Human implantation

Implantation of the human embryo

Alex Lopata

Department of Obstetrics and Gynaecology, University of Melbourne, Royal Women's Hospital, 132 Grattan Street, Carlton, Victoria 3053,

Australia

Embryonic development and implantation

For information on the growth of the early human embryo in the genital tract we still depend on the specimens collected by Hertig et al. (1956). In these classical studies embryos were obtained from the uterus or Fallopian tubes, after hysterectomies performed at estimated times following ovulation and conception times based on recorded coital dates. On about day 3 after ovulation a 12-cell embryo, possibly 60-72 h of age and described as a morula, was recovered from the uterus. By day 4 after presumed ovulation an early blastocyst, containing 53 trophoblastic and five embryonic cells was obtained. While by day 4.5 . after inferred ovulation, a zona-free blastocyst, containing 99 trophoblastic and eight embryonic cells, was recovered from the uterus (Hertig et al., 1956). In these and earlier studies the time of ovulation was estimated from menstrual cycle data as well as endometrial and corpus luteum morphology (Hertig et al., 1954).

More recently, collecting embryos and ova from the uterus by non-surgical procedures, after more precise detection of the expected times of ovulation, was reported by two groups (Croxatto et al., 1972; Buster et ai., 1985). Croxatto et al. used urinary luteinizing hormone (LH) and pregnanediol assays, and other parameters, to predict ovulation which, in many cases, was also assumed to correspond to the time of fertilization. In one woman they found a l2-16-cell morula 4 days after the LH peak and 3 days after coitus, and in a second case a 16-cell morula which was considered to be 4-5 days old. They also recovered an expanded zonal blastocyst, containing at least 186 cells, which was considered to be 5 days or slightly older. Buster et al. (1985) used serum LH assays and artificial inseminations to estimate ovulation and fertilization times. Embryos were collected by non-surgical uterine lavages performed 5 days after the LH peak. Five zona-enclosed blastocysts, with an estimated age ranging from 102.5-117 h, were recovered. However, embryos ranging from 2-16-cell stages were also obtained during a similar time interval. The latter findings suggest that delayed and arrested embryonic development, which is commonly observed in vitro, also occurs following in-vivo fertilization and attempted in-situ growth.

Human Reproduction Volume 11 Supplement 1 1996 © European Society for Human Reproduction and Embryology 175

A.Lopata

Based on the above studies it may be concluded that following in-vivo conception in the human, early blastocysts begin to form during the first few hours of the fifth day after ovulation (fertilization), and the majority of these embryos become expanded blastocysts by the middle or end of day 5. Although hatching blastocysts have not been collected from the uterus, presumably the human blastocyst sheds its zona during the sixth day of development. It is worth pointing out, in passing, that in the non-human primate, such as the marmoset monkey, hatching uterine blastocysts have been observed, and hatched blastocysts can be readily recovered from the uterus together with empty zonae containing a distinct opening.

If the human blastocyst emerges from its zona by the end of day 6, the trophectoderm would be expected to directly interact with the uterine luminal epithelium by day 7 after fertilization. It is not known whether the hatched human blastocyst is capable of initiating implantation immediately after shedding its zona, or whether the polar trophectoderm, which is the embryonic region that interacts with uterine epithelium in primates, requires a period of differentiation before implantation begins. Our studies in the marmoset monkey have shown that the hatched blastocyst requires a latent period of about 24 h before it attaches to a substrate such as Matrigel. Available information in the human suggests that implantation begins on day 7 or 8 after fertilization (Hertig, 1975; Lenton et al., 1982).

It has been known for a long time that human embryos obtained in an in-vitro fertilization (IVF) programme will form blastocysts by day 5 of development (Lopata et al., 1982). A small percentage of such blastocysts are viable since they have the potential to produce term pregnancies following transfer to the uterus (Bolton et al., 1991; Menezo et al., 1992; Olivennes et al., 1994). However, a majority of cultured blastocysts fail to implant probably due to abnormalities of embryonic growth in vitro. For example, it has been found that human blastocysts grown in simple culture media contained the following average number of cells in the trophectoderm and inner cell mass respectively: 39.1 and 19.3 on day 5; 44.7 and 37.6 on day 6; 80.6 and 45.6 on day 7 (Hardy, 1993). It is likely that human blastocysts of similar age, developing in vivo, would contain more than double the total number of cells reported above for each stage. A retarded level of cell proliferation in cultured blastocysts may partly account for the low implantation rate following their placement in the uterus.

