Changes in cell lineage specification elucidate evolutionary relations in Spiralia.
Comparative embryology of the various body plans, and an understanding of the molecular regulation of the establishment of these body plans, are powerful tools that can help us reconstruct the evolutionary relations between the animal phyla. The first Metazoa were undoubtedly radially symmetrical animals with two germ layers: ectoderm and endoderm. Examples of these diploblastic creatures can still be found today, in marine and fresh waters throughout the world. The evolution of a third germ layer, the mesoderm, led, in part, to the great Precambrian radiation of the animal kingdom. Elucidation of the developmental mechanisms underlying the formation of mesoderm could shed light on the evolutionary relations among different phyla. Whether this third germ layer evolved once or developed convergently in a number of ancestral diploblastic forms remains to be demonstrated.
Our recent research has focused on the development of two cell lines typical of the Spiralia (i.e., phyla with spiral cleavage): (1) the stem cell of the mesodermal bands (the mesentoblast) and (2) the trochoblasts. The mesentoblast was chosen because the origin of mesoderm is consistently similar in different Spiralia. The mesentoblast is formed from a single primary endodermal cell that is induced to follow a developmental program different from the other endodermal cells. After induction, the mesentoblast divides and gives rise to the stem cell of the left and right mesodermal bands. A comparative study of mesentoblast formation may be used to elucidate the evolutionary relations within the Spiralia.
Trochoblasts were analyzed because this cell line can be found in a number of Spiralia. These cells are ectodermally derived and form the prototroch - the larval locomotory organ typical of such spiralians as molluscs and annelids but absent in other spiralians like the nemerteans and flatworms.
In this paper, we discuss the molecular and developmental aspects of trochoblast and mesentoblast formation and their significance to the analysis of the phyletic relations between spiralian phyla.
The trochoblasts constitute the first fully specified cell line in a number of spiralian embryos (1). Detailed knowledge of trochoblast specification, however, is limited to Patella vulgate, the common limpet. Specification in Patella requires that the third cleavage is executed correctly (2); if this cleavage is inhibited, no trochoblast-specific gene expression will occur. Trochoblast specification is completed after the fourth cleavage; thereafter the trochoblasts divide only twice more and then differentiate into ciliated cells. From the fourth cleavage onward, specification is autonomous: i.e., cells isolated from the 16-cell embryo go through two cleavages, enter a division arrest, and become ciliated, just as in the intact embryo.
To investigate the molecular mechanism of trochoblast specification, we first focused on genes encoding tubulin as part of this process. Trochoblasts bear a large number of cilia which, in turn, are mainly composed of tubulin. In situ hybridization revealed that tubulin genes are expressed one cell cycle before the last division of the trochoblasts (3). One of the tubulin genes that we cloned from the Patella genome appeared.to be trochoblast specific. The promoter of this gene was coupled to the Lac-Z-reporter gene, and the construct was injected into 2-cell embryos. Expression of the reporter gene appeared to be limited to the trochal cells, and began about 30 minutes after the appearance of tubulin mRNA (3). Extensive mapping in the promoter showed that only a small region, between -108 and -1 with respect to the transcription start, is absolutely required for correct expression (4). In this region, two elements - located between -108 and -68 and between -52 and -42 - serve different functions in establishing correct spatiotemporal gene expression. Mutation of the -108/-68 element results in expression of the reporter gene in non-trochoblasts. Mutation in the -52/-42 region completely abolishes expression of the reporter gene. In addition to these two elements, two others located in the regions -418/-108 and + 1/+487, are required for correct expression; these latter elements can be located either before or after the -108/-1 region. We therefore consider the region -108/-1 to be the core of the promoter.
Nuclear proteins from different stages of development were isolated and a southwestern blot performed; the core region was used as a probe. Each stage shows a specific array of proteins binding to this core region (A. H. E. M. Klerkx and A. E. van Loon, unpub. data). We therefore assume that at different times in development, different proteins bind to the core region.
