Distinct roles of spinal commissural interneurons in transmission of contralateral sensory information

Abstract

Crossed reflexes are mediated by commissural pathways transmitting sensory information to the contralateral side of the body, but the underlying network is not fully understood. Commissural pathways coordinating the activities of spinal locomotor circuits during locomotion have been characterized in mice, but their relationship to crossed reflexes is unknown. We show the involvement of two genetically distinct groups of commissural interneurons (CINs) described in mice, V0 and V3 CINs, in the crossed reflex pathways. Our data suggest that the exclusively excitatory V3 CINs are directly involved in the excitatory crossed reflexes and show that they are essential for the inhibitory crossed reflexes. In contrast, the V0 CINs, a population that includes excitatory and inhibitory CINs, are not directly involved in excitatory or inhibitory crossed reflexes but downregulate the inhibitory crossed reflexes. Our data provide insights into the spinal circuitry underlying crossed reflexes in mice, describing the roles of V0 and V3 CINs in crossed reflexes.

Keywords: V0 interneurons; V3 interneurons; commissural interneuron; crossed reflex; interlimb coordination; locomotion; mouse genetics; sensory feedback; sensorymotor integration; spinal cord.

Figure 1.. Crossed reflex responses in wild type, V0^kill and V3^off mice.

(A) Schematic of the experimental setup. Nerve cuff electrodes to the left tibial or sural nerve along with one EMG recording electrode into the left Gs were implanted to determine the threshold current to activate local reflex. In addition, EMG recording electrodes to five muscles of the left leg were implanted to record crossed reflex responses. The scale bars in the recordings indicate 2 ms. Schematic was created with BioRender.com (agreement number: QQ250Q4M08). (B) EMG responses of all recorded muscles of the right leg to left tibial nerve stimulation at 1.2xT (i) and 5xT (ii) in representative one wild type (WT), one V0^kill and one V3^off mice. Shaded areas indicate nerve stimulation. (C) Same as in B, but for responses to left sural nerve stimulation. See also Figure S1, Figure S2, and Figure S7.

Figure 2.. Average EMG responses of muscles to contralateral nerve stimulation.

(A) Average traces of rectified and filtered EMG activities from right leg muscles as a response to left tibial nerve stimulation at 1.2xT (i) and 5xT (ii) in wild type, V0^kill and V3^off mice. The gray lines are averages from individual animals, whereas the black lines are averages across all animals. The shaded backgrounds indicate the time when the left nerves were stimulated. (B) Same as in a but for sural nerve stimulation. (C) Graphs showing the average area underneath the average rectified EMG traces from 12–50 msec delays (the delay was chosen to exclude influence from stimulation artifacts) from the stimulation onset for 1.2xT (top) and 5xT (bottom) stimulation of the tibial nerve. The error bars indicate the 95% confidence intervals. (D) Same as in c but for sural nerve stimulation. (E) Graphs illustrating the combined averages from all EMG traces for 1.2xT and 5xT stimulation of the tibial and the sural nerves combined. *: p<0.05, **: p<0.01, ***: p<0.001. See also Figure S7.

Figure 3.. The probability of muscle activation does not change in the presence or absence of V0 or the V3 CINs.

The graphs illustrated here represent data from 14 WT, 19 V0^kill, and 17 V3^off mice. On average 38.56 (standard deviation: 6.43) traces were used for each experiment. (A) Occurrence of observed responses across all recorded muscle in all three groups of mice when either nerve tibial or sural nerve was stimulated either at 1.2xT or 5xT, indicating higher rate of muscle activity was observed with higher stimulation strength. (B) Increasing the stimulation strength from 1.2xT to 5xT increases the occurrence of observation in muscles similarly in all three mouse groups. Data from both nerve stimulations are combined in this graph. See also Figure S7. (C) and (D) The effect of nerve stimulation at 1.2xT or 5xT is regardless of whether the tibial nerve (C) or the sural nerve (D) is stimulated. *: p<0.05, **: p<0.01, ***: p<0.001.

