Overview of Laboratory Research Program

The chorio-allantoic placenta separates the circulatory systems of the fetus and its mother, ensuring that all exchange of nutrients, wastes and gases takes place across its interface. During its brief lifespan, the placenta is therefore the sole provider of all fetal needs, and the sole route of disposal of all fetal wastes. Given that correct morphogenesis of the placenta is critical not only for its own growth and development but for that of the fetus as well, it is therefore surprising that so little attention has been focused on how the placenta is formed.

The goal of our research program is to elucidate the cellular, molecular and genetic components critical to placental morphogenesis in the mouse. Morphogenesis of the chorio-allantoic placenta involves growth and differentiation of allantoic mesoderm, chorionic ectoderm, and the ectoplacental cone as well as union between these initially well separated tissues. The allantois will become the umbilical cord whilst the chorion, together with the ectoplacental cone in the mouse, will become the mature chorionic disk. Despite the importance of these unions in placental morphogenesis, little is known about the mechanisms by which union occurs.

Current projects are designed to discover the genetic pathways, environmental cues, and cell-cell and molecular interactions that orchestrate differentiation of allantoic mesoderm into the umbilical vasculature, and union between the allantois, chorion and ectoplacental cone. Further, we have developed a method of allantoic explantation to discover the mechanism of early blood vessel formation and formation of the primordial blood plasma. Specific questions are How does allantoic mesoderm differentiate into endothelial, chorio-adhesive, mesenchymal and smooth muscle lineages?, How is distal-to-proximal polarity of differentiation established in the allantois?, What are the developmental mechanisms of chorio-allantoic union, and formation of the chorionic disk? Methods of transplantation, whole embryo culture, explants, genetics and molecular biology are being used in all aspects of this research program.

Background

When these studies were begun, little was known about the allantois other than what it looked like (Fig. 1) and where it came from (Gardner et al., 1985; Lawson et al., 1991). Over the past several years, significant progress has been made in identifying the developmental mechanisms involved in three of the allantoisÕ major activities: growth (Downs and Bertler, 2000), union with the chorion (Downs and Gardner, 1995; Downs, 2001), and vascularization (Downs and Harmann, 1997; Downs et al., 1998). In addition, a major achievement of our research program has been the characterization of allantoic development outside the fetus (Downs et al., 2001). Allantoic explants faithfully recapitulate the early stages of allantoic development, and will complement studies in the living conceptus to discover how allantoic mesodermal cell lineages coordinately segregate from each other and differentiate.

The allantois first appears as a bud-like projection consisting of mesodermal tissue (Fig. 1A). As the allantois grows, it acquires an outer layer of cells, called the mesothelium, and an inner core of vascularizing mesoderm (Fig. 1B). By 6-somite pairs, the allantois has united with the chorion (Fig. 1C). Shortly thereafter, the fused allantois becomes overtly vascularized and contains large numbers of primitive red blood cells (Fig. 1D). By approximately 11 dpc, the widespread vascular plexus becomes reduced to a single artery and vein (not shown).

Fig. 1. Growth and Development of the Murine Allantois (from Downs, 1998). (A) - (D) Histological sections of mouse conceptuses were prepared for visualization of the four phases of allantoic development. (A) Formation of the allantoic bud. Neural plate stage (7.25 - 7.5 dpc). (B) Growth of the allantois. Early somite stage (8.0 - 8.25 dpc). (C) Chorioallantoic fusion (8.5 dpc). 7-somite-stage. The arrow points to the chorioallantoic fusion junction. (D) Vascularization of the allantois. 9.5 dpc conceptus showing overt vascularization of the umbilical component of the chorioallantoic placenta. Arrowheads indicate embryonic red blood cells. Abbreviations: a, amniotic cavity; al, allantois; am, amnion; b, yolk sac blood island; ch, chorion; e, embryonic portion of the conceptus; m, mesothelium; p, ectoplacental cavity; ps, primitive streak; ys, yolk sac. x, exocoelomic cavity.