As trochoblasts are not exclusively formed in gastropods, but also in other molluscan classes, the Patella tubulin promoter was coupled to the Lac-Z gene and injected into embryos of representatives of other classes of molluscs. Embryos of a polyplacophoran (Acanthochiton) and a scaphopod (Dentalium) showed an expression pattern completely comparable with that in Patella (A. H. E. M. Klerkx, W. G. M. Damen, A. E. van Loon, and J. A. M. van den Biggelaar, unpub. data). Thus, the molecular mechanism for the regulation of a trochoblast-specific gene is conserved in representatives of different molluscan classes.
The spiralian taxon Annelida is presumed to include the closest relatives of the molluscs, and these worms form trochoblasts that originate from the same cell line as in the molluscs. The tubulin promoter gene construct therefore was injected into embryos of the polychaete annelid Platynereis. Seven of the resulting embryos survived to the trochophore stage. One of these showed expression, and that expression was limited to trochoblasts. Injections of the construct into a large number of embryos of another polychaete annelid, Nereis, have not resulted in Lac-Z-expression.
Nemerteans do not develop into larvae with a prototroch, but are supposed to be ancestral to the molluscs and annelids. We therefore examined the expression of the Patella tubulin promoter construct in nemertean embryos (Cerebratulus lacteus). Expression was found, but was not restricted to a specific domain of the 24-h larvae. Similarly, embryos of another spiralian taxon, the flatworms, do not develop trochophore larvae. A small number of embryos of the polyclad flatworm Hoploplana were injected with the construct, and no expression was found.
The molecular aspects of trochoblast-specific gene expression in molluscs have been conserved in the Polyplacophora, Scaphopoda, and Gastropoda. As the trochoblasts arise from the same cells in molluscs and annelids, we conclude that they are spiralian phyla with a close evolutionary relationship. The conservation, in molluscs and annelids, of the molecular mechanism regulating the expression of a trochoblast-specific gene needs further support. On the other hand, nemerteans and flatworms do not share the formation of trochoblasts, but nemerteans seem to have the molecular mechanism that is required for a cell-specific expression of the Patella trochoblast-specific gene. We therefore consider nemerteans, as well as flatworms, to be more distantly related to molluscs and annelids.
In many Spiralia, the most important contribution to the mesoderm derives from the mesentoblast, which produces the two mesodermal bands. In ancestral molluscs, the mesentoblast arises from a primary endodermal cell after an inductive interaction with micromeres in the animal hemisphere (5). This induction also establishes the plane of bilateral symmetry and dorsoventral polarity.
In embryos of Patella, the endodermal macromeres 3A-3D extend in the animal direction and make contact with the ectodermal micromeres of the opposite animal pole. Of these macromeres, only one maintains these contacts. This macromere becomes the mesentoblast precursor cell (5). Previous work on the development of the dorsoventral axis and mesentoblast formation in a number of gastropod families has shown that gastropod evolution has been accompanied by a heterochronic shift in mesentoblast formation (6, 7, [ILLUSTRATION FOR FIGURE 1 OMITTED]). In a series of gastropods, from Archeogastropoda to Pulmonata, mesentoblast formation is shifted from late cleavage stages to much earlier developmental stages.
In annelid embryos a similar extension of the macromeres 3A-3D occurs during the interval between the fifth and the sixth cleavage. But no single macromere is centralized. Despite the uniform contact that these cells have with the micromeres, one is induced to produce the mesentoblast. This inductive event also establishes the dorsoventral axis. A heterochronic shift in the specification of the dorsal quadrant, comparable to that found in gastropods, is also found in annelids (8).
