- Email: [email protected]
Development of bovine and porcine embryonic teratomas in athymic mice
ANIMAL REPRODUCTION SCIENCE ELSEVIER
Animal Reproduction
Science 45 ( 1996) 23 I-240
Development of bovine and porcine embryonic teratomas in athymic mice G.B. Anderson a,*, R.H. BonDurant b, L. Goff a, J. Groff ‘, A.L. Moyer a aDepartment of Animal Science, Universiry ofCalifornia. Davis, CA 95616. USA b Department ofPopulation Health and Reproduction, Universiry of California. Davis, CA 95616, USA ’Department of Pathology. Microbiology and Immunology. University of Calijornia, Davis, CA 95616. USA Accepted 2 May 1996
Abstract discs from bovine and porcine blastocysts of various the kidney capsule of athymic (nude) mice to evaluate growth of teratocarcinomas containing both differentiated tissues and undifferentiated stem cells. Inner cell masses were isolated immunosurgically from Day 8, Day 9 and Day 10 porcine blastocysts and from Day 8, Day IO and Day 12 bovine blastocysts. Embryonic discs were mechanically dissected from Day 11 and Day 12 porcine embryos and from Day 14 bovine embryos. Day 6 egg cylinders were dissected from BALB/C embryos and from hybrid embryos of a cross between BALB/C and an outbred strain of mouse. Two to four KM, embryonic discs or egg cylinders were transplanted under the kidney capsule of each athymic host. After 8 weeks, graft hosts were killed and their tumors removed, fixed and prepared for histological and immunohistochemical examination. Embryonic teratomas developed at high frequency from murine egg cylinders and from Day 11 and Day 12 porcine and Day 14 bovine embryos. Tumors were observed only infrequently from younger bovine and porcine blastocysts. Murine embryonic tumors were composed of numerous differentiated cell types of ectodermal, mesodermal and endodennal origins, but representation of the three embryonic germ layers was somewhat more restricted in bovine and porcine embryonic tumors. No undifferentiated stem cells were detected in tumors of any of the three species. These results demonstrate that teratomas will develop from bovine and porcine embryos when grafted to an immunocompromised host, but the presence of undifferentiated teratocarcinoma stem cells from these species has yet to be achieved. Inner
ages
were
Key~lorks:
cell masses
transplanted
(KM)
Cattle-embryology;
* Corresponding
and embryonic
under
EC cells; Pig-embryology;
author. Tel: 916/752-1682.
0378.4320/96/$15.00 Published PII SO378-4320(96)01581-3
Teratocarcinomas;
Fax: 916/752-0175.
by Elsevier Science B.V.
Teratomas
e-mail: [email protected]
232
G.B. Anderson et al./Animal
Reproduction
Science 45 (1996) 231-240
1. Introduction Embryonic stem (ES) cells are undifferentiated embryonic cells that retain in culture the capacity to differentiate into both somatic and germ cell lineages. Murine ES cells have been used extensively as models for embryogenesis in the study of cell lineage, regulation of gene expression and genetic manipulation (Robertson, 1991; Wilmut et al., 1991). Their potential use for genetic manipulation in agricultural species is widely appreciated (Anderson, 19921, but successful isolation of ES cells from embryos of the livestock species remains to be fully documented. Establishment of ES cells is usually accomplished by culturing intact blastocysts or isolated inner cell masses (ICM). An alternate approach used in mice to produce pluripotent embryonic cell lines is by inducing formation of a teratocarcinoma, a tumor of embryonic origin that contains not only various differentiated somatic tissues but also pluripotent stem cells. The undifferentiated stem cells, called embryonal carcinoma or EC cells, share numerous properties with ES cells, including the ability to differentiate into normal tissues when injected into a blastocyst (Brinster, 1974; Mintz and Illmensee, 1975; Papaioannou and Rossant, 1983). Teratocarcinomas occur spontaneously in some strains of mice but can be induced by transplantation of early embryos to extrauterine sites. Murine teratocarcinomas have been produced by transplantation of embryos to the testis (Stevens, 1984) or under the kidney capsule (Solter et al., 1970; Damjanov et al., 1983) of histocompatible hosts. The objective of this study was to examine species-specific characteristics of embryonic-cell differentiation by determining if bovine and porcine teratocarcinomas can be induced by transplantation of early embryos under the kidney capsule of an immunocompromised graft host. Athymic mice were used as graft hosts because of their capacity to accept xenografts.
2. Materials and methods 2.1. Collection and manipulation of embryos Porcine blastocysts were recovered from gilts by flushing the uterine horns with Dulbecco’s modified phosphate-buffered saline (PBS) supplemented with 1% calf serum (Gibco, Grand Island, NY) on Day 8,9, 10, 11 or 12 of the estrous cycle (Day 0 was the first day of estrus). Bovine blastocysts were recovered by non-surgically flushing the uteri of superovulated beef cows on Day 8, 10 or 12 of the estrous cycle (Day 0 was the day of est.&. Day 14 bovine embryos were recovered at slaughter from superovulated beef cows. Inner cell masses were isolated by immunosurgery (Piedrahita et al., 1990) from Day 8, Day 9 and Day 10 porcine blastocysts and Day 8, Day 10 and Day 12 bovine blastocysts. Embryonic discs from Day 11 and Day 12 porcine and Day 14 bovine embryos were isolated mechanically using two 27-gauge needles under a dissection microscope. Intact Day 8 bovine blastocysts obtained from procedures for in vitro oocyte maturation and fertilization and embryo culture (Behboodi et al., 1992)
G.B. Anderson et al./Animul
Reproduction Science 45 (1996) 231-240
233
were also tested for the capacity to establish embryonic grafts. Egg cylinder stage murine embryos were collected from Day 6 (Day 0 was the day of mating plug) BALB/C females previously mated to BALB/C males (inbred embryos) and from JU females (Eklund and Bradford, 1976) mated to BALB/C males (outbred embryos). Murine inbred embryos were from the same strain as the graft hosts; murine outbred embryos were from a cross between the graft host strain and an outbred strain. Egg cylinders were removed from extraembryonic tissues as described by Damjanov et al. (19871. 2.2. Transplantation
of embryonic grafts to athymic hosts
Athymic (nu/nu> BALB/C mice were used as hosts for embryonic grafts. Athymic mice were maintained on sterile bedding in filter-top, microisolator cages. Most graft hosts were 7-8 weeks of age at transplantation. Male and female hosts were randomly distributed among treatments. Hosts were anesthetized with an intraperitoneal injection of 0.3 cc of a 10% solution of sodium pentabarbital (Western Medical Supply, Arcadia, CA) and the right kidney was exposed through a lateral, oblique incision in the skin and body wall. Watchmaker forceps were used to make a small tear in the kidney capsule and two to four ICM, embryonic discs, egg cylinders or intact Day 8 in vitro-derived bovine blastocysts were deposited under the kidney capsule using a tapered 3 ~1 micropipette (Drummond Wiretrol, Drummond Scientific Co., Broomall, PA). The flank incision was closed with two 9 mm autoclips (Becton, Dickenson and Company, Parsippany, NY), which were removed 7 days after surgery. Seven to 11 athymic graft hosts were used per treatment. 2.3. Tumor recovery and evaluation Athymic graft hosts were killed by cervical dislocation 8 weeks after transplantation and kidneys were removed and examined for evidence of tumor development. Tumors were carefully excised from the kidney, weighed, measured at their largest diameter and fixed in 10% buffered formalin for paraffin embedding. Histological sections were prepared at 6 km and stained with hematoxylin and eosin using standard procedures. Histological sections were evaluated by light microscopy for the presence of various differentiated cells. Tumors were classified as benign teratomas or malignant teratocarcinomas according to criteria described by Solter et al. (1979) and Damjanov et al. (1987). Selected histological sections were also stained for alkaline phosphatase activity (Sigma Diagnostics Alkaline Phosphatase, Leukocyte Testing Kit), which is expressed by murine EC cells. Selected tumors were stained immunohistochemically for the presence of cytokeratins, vimentin, carcinoembryonic antigen (CEA), o-fetoprotein (a-F-P), and stage-specific embryonic antigen-l @SEA-I), all of which are markers for various differentiated and undifferentiated cell types. Some sections were also tested against anti-cytokeratin antibodies including AE 1/AE3, a combination monoclonal antibody that has a broad range of specificities, and CAM 5.2, which is often used to recognize simple epithelium.