Following natural conception, the human chorionic gonadotrophin (HCG) secreted by an intrauterine blastocyst could be detected in the serum by day 8 or 9 after the LH peak, in the majority of women studied (Lenton et at., 1982). This interval probably corresponds to 7 or 8 days after ovulation and fertilization. It has been proposed that implantation is established by this stage (Lenton et al., 1982), although its onset occurs a short, but unknown time earlier. In studies on blastocysts obtained by IVF and embryo culture, HCG released into the medium becomes detectable by day 7 of development (Lopata and Hay, 1989). The gonadotrophin is produced by both hatched and zona-enclosed blastocysts at about the same time of development, suggesting that a pre-programmed timing

176

Implantation of the human embryo

mechanism is switched on. Attachment of the blastocyst does not appear to be a prerequisite for the initiation and continuation of HCG secretion. Our studies on the distribution of mRNA for the (J.- and ~-subunits of chorionic gonadotrophin, in implantation stage marmoset blastocysts, have revealed that the invasive polar trophectoderm is the major source of the gonadotrophin (Lopata et aI., 1995). This region of the primate blastocyst is strategically located for releasing trophoblastic hormones into maternal blood vessels during the earliest as well as more advanced stages of implantation.

It would be of interest to know whether the cells of the inner cell mass play an inductive role, particularly in transforming the trophectoderm adjacent to them into an adhesive and invasive phenotype, and also in preferentially switching on the genes that express the subunits for chorionic gonadotrophin. The biological value of these series of localized trophoblastic transformations would be that the inner cell mass, and subsequently the embryonic disc and tissues derived from it, become closely associated with a region of trophoblast that pursues maternal blood and hence facilitates the exchange of nutrients and wastes. During this sequence of events embryonic differentiation and implantation occur at a brisk rate. By day 9 of development the uterine epithelium is penetrated completely, syncytial trophoblast becomes well established, and stromal fibroblasts begin to be transformed into decidual cells. The blastocyst then erodes the basement membrane, enters the stroma and the trophoblast begins to penetrate maternal blood vessels by day 10 or 11 after fertilization. At about the same time trophoblastic lacunae become filled with maternal blood (Rock and Hertig, 1948; Larsen, 1980). Decidualization, which is not conspicuous during the early stages of human implantation, becomes more pronounced shortly after day 11.

Clinical aspects of implantation

Complex tissue interactions must be taken into consideration when studying implantation. These interactions are illustrated in Figures 1-3. A number of tissue compartments are involved, apart from those present in the blastocyst. These include the uterine epithelium, and of course its basement membrane which contains bioactive molecules such as laminin, fibronectin and collagen type IV. Stimuli essential for implantation are exchanged between the blastocyst and uterine epithelium, and similar interactions probably also occur between the blastocyst and decidua, with the epithelium acting as an intermediary (Figure 2). It is also well known that the decidua is a dynamic tissue composed of many cell types including haematopoietic cells, reactive blood vessels with properties specific for the endometrium, and vascular endothelium. Moreover, the cell population and the composition of the extracellular matrix within the decidua, vary with its state of differentiation and stage of implantation (Glasser and McCormack, 1980; Glasser, 1990).

The main clinical problem with implantation is to understand the nature of the processes which govern the pregnancy rates obtained with IVF and embryo

177

A.Lopata

Corpus Luteum I

Oestrogen + Progesterone

Stromal ... ' . Inner Cell "

.'. '" Cells Laminin Mass ..... Uterine .....

Decidual B '" t Cells ....

Fibronectin M Epithelial

..oL Cells Tropliectoderm Endothelial Collagen IV (Trophoblast) Cells .... G Iycocalyt<

Ceiis Macrophages Proteoglycans Integrins

Lymphocytes

Figure 1. Interaction of tissues at implantation. In the embryonic compartment. shown in the box on the right. the inner cell mass and trophectoderm probably influence each other through paracrine and contact mechanisms. The polar trophectoderm interacts with uterine epithelium in primates through released signals such as cytokines and eicosanoids as well as via the components of the glycocalyx and cell-cell attachments. [ntegrins on the surface of the implanting trophectoderm are subsequently involved in the penetration process by binding to laminin. fibronectin and other compounds in the epithelial basement membrane and the matrix of the decidual stroma. The cellular and molecular composition of the stromal compartment. and the capacity of the epithelium to induce decidualization. arc powerfully influenced by the sex steroids secreted by the developing corpus luteum. Bidirectional arrows indicate feedback interactions between cellular and matrix compartments.

transfer. We can produce good blastocysts in vitro, although the implantation rate is only ~ 10% per embryo. Perhaps we have slightly improved on that figure in recent years, although only modestly. Of the embryos that do implant, 60% or more proceed to live births.

