Specificity of synaptic regeneration in the spinal cord of the larval sea lamprey.

1. Pairs of central neurones in large larval sea lampreys were impaled with micro-electrodes and studied for synaptic connexions in both unoperated control animals and animals which had recovered from complete spinal transection. Two identified classes of neurones served as post-synaptic targets: giant interneurones (g.i.s) and lateral cells (l.c.s). Several identified neurone types were tested as potential sources of presynaptic input. 2. When synaptic potentials had short, fixed latencies they also persisted during activation of the presynaptic cell at 3.3-33.3 Hz and were not eliminated after addition of lamprey saline containing high (20 mM) Ca2+. These presumably represented monosynaptic connexions. Variable-latency responses were eliminated by faster rates of stimulation of the presynaptic cell and were mediated via polysynaptic pathways. 3. In control animals, g.i.s received monosynaptic input from more caudal g.i.s in fourteen of thirty-two tested cell pairs. These excitatory post-synaptic potentials (e.p.s.p.s) were composite electrochemical responses. The amplitudes of the earlier electrical component averaged 1.66 +/- 0.24 mV (mean +/- S.E. of mean) and the amplitude of the later chemical component averaged 0.79 +/- 0.05 mV. 4. In operated larvae, eight of forty-seven g.i.-g.i. pairs separated by the transection scar were connected by monosynaptic composite e.p.s.p.s. In these pairs the electrical component averaged 0.84 +/- 0.17 mV (P less than 0.05 vs. control) and the chemical component averaged 1.56 +/- 0.40 mV. The average conduction velocity between these cells was less than that in control g.i.-g.i. pairs (0.93 +/- 0.11 vs. 1.61 +/- 0.25 m/s; P less than 0.01). 5. The l.c.s showed monosynaptic e.p.s.p.s after activation of a subset of the bulbar Müller neurones (B2-4) in seven of twelve pairs. In behaviourally recovered larvae three of twenty-two similar pairs separated by the transection scar were also connected via monosynaptic e.p.s.p.s. The average conduction velocity between these experimental neurones was also less than that in control bulbar-l.c. pairs (1.03 +/- 0.03 vs. 1.58 +/- 0.09 m/s; P less than 0.001). 6. Several types of neurones were either infrequently linked or never connected to g.i.s or to l.c.s in control larvae. In animals which had recovered from a spinal transection, no synaptic connexions were found from such neurones on to g.i.s or l.c.s respectively, in 124 tested cell pairs. In addition, dorsal cells (intraspinal primary sensory neurones) received no synaptic input upon stimulation of the spinal cord before or after transection.(ABSTRACT TRUNCATED AT 400 WORDS)