Molecular, Genetic, and Developmental Control of Cleavage

Along with the increase in cell numbers, mammalian cleavage is a period dominated by several critical developmental events. The earliest is the transition from maternally to zygotically produced gene products. Another is the polarization of individual blastomeres, which sets the stage for the developmental decision that results in the subdivision of the cleaving embryo into two distinct types of cells: the trophoblast and the inner cell mass (see Fig. 3.1). Most studies of the molecular biology and genetics of early mammalian development have been done on mice. Until more information on early primate embryogenesis becomes available, results obtained from experimentation on mice must be used as a guide.

Because of the lack of massive storage of maternal ribosomes and RNAs during oogenesis, development of the mammalian embryo must rely on the activation of zygotic gene products at a very early stage. Most maternal transcription products are degraded by the two-cell stage (Fig. 3.4). Some of these, however, stimulate the activation of the embryonic genome, which begins producing RNAs from a significant number of genes (>1500) by the time cleavage has advanced to the four-cell stage. There does not seem to be a sharp transition between the cessation of reliance on purely maternal gene products and the initiation of transcription from the embryonic genome. Some paternal gene products (e.g., isoforms of β-glucuronidase and β₂-microglobulin) appear in the embryo very early, while maternal actin and histone mRNAs are still being used for the production of corresponding proteins. As an indication of the extent to which the early embryo relies on its own gene products, development past the two-cell stage does not occur in the mouse if mRNA transcription is inhibited. In contrast, similar treatment of amphibian embryos does not disrupt development until late cleavage, at which time the embryos begin to synthesize the mRNAs required to control morphogenetic movements and gastrulation.

Mature eggs and sperms are transcriptionally inactive. A major reason for this is that their DNA is highly methylated. Methylation, which occurs on CpG dinucleotides, normally inactivates the associated gene. Such inactivation is often called epigenetic regulation because it does not alter the fundamental DNA sequence. Methylation can inactivate informational genes or their regulators (e.g., enhancers or promoters). Pronounced cycles of global methylation and demethylation occur during the life span of an individual (Fig. 3.5). Within 4 hours after fertilization, the paternally derived genome undergoes rapid, massive demethylation. Demethylation of the maternally derived genome occurs more gradually until the early morula, at which stage all the DNA is maximally demethylated. Remethylation ensues in the inner cell mass, until by the late blastocyst stage it returns to maximal levels. Within the germ cell line, the high methylation levels characteristic of the early embryo fall after the primordial germ cells have entered the genital ridge. During later gametogenesis, remethylation occurs. This remethylation imprints (see p. 43) maternal or paternal characteristics on the gametes and for some genes has profound effects on the embryos produced from these gametes. Epigenetic control is not confined to methylation patterns. Even as early as the zygote, different patterns of histone association with the chromatin account for pronounced differences in gene expression between the male and female pronuclei.

For the first couple of days after fertilization, transcriptional activity in the cleaving embryo is very low. Similarly, fertilized eggs and early mammalian embryos possess a limited capacity for the translation of mRNAs. The factor limiting translational efficiency may be the small number of ribosomes stored in the egg. As cleavage proceeds, products from maternally and paternally derived chromosomes are active in guiding development. Haploid embryos commonly die during cleavage or just after implantation. There is increasing evidence, however, that the control of early development involves more than simply having a diploid set of chromosomes in each cell.

One of the first manifestations of embryonic gene expression is the polarization of the blastomeres of the 8- and 16-cell embryo so that they have clearly recognizable apical and basal surfaces. The polarization of blastomeres leads to one of the most important steps in early mammalian development, namely, the decision that results in the appearance of two separate lines of cells—the trophoblast and the inner cell mass—from the early homogeneous blastomeres. In mice, up to the 8-cell stage all blastomeres are virtually identical. In the 8-cell embryo, the surfaces of the cells are covered with microvilli, and intercellular connections, mediated by E-cadherin, form. Shortly thereafter, differences are noted between polarized cells that have at least one surface situated on the outer border of the embryo and nonpolarized cells that are completely enclosed by other blastomeres. The polarized outer cells are destined to become trophoblast, whereas those cells located in the interior are destined to become the inner cell mass, from which the body of the embryo arises.