Achieving normal implantation is the problem, and low rates could be caused by various factors. Ovarian hyperstimulation may induce either endometrial or embryonic problems. I believe that the epithelial layer of the endometrium plays only a relatively unimportant role, being a kind of passive bystander, during the process of blastocyst penetration. At best the epithelium may have the ability to block implantation during some stages of the menstrual cycle. When this inhibitory influence is removed the epithelium becomes permissive of implantation, but it is the blastocyst that plays the subsequent active role. It is the polar trophoblast that intrudes between the epithelial cells, which are pushed aside, and that erodes the basal lamina and penetrates into the stroma and blood vessels. The duration of this permissive state has been defined as the implantation window. The permissive window appears to be tightly regulated in animals such as mice and rats, but is much more variable in humans, and probably in other primates. Some investigators have found that the implantation window in the human begins 4-5 days after fertilization and lasts about 4 days Navot et ai., 1989), while others have claimed that its time-span may even extend up to 8 days (Rogers et ai., 1989).

Based on embryo transfer experiments in animals, it has been assumed that precise synchrony between embryo and endometrium is important for implantation

178

Implantation of the human embryo

1 1 Stromal Transformation

Paracrine Endocrine Matrix

Figure 2. The uterine epithelium appears to be an obligatory intermediate in inducing stromal transformation (decidualization) during implantation. The epithelium is regarded as being an essential transducer that delivers the deciduogenic signal(s) to the stroma. The nature of the transmitted epithelial signal(s) are not known. Physiologically, the initial stimulus arises from the blastocyst, although decidualization can be induced in the absence of an embryo. The three classes of inductive stimuli from the blastocyst are illustrated by the arrows. The transformed stroma, in turn, directs cytodifferentiation and function of the epithelium in contact with it (represented by the paracrine, endocrine and matrix feedback arrows). The sex steroids have a critical priming role during these tissue interactions.

in the human. If the window of implantation lasts from 4-8 days, beginning on about day 19 of the cycle, all normal embryos will have formed blastocysts that have hatched during the receptive interval. Is it possible, therefore, to have embryo--endometrial asynchrony in women who have an implantation window spanning 4-8 days? As far as I am aware, there are no studies that correlate the likelihood of implantation with morphological findings on embryos and uterine epithelium, within transfer cycles, either at the light or electron microscopic level. So at present we cannot define normal embryonic and endometrial morphology, particularly between stimulated cycles, let alone the possible abnormalities.

It is clear that good quality embryos need to be produced for successful implantation. Virtually all patients can produce a good endometrium if levels of oestrogen and progesterone are sufficient without being too high. It follows, therefore, that any type of ovarian stimulation will produce an epithelium that will be effective, providing the blastocysts are of sufficiently good quality. Receptivity can only be induced at the site of the blastocyst within the uterine cavity. In the absence of an embryo, there is no other way of defining receptivity. Apart from my belief that the epithelium plays a passive role during blastocyst penetration, I also consider that there is lack of strong evidence showing that blastocysts firmly attach to the apical surface of the uterine epithelium. If they do so, it is a very transitory process, since the blastocyst is very mobile.

In effect, perhaps one of the main roles of the uterine epithelium is to transmit signals from the blastocyst to induce the decidual cell reaction (Figure 2). A blastocyst would implant even if uterine epithelium was removed, but it would not induce a decidual response which is the function of the epithelium. The implanting blastocyst produces signals, perhaps prostaglandins, leucotrienes,

179

A.Lopata

Decidual cell

Lam-Rec iii' :::il

Laminin J ...... L---1----,

I:

