variation in individual electrode design is inevitable. This requires extensive testing before electroporation can be performed at satisfactory efficiency.

Success rate.

Injection and subsequent electric pulsing is a situation of considerable stress to the embryos. This stress will result in a high percentage of lethality and malformation, even in the hands of experienced workers. With some routine in the method, we expect an average fraction of one third of satisfactorily electroporated embryos after 24 hr of reincubation. Longer reincubation periods evidently lead to higher losses. The most frequently encountered malformations are bent embryos. This malformation is due to electric pulsing, as it is not observed after injections only and is independent of DNA injection. If a high incidence of malformations is observed, voltage must be reduced and/or the distance of the electrodes to the embryo enlarged.

As described above, electroporation of cranial and caudal somites requires physical bending and close proximity of electrodes to the embryonic tissue. Dorsoventral electroporation involves piercing the extraembryonic membranes to position the electrode in the yolk underneath the embryo. These difficulties may lead to lower efficiency than observed in mediolateral electroporation due to higher lethality of embryos.

Reproducibility.

Like in all techniques requiring embryonic manipulation, every electroporation experiment is slightly different from others of the same series due to individual differences in material, embryos, and the actions of the manipulator. This holds especially true for the technique described here, as electrode position, conductivity of the medium, localization, and volume of the injected DNA, and the time course of the experiment will inevitably vary considerably in every experiment. It is crucial, therefore, to interpret data on the basis of a solid number of similar experiments.

Summary

Taken together, we present a powerful technique to electroporate avian somites in ovo at high efficiency and reproducible spatial specificity. Our method is suitable for lineage studies of one or more somitic lineages in the same somite, and allows simultaneous expression of at least two genes of interest in the same cells. In comparison to the existing techniques of retrovirus-mediated gene transfer in the chick embryo, electroporation presents three major advantages: the rapid expression of the electroporated gene (2–3 hr vs. 17–24 hr for viruses), the possibility to electroporate more than one gene in one cell (a possibility that is very restricted with retroviruses), no virus production step in culture. The electroporation technique, thus, opens new ways to address the morphologic and molecular basis of somite development.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Vectors

We have constructed a plasmid (pCLGFPA, Fig. 1, GenBank accession no. AJ575208) that ensures the expression of any gene of interest and the reporter gene GFP from the same vector, each driven by their own promoter/enhancer. The eGFP coding sequence (Clontech) is under the control of an SV40 promoter and enhancer region (derived from the pSl vector, Promega). The cDNA of the gene under study is inserted into the multiple cloning site and expressed under the control of a CMV enhancer and a chicken β-actin promoter. In this vector, we successfully tested the expression of insert sizes between 800 and 1,800 bp. It is however, unlikely that insert size would be a limiting factor. In pilot experiments, we noticed that the widely used CMV enhancer/promoter, which yields good results in neural tissues, works poorly in chick somites. However, we found that several other promotor/enhancer combinations work well in somites. In addition to those described above, good results have been obtained with pMiwIII (Araki and Nakamura, 1999), which contains a chicken β-actin promoter and a RSV enhancer region. We also tested successfully a vector where GFP expression is driven by EF1α-dependent Gal4VP16 (EF1αGal4UASGFP; Koster and Fraser, 2001).

thumbnail image

Figure 1. Electroporation vector pCLGFPA. The vector, derived from the pCMV tracer (Clontech), the pSI vector (Promega), and the pCAGGS vector (Niwa et al., 1991), contains a multiple cloning site where a cDNA of interest can be inserted downstream of a cytomegalovirus enhancer (CMV enh)/chick β-actin promotor (cβact. pr.). A bovine growth hormone polyadenylation (BGHPA) site provides efficient termination of the cloned cDNA. The enhanced green fluorescent protein (eGFP) gene is cloned between the SV40 promotor/enhancer (SV40 Pr.) and an SV40 polyadenylation site (SV40 PA). The corresponding authors will provide this vector freely upon request.

