Ex ovo electroporation for gene transfer into older chicken embryos



In ovo electroporation is an excellent method to ectopically induce or inhibit gene expression in chicken embryos and to study the in vivo function of genes during embryonic development. However, the application of electroporation in ovo to date is limited to an early stage of incubation (< stage 20). In older embryos (> stage 22), the vitelline and allantoic vessels have developed extensively and the in ovo manipulation of the embryo becomes exceedingly difficult. Therefore, in this study, we validate an ex ovo electroporation system, by which the time for performing electroporation can be extended up to at least day 7 of incubation. The application of this method will help to study gene function and regulation at later stages of development in the living chicken embryo. Developmental Dynamics 233:1470–1477, 2005. © 2005 Wiley-Liss, Inc.


The chicken embryo provides an excellent model system to study gene function and regulation during embryonic development in vivo. The technique of in ovo electroporation has been recognized as a powerful method to induce efficiently exogenous genes into the chicken embryo (Muramatsu et al.,1997; Momose et al.,1999; Nakamura and Funahashi,2001; for review, see Muramatsu et al.,1998; Itasaki et al.,1999; Yasuda and Nakamura,2000; Krull,2004). By in ovo electroporation, one or more cDNA plasmids, RNA for interference, or morpholinos can be introduced into embryonic cells to overexpress genes of interest or knock-down their expression in restricted regions (Momose et al.,1999; Bourikas and Stoeckli,2003; for review, see Kos et al.,2003; Krull,2004; Nakamura et al.,2004; Rao et al.,2004). Although transfected cDNA plasmid is not incorporated into the host chromosomal DNA, the expression of exogenous genes can last more than 8 to 11 days after electroporation (Luo et al.,2004; Luo and Redies,2004). Because the chicken embryo can be accessed and manipulated easily, in ovo electroporation is widely used to investigate questions of vertebrate development (Funahashi et al.,1999; Momose et al.,1999; Fujii et al.,2000; Tucker,2001; Katahira and Nakamura,2003; Pekarik et al.,2003; Matsuda and Cepko,2004; for review, see Itasaki et al.,1999; Hu et al.,2002; Kos et al.,2003; Nakamura et al.,2004). The method has been applied to studies of neurogenesis and neuronal differentiation (Bylund et al.,2003; Cheng et al.,2003; Eberhart et al.,2004), axon outgrowth (Fujii et al.,2000; Cramer et al.,2004), somitogenesis (Scaal et al.,2004), limb development (Yamamoto and Kuroiwa,2003; Shah et al.,2004), skeletal muscle function (Muramatsu et al.,2001; Yasui et al.,2001; for review, see McMahon and Wells,2004), eye development (Chen et al.,2004; Leconte et al.,2004; Yan et al.,2004), and gene therapy (for review, see Gehl,2003).

In general, in ovo electroporation is applied only to chicken embryos at early stages of incubation (i.e., embryos younger than stage 20 according to Hamburger and Hamilton,1951; Muramatsu et al.,1997; Momose et al.,1999; for review, see Itasaki et al.,1999; Krull,2004). At later stages (stage 22 and older), several experimental problems arise for in ovo electroporation (see Results and Discussion section), making it exceedingly difficult to perform electroporation. As a solution to these problems, we have used ex ovo (shell-less) culture for electroporation in chicken embryos (Treubert-Zimmermann et al.,2002; Luo et al.,2004; Luo and Redies,2004). For ex ovo culture, the whole content of the egg is transferred into a Petri dish system (Auerbach et al.,1974). Ex ovo culture of chicken embryos has been used, e.g., to study the effect of functionally blocking antibodies (Oberlender and Tuan,1994; Gänzler-Odenthal and Redies,1998), calcium-deficient hypertension (Koide and Tuan,1989), atherosclerosis (Koide et al.,1996), muscle maturation (Meinnel et al.,1989), intravenous tracer application (Heyers et al.,2003), and to test biosensors (Valdes et al.,2003). In the present work, we demonstrated by ex ovo culture that genes (cadherin7 [Cad7] and green fluorescent protein [GFP]) can be transferred successfully into a variety of central nervous system regions and muscle of chicken embryos of 4 to 7 days of incubation (embryonic day [E] 4 to E7; stage 23 to 30). Our results indicate that the ex ovo electroporation system can be used to study the function of genes at relatively late stages of chicken embryonic development.