Specification of the dorsoventral axis and mesentoblast formation in nemertean and flatworm embryos show similarities and differences compared to molluscs and annelids and to each other. The nemertean embryo is not divided into dorsal, ventral, right, and left quadrants, but into two dorsolateral and two ventrolateral quadrants (9). Despite this alternative quadrant arrangement with respect to the first cleavage planes, bandlets of mesenchymal cells seem to be derived from the same endomesodermal cell (4d) as in annelids and molluscs (10). Like molluscs and annelids, flatworm embryos are also divided into dorsal, ventral, and two lateral quadrants; the specification of the dorsal quadrant, however, must be different. After the formation of the fourth quartet of micromeres, the micromeres 4a-4d extend in the animal direction, in contrast to the macromeres 3A-3D in molluscan and annelid embryos. Finally, it is the micromere of the ventral quadrant (4b) that maintains the contacts with the animal micromeres (van den Biggelaar, unpub. obs.). Micromere 4d of the opposite dorsal quadrant then develops the mesentoblast. These differences in mesentoblast formation between annelids, molluscs, and nemerteans on the one hand, and flatworms on the other hand, again demonstrate that molluscs, annelids and nemerteans are more close related than with flatworms. The differences between nemerteans and the other two phyla (annelids and molluscs) with respect to the dorsoventral axis specification would argue that annelids and molluscs are more closely related to each other than either is to the nemerteans.
Resemblances in mesentoblast specification and the conservation of the regulatory mechanisms of a trochoblast-specific gene in three different classes of molluscs are consistent with the idea of a monophyletic origin of the molluscs. Trochoblast and mesentoblast specification (coupled to dorsoventral axis formation) in molluscs and annelids strengthens the idea of a close phylogenetic relationship between these phyla. Nemerteans and flatworms have distinct modes of dorsoventral axis formation and do not have a trochoblast cell line, excluding a close evolutionary relation with annelids and molluscs as well as with each other.
1. Wilson, E.B. 1904. Experimental studies in germinal localization. II Experiments on the cleavage-mosaic in Patella and Dentalium. J. Exp. Zool. 1: 197-268.
2. Damen, W.G. M, A. H. E. M. Klerkx, and A. E. van Loon. 1996. Micromere formation at third cleavage is decisive for trochoblast specification in the embryogenesis of Patella vulgata. Dev. Biol. 178: 238-250.
3. Damen, W. G. M., L.A. van Grunsven, and A.E. van Loon. 1994. Transcriptional regulation of tubulin gene expression in differentiating trochoblasts during early development of Patella vulgata. Development 120: 2835-2845.
4. Damen, W. G. M., and A. E. van Loon. 1996. Multiple cis-acting elements act cooperatively in directing trochoblast-specific expression of the alpha-tubulin-4 gene in Patella embryos. Dev. Biol. 176: 313-324.
5. van den Biggelaar, J. A.M. 1977. Development of dorsoventral polarity and mesentoblast determination in Patella vulgata. J. Morphol. 154: 157-186.
6. van den Biggelaar, J. A. M. 1996. The significance of the early cleavage pattern for the reconstruction of Gastropod phylogeny. Pp 155-160 in Origin and Evolutionary Radiation of the Mollusca, JD Taylor, ed. Oxford University Press, Oxford.
7. van den Biggelaar, J. A. M., and G. Haszprunar. 1996. Cleavage patterns and mesentoblast formation in the gastropoda: an evolutionary perspective. Evolution 50: 1520-1540.
8. Anderson, D.T. 1973. Embryology and Phylogeny in Annelids and Arthopods. Pergamon Press, Oxford.
9. Martindale, M. Q, and J. Q. Henry. 1995. Modifications of cell fate specification in equal-cleaving nemertean embryos: alternate patterns of spiralian development. Development 121: 3175-3185.
10. Henry, J. Q., and M. Q. Martindale. 1996. The origins of mesoderm in the equal-cleaving nemertean worm, Cerebratulus lacteus. Biol. Bull. 191: 290-292.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Genetic Regulatory Networks in Embryogenesis and Evolution|
|Author:||Loon, A.E. Van; Biggelaar, J.A.M. Van Den|
|Publication:||The Biological Bulletin|
|Date:||Dec 1, 1998|
|Previous Article:||Evolution of cleavage programs in relationship to axial specification and body plan evolution.|
|Next Article:||Axial patterning in the leech: developmental mechanisms and evolutionary implications.|