The relationship between the position of the blastomeres and their ultimate developmental fate was incorporated into the inside-outside hypothesis. The essence of this hypothesis is that the fate of a blastomere derives from its position within the embryo, rather than from its intrinsic properties. The outer blastomeres ultimately differentiate into the trophoblast, whereas the inner blastomeres form the inner cell mass. If marked blastomeres from disaggregated embryos are placed on the surface of another early embryo, they typically contribute to the formation of the trophoblast. Conversely, if the same marked cells are introduced into the interior of the host embryo, they participate in the formation of the inner cell mass (Fig. 3.6).

The cell polarity model offers an alternative explanation for the conversion of generic blastomeres to trophoblast or inner cell mass. According to this hypothesis, if the plane of cell division of a blastomere at the eight-cell stage is parallel to the outer surface of the embryo, the outer daughter cell develops a polarity, with its apical surface facing the zona pellucida (Fig. 3.7). The inner daughter cell remains apolar and goes on to form part of the inner cell mass. Experimental evidence suggests that a key element underlying a daughter cell’s becoming an outer cell is inheritance of a patch of outer cell membrane containing microvilli and the actin microfilament-stabilizing protein, ezrin. The proteins that produce polarity in the outer cells are postulated to direct their differentiation toward the trophoblastic lineage. Common to the inside-outside hypothesis and the cell polarity model is the recognition that a cell that does not contact the surface does not differentiate into trophoblast, but rather becomes part of the inner cell mass.

Even though by the 16-cell stage the embryo consists of clearly recognizable polar outer cells and nonpolar inner cells, cells of either type still can become transformed into cells of the other type. Thus, cells of the inner cell mass, if transplanted to the outer surface of another embryo, can become trophoblast, and at least some of the outer cells can turn into inner cell mass if transplanted into the interior. By the 32-cell stage, this capability for phenotypic transformation has become largely lost. Investigators have shown that cells of the inner cell mass of 16-cell embryos still retain the molecular machinery to turn into trophoblastic cells, because if the cells are exposed to the surface, they undergo the transformation into trophoblastic cells without new mRNA synthesis. Experiments of this type show that the developmental potential, or potency (the types of cells that a precursor cell can form) of many cells is greater than their normal developmental fate (the types of cells that a precursor cell normally forms).

The changes in phenotype of the inner and outer cells are accompanied by important molecular differences. Critical to the formation of cells of the trophoblast is the transcription factor, Cdx-2. Cdx-2 is essential for trophoblastic differentiation, and it also antagonizes the expression of molecules that are associated with the inner cell mass. Increased Cdx-2 levels both enhance the formation of molecules associated with polarization and increase the proportion of cells that undergo symmetrical cell division, thus increasing the number of trophoblastic cells. Cdx2 mutants fail to implant into the endometrial epithelium.

In contrast to cells of the trophoblast, which increasingly take on an epithelial character, cells of the inner cell mass express molecules that are associated with great developmental flexibility. Three such molecules are oct-4, Nanog, and Sox-2.

The oct4 gene codes for a specific transcription factor that binds the octamer ATTTGCAT on DNA. There is a close relationship between the expression of the oct4 gene and the highly undifferentiated state of cells. In mice, maternally derived oct-4 protein is found in developing oocytes and is active in the zygote. After the experimentally induced loss of oct-4 protein, development is arrested at the one-cell stage. This shows that maternally derived oct-4 protein is required to permit development to proceed to the two-cell stage, when transcription of the embryonic genes begins.

Oct-4 is expressed in all blastomeres up to the morula stage. As various differentiated cell types begin to emerge in the embryo, the levels of oct4 gene expression in these cells decrease until it is no longer detectable. Such a decrease is first noted in cells that become committed to forming extraembryonic structures and finally in cells of the specific germ layers as they emerge from the primitive streak (see Chapter 5). Even after virtually all cells of the embryo have ceased to express the oct4 gene, it is still detectable in the primordial germ cells as they migrate from the region of the allantois to the genital ridges. Because of its pattern of distribution, oct-4 protein is suspected to play a regulatory role in maintenance of the undifferentiated state and in establishing and maintaining the pluripotency of the germ cells.

