On the Origin of The Nervous System

+ See all authors and affiliations

Science  03 Jul 2009:
Vol. 325, Issue 5936, pp. 24-26
DOI: 10.1126/science.325_24

What did the first neurons and nervous systems look like, and what advantages did they confer on the animals that possessed them? In the seventh essay in Science's series in honor of the Year of Darwin, Greg Miller discusses some tantalizing clues that scientists have recently gained about the evolutionary origins of nervous systems.

The nervous systems of modern animals are amazingly diverse. A few hundred nerve cells are all a lowly nematode needs to find food and a mate. With about 100,000 neurons, a fruit fly can perform aerial acrobatics, dance to woo a mate, and throw kicks and punches to repel a rival. The sperm whale's 8-kilogram brain, the largest on the planet, is the navigation system for cross-ocean travel and 1000-meter dives and enables these highly social creatures to communicate. The human brain—one-sixth that size—is the wellspring of art, literature, and scientific inquiry.


But how did they all get started? What did the first neurons and nervous systems look like, and what advantages did they confer on the animals that possessed them? These were questions the father of evolution, Charles Darwin, was ill-equipped to address. Although comparative neuroanatomy dates back to ancient Greece, the tools of the trade had not been refined much by the mid-19th century. In Darwin's day, anatomists were limited to gross observations of brains; they knew relatively little about the workings of nerves themselves. Only around the time Darwin died in 1882 were scientists beginning to develop stains to label individual cells for more detailed postmortem neuroanatomical studies. Methods for investigating the electrical properties of individual neurons in living brain tissue were still decades away, to say nothing of techniques for investigating genes and genomes.

Using such modern tools, scientists have recently begun to gain some tantalizing clues about the evolutionary origins of nervous systems. They've found that some of the key molecular building blocks of neurons predate even the first multicellular organisms. By looking down the tree of life, they are concluding that assembling these components into a cell a modern neuroscientist would recognize as a neuron probably happened very early in animal evolution, more than 600 million years ago. Most scientists agree that circuits of interconnected neurons probably arose soon thereafter, first as diffuse webs and later as a centralized brain and nerves.

But the resolution of this picture is fuzzy. The order in which early branches split off the animal tree of life is controversial, and different arrangements imply different story lines for the origins and early evolution of nervous systems. The phylogeny is “a bit of a rat's nest right now,” says Sally Leys of the University of Alberta in Edmonton, Canada. Scientists also disagree on which animals were the first to have a centralized nervous system and how many times neurons and nervous systems evolved independently. Peering back through the ages for a glimpse of the first nervous systems is no easy trick.

In addition, there is some intellectual inertia that may need to be overcome, some researchers say. “If you look at any other organ or structure, people easily assume it could evolve multiple times, … but for some reason, people are stuck on [a single origin] of neurons,” says Leonid Moroz, an evolutionary neurobiologist at the Whitney Laboratory in St. Augustine, Florida.

How to build a neuron

Like nervous systems, nerve cells come in many varieties. Neurons are easy to recognize but somewhat slippery to define. They all share directionality—i.e., the ability to receive information at one end and transmit information at the other. Electrical excitability is another defining feature; a neuron can regulate the flow of ions across its outer membrane to conduct electrical impulses. Nearly all neurons form synapses, points of contact where chemical neurotransmitters convey messages between cells, and many neurons possess branches called dendrites for receiving synaptic inputs and long axons for conducting outgoing signals.

Arranged in circuits, neurons open up new behavioral possibilities for an animal. Electrical conduction via axons is faster and more precise than the diffusion of chemical signals, enabling quick detection and a coordinated response to threats and opportunities. With a few upgrades, a nervous system can remember past experiences and anticipate the future.

Although the advantages of going neural are clear, how it first happened is anything but. One hypothesis that still resonates today dates back to 1970. George Mackie of the University of Victoria in Canada envisioned something like the sheet of tissue that makes up the bell of a jellyfish as starting material. Cells in the sheet can both detect physical contact and contract in response. Mackie proposed that these multifunctional cells may have given rise to two new cell types: specialized sensory cells on the sheet's surface and muscle cells underneath. Initially, cells in the two layers touched, with ions passing through pores in the cells' membranes to conduct electrical impulses between them. With further specialization, the distance between the sensory and muscle cells grew and axons arose to bridge the gap. Eventually, “interneurons” appeared, forming synapses with sensory neurons at one end and with muscle cells at the other end.

It's a plausible scenario, says Detlev Arendt, a developmental biologist at the European Molecular Biology Laboratory in Heidelberg, Germany. But there are other possibilities, and they're not mutually exclusive, Arendt says. Moroz agrees: “Neurons may have appeared in multiple lineages in a relatively short time.”

Consistent with this idea, some crucial components of neurons exist in many cell types that predate the first neurons. Voltagegated ion channels, tiny pores that control the flow of ions across a neuron's outer membrane to create electrical signals, can be found in bacteria and archaebacteria, notes Robert Meech, a neurophysiologist at the University of Bristol in the United Kingdom. And in the 1960s, researchers found that when the single-celled Paramecium caudatum bumps into an obstacle, a voltage change sweeps from one end to the other, much like the “action potentials” that convey a signal down the length of a neuron. In Paramecium, this electrical blip reverses the beat of its cilia, temporarily altering its course. Electrical excitability, it seems, evolved long before neurons made it their specialty.

