Mouse strains and embryos

Adult mice were maintained under guidelines for hygiene, ventilation, and nutrition in the Animal Care Facility at the National Institute of Animal Health (NIAH) on 12hr light/12hr dark cycle. C3H/ HeJah mice were obtained from the Institute for Medical Research, University of Tokyo in 1963, propagated by sister-brother mating (Mannen et al., 1994), and used as a source of blastocysts for culture. C57BL/6Ncrj and Crj: CD-1(ICR) mice were purchased from Charles River Co., Japan. For natural matings, C3H/He females were placed in cages with males in the evening and were examined for vaginal plugs the following morning. The day on which a vaginal plug was detected was designated as day 0.5 of pregnancy (E0.5). For the induction of superovulation, females were given five I.U. of pregnant mare serum gonadotropin (Teikoku Zoki, Tokyo, Japan) by intraperitoneal injection, followed by five I.U. of human chorionic gonadotropin (Teikoku Zoki) after a 48 hr interval. On E3.5, C3H/He blastocysts were collected by flushing the uterine cavity with DMEM supplemented with 10% fetal calf serum (FCS). Pseudopregnant ICR females were prepared by mating with vasectomized males as described (Hogan et al., 1986).

Blastocyst culture and establishment of ES cell lines

All reagents were obtained from Sigma, St. Louis, MO, USA, unless otherwise specified. Single C3H/He blastocysts were cultured at 37°C in 35 mm plastic culture dishes containing ES medium in a humidified atmosphere of 95% air and 5% CO₂. ES medium was composed of DMEM (code #05919, Nissui, Tokyo, Japan) supplemented with 10% heat inactivated FCS (Cell Culture Lab., Oratories, OH, USA), glucose (4.5 g/l), 30 mM adenosine, 30 mM cytidine, 30 mM uridine, 30 mM guanosine, 10 mM thymidine, non-essential amino acids (Gibco BRL, Gaithersburg, MD, USA), 2-mercaptoethanol (10⁻⁴ M), leukemia inhibitory factor (LIF, 10³ unit/I, Amrad Co. Melbourne, Australia), penicillin G (31 mg/l), streptomycin sulfate (50 mg/l), and sodium bicarbonate (3.7g/l). Mouse STO fibroblasts (Martin and Evans, 1975) were treated with mitomycin C (Robertson, 1987) and seeded into the dishes on the second day of blastocyst culture. Blastocysts hatched after 2 days of culture and attached to the dish surface. After 5 days, the inner cell mass (ICM) of the blastocyst had proliferated to form a round aggregate at the center of the trophoectodermal cell sheet, which had attached and spread on the dish surface. The ICM was washed with PBS(−) and treated with trypsin-EDTA solution (0.001% trypsin and 1 mM EDTA) in PBS(−) at 37°C for a few minutes. After trypsin was inactivated by adding ES medium, the ICM was dissociated into small clumps by gentle pipetting, and the clumps were seeded into a new 35 mm dish with fresh STO feeder cells. This dissociation was designated as the first passage (P1). Culture medium was replaced every other day. Seven to ten days after dissociation of the ICM, small round colonies that were expected to contain stem cells were observed in some dishes, and these were trypsinized and seeded into new 35 mm dishes. ES cells were propagated as required, harvested, suspended in freezing medium at ca. 1 × 10⁷ cells/ml, and slowly frozen at −80°C. Freezing medium was composed of DMEM containing 10% FCS, 10% DMSO, 1% HEPES at pH 7.4. For teratoma formation, ca. 1 × 10⁷ ES cells were suspended in 1 ml PBS(−) and intraperitoneally injected into normal adult C3H/He females. After two to three months, solid tumor masses were recovered, fixed with Bouin's solution, embedded in paraffin, and examined histologically.

Immunocytochemistry

ES cells cultured on Lab-Tek chamber slides (Nunc Inc., Naperville, IL, USA) were fixed with a mixture of 95% ethanol and acetic acid (99:1, v/v) and were processed for immunocytochemistry as described (Kitani et al., 1991). Anti-C3H specific antibody (Kusakabe et al., 1988) was applied for 30 min. at room temperature. This antibody specifically recognizes one of the variants of the heat shock protein 70 family in this strain (Michikawa et al., 1993). After washing three times with cold PBS(−) for 5 min., slides were incubated for 30 min. at room temperature with a second antibody, FITC-conjugated anti-mouse IgG (No. 6250, TAGO, Burlingame, CA, USA) diluted 1:100. Then, slides were washed with cold PBS(−), and coverslips were mounted using 90% non-fluorescent glycerol in PBS(−). Normal mouse IgG₁ was used as negative control. Photomicrography was performed with an Olympus BH2-RFK fluorescence microscope.