Download figure to PowerPoint

DNA Preparation

We reproducibly obtained good results by using plasmids purified on EndoFree Plasmid purification columns (Quiagen). DNA pellets are dissolved in water to a concentration of 8–10 μg/μl. Highly concentrated plasmid is necessary to attain sufficient expression of constructs in electroporated cells. We use final concentrations of 5–7 μg/μl, which reveal no toxicity to cells, as cell morphology and behaviour is similar when lower concentrations (which yield poorer expression of constructs) are used. To visualize the injected solution, we add Fast Green dye to the DNA. To prevent its rapid diffusion from the site of injection, we thicken the DNA solution with carboxymethylcellulose. Final electroporation solution is 5–7 μg/μl DNA; high viscosity carboxymethylcellulose 0.33% (Sigma); Fast Green 1% (Sigma); MgCl2 1 mM; phosphate buffered saline (PBS) 1×, in water. For 15 μl of DNA preparation, we use 10 μl of purified DNA (8–10 μg/μl) = 5–7 μg/μl final; 2.5 μl of carboxymethylcellulose 1% stock = 0.33% final; 0.75 μl of Fast Green 20% stock = 1× final; 0.3 μl of MgCl2 50 mM stock = 1 mM final; 0.75μl of PBS 20× stock = 1× final; 0.70 μl water.

If several plasmids are coelectroporated, the respective increase in dilution of each solution needs to be considered. For instance, we mix constructs that do not contain GFP (for instance pMiwIII) and empty pCLGFPA at 4:1 with good results (i.e., high expression of the gene of interest and sufficient expression of GFP to easily identify the electroporated cells). When the combined function of two genes is examined, they are mixed at a 1:1 ratio, which results in slightly poorer expression of both transgenes compared with single electroporations. If coelectroporation of more than two is wanted, it might become necessary to adapt the concentration of DNA accordingly.

Electroporation Device

Commercially available electrodes (for instance from BTX Corp.) are expensive and not easily adaptable to our specific purpose; thus, we use self-made electrodes (Fig. 2) consisting of a platinum wire for the positive electrode (anode) and a tungsten wire for the negative electrode (cathode). The section of the wires is 0.5 mm in diameter. The wire is bent at right angle at the tip (ca. 2 mm in length) and insulated by nail polish except the upper side of the bent tip (see Fig. 2). It is important to have a perfect insulation of the lower side of the bent tip to prevent burning of embryonic tissue. The electrodes are held in a commercially available needle holder (e.g., Warner Instruments, Hamden CT), which is insulated with thermoretractable plastic, and connected (d) to the pulse generator. We are successfully using an Intracell Intracept TSS 10 pulser equipped with a pedal trigger.

thumbnail image

Figure 2. Schematic showing self-made electrodes suitable for in ovo electroporation. The electrode wire is insulated with nail polish (a, in red), except in the upper portion of the bent tip (b), and held in a commercially available needle holder (c, Warner Instruments, Hamden CT). The holder is insulated with thermoretractable plastic and connected with an insulated cable (d) to the pulse generator (not shown). Note that, for reasons of clarity, the image is not to scale: The holder (c) is 10 cm in length, and the free electrode (b) is 2 mm in length.

Download figure to PowerPoint

Embryonic Age

Best results are obtained with embryos that have between 20 and 32 somites (stages 13–18 HH, Hamburger and Hamilton, 1951). Younger embryos die frequently or display gross malformations, which are likely due to burned tissues from the electric pulsing (see below). Epithelial somites of older embryos are difficult to inject due to the onset of the ventral curvature of the embryonic tail. However, the dermomyotome of older somites can likewise be electroporated. This strategy can be useful to study late events in somite differentiation. In this case, the DNA solution is injected underneath the dermomyotomal sheet, and the electrodes are positioned under and above the embryo. By using this technique, the myotome seems refractory to electroporation, because we observed that no myocyte is labeled.