Decidual cell

Figure 3. Trophoblast invasion of decidua. Endometrial stromal cells that undergo decidual transformation acquire the capacity to synthesize and secrete basement membrane proteins, such as laminin, fibronectin and collagen type IV. The invasive trophoblast of the implanting embryo also expresses and releases some of these proteins as well as surface receptors (e.g. integrins) for various matrix proteins. Invasive trophoblast cells have also been found to produce plasminogen activator (u-PA), a serine proteinase that converts plasminogen to plasmin. High affinity cell slllt'ace receptors (uPA-R) and binding sites for plasminogen and plasmin serve to localize proteolytic activation to the pericellular area. The plasmin generated is able to degrade extracellular components and to activate latent metalloproteinases, such as procollagenase. The proteolytic activity of plasmin is regulated by PA inhibitors (PAI-I and PAI-2) and other proteinase inhibitors. The events illustrated bring together adhesive and proteolytic interactions during implantation. E1 = oestradiol: P 4 = progesterone.

platelet activating factor, histamine, or other bioactive agents such as cytokines, which may play an integrating role (Kennedy, 1983; Alecozay et at., 1991; Sharkey, 1995) in inducing the decidual cell reaction.

As the blastocyst begins to implant the trophoblast, and particularly invasive polar trophoblast, produces laminin and fibronectin (Figure 3). At the same time receptors for these glycoproteins appear on the trophoblast. The appearance of basement membrane compounds at the apical surface of uterine epithelium, may play a part in disorganizing the epithelium at the implantation site, in some species (Foidart et al., 1990). Among other mechanisms, the plasminogen activator and matrix metalloproteinase systems (Figure 3), are probably involved in disorganizing the uterine epithelium that interacts with the blastocyst during the penetration process (Lala and Graham, 1990).

180

Implantation of the human embryo

The marmoset as a model of human implantation

The marmoset is a very suitable experimental model for human implantation. Its reproductive cycle, which is normally 28 days, can be controlled by administering a prostaglandin F2u analogue during the mid-luteal phase or early pregnancy, to begin a new cycle (Summers et at., 1985). The marmoset ovulates about 8-9 days after this treatment, and ovulation can be readily detected by plasma progesterone assays. Since marmosets are very fecund, and are housed as mating pairs, ovulation nearly always corresponds to day one of pregnancy. Embryos ranging from the 4-cell to hatched blastocyst stages can be collected from the uterine cavity during precisely timed fertile cycles, since it is easy to irrigate the uterine cavity, and flush out the embryos. Cleavage stage embryos can be successfully grown to hatched blastocysts in a variety of culture media.

One of our models of implantation is to culture marmoset blastocysts on Matrigel, a basement membrane-like substance. Blastocysts attach to this matrix in the region of the polar trophectoderm, and subsequently invade it, and since the penetrating blastocysts retain their three-dimensional integrity the system provides an excellent model for experimental studies (Lopata et at., 1995). We have analysed the effects of leukaemia inhibitory factor (LIF) on such blastocysts around the time of implantation by permitting the embryos to invade the Matrigel for 4 days in the absence and presence of the cytokine, at various concentrations, in the culture medium. More cells and mitotic figures were observed in the inner cell mass of the treated embryos compared with controls. LIF had apparently accelerated the development of the inner cell mass and the amniotic cavity, but we could not determine its effect on the endoderm which had formed a complete yolk sac by day 4 of 'in-vitro implantation'. Our data also suggested that LIF promoted cytotrophoblast proliferation and to some extent inhibited the formation of syncytiotrophoblast. These observations remind us that LIF receptors have been identified on human b1astocysts, and our studies show how some of those receptors may be expressed on the differentiating tissues of the implanting primate embryo. The possible LIF interactions during implantation are illustrated diagrammatically in Figure 4.

Like the human blastocyst, the marmoset (m) blastocyst also produces chorionic gonadotrophin (CG), and probably other bioactive factors, during early stages of differentiation (Hearn et at., 1988). The production of mCG interested us because the gonadotrophin not only rescues the corpora lutea of early pregnancy in marmosets, but may also have an immunosuppressive action that contributes to survival of the blastocyst during implantation. There is also clear evidence that CG stimulates cytotrophoblast growth, so it is probably involved in early placenta formation, acting as an autocrine growth factor (Yagel et at., 1989).

We have also investigated the localization of CG in the implantation stage marmoset blastocyst. Bonduelle et al. (1988) reported that mRNA for the HCG ~-subunit is detectable at the 6-8-cell stage human embryo, where every cell is already transcribing the gene(s). My colleagues at the University of Adelaide cloned the genes for the u- and ~-subunits of marmoset CG (Simula et al.,

181

A.Lopata

o (J o o ,.: ,\;""

BM BASAL Figure 4. Oestrogen induces the synthesis of leukaemia inhibitory factor (LIF) in the uterine epithelium on the day of implantation. LIF is transported to the blastocyst where it interacts with LIF receptors (LIF-R) that arc expressed both in the trophectoderm and inner cell mass. mRNA = gene message for LlF-R: BM = basement membrane.