234
G.B. Anderson er al. / Animal Reproduction Science 45 (1996) 231-240
Cytokeratin expression is typical of epithelial differentiation in avian and mammalian species, whereas vimentin is typically expressed in cells of mesenchymal origin. Certain normal and malignant epithelial tissues and epithelial-derived cell lines may co-express vimentin (McNutt et al., 1985; Traub, 1985) Carcinoembryonic antigen is a complex glycoprotein expressed by various neoplasms, is normally produced in embryonic endodermal tissue (gut, liver and pancreas) and has been used as a marker for neoplasms of these tissues (Carney, 1988; Sheahan et al., 1990). a-fetoprotein is a glycoprotein synthesized by the fetal yolk sac, liver and gastrointestinal tract and has been used as a tumor marker for hepatic, testicular germ cells and yolk sac neoplasms (Jacobsen et al., 1981; Sheahan et al., 1990). Stage-specific embryonic antigen-l is expressed by murine undifferentiated cells including ICM, ES cells and EC cells (Mummery et al., 1990). A standard streptavidin-biotin-immunoperoxidase complex (ABC) technique was used for immunohistochemical staining procedures (Hsu et al., 1981). Embedded tissues were deparaffinized, cleared with xylene and dehydrated through a graded series of alcohol solutions. Tissues were blocked for endogenous peroxidase activity with 0.3% hydrogen peroxide in methanol for 30 min at room temperature and rinsed with PBS (pH 7.4). Non-specific staining was blocked with 10% heat-inactivated normal serum in PBS for 20 min at room temperature; normal serum was either horse or goat according to the secondary antibody species. The tissue sections were incubated with the primary antisera diluted in PBS (pH 7.4) at room temperature for 1 h, except for o-F-P, which required a 2-h incubation. Following incubation and each subsequent step, slides were rinsed in two changes of PBS for 5 min. The biotinylated secondary antibody diluted 1:500 in PBS was applied to tissue sections for 30 min, followed by avidin-biotinhorseradish peroxidase for 20 min. Biotinylated horse-anti-mouse or goat-anti-rabbit antibodies (Vector Laboratories, Burlingame, CA) were used as secondary antibodies against the primary monoclonal or polyclonal antibodies, respectively. Antibody localization was determined by the use of 3,3_diaminobenzidene (DAB) in imidazole buffer. Aminoethyl carbazole (AEC) was used as the chromogen for detection of SSEA-1. Slides were counterstained with Mayer’s hematoxylin, washed in tap water, dehydrated to xylene and coverslipped. Representative canine tissues were used as positive controls for AEl/AE3, CAM 5.2 and vimentin. Canine embryos, approximately 3 weeks of age and procured from clinical spay procedures, were used as positive controls for CEA and o-F-P. Feline gastrointestinal tissues were used as positive controls for SSEA-1 expression. For negative controls, PBS was used in place of the primary antisera. Antibodies included the rabbit-polyclonal anti-vimentin (ICN Biomedicals, Inc., Costa Mesa, CA), anti-a-F-P and anti-CEA (Dako Corporation, Carpinteria, CA) antibodies and the murine monoclonal antibody AEl/AE3 specific for the high and low molecular weight cytokeratins (Boehringer Mannheim Corp., Indianapolis, IN), CAM 5.2 (Becton Dickinson, San Jose, CA) specific for the simple epithelial cytokeratins 8 and 18, and anti-SSEA-1 (Kamiya Biomedical Co., Thousand Oaks, CA). The dilutions used were as follows: anti-vimentin, 1:800; anti-a-F-P, 1:800; anti-CEA, 1:100; antiSSEA-1 l:lOO, l:lOOO, l:lOOOO, 1:25000, 1:50000; AEl/AE3, 1:lOOO; CAM 5.2, 1:lO.
0 0
7 8
10 7
In viva-derived Day 8 Day IO
Day 12 b Day 14 5
_ 0
_ _ _
a Mean f standard error of the mean. b Tumors were lost during embedding. Embryonal carcinoma cells were not detected in any tumors.
2 5
0
Bovine ICM/embryonic In vitro-derived Day 8 10
1 7 9
_
7 9
_
7 6
7 9
discs/embryos
I
10 10 I1
1
Day 10 Day 11 Day 12
_
0 0
discs
Porcine ICM/embryonic Day 8 10 Day 9 10
7 8
7 9
Mesoderm
_ 5
_
0.02 0.09-1.01
_ _
_
0.01-0.20
1
_
_ 0.01-0.28
_
0.05-0.68 0.05-I .41
Range
a
0.02 * 0 0.65t0.18
_
0.0 1 0.10+0.04 0.03 + 0.01
_
0.21 + 0.07 0.29 f 0.09
xfSEM
under the kidney capsule
1 6-20
_
2-12 2-12
_
l-5 l-20
Range
13.8f 2.3
1.050
_ _
_
2.00 7.0 f 2.4 5.4+ 1.7
_
7.1 f 2.5 8.7 f 2.7
xfSEMa
Tumor diameter (mm)
discs and murine egg cylinders
Tumor weight(g)
0 0
_ _
6 6
Endoderm
Ektodenn
Total
With tumors
No. of tumors with cells derived from:
No. of hosts
of porcine and bovine inner cell masses (ICM) and embryonic
Murine egg cylinder Inbred 11 Outbred I1
Type of embryos
Table 1 Development of tumors after transplantation of athymic mice
236
G.B. Anderson et al./Animal Reproduction Science 45 (1996) 231-240
3. Results The incidence of tumor formation and the cell types derived from the three embryonic germ layers are summarized in Table 1. Day 6 murine egg cylinders readily formed
Fig. 1. Cyst-like structures frequently found in bovine and porcine but not murine tumors. Cells appeared histologically and immunohistochemically to be of epithelial origin. Sections were stained with hematoxylin and eosin. A, bovine; B, porcine (X 130).