In this study, we used an ex ovo approach to electroporate developing tissues of chicken embryos at stages later than E3.5. A schematic diagram of the different steps of ex ovo electroporation is shown in Figure 1 (see also Experimental Procedures section). In contrast to in ovo electroporation, the eggs were cracked with a metal edge at approximately E2.5 and the entire contents of the eggs, including the embryos, were transferred into a Petri dish system for further ex ovo culture. Between E4 and E7, we electroporated a mixture of expression plasmids containing the Cad7 and GFP genes into different areas of the chicken embryos (Fig. 2). Embryos were killed and processed for immunostaining at the desired stages (E6–E9).

Figure 1.

A schematic flow diagram of ex ovo electroporation. Fertilized eggs are placed in an egg incubator at 37°C (Day 0). After 2.5 days of incubation (Day 2.5), the eggs are cracked with a metal edge and the entire contents, including the embryos, were transferred into a Petri dish system. The Petri dishes (with lid, not shown) are placed into an incubator at 37°C for further culture. At the desired stage (e.g., Day 4), the plasmids are injected into the target area, immediately followed by electroporation. The transfected embryos are further incubated until the desired stage (e.g., Day 6). The success of electroporation is examined under an epifluorescence stereomicroscope, and the embryos are fixed and processed for immunostaining. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Figure 2.

Gene transfer into chicken embryos by ex ovo electroporation. AF: Schematic representation of plasmid injection (green) and electrodes placement [anode plate (+), cathode needle (−)] for gene transfer into the telencephalon (A), diencephalon (B), mesencephalon (C), cerebellum (D), spinal cord (E), and wing tissue (F), respectively. The dotted lines outline the ventricular system at embryonic day (E) 4. GL: The expression of green fluorescent protein (GFP; green) marked by arrows in whole-mount preparations at E6 (G–K) and E9 (L) after electroporation of the telencephalon (G), diencephalon (H), mesencephalon (I), cerebellum (J), spinal cord (K), and wing tissue (L), respectively. Embryos are viewed from the lateral side. The bright yellowish spots (e.g., arrrowhead in I) represent artefacts of illumination (light reflections). MR: Immunostaining for cadherin7 (Cad7) in transverse sections through the embryos depicted in G–L. Almost all GFP-positive cells (green) coexpress Cad7 (red, arrows). It should be noted that some red regions, which do not coexpress GFP, represent areas that express Cad7 endogenously (arrowheads in M–Q), for example, fiber tracts at the surface of the brain (M–P; Yoon et al.,2000) or Cad7-positive domains of the basal plate (P, Q; Ju et al.,2004). The areas boxed in M–P are magnified in S–V for adjacent sections. The dotted circle in R surrounds muscle tissue. The arrowheads in R indicate coexpressing cells outside muscle. The orientation of the sections is given in each panel (d, dorsal; v, ventral). SV: Higher magnification of the electroporated areas in sections adjacent to M–P, respectively. Cad7 (red) is expressed on the surface of most GFP-positive cells (green). The dashed lines indicate the ependymal lining. WX: Detection of apoptotic cells after electroporation with terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) assay. The distribution of apoptotic cells (white dots, indicated by arrows) has no relationship with electroporation (green cells). III, third ventricle; IV, fourth ventricle; cb, cerebellum; cart, cartilage; di, diencephalon; E, day of incubation; sg, spinal ganglion; tect, tectum; tel, telencephalon. Scale bars = 5 mm in G (applies to G,J) and in H (applies to H,I),K,L, 200 μm in M (applies to M–P) and in Q (applies to Q,R,W,X), 20 μm in S (applies to S–V).

Advantages of Performing Electroporation in Older Embryos Ex Ovo

The ex ovo method eliminates several problems associated with performing electroporation in ovo at later stages of chicken development. First, from stage 22, the vitelline membranes of the chicken embryo begin to stick gradually to the shell membranes, even if the eggs are turned around in the incubator. By opening a window in the shell for in ovo electroporation at these late stages, the vitelline blood vessels can be easily injured, often resulting in the death of the embryo. Second, at later stages of in ovo culture, vitelline and allantoic vessels cover the embryo and the accessing and manipulation of the embryo become more and more difficult. Third, older embryos move more strongly and their orientation is difficult to control. Finally, at late stages, the embryos have grown considerably and are difficult to get to through the relatively small windows in egg's shell. All these factors make it exceedingly difficult to microinject and electroporate embryos at later stages by in ovo culture. In contrast, with the ex ovo culture in Petri dishes, the yolk as well as the vitelline and allantoic vessels spread widely and the vessels can be detached from the embryo more easily, making the embryo more accessible. Moreover, the relatively shallow Petri dish (height of 16 mm) confines the movement of the embryos.