1995), so we had cDNA available for both subunits. From these sequences, we obtained riboprobes for the a- and ~-subunits, producing both sense and antisense probes, which were labelled with digoxygenin for in-situ hybridization. Analyses involving in-situ hybridization were then carried out on marmoset blastocysts invading Matrigel4 days after hatching (Lopata et aI., 1995), i.e. early implantation stages. As a positive control we used marmoset trophoblastic vesicles, which are known to secrete bioactive CG in tissue culture (Summers et al., 1987; Simula et al., 1995). Uterus, kidney and liver were used as negative controls.

The mRNA for CG-~ was predominantly expressed over the polar trophoblast region. Some signals were found in mural trophoblast, but far fewer. Inner cell mass had a low level of activity of the message for CG-~ in some cells, and there was virtually no activity in the endoderm.

With the CG-a subunit anti-sense riboprobe, hybridization also occurred over the polar trophoblast, but not as heavily. The transcript was uniformly distributed both in polar and mural trophoblast, with some a-subunit activity showing up distinctly in the inner cell mass. It appears that in the marmoset blastocyst, CG activity is located predominantly in the area where the trophoblast is attaching and implanting, whereas in differentiating tissues such as inner cell mass and endoderm, the genes for the gonadotrophin are probably being inactivated.

182

Implantation of the human embryo

References

Alecozay, A. A. , Harper, M.J.K., Schenken, R.S. and Hanahan, D.J. (1991) Paracrine interactions between platelet-activating factor and prostaglandins in hormonally-treated human luteal phase endometrium in vitro. 1. Reprod. Fertil., 91, 301-312.

Bolton, V.N., Wren, M.E. and Parsons, J.H. (1991) Pregnancies after in vitro fertilization and transfer of human blastocysts. Ferti!. Steril., 55, 830-832.

Bonduelle, M.-L., Dodd, R., Liebaers, I. et al. (1988) Chorionic gonadotrophin-beta mRNA. a trophoblast marker, is expressed in human 8-cell embryos derived from tripronucleate zygotes. Hum. Reprod., 3,909-914.

Buster, J.E., Bustillo, M., Rodi, I.A. et al .. (1985) Biologic and morphologic development of donated human ova recovered by nonsurgical lavage. Am. 1. Obstet. Gynecol., 153,211-217.

Croxatto, H.B., Diaz, S., Fuentealba, B. et al. (1972) Studies on the duration of egg transport in the human oviduct. I. The time interval between ovulation and egg recovery from the uterus in normal women. Fertil. Steril., 23, 447-458.

Foidart, J.M., Christiane, Y. and Emonard, H. (1990) Interaction between the human trophoblast cells and the extracellular matrix of the endometrium. Specific expression of alpha-galactose residues by invasive human trophoblast cells. In Denker, H.-Wand Aplin, J.D. (eds), Trophoblast Invasion and Endometrial Receptil'ity. Plenum Medical, New York, pp. 159-177.

Glasser, S.R. (1990) Biochemical and structural changes in uterine endometrial cell types following natural or artificial deciduogenic stimuli. In Denker, H.-W. and Aplin, J.D. (eds), Trophoblast Invasion and Endometrial Receptivity. Plenum Medical, New York, pp. 377-416.

Glasser, S.R. and McCormack, S.A. (1980) Role of stromal cell function diversity in altering endometrial cell responses. In Leroy, F., Finn, c.A., Psychoyos, A. and Hubinont, P.O. (eds), Blastocyst-Endometrial Relationships. S. Karger, Basel, pp. 102-114.

Hardy, K. (1993) Development of human blastocysts in vitro. In Bavister, B.D. (eds), Preimplantatiol1 Embryo Development. Springer Verlag, New York, pp. 184-199.

Hearn, J.P', Hodges, J.K. and Gems, S. (1988) Early secretion of chorionic gonadotrophin by marmoset embryos in vivo and in vitro. 1. Endocrinol., 119,249-255.

Hertig, A.T. (1975) Implantation of the human ovum. In Behrman, S.J. and Kistner, R.W (eds), Progress in Infertility. Little, Brown and Co., Boston, pp. 411-438.