G.B. Anderson et al./Animal
Reproduction
Science 45 (1996) 231-240
237
embryo-derived tumors when transplanted under the kidney capsule of athymic mice. Seven of 11 graft hosts that received BAL,B/C egg cylinders and nine of 11 hosts that received outbred egg cylinders developed tumors. Embryonic tumors were variable in size and weight, but mean tumor weights were approximately 0.2-0.3 g and mean diameters were approximately 7-9 mm. The incidence of embryo-derived tumor formation was age-dependent for porcine and bovine embryos. The earlier stage embryos tested (days 8 and 9 for porcine embryos and Days 8 and 10 for bovine embryos) failed to form detectable tumors. Only one of ten athymic hosts of Day 10 porcine ICM developed a tumor, which was small compared with tumors that developed from older porcine embryos. The majority of graft hosts of Day 12 porcine and Day 14 bovine embryonic discs developed tumors, some of which were similar in size and weight to murine embryo-derived tumors. Tumors were highly variable in gross morphology, differing in appearance depending on the number and size of observable cysts. The two tumors that developed from Day 12 bovine embryonic discs were only approximately 1 mm in diameter and were lost during paraffin embedding. Other tumors were evaluated for the presence of differentiated cell types from the three embryonic germ layers and undifferentiated stem cells. Most murine tumors contained cells derived from ectoderm, mesoderm and endoderm. Among the cell types detected of ectodermal origin were neural tissue and epithelial cells, some of which had keratinized. Mesodermally derived tissues included mesenchyme, connective tissue, fibroblasts, cartilage, bone, adipose tissue and muscle. Murine tumors also contained endodermally derived epithelium, and in some cases cells resembling hepatocytes. Porcine and bovine tumors appeared to be more restricted in the cell types represented in the tumors. Only one porcine tumor (from Day 12 embryonic discs) contained endodermally derived epithelium and hepatocyte-like cells; otherwise, porcine tumors contained cell types similar to those found in murine tumors. One porcine tumor appeared to contain cardiac muscle that underwent rhythmic contractions for several minutes after excision from the host. Bovine tumors consisted predominantly of mesodermal derivatives with a small amount of endodermal epithelium also present. Unlike murine and porcine tumors, no ectoderma1 derivatives were detected in any bovine tumor. Unlike murine tumors, bovine and porcine tumors frequently contained cyst-like structures lined with primitive stratified cells (Fig. 1). These cells were strongly positive for AEl/AE3, which reacts with a broad range of cytokeratins, suggesting that the cells were of epithelial origin. No tumor from any of me three species contained morphologically detectable EC cells. Immunohistochemical evaluation of selected sections confirmed the presence of various differentiated cell types. No clusters of cells high in alkaline phosphatase activity or SSEA-1 expression, which are characteristic of murine EC cells, were detected.
4. Discussion Embryonic age affected the frequency at which tumors were produced from transplanted bovine and porcine embryos. Only among the later-stage embryos tested (primarily Day 11 and Day 12 porcine and Day 14 bovine) were embryo-derived tumors
238
G.B. Anderson et al./Animal
Reproduction
Science 45 (1996) 231-240
detected. Murine embryos of various ages have been shown to produce teratomas and teratocarcinomas when grafted to extrauterine sites. Although tumors have been produced from murine embryos as young as 1 day old (Stevens, 1968), murine embryos at the Day 6 to Day 7 egg cylinder stage have proven to be most effective for induction of embryo-derived teratocarcinomas (Solter et al., 1980). Since neither bovine nor porcine embryos form egg cylinders, comparisons of developmental stages across the three species are imprecise. Based on development of the embryonic disc, Day 13 to Day 14 bovine embryos and Day 11 to Day 12 porcine embryos are equivalent to Carnegie Stage 5, which is the same stage for murine egg cylinder stage embryos (Butler and Juurlink, 1987). Thus, tumor development in the older stage bovine and porcine embryos tested was consistent with available results in mice, which have been studied extensively. Alternatively, embryonic age effects on tumor formation could have been influenced by differences in the number of cells transplanted. The embryonic stages that produced bovine and porcine tumors were older than the blastocyst-stage embryos most frequently used for attempted in vitro isolation of ES cells. The extensive presence of mesodermally derived tissues was consistent among murine, bovine and porcine tumors. Neural tissue has been reported to be the most prevalent differentiated tissue type found in murine teratocarcinomas (Damjanov and Solter, 1974), but no neural tissues, or any other ectodermal derivatives, were detected in bovine tumors. Murine tumors that developed in this study contained tissues representative of all three embryonic germ layers. In general, bovine tumors appeared to contain fewer different cell types than murine and porcine tumors. The basis for these species differences is unknown. Damjanov et al. (1987) reported that murine embryo-derived tumors weighing less than 1 g are invariably benign and those over 2 g are generally malignant. All bovine and porcine tumors in this study weighed less than 1 g and all murine tumors weighed less than 2 g. All tumors appeared to be benign teratomas by their lack of detectable EC cells. Martin (1982) reported that as many as 50% of experimentally induced murine embryonic tumors contain EC cells, the actual percentage depending on a number of genetic and epigenetic factors. The use of athymic graft hosts in this study may have contributed to the lack of EC cells detected in embryo-derived tumors. Athymic mice are known to be generally unfavorable for the production of malignant teratocarcinomas (Solter and Damjanov, 1979), although human pluripotent teratocarcinoma stem cell lines grafted subcutaneously to nude mice have yielded tumors containing EC cells (Andrews et al., 1980). Athymic mice were used in this study because of their capacity to accept xenografts and because of the lack of available histocompatible bovine and porcine graft hosts. Experimentally induced teratocarcinogenesis could prove to be a more-or-less species-specific trait, since spontaneous teratocarcinomas are generally found only in humans and mice, with rare cases being cited in other species. The inability to detect bovine and porcine EC cells might have been related to this low frequency of malignant embryonic tumors in livestock; perhaps EC cells would be detected if a larger number of tumors were examined. Furthermore, the retransplantation of embryo-derived tumor cells to a new graft host might have allowed the detection of EC cells missed during histological and histochemical evaluations, since tumors that contained EC cells would be expected to form new tumors. An alternative graft host
G.B. Anderson er ul./Animul
Reprducrion
Science 45 (1996) 231-240
239
might be the pre-immunocompetent bovine or porcine fetus, which should accept allografts and might favor development of embryo-derived teratocarcinomas.