Two other important genes in early development are Nanog and Sox2. Inner cells resulting from the division of cells in the eight-cell embryo begin to produce Sox-2, which binds onto DNA in partnership with oct-4 to regulate the expression of genes that control cellular differentiation. Nanog first appears in the late morula and along with Oct-4 functions to maintain the integrity of the inner cell mass. In the absence of Nanog function, cells of the inner cell mass differentiate into primitive endoderm (hypoblast), whereas lack of function of oct-4 causes inner cell mass cells to differentiate into trophoblast. Overall, but through different mechanisms, cells of both the trophoblast and inner cell mass are normally inhibited from becoming transformed into the other type.

Parental Imprinting

Experimentation, coupled with observations on some unusual developmental disturbances in mice and humans, has shown that the expression of certain genes derived from the egg differs from the expression of the same genes derived from the spermatozoon. Called parental imprinting, the effects are manifest in different ways. It is possible to remove a pronucleus from a newly inseminated mouse egg and replace it with a pronucleus taken from another inseminated egg at a similar stage of development (Fig. 3.8). If a male or female pronucleus is removed and replaced with a corresponding male or female pronucleus, development is normal. If a male pronucleus is removed and replaced with a female pronucleus (resulting in a zygote with two female pronuclei), however, the embryo itself develops fairly normally, but the placenta and yolk sac are poorly developed. Conversely, a zygote with two male pronuclei produces a severely stunted embryo, whereas the placenta and yolk sac are nearly normal.

Parental imprinting occurs during gametogenesis. Methylation of DNA, effected through specific imprinting centers, is one of the major means of imprinting and results in the differential expression of paternal and maternal alleles of the imprinted genes. Imprinted genes are transcriptionally silenced. The imprinted genes are maintained during development and possibly into adulthood, but a given imprint is not passed onto that individual’s progeny. Instead, the parental imprints on the genes are erased, and new imprints, corresponding to the sex of that individual, are established in the oocytes and sperm during gametogenesis.

Not all genes are parentally imprinted. Present estimates suggest that up to 2100 human genes are imprinted. Clinical Correlation 3.1 discusses some conditions and syndromes associated with disturbances in parental imprinting.

Clinical Correlation 3.1   Conditions and Syndromes Associated with Parental Imprinting

A striking example of paternal imprinting in humans is a hydatidiform mole (see Fig. 7.16), which is characterized by the overdevelopment of trophoblastic tissues and the extreme underdevelopment of the embryo. This condition can result from the fertilization of an egg by two spermatozoa and the consequent failure of the maternal genome of the egg to participate in development or from the duplication of a sperm pronucleus in an “empty” egg. This form of highly abnormal development is consistent with the hypothesis that paternal imprinting favors the development of the trophoblast at the expense of the embryo.

Several other syndromes are also based on parental imprinting. Beckwith-Wiedemann syndrome, characterized by fetal overgrowth and an increased incidence of childhood cancers, has been mapped to the imprinted region on chromosome 11, which contains the genes for insulinlike growth factor-II (IGF-II, which promotes cell proliferation) and H19 (a growth suppressor). It occurs when both alleles of the IGF2 gene express a paternal imprinting pattern. Another instructive example involves deletion of regions in the long arm of chromosome 15, specifically involving the gene UBE3A. Children of either sex who inherit the maternal deletion develop Angelman’s syndrome, which includes severe mental retardation, seizures, and ataxia. A child who inherits a paternal deletion of the same region develops Prader-Willi syndrome, characterized by obesity, short stature, hypogonadism, a bowed upper lip, and mild mental retardation.

X-Chromosome Inactivation

Another example of the inequality of genetic expression during early development is the pattern of X-chromosome inactivation in female embryos. It is well known from cytogenetic studies that one of the two X chromosomes in the cells of females is inactivated by extreme condensation. This is the basis for the sex chromatin, or Barr body, which can be shown in cells of females, but not in the cells of normal males. The purpose of X-chromosome inactivation is dosage compensation, or preservation of the cells from an excess of X-chromosomal gene products.

X-chromosome inactivation is initiated at the X-inactivation center, a unique locus on the X chromosome. XIST (X–inactive specific transcript), one of the genes in the X-inactivation center, produces a large RNA with no protein coding potential. XIST RNA remains in the nucleus and coats the entire inactive X chromosome, thus not allowing any further transcription from that chromosome. In the inactivated X chromosome, the XIST gene is unmethylated and expressed, whereas in the active X chromosome, this gene is methylated and silent.