Optimizing Conditions for Ex Ovo Culture in Petri Dishes

On the basis of our experience, eggs are best cracked at approximately E2.5 for ex ovo culture. Cracking before and after E2.5 increases the number of embryos dying during culture (Luo and Redies,2004). When cracking the egg, it is important to avoid injury of the vitelline membrane by the cracked shell's edges and to transfer the egg's content swiftly but gently into the Petri dish. The embryo usually floats on top of the egg yolk. Therefore, the crack should be made at the bottom of the egg, and the egg should be opened at the bottom over the Petri dish without turning, so that the embryo stays on top of the yolk after cracking. Generally, the efficiency of cracking increases gradually with experience.

One of the main disadvantages of ex ovo culture is that many embryos die during culture. To increase the survival rate, culture conditions should be optimized along the following lines: (1) To avoid bacterial and fungal infection of embryos, the first (smaller) Petri dish that directly holds the egg's content should be sterile. Egg cracking, embryo transfer into Petri dishes, and electroporation should be carried out in a sterile hood. The incubator should be kept clean of any egg content that may have been spoiled, to avoid sources of infection inside the incubator. (2) To prevent the desiccation of the embryos, the first Petri dish is placed into a larger Petri dish filled with sterile distilled water (Fig. 1). We reuse these larger (plastic) dishes several times after cleaning them in a washing machine for laboratory glassware, without autoclaving. The second Petri dish is covered by a lid, which allows some air movement into the dish. Moreover, the interior of the incubator is humidified by placing a container with water at the bottom of the incubator. (3) For a constant supply of fresh air into the incubator, a small air pump (as used for aquariums) is used to inject room air into the incubator via plastic tubing. Inside the incubator, a small computer fan moves the air around constantly. A sufficient supply of fresh air increases the survival of embryos postelectroporation. (4) To increase survival, relatively fresh eggs should be used. Local breeders of newly hatched chickens are a good source of fertilized eggs. There can be seasonal variations in the quality of the eggs. Care should be taken to avoid exposure of the eggs to mechanical shock, heat, or low temperatures during transport to the laboratory. In the laboratory, eggs are kept in a refrigerator at 10°C to prevent any premature development before incubation. Regular household refrigerators often have a working range at lower temperatures. We use a special refrigerator for wine bottles that can be set to 10°C. Fresh eggs can be kept at 10°C for 1–2 weeks, with the number of dying embryos beginning to increase after 1 week.

Even under optimized conditions, embryos die during the culture. In our experience, starting with 100 incubated eggs, 75 embryos are available for electroporation at E3.5, 60 embryos survive the first day after electroporation, 20 embryos are still alive at E8.5, and 3–4 embryos at E15 (Luo and Redies,2004).

Injection of Plasmids, Electrodes, and Target Areas of Electroporation

To perform electroporation, the solution containing the Cad7 and GFP plasmids was injected into the lumen of the ventricular system or wing tissue (Fig. 2A–F). The Cad7 plasmid was injected at a higher concentration than the GFP plasmid (ratio approximately 8:1) to ensure that all GFP-positive cells expressed Cad7. Our results show that the vast majority of cotransfected cells express both cDNAs (Fig. 2; see also Treubert-Zimmermann et al.,2002; Luo et al.,2004), as observed previously in other studies (Haas et al.,2001). Theoretically, it would be preferable to have both genes expressed from the same plasmid. However, such tandem constructs do not always result in high levels of transcription of both genes in our experience (Luo and Redies, unpublished observations).

For gene transfer into the telencephalon, diencephalon, mesencephalon, and cerebellum, the anode plate was placed on the dorsal side of the target area and the cathode needle was inserted into the mesenchyme at the ventral side of the brain (Fig. 2A–D). For gene transfer into the spinal cord and wing, the electrodes were placed on opposite sides of the target area outside the tissues, i.e., the cathode was not inserted into the tissues (Fig. 2E,F). After electrode placement, the electric pulses (Table 1) were applied immediately.