On the brink

A thorny tree.

Uncertainty about early animal evolution complicates any discussion of neural origins. This potential scenario is based on a recent genomics study.


To hunt for additional clues about neuron evolution, researchers have turned to some of the most primitive animals alive on Earth today: sponges. Many scientists think these marine and freshwater filter feeders are the living creatures most similar to the common ancestor of all animals. And to many researchers, sponges look like animals on the verge of a nervous breakthrough. Sponges don't have a nervous system, or even neurons, but they do have a surprising number of the building blocks that would be needed to put a nervous system together.

Researchers working to unravel the genome of the marine sponge Amphimedon queenslandica reported in PLoS ONE in 2007 that these animals contain the genetic blueprints for a set of proteins typically found on the receiving side of a synapse. In neurons, these proteins provide a scaffold that anchors neurotransmitter receptors into the cells' outer membrane. Yet electron microscope studies have failed to find synapses in sponges. And although their genomes contain genes for some neurotransmitter receptors, sponges appear to lack the type of receptors used for most excitatory neural communication in other animals. Thus, the function of these synaptic scaffolding proteins in a sponge is a mystery, says Kenneth Kosik, a neuroscientist at the University of California, Santa Barbara, who led the study.

Last year, the Amphimedon genome yielded another surprising find. A team led by Bernard Degnan of the University of Queensland in Brisbane, Australia, reported in Current Biology that cells in the sponge's larvae express a handful of genes that spur neural precursor cells to develop into full-fledged neurons in more complex animals. Inserting the sponge version of one of these genes into frog embryos and fruit fly larvae led to the birth of extra neurons. Degnan suspects that the cells that express them might be a type of protoneuron. These cells sit on the outer surface of the sponge larvae, and Degnan speculates that they may somehow help the free-floating larvae sense their environment and find a suitable place to settle down and metamorphose into their adult form.

Physiological experiments with sponges have also turned up signs of neural foreshadowing. Some sponges, for example, generate action potentials. In a 1997 Nature paper, Leys and Mackie reported that when the glass sponge Rhabdocalyptus dawsoni gets bumped or detects sediment in the water it filters for food, a voltage change sweeps across its body, and the cilia that pump water through the sponge's body shut down. (The glass sponge is essentially one giant cell with a continuous, weblike intra cellular space.) This electrical blip lasts about 5 seconds, compared with a millisecond or so for most action potentials in neurons, but it seems to accomplish the same basic goal as a reflex mediated by neurons: generating a coordinated response to an external stimulus.

All in all, says Leys, sponges provide a tantalizing picture of what an animal on the brink of evolving a nervous system might look like. Their cells have many of the right components, but some assembly is still required. And although they have a wider behavioral repertoire than most people realize, Leys says, their “reflexes” are far slower than those of animals with a nervous system.

But some researchers argue that sponges aren't the most primitive living animals. In one controversial study, published in the 10 April 2008 issue of Nature, a team of European and North American researchers reported that their analysis of 150 gene fragments from each of 77 animal taxa suggested that ctenophores, not sponges, are the lowest branch on the animal tree of life. Ctenophores, or comb jellies, are translucent, bloblike marine organisms that bear a passing resemblance to some jellyfish. Like true jellies, ctenophores have bona fide neurons and a simple netlike nervous system. Their position at the base of the animal family tree—if it stands up—would shake up many researchers' views on nervous system evolution.

Among the unpalatable implications, in the eyes of some researchers, is that if ctenophores came before sponges, the assorted nervous system components that have turned up in sponges may not be foreshadowing after all but rather the remnants of a nervous system that was lost after the sponge lineage split off from that of ctenophores.

What's new, jellyfish?

One way to build a neuron.

Neurons may have arisen from multifunctional cells—like those in a jelly fish's bell (far left)—that gradually became more specialized.


Another contender for most primitive animals with a proper nervous system are cnidarians, agroup that includes the truejellyfish, corals, and sea anemones. As in ctenophores, the nervous system of cnidarians is often described as a “nerve net.” This description is apt for anemones, which have a diffuse web of neurons with no discernible concentration of neurons in any one place. But some jellyfish, such as the bell-shaped Aglantha digitale, are more organized, with clearly defined bundles of nerves running around the base of the bell.

Cnidarian neurons generate action potentials and release neurotransmitters to communicate across synapses. The genome of the starlet sea anemone, Nematostella vectensis, reveals a surprisingly large array of genes encoding enzymes that synthesize or break down neurotransmitters, as well as receptors for these signaling molecules (Science, 6 July 2007, p. 86). The genome hints at a complexity of neural signaling similar to that seen in more complex animals, says Michel Anctil, a comparative neurobiologist at the University of Montreal in Canada: “The potential is there. How much they use it is what we have to figure out.”