Genetic studies show a complex ontogenetic history of X-chromosome inactivation (Fig. 3.9). In the female zygote, both X chromosomes are transcriptionally inactive, although not through the actions of XIST, because of the global inactivation of transcription in the early cleaving embryo. By the four-cell stage and into the morula stage, the paternally derived X chromosome becomes inactivated as the result of parental imprinting. Then, as the embryo forms the blastocyst, the paternally derived X chromosomes in the trophoblast and the hypoblast (see Fig. 5.1) remain inactivated, but within the cells of the inner cell mass both X chromosomes become active. As the cells of the inner cell mass begin to differentiate, the somatic cells undergo random permanent XIST-based X-chromosome inactivation of either the maternal or the paternal X chromosome. Within the germ cell line, activation of both X chromosomes occurs during the first meiotic division.

Developmental Properties of Cleaving Embryos

Early mammalian embryogenesis is considered to be a highly regulative process. Regulation is the ability of an embryo or organ primordium to produce a normal structure if parts have been removed or added.* At the cellular level, this means that the fates of cells in a regulative system are not irretrievably fixed, and the cells can still respond to environmental cues. Because the assignment of blastomeres to different cell lineages is one of the principal features of mammalian development, identifying the environmental factors that are involved is important.

Of the experimental techniques used to show regulative properties of early embryos, the simplest is to separate the blastomeres of early cleavage-stage embryos and determine whether each one can give rise to an entire embryo. This method has been used to show that single blastomeres from 2-cell and sometimes 4-cell embryos can form normal embryos, although blastomeres from later stages cannot do so. In mammalian studies, a single cell is more commonly taken from an early cleavage-stage embryo and injected into the blastocoele of a genetically different host. Such injected cells become incorporated into the host embryo, to form cellular chimeras or mosaics. When genetically different donor blastomeres are injected into host embryos, the donor cells can be identified by histochemical or cytogenetic analysis, and their fate (the tissues that they form) can be determined. Fate mapping experiments are important in embryology because they allow one to follow the pathways along which a particular cell can differentiate. Fate mapping experiments have shown that all blastomeres of an 8-cell mouse embryo remain totipotent; that is, they retain the ability to form any cell type in the body. Even at the 16-cell stage of cleavage, some blastomeres are capable of producing progeny that are found in both the inner cell mass and the trophoblastic lineage.

Another means of showing the regulative properties of early mammalian embryos is to dissociate mouse embryos into separate blastomeres and to combine the blastomeres of two or three embryos (Fig. 3.10). The combined blastomeres soon aggregate and reorganize to become a single large embryo, which goes on to become a normal-appearing tetraparental or hexaparental mouse. By various techniques of making chimeric embryos, it is possible to combine blastomeres to produce interspecies chimeras (e.g., a sheep-goat). It is likely that many human genetic mosaics (chimeras), most commonly recognized when some regions of the body are male and others are female, are the result of the fusion of two early fraternal twin embryos. Other possibilities for chimerism involve the exchange of cells through common vascular connections.

A significant question in early mammalian embryology is whether any of the three major body axes are represented in the egg or early embryo. Research on mouse embryos has resulted in dramatically different views. According to one view, the position of the second polar body, which after fertilization is typically found in line with the first cleavage plane, is a marker for the future anteroposterior axis. This would suggest that the egg before or just after fertilization possesses at least one predetermined axis, as is the case in many animals. Based on time-lapse photography, a contrary viewpoint posits that there is no predetermined axial plane within the egg, and the plane of the first cleavage division lies perpendicular to a line drawn between the final positions of the male and female pronuclei. Similarly, conflicting experimental data have not allowed researchers to determine whether there is any predetermined relationship between structures in the two- or four-cell embryo and the definitive body axes that become apparent at the time of early gastrulation. The bulk of evidence suggests that the early mammalian embryo is a highly regulative system and that the body axes do not become fixed until the end of cleavage or early gastrulation.

Experimental Manipulations of Cleaving Embryos

Much of the knowledge about the developmental properties of early mammalian embryos is the result of more recently devised techniques for experimentally manipulating them. Typically, the use of these techniques must be combined with other techniques that have been designed for in vitro fertilization, embryo culture, and embryo transfer (see Chapter 2).