5. Conclusion
Transplantation of bovine and porcine embryos under the kidney capsule of athymic mice resulted in the formation of tumors. The lack of EC cells detected in these embryo-derived tumors leaves unanswered the question of whether EC cells can be produced from livestock embryos. Failure to detect bovine and porcine EC cells in this study corroborates the generally negative results of experiments aimed at in vitro isolation of bovine and porcine ES cells.
Acknowledgements
We thank Dr. Esmail Behboodi for supplying the in vitro-derived bovine embryos used in the study. We also thank Dr. Ivan Damjanov for examining selected sections of bovine and porcine tumors and Diane K. Nayden for carrying out the immunohistochemical assays. Grant support from the National Pork Producers Council is gratefully acknowledged.
References Anderson, G.B., 1992. Isolation and use of embryonic stem cells from livestock species. Anim. Biotech.. 3: 165517.5. Andrews, P.W., Bronson, D.L., Benham, F., Strickland, S. and Knowles, B.B., 1980. A comparative study of eight cell lines derived from human testicular teratocarcinoma. Int. J. Cancer, 26: 269-280. Behboodi, E., Anderson, C.B. and BonDurant, R.H., 1992. Development of in vitro fertilized oocytes from pregnant and nonpregnant cows in oviductal epithelial and cumulus cell co-culture systems. Theriogenology, 38: 1077- 1084. Brinster, R.L., 1974. The effect of cells transferred into the mouse blastocyst on subsequent development. J. Exp. Med., 104: 1049-1056. Butler, H. and Juurlink, B.H.J., 1987. An Atlas for Staging Mammalian and Chick Embryos. CRC Press, Boca Raton, FL, pp. 89 and 145. Camey, W., 1988. Human tumor antigens and specific tumor therapy. Immunol. Today, 9: 363-364. Damjanov, 1. and Solter, D., 1974. Experimental teratoma. Curr. Top. Pathol., 59: 69-130. Damjanov. I., Bagasra, 0. and Solter. D., 1983. Genetic and epigenetic factors regulate the evolving malignancy of embryo-derived teratomas. In: L.M. Silver, G.R. Martin and S. Stricldand (Editors), Teratocarcinoma Stem Cells. Cold Spring Harbor Conference on Cell Proliferation, Vol. IO. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 501-517. Damjanov, 1.. Damjanov, A. and Solter, D., 1987. Production of teratocarcinomas from embryos transplanted to extra-uterine sites. In: E.J. Robertson (Editor), Teratocarcinomas and Embryonic Stem Cells, A Practical Approach. IRL Press, Oxford, pp. I - 18. Eklund, J. and Bradford, G.E., 1976. A note on short-term selection for birth weight in mice. Anim. Prod., 22: I27- 130. Hsu, S.M., Raine, L. and Fonger, H., 1981. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabelled antibody (PAP) procedures. J. Histochem. Cytochem., 29: 5777580.
240
G.B. Anderson et al./Animal
Reproduction Science 45 (1996) 231-240
Jacobsen, G.K., Jacobsen, M. and Clausen, P.P., 1981. Distribution of tumor-associated antigens in the various histologic components of germ cell tumors of the testis. Am. J. Surg. Pathol., 5: 257-266. Martin, G., 1982. Teratocarcinoma stem cells provide a model system for the study of early mammalian development in vitro. In: T. Maramatsu, G. Gachelin, A. Mascona and Y. lkawa (Editors), Teratocarcinoma and Embryonic Cell Interactions. Academic Press, New York, pp. 3- 16. McNutt, M.A., Balen, J.W., Gown, A.M., Hammor, S.P. and Vogel, A.M., 1985. Coexpression of intermediate filaments in human epithelial neoplasm. Ultrastruct. Pathol., 9: 3 l-43. Mintz, B. and Illmensee, K., 1975. Normal genetically mosaic mice produced from malignant teratocarcinoma cells. Proc. Natl. Acad. Sci., 72: 3585-3589. Mummery, C.L., Geyen, A., Fruend, E. and Shen, S., 1990. Characteristics of embryonic stem cell differentiation: A comparison with two embryonal carcinoma cell lines. Cell Diff. Dev., 30: 195-206. Papaioamrou, V.E. and Rossant, J., 1983. Effects of the embryonic environment on proliferation and differentiation of embryonal carcinoma cells. Cancer Surveys, 2: 165- 183. Piedrahita, J.A., Anderson, G.B. and BonDurant, R.H., 1990. Influence of feeder layer type on the efficiency of isolation of porcine embryo-derived cell lines. Theriogenology, 34: 865-877. Robertson, E.J., 1991. Using embryonic stem cells to introduce mutations into the mouse germ line. Biol. Reprod., 44: 238-245. Sheahan, K., O’Brien, M.J., Burke, B., Dervan, P.A., O’Keone, J.C., Gottlieb, L.S. and Zamcheck, N., 1990. Differential reactivities of carcinoembryonic antigen (CEA) and CEA-related monoclonal and polyclonal antibodies in common epithelial malignancies. Am. J. Clin. Pathol., 94: 157- 164. Solter, D. and Damjanov, I., 1979. Teratocarcinomas rarely develop from embryos transplanted into athymic mice. Nature, 278: 5699-5707. Solter, D., Skreb, N. and Damjanov, I., 1970. Extrauterine growth of mouse egg cylinders results in malignant teratoma. Nature, 227: 503-504. Solter, D., Shevinsky, D., Knowles, B. and Strickland, S., 1979. The induction of antigenic changes in a teratocarcinoma stem cell line (F9) by retinoic acid. Dev. Biol., 70: 5 15-52 I. Solter, D., Dominis, M. and Damjanov, I., 1980. Embryo-derived teratocarcinoma. Il. Teratocarcinogenesis depends on the type of embryonic graft. Int. J. Cancer, 25: 341-343. Stevens, L.C., 1968. The development of teratomas from intratesticular grafts of tubal mouse eggs. J. Embryol. Exp. Morphol., 20: 329-341. Stevens, L.C., 1984. Spontaneous and experimentally induced testicular teratomas in mice. Cell Diff., 15: 69-74. Traub, P., 1985. Intermediate Filaments. Springer, New York. Wilmut, I., Hooper, M.L. and Simons, J.P., 1991. Genetic manipulation of mammals and its application in reproductive biology. J. Reprod. Fertil., 92: 245-279.
