Lets return to the problem of trying to understand the evolutionary origin of the Metazoa . We have looked at a variety of organisms below this level; now lets consider a top-down approach, and propose that:

The first "real animals" were worms.

What distinguishes worms from other animals? The lack of a rigid (usually jointed) skeleton. What is a skeleton?
Muscles can only contract. They do not actively elongate. An external, antagonistic force is needed to stretch them out again. This antagonistic force is usually conducted by a skeleton. The primary function of a skeleton is to provide for muscle antagonism , usually between different muscles that can be independently controlled, but sometimes between muscle and elastic antagonists. Soft-bodied animals must have solved this problem, without a rigid skeleton.

But worms share the following characteristics with other "real animals":
  1. Use muscles for locomotion, by creating "waves" of muscle contraction.
    I want to focus on locomotion, and the fact that there is an enormous gap between the use of actin and myosin for locomotion by an amoeboid cell like Naegleria , and the use of actin and myosin in muscle cells of worms and other multicellular animals.
    In muscle cells, there are relatively stable, parallel arrays of actin filaments, and myosin interacts with these arrays to produce contraction (shortening) of the cell.
    In a cell that moves like an amoeba, actin filaments are continuously appearing and disappearing, and the combined action of actin, myosin, and many actin-regulating components produces locomotion of the cell, rather than contraction. No extracellular skeleton is needed.
    These are quite different mechanisms, but contraction appears to be simpler. Evolving from an amoeboid cell to a contractile cell was probably easy, and might have occured in several independent lines.
    If we look at present-day animals, we conclude that contractile cells are not sufficient for locomotion. Two other features are needed: a skeleton and a control system.
    Muscles are typically controlled by a nervous system so that cycles of shortening and elongation can be generated. However, the locomotion of worms typically involves rather simple propagated waves of contraction, and it is reasonable to assume that at an early stage in the evolution of animal locomotion, wave propagation could have been coordinated by electrical signals propagated by the muscle cells themselves (as in mammalian heart muscle), or by an overlying epithelial layer.
  2. Use acetylcholine as a neurotransmitter. (This has not been found in Porifera or Cnidaria, even though acetylcholinesterase may be present in these and in some unicellular eukaryotes.)
  3. Designed for forward locomotion with a well-defined leading end: Cephalization
  4. Many are designed for locomotion on a surface, leading to bilateral symmetry .
  5. Cell specialization, at least a specialization of endodermal, gut , cells for feeding and ectodermal cells for outer covering.
  6. Multicellular
  7. Typical pattern of sexual reproduction, with development from a unicellular zygote.

If we were able to reconstruct the phylogeny of the origin of Metazoa, we would want to know the sequence of appearance of these features, and whether there was one line that led to this combination of features (monophyletic ) or more than one line (polyphyletic ).

How can we fill in the gap between an amoeba and a worm, and try to answer the major questions about this transition? There are various approaches:

Approach #1. Examining the embryological development of existing animals

More than a century ago, it was proposed that the evolutionary history of a species could always be determined by studying its embryology. This view is often called "recapitulation theory", and summarized by the phrase "ontogeny recapitulates phylogeny". It is based on two premises: 1) that animals must go through the same stages in their embryological development as their ancestors went through in their embryological development. 2) therefore, evolution can only proceed by adding on stages beyond the previous adult stage. These premises are clearly wrong. We know enough by now to know that evolution does not follow any restrictive rules such as this, and that embryological development can be altered by evolution just like any other characteristics.
So why was this theory ever believed at all? It is most closely associated with a German zoologist, Ernst Haeckel. It was primarily based on resemblences between the larval forms of different groups of Crustaceans, which easily suggest that evolution occured by addition of new adult stages. In addition, it appeared to be applicable to the resemblences between embryos of different groups of vertebrates, especially the formation of gill slits in early embryos of higher vertebrates that have no gill slits as adults. However, there is no theoretical basis for generalizing from these examples.
Nevertheless, Haeckel made this generalization, and proposed that the early embryology of animals such as sea urchins revealed their phylogeny.

Early embryology of the sea urchin embryo:

blastula gastrula mesodermal invagination

He proposed that the earliest multicellular animals were "Blasteas", simple spherical organisms. These were followed by "Gastreas", with simple guts. One line of evolution, leading to modern Cnidarians, diverged at the Gastrea level and never acquired a third cell layer or bilateral symmetry. These were known as "diploblastic" organisms. Another line led to bilaterally symmetric, triploblastic, animals.
This is a "colonial", or "cleavage" theory, with the Blastea evolving by cell multiplication, with Volvox as a model for this process. Both the Blastea and the Gastrea were assumed to be covered with cilia for locomotion. As a further speculation, we might propose that contractile cells evolved to close the mouth of the gut cavity during feeding: thus a preadaptation for later use of contractile cells in locomotion.