Classic strategies for investigating the developmental properties of embryos are (1) removing a part and determining the way that the remainder of the embryo compensates for the loss (such experiments are called deletion or ablation experiments) and (2) adding a part and determining the way that the embryo integrates the added material into its overall body plan (such experiments are called addition experiments). Although some deletion experiments have been done, the strategy of addition experiments has proved to be more fruitful in elucidating mechanisms controlling mammalian embryogenesis.

Blastomere deletion and addition experiments (Fig. 3.11) have convincingly shown the regulative nature (i.e., the strong tendency for the system to be restored to wholeness) of early mammalian embryos. Such knowledge is important in understanding why the exposure of early human embryos to unfavorable environmental influences typically results in either death or a normal embryo.

One of the most powerful experimental techniques has been the injection of genetically or artificially labeled cells into the blastocyst cavity of a host embryo (see Fig. 3.11B). This technique has been used to show that the added cells become normally integrated into the body of the host embryo, thus providing additional evidence for embryonic regulation. An equally powerful use of this technique has been in the study of cell lineages in the early embryo. By identifying the progeny of the injected marked cells, investigators have been able to determine the developmental potency of the donor cells.

A technique that provides great insight into the genetic control mechanisms of mammalian development is the production of transgenic embryos. Transgenic embryos (commonly mice) are produced by directly injecting foreign DNA into the pronuclei of zygotes (Fig. 3.12A). The DNA, usually recombinant DNA for a specific gene, can be fused with a different regulatory element that can be controlled by the investigator.

Transgenic mice can be created by injecting the rat growth hormone gene coupled with a metallothionein promoter region (MT-I) into the pronuclei of mouse zygotes. The injected zygotes are transplanted into the uteri of foster mothers, which give birth to normal-looking transgenic mice. Later in life, when these transgenic mice are fed a diet rich in zinc, which stimulates the MT-I promoter region, the rat growth hormone gene is activated and causes the liver to produce large amounts of the polypeptide growth hormone. The function of the transplanted gene is obvious; under the influence of the rat growth hormone that they are producing, the transgenic mice grow to a much larger size than their normal littermates (Fig. 3.13).

In addition to adding genes to embryos, several powerful techniques have been developed to inactivate specific genes or gene products. At the DNA level, it is now common to knock out a gene of interest as a way to determine its function in normal development. Some genes have multiple functions at various times and in various tissues throughout embryogenesis. Their function in early development may be so critical that in the absence of its function the embryo dies even as early as gastrulation. To deal with this problem, techniques have been devised to interfere with tissue-specific promotors, so that the function of a gene in a given organ (e.g., the eye) can be disrupted in the primordium of that structure alone. Other techniques operate at the RNA level. For example, noncoding RNAi (RNA interference), injected into an embryo, knocks down, rather than blocks, gene expression. At the protein level, genetically engineered nonfunctional receptor molecules injected into an embryo can displace their normal counterparts and bind a signaling molecule without the ability to transduce the signal into the interior of the cell. There are situations in which each of these techniques is particularly useful in investigating a question in development.

Some types of twinning represent a natural experiment that shows the highly regulative nature of early human embryos, as described in Clinical Correlation 3.2.

Clinical Correlation 3.2   Twinning

Some types of twinning represent a natural experiment that shows the highly regulative nature of early human embryos. In the United States, about 1 pregnancy in 90 results in twins, and 1 in 8000 results in triplets. Of the total number of twins born, approximately two thirds are fraternal, or dizygotic, twins and one third are identical, or monozygotic, twins. Dizygotic twins are the product of the fertilization of 2 ovulated eggs, and the mechanism of their formation involves the endocrine control of ovulation. Monozygotic twins and some triplets are the product of a single fertilized egg. They arise by the subdivision and splitting of a single embryo. Although monozygotic twins could theoretically arise by the splitting of a 2-cell embryo, it is commonly accepted that most arise by the subdivision of the inner cell mass in a blastocyst, or possibly even splitting of the epithelial epiblast a few days later (Fig. 3.14). Because most monozygotic twins are normal, the early human embryo can obviously be subdivided, and each component regulates to form a normal embryo. Inferences on the origin and relationships of multiple births can be made from the arrangement of the extraembryonic membranes at the time of birth (see Chapter 7).

