LOCAL SEROTONERGIC MODULATION OF CALCIUM-DEPENDENT POTASSIUM CHANNELS CONTROLS INTERSEGMENTAL COORDINATION IN THE LAMPREY SPINAL-CORD

1. The intersegmental coordination during undulatory locomotion in lamprey is characterized by a constant phase lag between consecutive segments, that is, the ratio between the intersegmental time lag and the cycle duration remains constant. It is shown that the spinal 5-HT (serotonin) system can, in a graded fashion, control the phase lag value from a rostrocaudal to a caudorostral lag corresponding to a reversed direction of swimming. These effects can be explained by a 5-HT-induced depression of Ca2+-dependent K+ channels (K(Ca) channels) in network neurons. 2. The actions of the spinal 5-HT system were analyzed in the lamprey spinal cord preparation in vitro. Fictive swimming was induced by bath application of N-methyl-D-aspartate (NMDA). The intersegmental phase lag between ventral root burst activities was measured along the ipsilateral side of the spinal cord. The chamber with the preparation was partitioned into two pools so that the rostral and caudal halves of the preparation could be perfused independently with solutions containing the same level of NMDA (100-150-mu-M) with or without additional 5-HT or a 5-HT uptake blocker (citalopram). 3. Addition of 5-HT to one of these partitioned pools changed the intersegmental phase lag in this pool, whereas the cycle duration remained unchanged. It was determined by the activity in the "non-5-HT" pool. Addition of 5-HT to the caudal pool resulted in an increased rostrocaudal phase lag. When 5-HT was added to the rostral pool, on the other hand, the phase lag shifted direction to a backward coordination. The higher the 5-HT concentration (in a range from 0.1 to 1.0-mu-M), the larger the phase lag shift, up to approximately 2% of the cycle duration per segment. Similar phase lag changes were induced when a 5-HT uptake blocker (citalopram) was added to the rostral pool of the partitioned chamber. Endogenously released 5-HT may thus induce a similar intersegmental phase lag change. 4. In this partitioned pool experiment, the spinal cord section in the pool containing NMDA will entrain that in the 5-HT-containing pool. If the entire spinal cord had been exposed to 5-HT and NMDA, the cycle duration would have been increased markedly. This experimental setup allowed us to study the effects on intersegmental coordination without any concomitant effects on cycle duration. The 5-HT cells along the spinal cord can most likely be activated locally in a number of segments; if so, they will cause an increased phase lag in this area in relation to a rostrally or caudally located leading segment. 5. The cellular effects of the same level of 5-HT that influences the intersegmental phase lag were explored by intracellular recording. Fictive swimming was induced by NMDA (100-150-mu-M) only in the rostral pool of the partitioned chamber while intracellular potentials were recorded from motoneurons in the caudal pool perfused with a physiological solution containing 5-mu-M strychnine. The axons of excitatory interneurons extending over several segments will provide a synaptic excitatory drive to motoneurons and interneurons in this pool near the partition, whereas the inhibitory propriospinal drive was blocked by strychnine. Addition of 5-HT to the caudal pool at a concentration sufficient for a phase lag reversal (1.0-mu-M) did not influence the amplitude of the rhythmic excitatory postsynaptic potentials that originated from propriospinal excitatory interneurons in the rostral segments. On the other hand, 5-HT at the same concentration significantly reduced the amplitude of the late postspike afterhyperpolarization (AHP). The AHP is due to activation of K(Ca) channels. As previously shown, the summation of AHPs is one important burst-terminating factor in the segmental network; when the AHPs are reduced, the burst duration will increase. The spinal network will then inevitably reduce the intrinsic burst frequency. The serotonergic suppression of K(Ca) channels and AHPs may also account for the effects exerted on the intersegmental coordination.