Table 1. Efficiency of Ex Ovo Electroporation in Different Areas of the Chicken Embryoa
TissueNumber of electroporated embryos (E4)Number of surviving embryos (E6)Number of GFP-positive embryos and efficiency of electroporation (%)bElectroporation parameters
  • a

    E, day of incubation; GFP, green fluorescent protein; V, volts.

  • b

    Expressed as a percentage of surviving embryos.

Telencephalon1285 (62%)18 V, 60 msec, 6 pulses
Diencephalon1265 (83%)18 V, 60 msec, 6 pulses
Mesencephalon1387 (87%)18 V, 60 msec, 6 pulses
Cerebellum966 (100%)14 V, 60 msec, 6 pulses
Spinal cord1199 (100%)18 V, 60 msec, 6 pulses
Wing6 (E7)5 (E9)5 (100%)25 V, 60 msec, 6 pulses

Exogenous Gene Expression After Electroporation

In the present study, the expression of GFP in whole-mount preparations began at around 4 hr after electroporation and it was strong enough for photography after 6 hr (data not shown). Other studies carried out in ovo at younger stages show that the expression of GFP has reached levels high enough for photography at around 2–3 hr after in ovo electroporation (Funahashi et al.,1999; Momose et al.,1999). This time difference of GFP expression between the literature and our data may result from the use of older embryos, which have, in general, slower rates of proliferation and differentiation than early embryos, perhaps resulting in slower rates of GFP protein synthesis. We routinely visualized GFP fluorescence in all whole-mount specimens under an epifluorescence stereomicroscope to evaluate the success of electroporation before fixation and further processing (arrows in Fig. 2G–L).

In immunostained sections of electroporated areas, almost all GFP-positive cells (green) coexpressed Cad7 (red, Fig. 2M–V). Our results show that the ex ovo electroporation method presented here can be used to introduce genes not only into the tectum and cerebellum (Fig. 2I,J,O,P,U,V; Treubert-Zimmermann et al.,2002; Luo et al.,2004) but also into the telencephalon, diencephalon, and spinal cord (Fig. 2G,H,K,M,N,Q,S,T). Furthermore, other tissues such as muscle (dotted circle in Fig. 2R) and peripheral nerve (data not shown) can be also transfected by ex ovo electroporation.

It should be noted that, in sections immunostained for Cad7 (Fig. 2M–V), some regions express Cad7 endogenously (red); hence, they do not coexpress GFP (green; arrowheads in Fig. 2M–Q). The level of exogenously induced Cad7 expression (arrows in Fig. 2M–Q) is at least as high as endogenous Cad7 expression (arrowheads in Fig. 2M–Q).

Efficiency of Electroporation

The efficiency of ex ovo electroporation, expressed as the percentage of embryos with GFP fluorescence, can reach high levels (80–100%; Table 1). In our experience, a maximum of approximately 50% of cells in a target area were transfected. It should be noted that this percentage was not consistent from embryo to embryo.

It is known that the efficiency of gene transfer depends on the size of the electrodes, their positioning around the target tissue, and the selection of the electric pulse sequence (Momose et al.,1999). As in electrophoresis, the plasmid cDNA moves to the anode during electroporation. Consequently, the target tissue should be placed between the anode and the injected plasmid. In our study, we chose a rectangle plate (3 × 7 mm) as an anode (Fig. 1). This plate anode ensures that the electric field is large enough and covers the entire target area for an efficient gene transfer. For the cathode, a tungsten needle is inserted into the tissues beside the target area and opposite to the anode (Fig. 2A–D). For embryos of E2.5 or younger, both electrodes are usually placed outside the embryo. The cathode placement inside the embryonic tissue is possible in older embryos due to their larger size. The insertion of the cathode into the tissue allows relative low voltage for electroporation and produces a higher efficiency of gene transfer (Momose et al.,1999). It should be noted that the cathode needle should not be inserted into the ventricular system because hydrogen bubbles, generated during electroporation, can produce a toxic effect on the embryos and result in their death. Furthermore, to obtain efficient gene transfer, the voltage of the electric pulse should be higher with increasing distance between the anode and the cathode (Table 1; Fig. 2A–D).