Apparently, among many sets of twins, one member does not survive to birth. This is a reflection of the increasing recognition that perhaps most conceptuses do not survive. According to some estimates, as many as one in eight live births is a surviving member of a twin pair. Quadruplets or higher orders of multiple births occur very rarely. In previous years, these could be combinations of multiple ovulations and splitting of single embryos. In the modern era of reproductive technology, most multiple births, sometimes up to septuplets, can be attributed to the side effects of fertility drugs taken by the mother.

The separation of portions of an embryo is sometimes incomplete, and although two embryos take shape, they are joined by a tissue bridge of varying proportions. When this occurs, the twins are called conjoined twins (sometimes colloquially called Siamese twins). The extent of bridging between the twins varies from a relatively thin connection in the chest or back to massive fusions along much of the body axis. Examples of the wide variety of types of conjoined twins are illustrated in Figures 3.15 and 3.16. With the increasing sophistication of surgical techniques, twins with more complex degrees of fusion can be separated. A much less common variety of conjoined twin is a parasitic twin, in which a much smaller but often remarkably complete portion of a body protrudes from the body of an otherwise normal host twin (Fig. 3.17). Common attachment sites of parasitic twins are the oral region, the mediastinum, and the pelvis. The mechanism of conjoined twinning has not been directly shown experimentally, but possible theoretical explanations are the partial secondary fusion of originally separated portions of the inner cell mass or the formation of two primitive streaks in a single embryo (see Chapter 5).

One phenomenon often encountered in conjoined twins is a reversal of symmetry of the organs of one of the pair (see Fig. 3.16B). Such reversals of symmetry are common in duplicated organs or entire embryos. More than a century ago, this phenomenon was recorded in a large variety of biological situations and was incorporated into what is now called Bateson’s rule, which states that when duplicated structures are joined during critical developmental stages, one structure is the mirror image of the other. Despite the long recognition of this phenomenon, only in recent years has there been any understanding of the mechanism underlying the reversal of symmetry.

Stem Cells and Cloning

A major development in biomedical research at the turn of the 21st century was the realization that certain cells (stem cells) in both human embryos and adults have the capacity to develop into a variety of cell and tissue types in response to specific environments. In embryos, stem cells can be derived from the inner cell mass (embryonic stem cells [ES cells]) or primordial germ cells (embryonic germ cells). In adults, stem cells have been isolated from tissues as diverse as bone marrow, skeletal muscle, brain tissue, and fat. Regardless of their origin, stem cells are maintained and propagated in an undifferentiated state in culture. Characteristically, stem cells express oct4, Sox2, and Nanog (see p. 42), which are involved in maintaining the undifferentiated state.

In response to specific combinations of exogenous agents (e.g., cocktails of growth factors) added to the culture medium, stem cells can be induced to differentiate into specific adult cell types, for example, red and white blood cells, neurons, skeletal and cardiac muscle, or cartilage. When introduced into living tissues, poorly defined local factors can direct the differentiation of adult or embryonic stem cells into specific adult cell types. These techniques have tremendous potential for the treatment of a variety of conditions, including diabetes, parkinsonism, blood diseases, and spinal cord injury, but many complicating factors (e.g., immune rejection of the implanted cells) must be dealt with before these techniques become practical and safe for human application.

An important development in stem cell technology has been the production of induced pluripotent stem cells (IPS cells) from somatic cells of adults. If genes characteristic of embryonic stem cells (e.g., Oct4, Sox2, and Nanog) are introduced into a differentiated adult cell (e.g., a fibroblast), the cell will then assume the properties of an embryonic stem cell. Like an embryonic stem cell, an artificially created stem cell that is exposed to an appropriate environment will be capable of differentiating into a wide variety of other adult cell types. This technique has great potential for patient-specific therapy. For example, in the treatment of a genetic disease characterized by the inability to manufacture a specific molecule, cells of a patient could be converted into IPS cells, subjected to corrective gene therapy, and then reintroduced into that person’s body. Under ideal conditions, the introduced IPS cells would then begin to produce the deficient molecule. Cloning, which is often confused with stem cell technology, consists of fusing or introducing an adult cell or nucleus into an enucleated oocyte and allowing the hybrid cell to develop into an embryo and ultimately to mature into an adult. Although forms of cloning have been successfully accomplished since the 1960s, the creation of the sheep Dolly in 1996 had the greatest influence on the public imagination. Cloning is not easily accomplished, and there is a significant incidence of abnormal development among cloned individuals.