Hertig, A.T., Rock, J. and Adams, E.C. (1956) A description of 34 human ova within the first 17 days of development. Am. 1. Anat., 98, 435-494.

Hertig, A.T., Rock, J., Adams, E.C. and MUlligan, W.J. (1954) On the preimplantation stages of the human ovum: a description of 4 normal and 4 abnormal specimens ranging from the second to the fifth day of development. Contrib. Embryol., 35, 199-220.

Kennedy, T.G. (1983) Embryonic signals and the initiation of blastocyst implantation. Aust. 1. BioI. Sci., 36, 531-543.

Lala, P.K. and Graham, C.H. (1990) Mechanisms of trophoblast invasiveness and their control: the role of proteases and protease inhibitors. Cancer Metast. Rev., 9, 369-379.

Larsen, J.F. (1980) Human implantation and clinical aspects. In Leroy, F., Finn, c.A., Psychoyos, A. and Hubinont, P.O. (eds), Blastocyst-Endometrium Relationships. S. Karger, Basel. pp. 284-296.

Lenton, E.A., Neal, L.M. and Sulaiman, R. (1982) Plasma concentrations of human chorionic gonadotrophin from the time of implantation until the second week of pregnancy. Ferri!. Steri/ .. 37,773-778.

Lopata, A., Berka, 1 .. Simula, A. et al. (1995) Differential distribution of mRNA for the alpha and beta subunits of chorionic gonadotrophin in the implantation stage blastocyst of the marmoset monkey. Placenta. 16, 335-346.

Lopata, A. and Hay, D.L. (1989) The surplus human embryo: its potential for growth, blastulation, hatching, and human chorionic gonadotropin production in culture. Ferti!. Steril., 51, 984-991.

Lopata, A., Kohlman, DJ., Bowes, L.G. and Watkins, WB. (1995) Culture of marmoset blastocysts on Matrigel: a model of differentiation during the implantation period. Anat. Rec., 241,469-486.

Lopata, A., Kohlman. D.J. and Kellow, G.N. (1982) The fine structure of human blastocysts developed in culture. In Burger, M.M. and Weber, R. (eds), Embryonic Development, Part B: Cellular Aspects. Alan R.Liss Inc .. New York, pp. 69-85.

183

A.Lopata

Menezo , Y, Hazout, A., Dumont, M. et al. (1992) Co-culture of embryos on Vero cells and transfer of blastocysts in humans. Hum. Reprod., 7 (Suppl. I), 101-106.

Navot. D., Anderson. T.L., Droesch, K. et al. (1989) Hormonal manipulation of endometrial maturation. 1. Ciin. Endocrinol. Metab., 68, 801-807.

Olivenncs, F., Hazout, A., Lelaidier, C. et al. (1994) Four indications for embryo transfer at the blastocyst stage. Hum. Reprod., 9, 2367-2373.

Rock, 1. and Hertig, A., T. (1948) The human conceptus during the first two weeks of gestation. Am. 1. Obstet. Gynecol., 55, 6-17.

Rogers, P., Murphy, C., Cameron, 1. et al. (1989) Uterine receptivity in women receiving steroid replacement therapy for premature ovarian failure: ultrastructural and endocrinological parameters. Hum. Reprod., 4, 349-354.

Sharkey, A. (1995) Cytokines and embryo/endometrial interactions. Reprod. Med. Rev., 4,87-100. Simula, A.P., Amato, F., Faast, R. et al. (1995) Luteinizing hormone!chorionic gonadotropin

bioactivity in the common marmoset (Callithrix jacchus) is due to a chorionic gonadotropin molecule with a structure intermediate between human chorionic gonadotropin and human luteinizing hormone. Bioi. Reprod., 53, 380-389.

Summers, P.M., Taylor. C.T. and Hearn, 1.P. (1987) Characteristics of trophoblastic tissue derived from ill vitro culture of preimplantation embryos of the common marmoset monkey. Placenta, 8,411-422.

Summers, P.M., Wennink, C.1. and Hodges, 1.K. (1985) Cloprostenol-induced luteolysis in the marmoset monkey (Callithrix jacchus). 1. Reprod. Fertil., 73, 133-138.

Yagel, S., Casper, R.F., Powell, W. et al. (J 989) Characterization of pure human first-trimester cytotrophoblast cells in long term culture: growth pattern, markers, and hormone production. Am. 1. Obstet. GVllecol.. 160, 938-945.

184