The restriction of electroporation to a small area results from the combination of plasmid injection and the placement and size of the electrodes. In our experiments, the site of electroporation is determined by the highly localized plasmid injection through a micropipette and the placement of the tip of the needle cathode. The large plate anode contributes little to the regional restriction of electroporation. Single-cell electroporation can be achieved if a micropipette is used for both plasmid application and electroporation (Haas et al.,2001).

The incidence of cell death after electroporation was also investigated in the present study. We did not see any difference in apoptosis between the electroporated target areas and the corresponding control areas on the other side of the embryo (Fig. 2W,X).

Disadvantage of Ex Ovo Culture for Embryos Older Than E10

Before E10, chicken embryos mobilize calcium exclusively from the yolk and the development of chicken embryos appears to be indistinguishable between ex ovo and in ovo culture (De Gennaro et al.,1980). In fact, in our study, we have not found any differences in the size of the embryos up to stage E10. Moreover, the number of apoptotic cells in different brain areas at E6 was exceedingly low and similar in embryos cultured in ovo and ex ovo (data not shown).

After E10, the shell is the main source for extraembryonic calcium supply. Thereafter, chicken embryos grown in ex ovo culture become calcium deficient (Tuan and Ono,1986). An extra supply of eggshell, placed onto the chorioallantoic membrane, may partially reconstitute the supply of calcium for late-stage embryos in the ex ovo cultures (Tuan,1983; Dunn et al.,1987). Moreover, in ex ovo cultures, the level of serum insulin-like growth factor-I does not rise in mid-embryogenesis, as is observed for in ovo cultures (Robcis et al.,1991).

In general, the main disadvantage of ex ovo cultures is growth retardation after E10, followed by death of the embryos in the third (final) week of incubation. The survival and growth of the chicken embryos are also affected by the geometry of the culture vessel and the gas exchange through the culture chamber walls (Dunn et al.,1981; Jakobson et al.,1989; Dugan et al.,1991). Incubation chambers, which are constructed to resemble the egg shell more closely, lead to an increased survival rate (Dunn et al.,1981; Jakobson et al.,1989; Dugan et al.,1991). However, in such chambers, the embryo can move more freely and might be less accessible.


In the present study, we transfected Cad7 and GFP expression plasmids into different tissues of older chicken embryos. By using ex ovo electroporation, we have demonstrated previously that cadherins play a role in the axonal pathfinding of optic tectum and the migration of Purkinje cells to cerebellum cortex during chicken embryonic development (Treubert-Zimmermann et al.,2002; Luo et al.,2004). The transfer of other genes or cotransfer of more than two genes by this method is possible. Thus, the ex ovo electroporation method described here is a valuable tool to study gene function and regulation in older chicken embryos.


Animal, Plasmid, and Antibody

The experiments were carried out using White Leghorn chicken (Gallus domesticus). Fertilized eggs were incubated in a forced-draft egg incubator (BSS160, Ehret, Germany) at 37°C. Embryos were staged according to Hamburger and Hamilton (1951).

The plasmid pCAGGS-GFP was a kind gift of Dr. H. Ogawa, National Institute of Basic Biology, Okazaki, Japan (Momose et al.,1999). This plasmid was produced by excising the GFP coding sequences from plasmid pCMX-SAH/Y145F (Ogawa and Umesono,1998) and inserting it into blunted EcoRI site of the plasmid pCAGGS, a eukaryotic expression vector carrying a CMV/chicken β-actin promoter (Niwa et al.,1991). Full-length Cad7 cDNA from plasmid pBluescript-cad7 (kind gift of Dr. S. Nakagawa and Dr. M. Takeichi, Kyoto University, Kyoto, Japan) was cloned into blunted EcoRI sites of the plasmid pCAGGS after removing the GFP insert. Plasmids were purified from XL1-blue strains of Escherichia coli (Stratagene, La Jolla, CA) using Qiagen columns (Qiagen, Hilden, Germany) and solubilized in Gey's balanced salt solution (GBSS, Sigma).

For immunostaining of sections, a mouse monoclonal antibody against Cad7 (CCD7-1; Nakagawa and Takeichi,1998) was used, followed by Cy3-labeled secondary antibody against mouse IgG (Jackson ImmunoResearch). Dye Hoechst 33258 (Molecular Probes) was used for staining of cell nuclei.