Cloning and stem cell technology have brought to light significant ethical and societal issues. For example, human embryonic stem cells have been introduced into mouse blastocysts in an attempt to determine the influences that control their differentiation. It will be fascinating to see how these issues, all sides of which have profound implications, are resolved.

Genetic engineering of specific genes is possible in ES cells. When such genetically manipulated cells are introduced into blastocysts, they can become incorporated into the host embryo (see Fig. 3.12B). If the progeny of a genetically engineered ES cell become incorporated into the germline, the genetic trait can be passed to succeeding generations.

Zona Pellucida

During the entire period from ovulation until entry into the uterine cavity, the ovum and the embryo are surrounded by the zona pellucida. During this time, the composition of the zona changes, through contributions from the blastomeres and the maternal reproductive tissues. These changes facilitate the transport and differentiation of the embryo. After the embryo reaches the uterine cavity, it begins to shed the zona pellucida in preparation for implantation. This is accomplished by a process called blastocyst hatching. A small region of the zona pellucida, usually directly over the inner cell mass in the primate, dissolves, and the blastocyst emerges from the hole. In rodents, blastocyst hatching is accomplished through the action of cysteine protease enzymes that are released from long microvillous extensions (trophectodermal projections) protruding from the surfaces of the trophoblastic cells. Over a narrow time window (4 hours in rodents), the zona pellucida in this area is digested, and the embryo begins to protrude. In the uterus, the trophectodermal projections then make contact with the endometrial epithelial cells as the process of implantation begins. Enzymatic activity around the entire trophoblast soon begins to dissolve the rest of the zona pellucida. Only a few specimens of human embryos have been taken in vivo from the period just preceding implantation, but in vitro studies on human embryos suggest a similar mechanism, which probably occurs 1 to 2 days before implantation (see Fig. 3.3C). Box 3.1 summarizes the functions of the zona pellucida.

Implantation into the Uterine Lining

Approximately 6 to 7 days after fertilization, the embryo begins to make a firm attachment to the epithelial lining of the endometrium. Soon thereafter, it sinks into the endometrial stroma, and its original site of penetration into the endometrium becomes closed over by the epithelium, similar to a healing skin wound.

Successful implantation requires a high degree of preparation and coordination by the embryo and the endometrium (Table 3.1). The complex hormonal preparations of the endometrium that began at the close of the previous menstrual period all are aimed at providing a suitable cellular and nutritional environment for the embryo. Even before actual contact is made between the embryo and endometrium, the uterine epithelium secretes into the uterine fluid certain cytokines and chemokines that facilitate the implantation process. At the same time, cytokine receptors appear on the surface of the trophoblast. Dissolution of the zona pellucida signals the readiness of the embryo to begin implantation.

Table 3.1

Stages in Human Implantation

Age (Days)	Developmental Event in Embryos
5	Maturation of blastocyst
5	Loss of zona pellucida from blastocyst
6?	Attachment of blastocyst to uterine epithelium
6-7	Epithelial penetration
	Trophoblastic plate formation and invasion of uterine stroma by blastocyst
9-11	Lacuna formation along with erosion of spiral arteries in endometrium
12-13	Primary villus formation
13-15	Secondary placental villi, secondary yolk sac formation
16-18	Branching and anchoring villus formation
18-22	Tertiary villus formation

Modified from Enders AC: Implantation, embryology. In Encyclopedia of human biology, vol 4, New York, 1991, Academic Press.

The first stage in implantation consists of attachment of the expanded blastocyst to the endometrial epithelium. The apical surfaces of the hormonally conditioned endometrial epithelial cells express various adhesion molecules (e.g., integrins) that allow implantation to occur in the narrow window of 20 to 24 days in the ideal menstrual cycle. Correspondingly, the trophoblastic cells of the preimplantation blastocyst also express adhesion molecules on their surfaces. The blastocyst attaches to the endometrial epithelium through the mediation of bridging ligands. Some studies have stressed the importance of the cytokine leukemia-inhibiting factor (LIF) on the endometrial surface and LIF receptors on the trophoblast during implantation. In vivo and in vitro studies have shown that attachment of the blastocyst occurs at the area above the inner cell mass (embryonic pole), a finding suggesting that the surfaces of the trophoblast are not all the same.