Generation of chimeras

Chimeric mice were produced by injecting 15-25 C3H/He ES cells into each blastocyst cavity (Bradley, 1987) or 10-15 cells into the perivitelline space of 8-cell stage embryos (Tokunaga and Tsunoda, 1992) from superovulated C57BL/6 females. Passage numbers of injected ES cells were between P6 and P12. Injected embryos were cultured overnight in M16 medium, and viable embryos were transferred into the uterine cavity of pseudopregnant ICR females as described (Hogan et al., 1986). Chimeras were identified by the agouti coat color characteristic of C3H/He mice, and they were mated with C57BL/6 mice to assess germline-transmission of the injected C3H/He ES cells.

Analysis of karyotype and sex chromosomes

For karyotype analysis, ES cells were allowed to incorporate 5-bromodeoxyuridine (100 μg/ml) for 8-9 hr. During the last 1 hr, Colcemid (40 ng/ml) also was present. Harvested cells were treated 0.075 M KCl, fixed with methanol: acetic acid (3:1,v/v), and spread onto clean slides. Chromosome slides were stained with freshly prepared acridine orange and examined with a fluorescence microscope.

To determine the sex of the ES cell lines, polymerase chain reaction (PCR) was used to amplify a mouse Y chromosome-specific repetitive sequence (pY353/B) (Bradbury et al., 1990) from ES cell DNAs. For in situ hybridization, ES cell preparations were spread on glass slides and were hybridized with a digoxigenin-labeled pY353/B probe. Hybridization was detected by an alkaline phosphatase labeled anti-digoxigenin antibody (Boehringer, 1093274) according to the manufacturer's instructions. Photomicrography was performed with a Nikon Nomarski microscope.

Blastocyst culture and establishment of ES cell lines from C3H/He mice

Table 1 shows the incremental efficiencies of establishment of ES cell lines from C3H/He blastocysts. Two types of ovulation were used. From superovulated C3H/He females, a total of 124 blastocysts were cultured in 11 experiments. Successful growth of 100 inner cell masses (ICMs) (81%), 33 primary ES-like colonies (27%), and 1 permanent cell line (0.8%) was observed. From naturally ovulated C3H/He females, a total of 91 blastocysts were cultured in 5 experiments. In this group successful growth of 77 ICMs (85%), 6 primary ES-like colonies (7%) and 2 permanent cell lines (2%) was obtained. There appeared to be little difference in the efficiencies of establishment of cell lines between these two types of ovulation.

Table 1.

Establishment of permanent cell lines from C3H/He strain mice

These cell lines established from C3H/He blastocysts (termed H-1, H-2 and H-3) were confirmed as ES cell lines by the following observations. All cell lines formed colonies of tightly packed cells with prominent nucleoli were formed (Fig. 1A). This is a typical morphological feature of ES cells in culture (Robertson, 1987). The cells were maintained on an STO feeder layer by passaging them every 2-3 days at a split ratio of 1:10. Each ES cell line was considered to be clonal, because it was derived from a single blastocyst culture (See Materials and Methods). The cell lines were neoplastic when transplanted into C3H/He mice. After two to three months of intraperitoneal injection of ca. 1 × 10⁷ cells into normal C3H/He mice, all of the cell lines developed into solid tumor masses. These tumor masses contained undifferentiated as well as differentiated cells. The differentiated cells appeared to derive from all three embryonic germ layers including neural cells and cornified epithelial cells (ectodermal origin), striated muscle cells and chondrocytes (mesodermal origin), ciliated columnar epithelial cells (endodermal origin), and other unidentified tissues (Fig. 2A, B). All of the C3H/He ES cell lines were immunocytochemically reactive with a anti-CSA antibody that recognizes with a C3H-specific antigen (Kusakabe et al., 1988) (Fig.1B). STO feeder cells were not immunoreactive. A finely granular immunofluorescence was detected in the cytoplasm of the ES cells, in accordance with the known localization of this antigen in mitochondria (Michikawa et al., 1993).

Fig. 1.

(A) Phase contrast micrograph of the H-1 ES cells established from C3H/He blastocysts. The ES cell line showed typical morphological characteristics of ES cells in culture, e.g., formation of compact colonies from tightly packed small cells with prominent nucleoli. (B) The H-1 ES cell line was immunocytochemically stained with anti-CSA antibody, while STO feeder cells were not stained. (C) Phase contrast micrograph of (B).Bar=100 μm.

Fig. 2.

Histological section of a solid tumor formed by H-1 ES ceils after intraperitoneal injection into normal C3H/He mice. Undifferentiated stem cells and various differentiated cells are observed. (A) Neural cells (n), cornified epithelial cells (e), hair follicles (hf) and tracheal epithelial cells (tr). (B) Striated muscle cells (m) and chondrocytes (ch). Bar=100 μm.