Egg Preparation

Eggs are handled routinely as described in the literature (e.g., Tickle, 1993; Selleck, 1996; Saunders, 1996). Briefly, eggs are incubated at 95% humidity and 38.5°C to the required embryonic stage (see above). The upper side of the egg shell is reinforced by broad adhesive tape; 3–4 ml of albumen is withdrawn. Eggs are windowed; to visualize the embryo, India ink (Pelikan, diluted 1:10 with Ringer solution containing penicillin–streptomycin) is injected into the yolk underneath the embryo. The vitellin membrane is carefully removed at the site of manipulation.

Injection of DNA

Glass capillary (borosilicate glass capillaries (Clark) outside diameter = 1.5 mm, inside diameter = 1.17 mm) are drawn on a Sutter P-97 Puller. The tip is broken to obtain an opening of the capillary as fine as possible. The capillary is attached to a rubber tube. The DNA solution is aspired in the capillary and injected in the somite by mouth (Fig. 3a). To perform the injection, the capillary is inserted parallel to the neural tube at the level of the anterior segmental plate, and carefully pushed into somites I through IV (somite staging according to Christ and Ordahl, 1995). At this somitic stage, somites are composed of an outer epithelium, which surrounds a central cavity, the somitocoele. DNA is injected into the somitocoele of somite IV by blowing, then the capillary is retracted to somite III, and DNA is injected, likewise in somites II and I. The volume of DNA injected varies according to the needs of the experiment. To achieve maximal transfection, we inject a quantity that fills the entire somitocoele. To transfect a smaller subset of cells, we inject a smaller quantity of DNA. After injection, electroporation should follow as quickly as possible; although the high viscosity of the DNA preparation prevents too rapid diffusion, somites should be electroporated within 30 sec after injection.

thumbnail image

Figure 3. Schematic representation of the electroporation procedure in ovo. a: Experimental setup viewed from above. On the left, image showing an embryonic day (E) 2 chicken embryo in dorsal view, cranial to the top. The frame outlines the manipulated region. The graphic in the middle illustrates the injection of several epithelial somites with DNA preparation (in green) using a finely drawn capillary, and the positioning of the electrodes left and right to the embryo. By electric pulsing, DNA will integrate, thus, into the medial somitic cells in direction of the anode. The graphic on the right illustrates the expression of the plasmid (in green) in the medial aspect of the somites approximately 3 hr after electroporation (for details see text). b: Experimental setup shown in a viewed in transverse section. The extraembryonic membranes left and right to the embryo are gently pressed down by the insulated lower side of the electrodes. Thus, the anode, the somitic cells to be electroporated, the DNA solution, and the cathode are aligned on a straight line (dotted line). After pulsing, the injected DNA integrates into the cells adjacent to the anode (in green). c: Illustration of craniocaudal somite electroporation. The graphic shows a simplified lateral view of a schematic chick embryo, the head to the right, dorsal to the top. To electroporate the caudal aspect of somite IV (Christ and Ordahl, 1995), the embryonic craniocaudal axis is gently bent by the insulated lower side of the electrodes. Thus, the anode, the somitic cells to be electroporated, the DNA solution, and the cathode are aligned on a straight line (dotted line). After pulsing, the injected DNA integrates into the cells adjacent to the anode (in green). After the experiment, the embryo will return to its normal straight position.