Recapitulation theory is no longer accepted. Even if it were, the argument would be weak, because most animal species do not develop in the manner shown for sea urchin embryos. In fact, most animal species are insects, and many insect embryos develop by cellularization.

This does not say that all colonial theories are wrong. It just says that embryology does not provide evidence to prove it. For colonial theories, involving a ciliated or flagellated Blastea-like form, characteristics 765 would probably appear before the locomotory characteristics 4321. However, colonial theories involving an earlier appearance of contractile locomotion are also imaginable.

Approach #2: Fossils
Fossils have not been a very important source of information about the transition that we are discussing, because the earliest multicellular animals were small and soft-bodied and did not leave a fossil record. They may have appeared between -600 and -800 Myr, or even earlier. Extensive biomineralization by animals began at the beginning of the Cambrian era ( -540 Myr) and by -520 Myr (the Burgess shale) already show a great diversity of animal types that are unlikely to be the most primitive forms. Many of the present day phyla appear to be represented at this time. Some fossils have been found from the late PreCambrian (known as the Ediacaran faunas, from about -575 Myr). In this assemblage, radial symmetry appears to be predominant. This could support the idea that radially symmetric, diploblastic animals that were ancestors of present Cnidarians evolved at a very early stage. However, it could also mean that sessile animals evolved protective coverings that were preserved as fossils before such coverings were evolved by more active, mobile, worm-like animals.
For further information, the following review is a good starting point:
S. C. Morris (1993) Nature 361: 219-225.

Approach #3: Molecular evolution
Molecular evolution means using the sequences of proteins and nucleic acids that are common to several organisms as a source of information about their relationships. This can be most useful if the molecules are used for identical and equally important functions in all of the organisms, so that natural selection does not cause changes in sequence. It would be even more useful if random changes in sequence occurred at a constant rate, independent of context, so that the degree of sequence difference was an accurate measure of the time since two organisms had a common ancestor. It is difficult to prove that this was the case, and sometimes possible to disprove it. There are various ways to measure and interpret sequence similarity, and not all give the same answers. This is still an evolving field.
For large-scale phylogeny, molecules must be chosen that are highly conserved, so that they remain reasonably similar over time periods of hundreds of millions of years. For this purpose, the most useful molecules appear to be the ribosomal RNA molecules. Here are some examples of conclusions that have been reached by studying these molecules:

a) Christen et al. (1991), EMBO Journal 10:499-503 used partial sequences of the largest rRNA, 28S rRNA. The new sequence information in this paper is primarily from Cnidarians and sponges. This information indicates a relatively close relationship between Cnidarians and sponges, and relatively equidistant relationships between this group, other Metazoa, and plants. Fungi and ciliates were more distantly related.

b) Wainwright et al. (1993), Science 260: 340-342 used a smaller rRNA, 16S rRNA, and reached a different conclusion, with Cnidarians and other metazoa more closely related to each other than to sponges, and all of these animal groups more closely related to fungi than to plants. Choanoflagellates were not much more closely related to sponges than to other animals, and showed a relatively strong relationship to fungi.

Both these, and many other studies place Dictyostelium near the "root" of the eukaryotic evolutionary tree.

Clearly, these studies do not yet provide the final word on phylogeny. However, only a relatively small number of organisms have been used to date to obtain this sequence information, and there is lots of potential for further information.

Approach #4) Molecular genetics
The newest ideas come from work that was initiated by Ed Lewis, indicating that the development of the body segments of Drosophila is regulated by a series of genes located on a chromosome in the same sequence as the segments that they regulate. Subsequently, this has been found to be true for several other species, including unsegmented species, and is now interpreted as a mapping of an anterior-posterior gradient of developmental regulation to a linear sequence of regulatory Hox genes on the chromosome. Some zoologists have suggested that this is the definitive characteristic of all multicellular animals, and they have given it a name : the zootype . The number of organisms in which this arangement has been found is still small, but growing, and it does include two Cnidarians, Drosophila, Caenorhabditis, a leech, and most Chordate classes. It suggests that here is an important new source of information about similarities and differences between present-day organisms.
If you want to explore this idea further, the reference is
J. M. W. Slack et al., (1993) Nature 361: 490-492.
An earlier discussion of interest is J. Lewis, Nature 341:382-383 (1989).
None of these methods have given us an answer to the question of whether the ancestors of worms used actin-myosin based motility or cilia before they evolved contractile cells. We cannot answer the question of whether all contractile cells are homologous , because they evolved once in a common ancestor of all Metazoa that have muscles, or several times (parallel evolution) (in which case muscles of different groups may be analogous ).
-->Forward to next section