Literature

"The rise of the central nervous system [CNS] in animal evolution has puzzled scientists for centuries." According to evolutionary theory, "Vertebrates, insects and worms evolved from the same ancestor, but their CNSs are different and were thought to have evolved only after their lineages had split during evolution." To gain an insight into what CNS of the last common ancestor of these groups might have been like, researchers selected for study the nervous system of a marine annelid worm called Platynereis dumerilii. "Platynereis can be considered a living fossil," says Arendt (who led the team), "it still lives in the same environment as the last common ancestors used to and has preserved many ancestral features, including a prototype invertebrate CNS."
"Our findings were overwhelming," says Alexandru Denes (who carried out the research in Arendt's lab). "The molecular anatomy of the developing CNS turned out to be virtually the same in vertebrates and Platynereis." The research paper expresses it thus: "Our comparative study of mediolateral neural patterning and neuron-type distribution in the developing trunk CNS of the annelid Platynereis revealed an unexpected degree of similarity to the mediolateral architecture of the developing vertebrate neural tube."
The implication is that something very like the vertebrate CNS existed in the putative common ancestor: "Taken together, our data make a very strong case that the complex molecular mediolateral architecture of the developing trunk CNS, as shared between Platynereis and vertebrates, was already present in their last common ancestor, Urbilateria."
So, the implication is that a highly complex nervous system existed in Precambrian organisms. "Such a complex arrangement could not have been invented twice throughout evolution, it must be the same system," adds Gaspar Jekely. "It looks like Platynereis and vertebrates have inherited the organisation of their CNS from their remote common ancestors."
This is interesting logic, because it links complexity with improbability. The greater the complexity, the less likely it is for processes governed by law or chance to produce it. Science built on naturalism has no other options available: complexity has to be the result of improbable interactions in nature. Design inferences are not permitted alongside inferences of law or chance. It is not acceptable to think that if it is too complex to have been 'invented' twice by evolution, it is also reasonable to conclude that it could not have been 'invented by evolution' even once.
Based on the research, the authors acknowledge that their findings "would imply that it was initially present also in the evolutionary lines leading to Drosophila, the nematode Caenorhabditis, and the enteropneust Saccoglossus. Yet it is clear from the available data that these animals are missing or have modified at least part of this pattern." Loss of complexity is a part of their story, which is a surprising result.
One cannot help notice that studies of this type keep pushing back the origin of complexity into the Precambrian, dominated until the Ediacaran by single-celled animals. Arguing that all this complexity is the product of mutation and natural selection becomes less and less convincing as we learn more about the data needing an explanation.

Molecular Architecture of Annelid Nerve Cord Supports Common Origin of Nervous System Centralization in Bilateria
Alexandru S. Denes, Gaspar Jekely, Patrick R.H. Steinmetz, Florian Raible, Heidi Snyman, Benjamin Prud'homme, David E.K. Ferrier, Guillaume Balavoine, and Detlev Arendt
Cell, Vol 129, 277-288, 20 April 2007

Summary: To elucidate the evolutionary origin of nervous system centralization, we investigated the molecular architecture of the trunk nervous system in the annelid Platynereis dumerilii. Annelids belong to Bilateria, an evolutionary lineage of bilateral animals that also includes vertebrates and insects. Comparing nervous system development in annelids to that of other bilaterians could provide valuable information about the common ancestor of all Bilateria. We find that the Platynereis neuroectoderm is subdivided into longitudinal progenitor domains by partially overlapping expression regions of nk and pax genes. These domains match corresponding domains in the vertebrate neural tube and give rise to conserved neural cell types. As in vertebrates, neural patterning genes are sensitive to Bmp signaling. Our data indicate that this mediolateral architecture was present in the last common bilaterian ancestor and thus support a common origin of nervous system centralization in Bilateria.

See also:
The origin of the brain lies in a worm
Researchers discover that the centralised nervous system of vertebrates is much older than expected
EMBL Press Release 20 April 2007