Getting a head

Just as sponges, comb jellies, and sea anemones may hold clues to how the first nerves and nerve nets arose, other creatures may shed light on the evolution of more complex neural circuitry. “I think everybody agrees that nervous systems were at first diffuse and then evolved to be centralized,” with a concentration of neurons in the front end of the animal—that is, a brain—and a nerve cord connecting it to the rest of the body, says Arendt. “But there's no consensus yet on exactly when this happened.” Arendt and others have argued that a centralized nervous system existed in the ancestor of all bilaterally symmetrical animals, or bilaterians.

Studies dating back to the early 1990s have found that several genes involved in shaping the central nervous system as it develops in fruit flies are expressed in a similar pattern and play a similar role in nervous system development in vertebrates and in distantly related invertebrates such as annelid worms. That implies that these genes were already present in the last common ancestor of all these creatures—the ancestor of all bilaterians—and suggests to Arendt and others that this ancestor had a centralized nervous system.

But not everyone is so sure. Christopher Lowe of the University of Chicago in Illinois and colleagues have found that Saccoglossus kowalevskii, a wormlike creature that belongs to the group of invertebrates most closely related to vertebrates, the hemichordates, has many of the same neural development genes expressed in a similar pattern.Yet Saccoglossus has a mostly diffuse nervous system. If having the genes and expressing them in a given pattern isn't sufficient to establish a centralized nervous system, that leaves open the possibility that the bilaterian ancestor had a diffuse nervous system, Lowe says. “I would argue that we have a range of possibilities,” he says. Because most but not all modern bilaterians have a centralized nervous system, there will be awkward implications no matter what. If the bilaterian ancestor had a diffuse nervous system, centralized nervous systems must have originated multiple times in multiple bilaterian lineages—a far less parsimonious scenario than a single origin. On the other hand, if the ancestor had a centralized nervous system, several lineages, including that of Saccoglossus, must have later reverted to a diffuse nervous system—an apparent down-grade that's hard to explain.

The puzzles don't end there. Fastforwarding a bit in evolutionary time raises a new set of questions. What is the origin of the myelin insulation that speeds conduction down axons and ensures the fidelity of neural signals? Or of the glial cells that are proving to have important roles in brain function and appear to be more numerous in complex nervous systems?

For that matter, how many ways are there to build a complex brain? Aristotle's notion of a scala naturae, or natural ladder, influenced the thinking of researchers well into the 20th century, says R. Glenn Northcutt, an expert on vertebrate brain evolution at the University of California, San Diego. “It was assumed that all vertebrates and invertebrates could be arranged in a linear series, with man and the angels at the top,” Northcutt says, but “we now know that's just nonsense.” Most researchers now agree that equally complex—but anatomically different—brains have evolved in birds, mammals, and other animal lineages, Northcutt says: “At least four or five times independently, … major radiations of vertebrates have evolved complex brain structure.” But whether brains that are put together differently operate on similar principles is still an open question. And then there is the enduring question of what, if anything, is special about the human brain. Perhaps the emerging clues about the long evolutionary path we've taken will one day help us decide where we are.


A. S. Denes et al., "Molecular Architecture of Annelid Nerve Cord Supports Common Origin of Nervous System Centralization in Bilateria." Cell 129, 277 (2007).

C. W. Dunn et al., "Broad phylogenomic sampling improves resolution of the animal tree of life." Nature 452, 745 (2008).

S. P. Leys and G. O. Mackie, "Electrical recording from a glass sponge." Nature 387, 29 (1997).

C. J. Lowe et al., "Anteroposterior Patterning in Hemichordates and the Origins of the Chordate Nervous System." Cell 113, 853 (2003).

C. J. Lowe et al., "Dorsoventral Patterning in Hemichordates: Insights into Early Chordate Evolution." PLoS Biology 4(9), e291 (2006).

G. O. Mackie, "Neuroid Conduction and the Evolution of Conducting Tissues." The Quarterly Review of Biology 45, 319 (1970).

R. W. Meech and G. O. Mackie, "Evolution of excitability in lower metazoans," in Invertebrate Neurobiology, G. North and R. Greenspan, Eds. (Cold Spring Harbor Laboratory Press, New York, 2007).

H. Philippe et al., "Phylogenomics Revives Traditional Views on Deep Animal Relationships." Current Biology 19, 706 (2009).

N. H. Putnam et al., "Sea Anemone Genome Reveals Ancestral Eumetazoan Gene Repertoire and Genomic Organization." Science 317, 86 (2007).

G. Richards et al., "Sponge Genes Provide New Insight into the Evolutionary Origin of the Neurogenic Circuit." Current Biology 18, 1156 (2008).

O. Sakarya et al., "A Post-Synaptic Scaffold at the Origin of the Animal Kingdom." PLoS ONE 2(6), e506 (2007).

K. Tessmar-Raible et al., "Conserved Sensory-Neurosecretory Cell Types in Annelid and Fish Forebrain: Insights into Hypothalamus Evolution." Cell 129, 1389 (2007).

Related Content


Navigate This Article