Animal Reproduction
Science 45 ( 1996) 23 I-240
Development of bovine and porcine embryonic teratomas in athymic mice G.B. Anderson a,*, R.H. BonDurant b, L. Goff a, J. Groff ‘, A.L. Moyer a aDepartment of Animal Science, Universiry ofCalifornia. Davis, CA 95616. USA b Department ofPopulation Health and Reproduction, Universiry of California. Davis, CA 95616, USA ’Department of Pathology. Microbiology and Immunology. University of Calijornia, Davis, CA 95616. USA Accepted 2 May 1996
Abstract discs from bovine and porcine blastocysts of various the kidney capsule of athymic (nude) mice to evaluate growth of teratocarcinomas containing both differentiated tissues and undifferentiated stem cells. Inner cell masses were isolated immunosurgically from Day 8, Day 9 and Day 10 porcine blastocysts and from Day 8, Day IO and Day 12 bovine blastocysts. Embryonic discs were mechanically dissected from Day 11 and Day 12 porcine embryos and from Day 14 bovine embryos. Day 6 egg cylinders were dissected from BALB/C embryos and from hybrid embryos of a cross between BALB/C and an outbred strain of mouse. Two to four KM, embryonic discs or egg cylinders were transplanted under the kidney capsule of each athymic host. After 8 weeks, graft hosts were killed and their tumors removed, fixed and prepared for histological and immunohistochemical examination. Embryonic teratomas developed at high frequency from murine egg cylinders and from Day 11 and Day 12 porcine and Day 14 bovine embryos. Tumors were observed only infrequently from younger bovine and porcine blastocysts. Murine embryonic tumors were composed of numerous differentiated cell types of ectodermal, mesodermal and endodennal origins, but representation of the three embryonic germ layers was somewhat more restricted in bovine and porcine embryonic tumors. No undifferentiated stem cells were detected in tumors of any of the three species. These results demonstrate that teratomas will develop from bovine and porcine embryos when grafted to an immunocompromised host, but the presence of undifferentiated teratocarcinoma stem cells from these species has yet to be achieved. Inner
ages
were
Key~lorks:
cell masses
transplanted
(KM)
Cattle-embryology;
* Corresponding
and embryonic
under
EC cells; Pig-embryology;
author. Tel: 916/752-1682.
0378.4320/96/$15.00 Published PII SO378-4320(96)01581-3
Teratocarcinomas;
Fax: 916/752-0175.
by Elsevier Science B.V.
Teratomas
e-mail: [email protected]
232
G.B. Anderson et al./Animal
Reproduction
Science 45 (1996) 231-240
1. Introduction Embryonic stem (ES) cells are undifferentiated embryonic cells that retain in culture the capacity to differentiate into both somatic and germ cell lineages. Murine ES cells have been used extensively as models for embryogenesis in the study of cell lineage, regulation of gene expression and genetic manipulation (Robertson, 1991; Wilmut et al., 1991). Their potential use for genetic manipulation in agricultural species is widely appreciated (Anderson, 19921, but successful isolation of ES cells from embryos of the livestock species remains to be fully documented. Establishment of ES cells is usually accomplished by culturing intact blastocysts or isolated inner cell masses (ICM). An alternate approach used in mice to produce pluripotent embryonic cell lines is by inducing formation of a teratocarcinoma, a tumor of embryonic origin that contains not only various differentiated somatic tissues but also pluripotent stem cells. The undifferentiated stem cells, called embryonal carcinoma or EC cells, share numerous properties with ES cells, including the ability to differentiate into normal tissues when injected into a blastocyst (Brinster, 1974; Mintz and Illmensee, 1975; Papaioannou and Rossant, 1983). Teratocarcinomas occur spontaneously in some strains of mice but can be induced by transplantation of early embryos to extrauterine sites. Murine teratocarcinomas have been produced by transplantation of embryos to the testis (Stevens, 1984) or under the kidney capsule (Solter et al., 1970; Damjanov et al., 1983) of histocompatible hosts. The objective of this study was to examine species-specific characteristics of embryonic-cell differentiation by determining if bovine and porcine teratocarcinomas can be induced by transplantation of early embryos under the kidney capsule of an immunocompromised graft host. Athymic mice were used as graft hosts because of their capacity to accept xenografts.
2. Materials and methods 2.1. Collection and manipulation of embryos Porcine blastocysts were recovered from gilts by flushing the uterine horns with Dulbecco’s modified phosphate-buffered saline (PBS) supplemented with 1% calf serum (Gibco, Grand Island, NY) on Day 8,9, 10, 11 or 12 of the estrous cycle (Day 0 was the first day of estrus). Bovine blastocysts were recovered by non-surgically flushing the uteri of superovulated beef cows on Day 8, 10 or 12 of the estrous cycle (Day 0 was the day of est.&. Day 14 bovine embryos were recovered at slaughter from superovulated beef cows. Inner cell masses were isolated by immunosurgery (Piedrahita et al., 1990) from Day 8, Day 9 and Day 10 porcine blastocysts and Day 8, Day 10 and Day 12 bovine blastocysts. Embryonic discs from Day 11 and Day 12 porcine and Day 14 bovine embryos were isolated mechanically using two 27-gauge needles under a dissection microscope. Intact Day 8 bovine blastocysts obtained from procedures for in vitro oocyte maturation and fertilization and embryo culture (Behboodi et al., 1992)
G.B. Anderson et al./Animul
Reproduction Science 45 (1996) 231-240
233
were also tested for the capacity to establish embryonic grafts. Egg cylinder stage murine embryos were collected from Day 6 (Day 0 was the day of mating plug) BALB/C females previously mated to BALB/C males (inbred embryos) and from JU females (Eklund and Bradford, 1976) mated to BALB/C males (outbred embryos). Murine inbred embryos were from the same strain as the graft hosts; murine outbred embryos were from a cross between the graft host strain and an outbred strain. Egg cylinders were removed from extraembryonic tissues as described by Damjanov et al. (19871. 2.2. Transplantation
of embryonic grafts to athymic hosts
Athymic (nu/nu> BALB/C mice were used as hosts for embryonic grafts. Athymic mice were maintained on sterile bedding in filter-top, microisolator cages. Most graft hosts were 7-8 weeks of age at transplantation. Male and female hosts were randomly distributed among treatments. Hosts were anesthetized with an intraperitoneal injection of 0.3 cc of a 10% solution of sodium pentabarbital (Western Medical Supply, Arcadia, CA) and the right kidney was exposed through a lateral, oblique incision in the skin and body wall. Watchmaker forceps were used to make a small tear in the kidney capsule and two to four ICM, embryonic discs, egg cylinders or intact Day 8 in vitro-derived bovine blastocysts were deposited under the kidney capsule using a tapered 3 ~1 micropipette (Drummond Wiretrol, Drummond Scientific Co., Broomall, PA). The flank incision was closed with two 9 mm autoclips (Becton, Dickenson and Company, Parsippany, NY), which were removed 7 days after surgery. Seven to 11 athymic graft hosts were used per treatment. 2.3. Tumor recovery and evaluation Athymic graft hosts were killed by cervical dislocation 8 weeks after transplantation and kidneys were removed and examined for evidence of tumor development. Tumors were carefully excised from the kidney, weighed, measured at their largest diameter and fixed in 10% buffered formalin for paraffin embedding. Histological sections were prepared at 6 km and stained with hematoxylin and eosin using standard procedures. Histological sections were evaluated by light microscopy for the presence of various differentiated cells. Tumors were classified as benign teratomas or malignant teratocarcinomas according to criteria described by Solter et al. (1979) and Damjanov et al. (1987). Selected histological sections were also stained for alkaline phosphatase activity (Sigma Diagnostics Alkaline Phosphatase, Leukocyte Testing Kit), which is expressed by murine EC cells. Selected tumors were stained immunohistochemically for the presence of cytokeratins, vimentin, carcinoembryonic antigen (CEA), o-fetoprotein (a-F-P), and stage-specific embryonic antigen-l @SEA-I), all of which are markers for various differentiated and undifferentiated cell types. Some sections were also tested against anti-cytokeratin antibodies including AE 1/AE3, a combination monoclonal antibody that has a broad range of specificities, and CAM 5.2, which is often used to recognize simple epithelium.