Generation of chimeric mice and germline-transmission

The potential of the ES cells to form germline chimeras was tested by two methods. After blastocyst-injection, chimeric mice were obtained only from H-3 cells (Table 2A). Injection of H-3 cells into 89 C57BL/6 blastocysts resulted in 30 offspring, and 4 of them showed only slight chimerism in abdominal coat color. Each male chimera was mated with four or five C57BL/6 females, and 30 to 36 offspring were obtained from each chimeric father. None of the pups showed C3H/He agouti coat color, indicating that no germline-transmission had occurred in these male chimeras. This result is in accordance with the fact that female ES cells do not differentiate into sperm in chimeric males. H-1 and H-2 cells, did not generate chimeric mice after they were injected into 60 and 48 C57BL/6 blastocysts, respectively (Table 2A).

Table 2A.

Generation of chimeric mice by blastocyst-injection

The eight-cell embryo injection method also was tested. it has been reported that male ES cells may convert the sex of female host embryos, resulting in a male distorted sex-ratio in the postnatal chimeras (Bradley et al., 1984). in this way, chimeric males may transmit their genotype to a number of offspring. Tokunaga and Tsunoda (1993) have reported efficient production of germline chimeras by injecting male ES cells into 8-cell stage embryos. Since our preliminary experiments indicated that only H-1 ES cells were male (see below), we applied the 8-cell embryo injection method to these ES cells. After H-1 ES cells at passage number 10 were injected into the perivitelline space of 8-cell embryos from C57BL/6 (41 embryos in three experiments), 1 male chimera with overt coat color chimerism was obtained among 10 postnatal offspring (Table 2B). This chimera had an agouti coat color that was extensively intermingled throughout whole body (Fig. 3A), suggesting that the distribution of H-1 ES cells was widespread in various internal organs. This male chimera was mated with 8 C57BL/6 females. A total of 55 offspring were born, and all had agouti coat color (Fig. 3B). These results indicate that H-1 ES cells had exclusively contributed to the differentiation of germ cells in this male chimera.

Table 2B.

Generation of Chimeric mice by 8-cell embryo-injection

Fig. 3.

(A) A male chimera (left) obtained after injection of H-1 cells into the perivitelline space of 8-cell embryo from C57BL/6. This male chimera was mated with C57BL/6 female (right) to examine germline-transmission of the H-1 cells. (B) All of the pups from this mating showed agouti coat color typical of the C3H/He strain, indicating germline-transmission of H-1 cells.

Analysis of karyotype and sex chromosomes

Male ES cells have been used for most gene knockout studies (Thomas and Capecchi, 1987; Doetschman et al., 1987), because male ES cell lines are karyotypically more stable than female lines, and germline transmission of male ES cells can be assessed more easily than that of female ES cells. The sex type of these ES cell lines was determined by PCR using Y chromosome specific primers and in situ hybridization. With the PCR method, a male-specific band was amplified from H-1 cell DNA but not H-2 or H-3 cell DNA (data not shown). Figure 4 shows in situ hybridization with a digoxigenin-labeled pY353/B probe. Male specific dots were detected in the nuclei of H-1 cells (Fig. 4A). No signal was observed in the nuclei of H-3 (Fig. 4B) or H-2 cells (data not shown).

Fig. 4.

In situ hybridization of cell preparations from ES cells with digoxigenin-labeled PCR product of pY353/B. Specific signals were visualized by alkaline phosphatase labeled anti-digoxigenin antibody system and photographed with Nomarski optics. (A) In the preparation from H-1 ES cells, specific signals were detected in the nuclei. (B) In the preparation of H-3 ES cells, no signal was observed in the nuclei.

A normal euploid karyotype is an essential prerequisite for ES cells to differentiate into functional germ cells in chimeric mice. The low efficiencies with which our C3H/He ES cells formed germline chimera may be due to low proportions of normal euploid cells in these cell lines. Accordingly, the karyotype was examined at the 10th passage by an R-banding technique after BrdU incorporation (Table 3). The sex chromosome constitution was XY in H-1, and XX in H-2 and H-3 confirming the results obtained by PCR and in situ hybridization. In H-1, 32 out of 42 cells examined had 40 chromosomes, but 14 cells contained a long marker chromosome most probably derived from chromosome 3. The karyotype was apparently normal in 18 H-1 cells, and 9 hyperploid cells were characterized by trisomy 11. Approximately half of the cells in mitosis were tetraploid in H-2. Only 13 of 68 cells in the diploid range had 40 chromosomes, and most of the remaining cells had an extra chromosome 11. In H-3, more than half of the cells examined had 41 chromosomes. Here again, trisomy 11 was observed most frequently. Thus, it may be concluded that the high proportion of abnormal karyotypes in these cell lines was at least partly responsible for the low frequencies of germline chimeras after injection into blastocysts or the perivitelline space of 8-cell embryos. Recloning of these C3H/He ES cells is currently in progress.

Table 3.

Karyotype of ES cell lines from C3H/He strain mice