The Development of a Simple Nervous System Tracing the pedigree of nerve cels in the embryonic growth ofdwarfand giant leeches yields preliminary cl ues to the functioning of the ad ult nervo us system by Gunther S. Stent and David A. Weisblat T he nervous system presents two of the most challenging questions of contemporary biology. How do networks of neurons, or nerve cells, gen erate animal behavior? And how do the neurons and their specifc connections arise during the development of an ani mal from a fertilized egg? The second question cannot be considered indepen dently of the frst, since the anatomy and the functioning of the adult nervous sys tem represent the end point of neural development. Detailed anatomical and functional knowledge is needed even to ask well-focused questions in devel opmental neurobiology; to gain such knowledge animals with a simple ner vous system are- particularly suitable. One such animal is the leech, the blood· sucking relative of the earthworm. For many people the mere mention of leeches evokes the revulsion expressed by Charlie Allnutt (played by Hum phrey Bogart) in The Afican Qeen: emerging from a swamp and fnding the parasites clinging to his skin, All nutt shouts, "I hate leeches!" Neverthe less, leeches have their good points; un known to Allnutt, while drawing his blood the leeches would have injected anticoagulant substances into his blood stream that might have reduced his risk of certain cardiovascular disorders. More to the point, the leech's simple body plan makes it an attractive animal to the experimental biologist. The tubu lar body of the leech is built of 32 sim ilar segments and ofers the possibility of understanding the entire animal by studying just one of its segments. Of the 32 segments the frontmost four make up the specialized structures of the head, including a pair of eyes on the dorsal, or upper, surface and a .front sucker on the ventral, or lower, surface. The rearmost seven segments make up the specialized structures of the tail, including the anus and a large rear sucker. The anatomy of the intervening 21 mid-body segments is highly stereotyped. Each segment has a 136 complete set of visceral organs, includ· ing circulatory vessels, kidneys and gut. The skin of each segment is subdivided into a fxed number of annuli, or rings; the middle annulus bears an array of sensory organs distributed around the circumference of the body tube. The body wall of each segment is girded by circular muscles that can constrict the body tube. Deeper in the wall lie longi tudinal muscles; their contraction short ens the body tube. T he nervous system of the leech re fects the segmental body plan. It consists of 32 ganglia, interconnected to form the ventral nerve cord. Each gan glion consists of some 200 bilateral ly symmetrical pairs of neurons and a few unpaired neurons. The ganglion is linked to the body wall and the internal organs by two bilateral pairs of segmen tal nerves and to the neighboring ganglia by bundles of connective nerves. Ex haustive studies of the nervous system of the leech, carried out mainly by John G. Nicholls and his students at Stanford University, have shown that the anato my of the segmental ganglia is sufcient ly stereotyped from segment to segment and sufciently invariant from specimen to specimen for a large fraction of the neurons to be reliably identifed [see "The Nervous System of the Leech," by John G. Nicholls and David Van Essen; SCIENTIFIC AMERICAN, January, 1974]. A neuron of the leech's segmental ganglia can be penetrated with micro electrodes to record the cell's electrical activity. Similarly, a stain or a fuores cent dye can be injected through a mi cropipette to reveal the anatomical de tails of an individual neuron. By means of these techniques it has been possi ble to establish the pattern of neuronal connections and thereby to account for some simple acts of refexive be havior, such as the shortening of the body tube in response to tactile stimu lation, and even for some moderately complex integrated movements, such as the heartbeat. The most complex behavior that has been described in terms of identifed neurons and their connections is swim· mingo Work done in our laboratory at the University of California at Berkeley from 1971 to 1977 by William B. Kris tan, Jr., Carol Ort, Otto Friesen, Marga ret Poon and Ronald Calabrese showed that the contractile rhythm of the lon gitudinal muscles responsible for the swimming movement of the leech is generated by a set of 12 bilateral pairs of rhythmically active motor neurons in each segmental ganglion. The rhythm of the motor neurons is imposed on them by four bilateral pairs of interneurons (intermediate nerve cells), which form the central swim oscillator. Since the functional elements of the nervous system of the leech are known in some detail, the system provides the kind of clearly defned conceptual end point needed for asking specifc ques tions about developmental processes. For instance, is the identifed swim cir cuitry organized from the outside in, so that the longitudinal muscles are frst connected to the motor neurons, the mo tor neurons are then connected to the interneurons and the interneurons are fnally joined to form the central oscil latory network? Or does development proceed in the reverse, inside-out order? Or are all three levels of connections established concurrently? Indeed, is it possible that the neural circuitry does not develop by establishing specifc con nections but rather by selecting from an initially overconnected neuronal net work those connections that are func tionally appropriate? S even years ago, in the hope of answer ing some of these questions, we em barked on a long-term study of the de velopment of the leech's nervous sys tem. Our frst task was fnding a suitable species that could be cultivated in the © 1981 SCIENTIFIC AMERICAN, INC