 

Significant Research Accomplishments

1. Staging mouse gastrulae by morphological landmarks (Downs and Davies, 1993). In the course of localizing the myc family of proto-oncogenes to the mouse gastrula (Downs et al., 1989), it became clear that specific developmental processes might occur during a very narrow developmental window but go unnoticed due to the inaccurate system of clumping embryos into "days postcoitum". This was especially worrisome as many morphological differences appeared between conceptuses within individual litters of mice. To provide consistency and reproducibility both within and between experiments, well-known morphological landmarks were consolidated into a simple staging system for use on gastrulating mouse embryos in the dissection microscope. The system has been invaluable not only for our own research on allantoic development as, in its absence, it is unlikely that the developmental processes described below would have been discovered, but for many other studies world-wide.

2. Chorioallantoic fusion is dependent upon the developmental maturity of the allantois (Downs and Gardner, 1995; Downs, 2001). The first question addressed concerning allantoic development was How does the allantois fuse with the chorion? The allantois grows into the exocoelomic cavity (Fig. 1B) and eventually makes contact with the chorion to form the chorio-allantoic placenta (Fig. 1C). Are the molecules involved in this morphogenetic event expressed constitutively on both the allantois and chorion, expressed upon contact between the two tissues, or expressed gradually on one or both of these tissues? To distinguish between these possibilities, a novel microsurgical technique was devised in which distal halves of allantoises were placed into the exocoelomic cavity of hosts whose own allantois had been removed. Operated conceptuses were then cultured for varying time periods and examined. Comparison of operated with intact unoperated conceptuses revealed that chorioallantoic fusion (i) begins at approximately 3-4-somite pairs and is maximal by 6-7-somites, (ii) is mediated by the mesothelial surfaces of the allantois and the chorion, (iii) is dependent upon the developmental maturity of the allantois, and (iv) may involve an internal timing mechanism (discussed in Downs, 1998). Results predicted that the molecules involved in chorioallantoic attachment would be expressed gradually on the surface of the allantois and constitutively on the chorion. These results were preliminarily verified by serendipitous knockouts of VCAM1 and a4-integrin (Gurtner et al., 1995; Kwee et al., 1995; Yang et al., 1995), a receptor/counter-receptor complex in whose absence chorioallantoic fusion does not occur. VCAM1 was expressed in the allantois, and its counter-receptor, a4integrin, was expressed in the mesodermal component of the chorion. Current studies have fine-mapped these expression patterns and the results accord perfectly with predictions of classical embryology, confirming a mechanism of gradual acquisition of adhesiveness by the allantois (Downs, 2001)

3. The murine allantois differentiates with distal-to-proximal directionality (Downs and Harmann, 1997). The allantois appears as a small mesodermal bud but, within 12 hours, it acquires two major cell populations, an outer layer of mesothelium and an inner core of vascularizing mesoderm. To discover when and where allantoic mesoderm differentiates and the extent to which allantoic cells are pluripotent, the nascent allantois of genetically-labeled conceptuses was subdivided into three regions from which very small groups of cells were removed and transplanted back to the allantois and into ectopic fetal sites. Transplantations into the allantois measured allantoic fate, and demonstrated that allantoic mesoderm contributes only to the allantois, and not to the fetus. These results also cast doubt on the hypothesis that the germ line is set aside in the allantois and then re-enters the fetus at the appropriate time. Transplantations to ectopic fetal sites measured developmental potency and showed that differentiation of allantoic mesoderm is directional, with allantoic cells more distal to the embryo initially farther along in their developmental program than those more proximal (Fig. 2). Moreover, distalmost allantoic cells are also the oldest ones. Together, these findings strongly suggested that differentiation of allantoic mesoderm is dependent upon both cell position and cell age. Further, they predicted that directionality of differentiation ought to begin in the distal region and be evident as changes in morphology and expression of differentiated cell markers. Because colonization of donor allantoic mesoderm was predominantly into the fetal hosts’ dorsal aorta, a major blood vessel that is formed by vasculogenesis, we also tentatively concluded that the allantois vascularizes intrinsically by vasculogenesis rather than extrinsically, by angiogenesis. These findings raised the possibility that, as in the yolk sac, vasculogenesis in the allantois may be accompanied by erythropoiesis. These hypotheses were addressed in Point 4, below.