234
G.B. Anderson er al. / Animal Reproduction Science 45 (1996) 231-240
Cytokeratin expression is typical of epithelial differentiation in avian and mammalian species, whereas vimentin is typically expressed in cells of mesenchymal origin. Certain normal and malignant epithelial tissues and epithelial-derived cell lines may co-express vimentin (McNutt et al., 1985; Traub, 1985) Carcinoembryonic antigen is a complex glycoprotein expressed by various neoplasms, is normally produced in embryonic endodermal tissue (gut, liver and pancreas) and has been used as a marker for neoplasms of these tissues (Carney, 1988; Sheahan et al., 1990). a-fetoprotein is a glycoprotein synthesized by the fetal yolk sac, liver and gastrointestinal tract and has been used as a tumor marker for hepatic, testicular germ cells and yolk sac neoplasms (Jacobsen et al., 1981; Sheahan et al., 1990). Stage-specific embryonic antigen-l is expressed by murine undifferentiated cells including ICM, ES cells and EC cells (Mummery et al., 1990). A standard streptavidin-biotin-immunoperoxidase complex (ABC) technique was used for immunohistochemical staining procedures (Hsu et al., 1981). Embedded tissues were deparaffinized, cleared with xylene and dehydrated through a graded series of alcohol solutions. Tissues were blocked for endogenous peroxidase activity with 0.3% hydrogen peroxide in methanol for 30 min at room temperature and rinsed with PBS (pH 7.4). Non-specific staining was blocked with 10% heat-inactivated normal serum in PBS for 20 min at room temperature; normal serum was either horse or goat according to the secondary antibody species. The tissue sections were incubated with the primary antisera diluted in PBS (pH 7.4) at room temperature for 1 h, except for o-F-P, which required a 2-h incubation. Following incubation and each subsequent step, slides were rinsed in two changes of PBS for 5 min. The biotinylated secondary antibody diluted 1:500 in PBS was applied to tissue sections for 30 min, followed by avidin-biotinhorseradish peroxidase for 20 min. Biotinylated horse-anti-mouse or goat-anti-rabbit antibodies (Vector Laboratories, Burlingame, CA) were used as secondary antibodies against the primary monoclonal or polyclonal antibodies, respectively. Antibody localization was determined by the use of 3,3_diaminobenzidene (DAB) in imidazole buffer. Aminoethyl carbazole (AEC) was used as the chromogen for detection of SSEA-1. Slides were counterstained with Mayer’s hematoxylin, washed in tap water, dehydrated to xylene and coverslipped. Representative canine tissues were used as positive controls for AEl/AE3, CAM 5.2 and vimentin. Canine embryos, approximately 3 weeks of age and procured from clinical spay procedures, were used as positive controls for CEA and o-F-P. Feline gastrointestinal tissues were used as positive controls for SSEA-1 expression. For negative controls, PBS was used in place of the primary antisera. Antibodies included the rabbit-polyclonal anti-vimentin (ICN Biomedicals, Inc., Costa Mesa, CA), anti-a-F-P and anti-CEA (Dako Corporation, Carpinteria, CA) antibodies and the murine monoclonal antibody AEl/AE3 specific for the high and low molecular weight cytokeratins (Boehringer Mannheim Corp., Indianapolis, IN), CAM 5.2 (Becton Dickinson, San Jose, CA) specific for the simple epithelial cytokeratins 8 and 18, and anti-SSEA-1 (Kamiya Biomedical Co., Thousand Oaks, CA). The dilutions used were as follows: anti-vimentin, 1:800; anti-a-F-P, 1:800; anti-CEA, 1:100; antiSSEA-1 l:lOO, l:lOOO, l:lOOOO, 1:25000, 1:50000; AEl/AE3, 1:lOOO; CAM 5.2, 1:lO.
0 0
7 8
10 7
In viva-derived Day 8 Day IO
Day 12 b Day 14 5
_ 0
_ _ _
a Mean f standard error of the mean. b Tumors were lost during embedding. Embryonal carcinoma cells were not detected in any tumors.
2 5
0
Bovine ICM/embryonic In vitro-derived Day 8 10
1 7 9
_
7 9
_
7 6
7 9
discs/embryos
I
10 10 I1
1
Day 10 Day 11 Day 12
_
0 0
discs
Porcine ICM/embryonic Day 8 10 Day 9 10
7 8
7 9
Mesoderm
_ 5
_
0.02 0.09-1.01
_ _
_
0.01-0.20
1
_
_ 0.01-0.28
_
0.05-0.68 0.05-I .41
Range
a
0.02 * 0 0.65t0.18
_
0.0 1 0.10+0.04 0.03 + 0.01
_
0.21 + 0.07 0.29 f 0.09
xfSEM
under the kidney capsule
1 6-20
_
2-12 2-12
_
l-5 l-20
Range
13.8f 2.3
1.050
_ _
_
2.00 7.0 f 2.4 5.4+ 1.7
_
7.1 f 2.5 8.7 f 2.7
xfSEMa
Tumor diameter (mm)
discs and murine egg cylinders
Tumor weight(g)
0 0
_ _
6 6
Endoderm
Ektodenn
Total
With tumors
No. of tumors with cells derived from:
No. of hosts
of porcine and bovine inner cell masses (ICM) and embryonic
Murine egg cylinder Inbred 11 Outbred I1
Type of embryos
Table 1 Development of tumors after transplantation of athymic mice
236
G.B. Anderson et al./Animal Reproduction Science 45 (1996) 231-240
3. Results The incidence of tumor formation and the cell types derived from the three embryonic germ layers are summarized in Table 1. Day 6 murine egg cylinders readily formed
Fig. 1. Cyst-like structures frequently found in bovine and porcine but not murine tumors. Cells appeared histologically and immunohistochemically to be of epithelial origin. Sections were stained with hematoxylin and eosin. A, bovine; B, porcine (X 130).