Ex Ovo Electroporation

The process of ex ovo culture and electroporation is shown schematically in Figure 1. Some of the experimental details have already been mentioned above (Results and Discussion section). Briefly, the fertilized eggs were incubated in a forced-draft egg incubator at 37°C. At approximately E2.5, the incubated eggs were cracked with a sharp metal edge and the whole content of each egg, including the embryo, was transferred into a Petri dish (diameter of 90 mm and height of 16 mm). This Petri dish was placed without lid into a second Petri dish (diameter of 150 mm and height of 20 mm) that was filled with approximately 20 ml of sterile distilled water. The second Petri dish was covered by a lid and returned into an aerated, forced-draft incubator (B6120, Heraeus, Germany) for further culture at 37°C.

For plasmid injection at E4 or E7, the vitelline and amnion membranes over the target area were carefully torn with fine forceps. Then, 0.1–0.5 μl of GBSS solution containing a mixture of Cad7 plasmid (2 μg/μl) and GFP plasmid (0.25 μg/μl) and 0.1% Fast Green (Sigma) was injected into the ventricular system or the tissues close to the target area (Fig. 2A–F) by means of a glass capillary held by a mouth piece (A5177, Sigma). The glass capillary was pulled on a microelectrode puller (PUL-100 Micropipette Puller, World Precision Instruments, Berlin, Germany) from glass tubes of 1.0-mm diameter (TW100F-4, World Precision Instruments). For electroporation, an electrode with a tungsten needle cathode and a rectangle plate anode (3 × 7 mm; CUY 661-3×7, Nepa Gene, Chiba, Japan; Fig. 1) was positioned around the target area (Fig. 2A–F). Electric pulses (Table 1) were applied immediately after injection of plasmids by an electroporator (CUY 21, Nepa Gene). Electroporated embryos were promptly returned to the incubator. Two days after electroporation, GFP fluorescence was visualized in whole-mount preparations under a fluorescent microscope (Leica, MZFLIII, Germany) equipped by a computer-based digital camera (PCO CCD Imagine, PixelFly, Kelheim, Germany). Embryos were then fixed with 4% formaldehyde in HEPES-buffered salt solution (HBSS) for 4–6 hr on ice.


For the detection of Cad7 and GFP expression, immunohistochemical staining was performed, as described previously (Luo and Redies,2004). Cryostat sections of 20 μm thickness were post-fixed in 4% formaldehyde in HBSS for 30 min on ice and washed three times with Tris-buffered saline (TBS, pH 7.6) containing 1 mM Ca2+ and Mg2+. After the sections were preincubated with blocking solution (5% skimmed milk, 0.3% Triton X-100 in TBS) at room temperature for 60 min, antibody against Cad7 was applied at room temperature for 60 min. Then, sections were incubated with Cy3-labeled secondary antibody against mouse IgG (1:300 dilution). Finally, cell nuclei were stained with dye Hoechst 33258 (Molecular Probes). Fluorescence was imaged under a fluorescence microscope (Olympus BX40, Hamburg, Germany) equipped with a digital camera (Olympus DP70). Digitized images were adjusted in contrast and brightness with the Photoshop software, if necessary (Adobe Systems, Mountain View, CA).

Analysis of Apoptosis

Apoptosis was compared between transfected and untransfected sides in electroporated embryos (six E6 embryos) and between embryos cultured ex ovo and in ovo at E6 (one embryo each). To detect apoptotic cells, cryostat sections were stained using the In Situ Cell Death Detection Kit/TMR Red (Roche, Germany) according to the manufacturer's instructions. Briefly, after post-fixation in 4% formaldehyde, cryostat sections were washed with PBS (pH 7.4). Slides were then incubated in permeabilization solution (0.1% Triton X-100 in 0.1% sodium citrate solution) for 2 min on the ice. The terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) reaction mixture was added to the slides and the sections were incubated at 37°C for 60 min in the dark. Finally, cell nuclei were stained with dye Hoechst 33258.


We thank Dr. H. Ogawa for his kind gift of pCAGGS-GFP plasmid, Dr. S. Nakagawa and Dr. M. Takeichi for Cad7 plasmid, Mrs. U. Rother, Mrs. M. Theune, and Mrs. G. Carlson for technical assistance, and Mr. J. Geiling for drawing the illustrations shown in Figure 1.