The next stage of implantation is penetration of the uterine epithelium. In primates, the cellular trophoblast undergoes a further stage in its differentiation just before it contacts the endometrium. In the area around the inner cell mass, cells derived from the cellular trophoblast (cytotrophoblast) fuse to form a multinucleated syncytiotrophoblast. Although only a small area of syncytiotrophoblast is evident at the start of implantation, this structure (sometimes called the syntrophoblast) soon surrounds the entire embryo. Small projections of syncytiotrophoblast insert themselves between uterine epithelial cells. They spread along the epithelial surface of the basal lamina that underlies the endometrial epithelium to form a flattened trophoblastic plate. Within a day or so, syncytiotrophoblastic projections from the small trophoblastic plate begin to penetrate the basal lamina. The early syncytiotrophoblast is a highly invasive tissue, and it quickly expands and erodes its way into the endometrial stroma (Fig. 3.18A and B). Although the invasion of the syncytiotrophoblast into the endometrium is obviously enzymatically mediated, the biochemical basis in humans is not well understood. By 10 to 12 days after fertilization, the embryo is completely embedded in the endometrium. The site of initial penetration is first marked by a bare area or a noncellular plug and is later sealed by migrating uterine epithelial cells (Fig. 3.18C and D).

As early implantation continues, projections from the invading syncytiotrophoblast envelop portions of the maternal endometrial blood vessels. They erode into the vessel walls, and maternal blood begins to fill the isolated lacunae that have been forming in the trophoblast (see Fig. 3.18C and D). Trophoblastic processes enter the blood vessels and even share junctional complexes with the endothelial cells. By the time blood-filled lacunae have formed, the trophoblast changes character, and it is not as invasive as it was during the first few days of implantation. Leakage of blood from the uterus at this stage can produce “spotting,” which is sometimes misinterpreted to be an abnormal menstrual period.

While the embryo burrows into the endometrium, and some cytotrophoblastic cells fuse into syncytiotrophoblast, the fibroblastlike stromal cells of the edematous endometrium swell, with the accumulation of glycogen and lipid droplets (see Fig. 7.6). These cells, called decidual cells, are tightly adherent and form a massive cellular matrix that first surrounds the implanting embryo and later occupies most of the endometrium. Concurrent with the decidual reaction, as this transformation is called, the leukocytes that have infiltrated the endometrial stroma during the late progestational phase of the endometrial cycle secrete interleukin-2, which prevents maternal recognition of the embryo as a foreign body during the early stages of implantation. An embryo is antigenically different from the mother and consequently should be rejected by a cellular immune reaction similar to the type that rejects an incompatible heart or kidney transplant. A primary function of the decidual reaction apparently is to provide an immunologically privileged site to protect the developing embryo from being rejected, but a real understanding of how this is accomplished has resisted years of intensive research.

Frequently, a blastocyst fails to attach to the endometrium, and implantation does not occur. Failure of implantation is a particularly vexing problem in in vitro fertilization and embryo transfer procedures, for which the success rate of implantation of transferred embryos remains at about 25% to 30% (see Clinical Correlation 2.1).

Embryo Failure and Spontaneous Abortion

Many fertilized eggs (>50%) do not develop to maturity and are spontaneously aborted. Most spontaneous abortions (miscarriages) occur during the first 3 weeks of pregnancy. Because of the small size of the embryo at that time, spontaneous abortions are often not recognized by the mother, who may equate the abortion and attendant hemorrhage with a late and unusually heavy menstrual period.

Examinations of early embryos obtained after spontaneous abortion or from uteri removed by hysterectomy during the early weeks of pregnancy have shown that many of the aborted embryos are highly abnormal. Chromosomal abnormalities represent the most common category of abnormality in abortuses (about 50% of the cases). When viewed in the light of the accompanying pathological conditions, spontaneous abortion can be viewed as a natural mechanism for reducing the incidence of severely malformed infants.