G.B. Anderson et al./Animal
Reproduction
Science 45 (1996) 231-240
237
embryo-derived tumors when transplanted under the kidney capsule of athymic mice. Seven of 11 graft hosts that received BAL,B/C egg cylinders and nine of 11 hosts that received outbred egg cylinders developed tumors. Embryonic tumors were variable in size and weight, but mean tumor weights were approximately 0.2-0.3 g and mean diameters were approximately 7-9 mm. The incidence of embryo-derived tumor formation was age-dependent for porcine and bovine embryos. The earlier stage embryos tested (days 8 and 9 for porcine embryos and Days 8 and 10 for bovine embryos) failed to form detectable tumors. Only one of ten athymic hosts of Day 10 porcine ICM developed a tumor, which was small compared with tumors that developed from older porcine embryos. The majority of graft hosts of Day 12 porcine and Day 14 bovine embryonic discs developed tumors, some of which were similar in size and weight to murine embryo-derived tumors. Tumors were highly variable in gross morphology, differing in appearance depending on the number and size of observable cysts. The two tumors that developed from Day 12 bovine embryonic discs were only approximately 1 mm in diameter and were lost during paraffin embedding. Other tumors were evaluated for the presence of differentiated cell types from the three embryonic germ layers and undifferentiated stem cells. Most murine tumors contained cells derived from ectoderm, mesoderm and endoderm. Among the cell types detected of ectodermal origin were neural tissue and epithelial cells, some of which had keratinized. Mesodermally derived tissues included mesenchyme, connective tissue, fibroblasts, cartilage, bone, adipose tissue and muscle. Murine tumors also contained endodermally derived epithelium, and in some cases cells resembling hepatocytes. Porcine and bovine tumors appeared to be more restricted in the cell types represented in the tumors. Only one porcine tumor (from Day 12 embryonic discs) contained endodermally derived epithelium and hepatocyte-like cells; otherwise, porcine tumors contained cell types similar to those found in murine tumors. One porcine tumor appeared to contain cardiac muscle that underwent rhythmic contractions for several minutes after excision from the host. Bovine tumors consisted predominantly of mesodermal derivatives with a small amount of endodermal epithelium also present. Unlike murine and porcine tumors, no ectoderma1 derivatives were detected in any bovine tumor. Unlike murine tumors, bovine and porcine tumors frequently contained cyst-like structures lined with primitive stratified cells (Fig. 1). These cells were strongly positive for AEl/AE3, which reacts with a broad range of cytokeratins, suggesting that the cells were of epithelial origin. No tumor from any of me three species contained morphologically detectable EC cells. Immunohistochemical evaluation of selected sections confirmed the presence of various differentiated cell types. No clusters of cells high in alkaline phosphatase activity or SSEA-1 expression, which are characteristic of murine EC cells, were detected.
4. Discussion Embryonic age affected the frequency at which tumors were produced from transplanted bovine and porcine embryos. Only among the later-stage embryos tested (primarily Day 11 and Day 12 porcine and Day 14 bovine) were embryo-derived tumors
238
G.B. Anderson et al./Animal
Reproduction
Science 45 (1996) 231-240
detected. Murine embryos of various ages have been shown to produce teratomas and teratocarcinomas when grafted to extrauterine sites. Although tumors have been produced from murine embryos as young as 1 day old (Stevens, 1968), murine embryos at the Day 6 to Day 7 egg cylinder stage have proven to be most effective for induction of embryo-derived teratocarcinomas (Solter et al., 1980). Since neither bovine nor porcine embryos form egg cylinders, comparisons of developmental stages across the three species are imprecise. Based on development of the embryonic disc, Day 13 to Day 14 bovine embryos and Day 11 to Day 12 porcine embryos are equivalent to Carnegie Stage 5, which is the same stage for murine egg cylinder stage embryos (Butler and Juurlink, 1987). Thus, tumor development in the older stage bovine and porcine embryos tested was consistent with available results in mice, which have been studied extensively. Alternatively, embryonic age effects on tumor formation could have been influenced by differences in the number of cells transplanted. The embryonic stages that produced bovine and porcine tumors were older than the blastocyst-stage embryos most frequently used for attempted in vitro isolation of ES cells. The extensive presence of mesodermally derived tissues was consistent among murine, bovine and porcine tumors. Neural tissue has been reported to be the most prevalent differentiated tissue type found in murine teratocarcinomas (Damjanov and Solter, 1974), but no neural tissues, or any other ectodermal derivatives, were detected in bovine tumors. Murine tumors that developed in this study contained tissues representative of all three embryonic germ layers. In general, bovine tumors appeared to contain fewer different cell types than murine and porcine tumors. The basis for these species differences is unknown. Damjanov et al. (1987) reported that murine embryo-derived tumors weighing less than 1 g are invariably benign and those over 2 g are generally malignant. All bovine and porcine tumors in this study weighed less than 1 g and all murine tumors weighed less than 2 g. All tumors appeared to be benign teratomas by their lack of detectable EC cells. Martin (1982) reported that as many as 50% of experimentally induced murine embryonic tumors contain EC cells, the actual percentage depending on a number of genetic and epigenetic factors. The use of athymic graft hosts in this study may have contributed to the lack of EC cells detected in embryo-derived tumors. Athymic mice are known to be generally unfavorable for the production of malignant teratocarcinomas (Solter and Damjanov, 1979), although human pluripotent teratocarcinoma stem cell lines grafted subcutaneously to nude mice have yielded tumors containing EC cells (Andrews et al., 1980). Athymic mice were used in this study because of their capacity to accept xenografts and because of the lack of available histocompatible bovine and porcine graft hosts. Experimentally induced teratocarcinogenesis could prove to be a more-or-less species-specific trait, since spontaneous teratocarcinomas are generally found only in humans and mice, with rare cases being cited in other species. The inability to detect bovine and porcine EC cells might have been related to this low frequency of malignant embryonic tumors in livestock; perhaps EC cells would be detected if a larger number of tumors were examined. Furthermore, the retransplantation of embryo-derived tumor cells to a new graft host might have allowed the detection of EC cells missed during histological and histochemical evaluations, since tumors that contained EC cells would be expected to form new tumors. An alternative graft host
G.B. Anderson er ul./Animul
Reprducrion
Science 45 (1996) 231-240
239
might be the pre-immunocompetent bovine or porcine fetus, which should accept allografts and might favor development of embryo-derived teratocarcinomas.