Download figure to PowerPoint

Electroporation Procedure

A total of 500 μl of Ringer solution containing penicillin–streptomycin are dropped on the newly injected manipulation site. To electroporate medially or laterally, electrodes are positioned left and right of the embryo at a distance of roughly the width of the embryo. Importantly, the free electrodes must not touch embryonic tissue, as this would cause severe burns. The insulated electrodes are pressed carefully onto the extraembryonic membranes, so that somites are positioned in line with the blank tips of the electrodes. The positioning of the electrodes is regulated so that the anode, the somitic compartment to be electroporated, the DNA solution, and the cathode are aligned (see Fig. 3b). Upon passing current, the negatively charged DNA, thus, will enter the somitic cells adjacent to the anode. To electroporate cranially or caudally, the embryo needs to be slightly bent along its craniocaudal axis by the insulated tips of the electrodes (Fig. 3c). To electroporate dorsally or ventrally, one electrode is placed dorsal to the embryo, the other is positioned ventral to the embryo, in the yolk (not shown). By using an Intracel Intracept TSS 10 pulser, we apply five square pulses of 60–80 V, 20-msec width. It is important to note that these parameters are to be optimized for each experimental setup. Each electroporation set-up delivers voltage/current with different fidelities. Moreover, the voltage that is required for efficient electroporation is highly dependent on the size of the free electrodes, i.e., the surface of bare wire on the otherwise insulated electrode tips. While 20–30 V are sufficient if electrodes are 10 mm in length, 60–80 V are required when free electrode length is reduced to 1–2 mm, such as those that we use. Lower voltage or electrodes positioned further apart from the embryo will decrease electroporation efficiency, higher voltage or closer position of electrodes will increase the risk of burned tissues, which result in malformations at the site of manipulation or death. Again, pilot experiments need to be performed on individual electrodes with a reporter construct such as GFP to optimize voltage in the individual experimental set-up.

To electroporate two different constructs at different locations in a somite, the first construct is electroporated as described. The remaining DNA solution is allowed to diffuse out of the somitocoele for at least 5 min. Then, the second DNA preparation is injected and rapidly electroporated in the second direction. Obviously, this double manipulation increases the risk of burns to the tissues.

Electrodes need to be gently washed in water between each electroporation, so that coagulated yolk or albumen does not gradually insulate them. Finally, the egg is resealed with tape and reincubated for the time required.

Targeting Different Somitic Lineages

For a review on somite compartments and somite-derived lineages, see, e.g., Christ and Ordahl (1995), Brand-Saberi et al. (1996), Marcelle et al. (2002). By using the electroporation protocol described above, targeting the medial portion of the epithelial somite will transfect the medial lip of the dermomyotome, which contains precursors of the epaxial myotome and the dorsomedial dermis (Fig. 4a–c). Targeting the dorsal quadrant will transfect the central dermomyotome and its myotomal and dermal derivatives (not shown). Targeting the lateral portion will transfect the lateral lip of the dermomyotome from which cells of the hypaxial myotome as well as limb muscle cells are recruited (Fig. 4e,f). Likewise, the cranial and caudal borders of the somite can be electroporated (Fig. 4d). The technique of electroporation that is described here works poorly in mesenchymal tissues, such as the segmental plate.

Verification of Transgene Expression

We have been observing GFP expression in somites in as little as 3 hr after electroporation. When GFP and the gene of interest are not on the same plasmid, it is advisable to determine the expression level of the latter by in situ hybridization (Fig. 4c) or immunohistochemistry. In most cases, however, we have observed that cells incorporate both plasmids (Fig. 4c). Importantly, electroporated embryos should display normal overall morphology, as electroporation-induced malformations (burns, bent embryo, etc.) may lead to artefactual variations in gene expression.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

We thank the members of Dr. Olivier Pourquié's lab, in particular Dr. Mike McGrew, for many helpful comments and the gift of pMiwIII vector. Figure 2 was designed by using Blender 3D 2.23. Work in our laboratory is supported by grants from the Actions Concertées Incitatives (ACI), the Association Française contre les Myopathies (AFM), the Fondation pour le Recherche Médicale (FRM), and the Association pour le Recherche sur le Cancer (ARC). M.S. was supported by a Marie-Curie Individual Fellowship MCFI-2000-01593. C.L. and J.G. are Fellows from the Ministère de l'Education Nationale, de la Recherche et des Technologies (MENRT).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES