Intersegmental Interactions Give Rise to a Global Network

Abstract

Animal motor behaviors require the coordination of different body segments. Thus the activity of the networks that control each segment, which are distributed along the nerve cord, should be adequately matched in time. This temporal organization may depend on signals originated in the brain, the periphery or other segments. Here we evaluate the role of intersegmental interactions. Because of the relatively regular anatomy of leeches, the study of intersegmental coordination in these animals restricts the analysis to interactions among iterated units. We focused on crawling, a rhythmic locomotive behavior through which leeches move on solid ground. The motor pattern was studied ex vivo, in isolated ganglia and chains of three ganglia, and in vivo. Fictive crawling ex vivo (crawling) displayed rhythmic characteristics similar to those observed in vivo. Within the three-ganglion chains the motor output presented an anterior-posterior order, revealing the existence of a coordination mechanism that occurred in the absence of brain or peripheral signals. An experimental perturbation that reversibly abolished the motor pattern in isolated ganglia produced only a marginal effect on the motor activity recorded in three-ganglion chains. Therefore, the segmental central pattern generators present in each ganglion of the chain lost the autonomy observed in isolated ganglia, and constituted a global network that reduced the degrees of freedom of the system. However, the intersegmental phase lag in the three-ganglion chains was markedly longer than in vivo. This work suggests that intersegmental interactions operate as a backbone of correlated motor activity, but additional signals are required to enhance and speed coordination in the animal.

Keywords: central pattern generator; intersegmental coordination; motor control; recurrent inhibition; rhythmic motor pattern.

FIGURE 1

Crawling motor pattern. (A) The drawing depicts a leech crawling step that results from coordinated waves of elongation (i,ii) and contraction (iii,iv) phases. (B) Crawling is induced in the isolated nervous system by dopamine and can be monitored through intracellular recordings of the CV motoneuron and extracellular recordings of the DE-3 motoneuron in the DP nerve, whose activities correspond to the elongation and contraction phases of crawling, respectively. (C) Schematic network interaction underlying crawling. (D) Diagrams of different hypothetical pathways controlling intersegmental coordination. Each gray box represents a leech segment bearing a crawling central pattern generator (CPG), constituted by a half-center oscillator (C and E) responsible for the excitation of the motoneurons (MNs) active in each phase. Intersegmental coordination depends on: (i), a command neuron located in the cephalic ganglion that sequentially activates each segmental CPG; (ii), the interaction of local circuits (for simplicity we limited the connections to one direction, but they can operate in both directions); (iii), sensory feedback from the periphery.

FIGURE 2

Crawling in reduced experimental configurations. (Ai) Left, diagram depicting the recording configuration; right, representative extracellular recording of a DP nerve in an isolated ganglion (G0) during a dopamine-induced crawling episode. (Aii) As in (Ai) for a chain of three ganglia (G1-G2-G3) where a DP nerve was recorded in each one. (B) Box plots describing the cycle frequency, duty cycle and firing frequency in G0, G1, G2, and G3; n = 12 ganglia from 11 leeches for G0, 15 ganglia from 11 leeches for G1, 16 ganglia from 14 leeches for G2, 13 ganglia from 10 leeches for G3. Comparison of G0 vs. G2 was performed by Wilcoxon rank-sum test; ^# indicates p = 0.027. For comparison of G1, G2, and G3 we used Kruskal-Wallis test; p > 0.05 for the three variables. (C) Dot plot describing the coefficient of variation (CV) of the cycle frequency, duty cycle and firing frequency in G0 and G2 [n as in panel (B)]. Each dot presents the CV value of the 10 cycles analyzed in each preparation. ^# indicates p < 0.001 (Wilcoxon rank-sum test).

FIGURE 3

Correlated activity among ganglia during crawling. (A) Filtered versions of DP activity in a chain of ganglia (light, medium and dark green traces) superimposed on the original extracellular recordings (in gray) during a crawling episode. (B) Cross-correlation of the activity between the DP pairs shown in panel (A). (C) Box plots comparing the cross-correlation index in experimental (exp) and scrambled (scr) nerve pairs. ^## indicates p < 0.005 and ^###p < 0.001 (Wilcoxon rank-sum test); n = 8 chains from 7 leeches for exp and 8 recording combinations for scr. (D) Box plots showing the phase lag of each of the experimental nerve pairs. Friedman test p < 0.005, ^##p < 0.01 (Conover’s post-hoc test). n = 8 chains from 7 leeches.

FIGURE 4

Effect of NS neuron upon crawling. (Ai) Extracellular recording of a DP nerve in an isolated ganglion during a crawling episode, while a square –5 nA current pulse was injected in an NS neuron [I NS(0)]. Horizontal arrows indicate the intervals considered in the analysis. (Aii) As in (Ai) for a chain of ganglia where a DP was recorded in each ganglion and the pulse was applied in one NS of G2 [I NS(2)]. (B) Box plots describing the relative cycle frequency, duty cycle and spike frequency in G0, G1, G2, and G3. n = 11 pulses in 7 crawling episodes in 6 leeches for G0; n = 18 pulses in 9 crawling episodes in 8 leeches for G1; 27 pulses in 12 crawling episodes in 10 leeches for G2; and 24 pulses in 10 crawling episodes in 7 leeches for G3. One sample Wilcoxon signed-rank test was applied to evaluate whether the values were different than 1. *** indicates p < 0.001 and * indicates p < 0.05. Comparison of G0 vs. G2 was performed by Wilcoxon rank-sum test; ^### indicates p < 0.00001. For comparison of G1, G2, and G3 we used Kruskal-Wallis test; p > 0.05 for cycle frequency and duty cycle and p < 0.00001 for cycle frequency; & indicates p < 0.00001 (Dunn’s post-hoc test).

FIGURE 5

Effect of NS neuron upon the rhythmic activity of T cells during crawling. (A) Extracellular recordings of a DP nerve, and intracellular recordings of NS and T cells in an isolated ganglion, while a square –5-nA current pulse was injected in the NS neuron in the course of a crawling episode. The lowest trace is a filtered version of the T cell recording (Tf). The triangles indicate the timing of the IPSPs. (B) Frequency of the IPSPs measured before, during and after (pre, pulse, post, respectively) the pulse was applied. p < 0.01 for comparison of pre, pulse and test (Friedman test), ^# indicates p < 0.05, and ^##p < 0.01 (Conover post-hoc test). n = 7 pulses in 5 crawling episodes in 4 leeches.

FIGURE 6

Effect of NS upon DP basal activity. (Ai) Extracellular recordings of spontaneous DP activity performed in an isolated ganglion while square hyperpolarizing current pulses of increasing amplitude were injected in the NS neuron [I NS(0)]. For simplicity we only show 0, –4, and –8 nA current pulses (upper, middle and lower trace, respectively). Horizontal arrows indicate the intervals considered in the analysis. (Aii) As in (Ai) for a chain of ganglia; the current was injected in one NS neuron of the middle ganglia [I NS(2)]. (B) Box plots comparing DP basal spike frequency in isolated and chain ganglia. ^## indicates p < 0.002 for G0 vs. G2 (Wilcoxon rank-sum test). p > 0.05 (Kruskal-Wallis test) for G1, G2, and G3. n = 20 ganglia from 12 leeches for G0, 26 ganglia from 16 to 17 and 16 leeches for G1, G2, and G3, respectively. (C) Scatter plot showing the pulse ratio in isolated and middle ganglia for hyperpolarizing pulses of increasing amplitude. All points are significantly different than 1 (p < 0.001, One sample Wilcoxon signed-rank test against 1); ^# indicates p = 0.02 and ^## indicates p = 0.002 (Wilcoxon rank-sum test). n = 17–19 ganglia from 11 leeches for G0 and 24–25 ganglia from 15 to 16 leeches for G2. (D) Box plots of pulse ratio in chains of ganglia for –8 nA hyperpolarizing pulses applied in G2. ** indicates p < 0.01, ***p < 0.0001 (One sample Wilcoxon signed-rank test against 1). p < 0.00001 for comparison of G1, G2, and G3 (Kruskal-Wallis test), ^###p < 0.001 (Dunn’s post-hoc test). n = 22 ganglia from 13 leeches for G1, 25 ganglia from 16 leeches for G2 and 24 ganglia from 14 leeches for G3.

FIGURE 7

Correlated activity among segments during crawling. (A) Snapshot from a video during the elongation phase of crawling. Thirteen white dots were painted over the dorsal midline determining 14 fragments (f1–f14) along the longitudinal axis, including the head (H) and tail (T) edges as markers. (B) Length change of fragments between head edge (H) to successive dots, including the head to tail (H-T). For clarity we show every other fragment. (C) Cycle frequency measured in isolated ganglia (G), in three-ganglion chains and in intact animals (behav); n = 12 ganglia from 11 leeches (G), 16 chains from 14 leeches (chain) and 4 intact animals (behav). ^## indicates p < 0.005 and ^### indicates p < 0.0005 (Wilcoxon rank-sum test). (D) Relative position of midbody ganglia 1–21; each symbol represents one leech (n = 3). The line presents a polynomial fit (R = 0.99). (E) The gray circles show the ganglion number as a function of the mean relative position along the leech longitudinal axis; and the black triangles show the relative position of each of the points drawn on the leech shown in panel (A). (F) Length changes of fragments f5–f8. (G) Cross-correlation of pairs of traces shown in panel (F), identified on the right. (H) Phase lag measured in isolated three-ganglion chains and in the animal. n = 8 chains from 7 leeches for chains, and 4 intact animals for behav. ^### indicates p < 0.001 (Wilcoxon rank-sum test).