5. Conclusion
Transplantation of bovine and porcine embryos under the kidney capsule of athymic mice resulted in the formation of tumors. The lack of EC cells detected in these embryo-derived tumors leaves unanswered the question of whether EC cells can be produced from livestock embryos. Failure to detect bovine and porcine EC cells in this study corroborates the generally negative results of experiments aimed at in vitro isolation of bovine and porcine ES cells.
Acknowledgements
We thank Dr. Esmail Behboodi for supplying the in vitro-derived bovine embryos used in the study. We also thank Dr. Ivan Damjanov for examining selected sections of bovine and porcine tumors and Diane K. Nayden for carrying out the immunohistochemical assays. Grant support from the National Pork Producers Council is gratefully acknowledged.
References Anderson, G.B., 1992. Isolation and use of embryonic stem cells from livestock species. Anim. Biotech.. 3: 165517.5. Andrews, P.W., Bronson, D.L., Benham, F., Strickland, S. and Knowles, B.B., 1980. A comparative study of eight cell lines derived from human testicular teratocarcinoma. Int. J. Cancer, 26: 269-280. Behboodi, E., Anderson, C.B. and BonDurant, R.H., 1992. Development of in vitro fertilized oocytes from pregnant and nonpregnant cows in oviductal epithelial and cumulus cell co-culture systems. Theriogenology, 38: 1077- 1084. Brinster, R.L., 1974. The effect of cells transferred into the mouse blastocyst on subsequent development. J. Exp. Med., 104: 1049-1056. Butler, H. and Juurlink, B.H.J., 1987. An Atlas for Staging Mammalian and Chick Embryos. CRC Press, Boca Raton, FL, pp. 89 and 145. Camey, W., 1988. Human tumor antigens and specific tumor therapy. Immunol. Today, 9: 363-364. Damjanov, 1. and Solter, D., 1974. Experimental teratoma. Curr. Top. Pathol., 59: 69-130. Damjanov. I., Bagasra, 0. and Solter. D., 1983. Genetic and epigenetic factors regulate the evolving malignancy of embryo-derived teratomas. In: L.M. Silver, G.R. Martin and S. Stricldand (Editors), Teratocarcinoma Stem Cells. Cold Spring Harbor Conference on Cell Proliferation, Vol. IO. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 501-517. Damjanov, 1.. Damjanov, A. and Solter, D., 1987. Production of teratocarcinomas from embryos transplanted to extra-uterine sites. In: E.J. Robertson (Editor), Teratocarcinomas and Embryonic Stem Cells, A Practical Approach. IRL Press, Oxford, pp. I - 18. Eklund, J. and Bradford, G.E., 1976. A note on short-term selection for birth weight in mice. Anim. Prod., 22: I27- 130. Hsu, S.M., Raine, L. and Fonger, H., 1981. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabelled antibody (PAP) procedures. J. Histochem. Cytochem., 29: 5777580.
240
G.B. Anderson et al./Animal
Reproduction Science 45 (1996) 231-240
Jacobsen, G.K., Jacobsen, M. and Clausen, P.P., 1981. Distribution of tumor-associated antigens in the various histologic components of germ cell tumors of the testis. Am. J. Surg. Pathol., 5: 257-266. Martin, G., 1982. Teratocarcinoma stem cells provide a model system for the study of early mammalian development in vitro. In: T. Maramatsu, G. Gachelin, A. Mascona and Y. lkawa (Editors), Teratocarcinoma and Embryonic Cell Interactions. Academic Press, New York, pp. 3- 16. McNutt, M.A., Balen, J.W., Gown, A.M., Hammor, S.P. and Vogel, A.M., 1985. Coexpression of intermediate filaments in human epithelial neoplasm. Ultrastruct. Pathol., 9: 3 l-43. Mintz, B. and Illmensee, K., 1975. Normal genetically mosaic mice produced from malignant teratocarcinoma cells. Proc. Natl. Acad. Sci., 72: 3585-3589. Mummery, C.L., Geyen, A., Fruend, E. and Shen, S., 1990. Characteristics of embryonic stem cell differentiation: A comparison with two embryonal carcinoma cell lines. Cell Diff. Dev., 30: 195-206. Papaioamrou, V.E. and Rossant, J., 1983. Effects of the embryonic environment on proliferation and differentiation of embryonal carcinoma cells. Cancer Surveys, 2: 165- 183. Piedrahita, J.A., Anderson, G.B. and BonDurant, R.H., 1990. Influence of feeder layer type on the efficiency of isolation of porcine embryo-derived cell lines. Theriogenology, 34: 865-877. Robertson, E.J., 1991. Using embryonic stem cells to introduce mutations into the mouse germ line. Biol. Reprod., 44: 238-245. Sheahan, K., O’Brien, M.J., Burke, B., Dervan, P.A., O’Keone, J.C., Gottlieb, L.S. and Zamcheck, N., 1990. Differential reactivities of carcinoembryonic antigen (CEA) and CEA-related monoclonal and polyclonal antibodies in common epithelial malignancies. Am. J. Clin. Pathol., 94: 157- 164. Solter, D. and Damjanov, I., 1979. Teratocarcinomas rarely develop from embryos transplanted into athymic mice. Nature, 278: 5699-5707. Solter, D., Skreb, N. and Damjanov, I., 1970. Extrauterine growth of mouse egg cylinders results in malignant teratoma. Nature, 227: 503-504. Solter, D., Shevinsky, D., Knowles, B. and Strickland, S., 1979. The induction of antigenic changes in a teratocarcinoma stem cell line (F9) by retinoic acid. Dev. Biol., 70: 5 15-52 I. Solter, D., Dominis, M. and Damjanov, I., 1980. Embryo-derived teratocarcinoma. Il. Teratocarcinogenesis depends on the type of embryonic graft. Int. J. Cancer, 25: 341-343. Stevens, L.C., 1968. The development of teratomas from intratesticular grafts of tubal mouse eggs. J. Embryol. Exp. Morphol., 20: 329-341. Stevens, L.C., 1984. Spontaneous and experimentally induced testicular teratomas in mice. Cell Diff., 15: 69-74. Traub, P., 1985. Intermediate Filaments. Springer, New York. Wilmut, I., Hooper, M.L. and Simons, J.P., 1991. Genetic manipulation of mammals and its application in reproductive biology. J. Reprod. Fertil., 92: 245-279.