 

 Fig. 2. Directionality of Allantoic Differentiation. Schematic diagram depicting the proximodistal axis of differentiation in the headfold-stage allantois (from Downs, 1998). The allantois is continuous with the primitive streak (PS). Lawson et al. (1991) have demonstrated that cells from the epiblast ingress into the primitive streak where they are transformed into extraembryonic mesodermal cells. These cells translocate distally into the allantoic base (indicated by the arrow extending from the PS to the distal allantoic region). As more extraembryonic mesodermal cells are added from the streak to the allantois, the basal cells move distally, first into the mid-region where they become differentiated into angioblasts, and secondly into the distal region, where some cells may continue to differentiate into angioblasts and others into cells specialized for fusion with the chorion or endothelium at the chorioallantoic fusion junction (described in Downs and Harmann, 1997).

 

4. Vascularization in the murine allantois occurs by vasculogenesis that is not accompanied by erythropoiesis (Downs et al., 1998). To discover whether formation of allantoic blood vessels was intrinsic to allantoic mesoderm ("vasculogenesis") or was the result of invasion of blood vessels from the adjacent yolk sac and nearby fetus ("angiogenesis"), a combination of microsurgery, morphology, culture of allantoises in isolation and immuno- and histochemical analyses were used. In addition to discovering that allantoic blood vessels are formed de novo, by vasculogenesis involving transformation of mesoderm into the endothelial cell lineage, we also demonstrated that allantoic vasculogenesis is not accompanied by erythropoiesis. This finding enabled us to eliminate the erythroid lineage from further consideration of allantoic differentiation. We also confirmed by morphology and localization of Flk-1, an early marker of endothelial cells, that transformation of allantoic mesoderm into the outer mesothelial cell lineage and the inner core endothelial cell lineage occurs with distal-to-proximal directionality as early as the neural plate stage (approximately 7.75 days postcoitum, dpc). Then, as development proceeds, the allantoic vasculature spreads to the base of the allantois, eventually uniting with those of the yolk sac and fetus, forming with them a continuous circulatory network designed to maximize oxygen intake required for the development and wide-scale organogenesis of the fetus. We hypothesized that a key player in differentiation of allantoic mesoderm may be the overlying mesothelium. This is because flk-1 is expressed only in the core, whereas flk’s ligand, VEGF, is expressed in mesothelium, suggestive of a paracrine mechanism of cell signaling between mesothelium and core mesoderm in the formation of endothelium. We have recently supported this hypothesis by localizing VEGF to allantoic mesothelium (Downs et al., 2001).

5. Growth in pre-fusion murine allantoises involves cell proliferation and activity of the primitive streak (Downs and Bertler, 2000). As described above, the murine allantois is composed of extraembryonic mesoderm which, prior to fusion with the chorion, differentiates into at least two cell lineages: a chorio-adhesive cell lineage called mesothelium, and the endothelium of the umbilical vasculature. Differentiation into these lineages occurs with distal-to-proximal directionality. How the allantois grows is less clear, but cell proliferation and addition of mesoderm from the underlying primitive streak appear to play important roles. Recently, we investigated the relationship between growth and differentiation in the murine allantois. Techniques of histology and microsurgery were used to examine pre-fusion allantoises at 9 developmental timepoints which differed by approximately 2 hours. Cell counts revealed that allantoic size increased over time. Two hours of exposure to colcemid enhanced mitotic figures, which were used to calculate the relative number of proliferating cells (mitotic index, MI) in pre-fusion allantoises at each developmental timepoint. Cell proliferation was highest in nascent allantoises and showed signs of slowing by 2 somite pairs. By 5-6-somite pairs, when most allantoises are attaching to the chorion, the overall MI decreased significantly. No regional differences in the mitotic index were observed at any developmental stage. Total cell numbers and the mitotic index were then used to discover the extent of streak contribution to pre-fusion allantoises. Cell proliferation and streak activity were highest in nascent allantoises, after which growth occurred predominantly by cell proliferation. Formation of allantoic regenerates by microsurgical removal and culture in intact conceptuses provided independent confirmation that, as the allantois matured, the primitive streak ceased to be a major contributor to its growth. Thus, the allantois grows by both mitosis and addition of mesoderm from the streak. That the periods of highest cell proliferation and streak activity coincided raises intriguing questions concerning their interplay in the control of growth in the murine allantois.

6. Allantoic explants recapitulate early allantoic development (Downs et al., 2001). Visual analysis and experimental manipulation of the allantois in whole embryo culture as described above have been essential for elucidating allantoic development. Nonetheless, use of the whole conceptus to study allantoic development is limited. Ideally, manipulation of allantoises in isolation would be an extremely useful adjunct to discover, for example, the interplay of allantoic cell lineages with each other, and the effect of varying concentrations of growth factors on allantoic vasculogenesis. Toward this end, a suitable explant system in which allantoic development is faithfully recapitulated was developed. Three cell lineages, endothelial, mesothelial and mesenchymal, were involved in expansion of the allantois in vitro. During the first 24 hours in culture, explants vascularized stereotypically and maintained distal-to-proximal directionality of differentiation. With daily feeding, explants proliferated for 48 hours more, during which the vasculature underwent remodeling. Explanted allantoic cells could be returned to intact allantoises where they integrated into appropriate cell lineages, demonstrating that culture did not adversely affect differentiation. In low serum, allantoic angioblasts neither robustly vascularized nor survived beyond 24 hours but could be rescued with exogenous VEGF. Thus, allantoises display and respond to many of the same signaling factors used by the vitelline and cardiovascular systems. Given that allantoises can be isolated free of contamination and that allantoic vasculogenesis is not accompanied by erythropoiesis, we proposed the murine allantois as an important new and complementary tool to existing in vitro systems not only for shedding light on allantoic development, but for elucidating early details of blood vessel formation.

7. Use of the allantois in clinical settings

Because the principal function of the allantois is to form the vascular connection between the mother and fetus, the umbilical vessels provide a direct gateway to the fetal circulation and a potentially invaluable system for continuous delivery of blood-borne therapeutic factors to the fetus during gestation. Combined use of genetically-engineered allantoic explants with transplantation into the fetal umbilical cord have been proposed as a method for fetal gene therapy in utero (Downs, 1998). Umbilical therapy may be critical in cases where the therapeutic factor may be toxic to the mother, or cannot cross the placental barrier, or whose half-life is so short as to make intermittent injection into the umbilical cord impractical and costly. Delivery of recombinant factors in fetal life, while the immune system is developing, may permit some therapeutic factors to be tolerated by the immune system later in life. One potentially attractive benefit to umbilical therapy is that because the placenta will be shed at birth, there will be no unforeseen deleterious effects of this approach on the adult. Our results have demonstrated that, with little effort, the murine allantois can be extensively manipulated without compromising development of the fetus. These results, along with others which have demonstrated gene expression from endothelial cell-specific promoters, suggest that the umbilical cord has the potential to express therapeutic factors via the endothelial cells that comprise its blood vessels with subsequent delivery of the factor to the fetal circulation. The mouse may be an ideal system for testing the soundness of umbilical therapy.

Significance

A major goal of biology is to identify stem and progenitor cells in both embryos and adults in order to discover the mechanisms that effect differentiation down one developmental pathway and not another. Nowadays most efforts are directed to studies outside of the fetus where, taken from their normal and potentially restrained context, stem and progenitor cells may behave differently, exhibiting wider developmental potential than would normally be found in the intact organism. The goal of our research program is to discover, through use of novel microsurgery and molecular analyses within and outside of living mouse embryos, the intrinsic and extrinsic cues leading to differentiation of mesoderm into the endothelium and supporting cells of blood vessels in the murine allantois. Further, what happens in the womb will affect all of us throughout our lives. Despite the importance of the placenta in fetal health and well-being, study of development of the chorio-allantoic placenta is largely neglected in most modern embryological studies, with focus on the mammalian fetus, as if it were a frog or a chick. With our growing blueprint of normal allantoic development, transgenic approaches can be used rationally to understand the role of genes in differentiation of mesoderm and placentation.