Figure 4.. Temporal differences in muscle activation pattern in wild type, V0^kill, and V3^off mice.

The graphs illustrated here represent data from 14 WT, 19 V0^kill, and 17 V3^off mice. On average 27.69 (standard deviation: 9.48) traces were used for each experiment. (A) In general, the latencies of muscle activation was larger in V0^kill mice than wild type and V3^off mice. (B) Increasing the stimulation strength from 1.2xT to 5xT caused an increase in delay across all animal groups. Data from both nerve stimulations are combined in this graph. (C) The stimulation strength dependent increase in latency shown in B was clearly present in wild type and V0^kill mice but not in V3^off mice, when tibial nerve was stimulated (i). The delays did not show any strength dependent change in any mouse groups when sural nerve was stimulated (ii). *: p<0.05, **: p<0.01, ***: p<0.001. See also Figure S7.

Figure 5.. V3 CINs but not V0 CINs are necessary for the inhibitory crossed reflex.

(A) Schematic of the experimental setup. The experimental setup is as presented in Figure 1A with the addition of one nerve stimulation electrode to the right sural nerve. The stimulation of the right sural nerve allowed us to elicit local reflex response (green trace) and stimulation of the left sural nerve the crossed reflex response (purple trace). The simultaneous stimulation of the left and right sural nerves with varying delays was used to detect inhibitory influence (black trace). Lower activity in black traces then green traces (blue arrows) indicated inhibitory crossed reflex action. Schematic was created with BioRender.com (agreement number: TZ250Q4IW4). The shaded backgrounds indicate the time of contralateral nerve stimulation and hatched background indicate 10 ms delay after stimulation for reference. (B) Inhibitory effect was detected in the majority of EMG recordings in wild type and V0^kill mice but not in V3^off mice. Arrows in A and B indicate the onset of the ipsilateral sural nerve stimulation. (C) To quantify the inhibitory crossed reflex, we averaged the green and black traces in the area indicated by the bold line (12 -18 msec delay from stimulation onset) and took the ratio of these averages. Numbers smaller than 1 indicate inhibition. (D) Graphs of ratios described in C for contralateral tibial (i) or sural (ii) nerve stimulation. The graph in iii is when the data in i and ii is pooled together. These graphs illustrate statistically, inhibition was preserved in V0^kill mice, but not in V3^off mice. dnr: reflex response to double nerve stimulation, lr: local reflex response. *: p<0.05, ***: p<0.001. See also Figure S3, Figure S4, and Figure S7.

Figure 6.. V3 CINs but not the V0 CINs are necessary for inhibitory crossed reflex during locomotion.

(A) Example EMG recordings from right St and TA muscles as a response to the left sural nerve stimulation during resting (top), and during locomotion when the muscle was either inactive (middle) or active (bottom) prior to the stimulation in a wild type mouse. (B) Bar diagrams illustrating the probability of occurrence of a silent period in all recorded muscles of the right leg within the 5 msec window immediately after left sural nerve stimulation when there was no muscle activity (top) or there was activity (bottom) prior to nerve stimulation. (C) Box diagrams showing the average (+/− standard deviation) on and offsets of right muscle activities as response to the left sural nerve stimulation at 5xT when there were no activity (i) or activity (ii) before stimulation (contralateral nerve stimulation indicated by the shaded area) during locomotion. Hatched area indicates 10 msec time window after nerve stimulation. Black bar: wild type, red bar: V0^kill, and blue bar: V3^off. (D) Diagrams illustrating the onset latencies of muscle activities in Black bar: wild type (black), V0^kill (red), and V3^off (blue) mice during locomotion. Circles are averages from individual mice, and the horizontal lines indicate group averages (+/− standard deviation). *: p<0.05. See also Figure S5, Figure S6, Figure S7, Video S1, Video S2 and Video S3.

Figure 7.. Commissural pathways involved in crossed reflexes.

Our findings suggest that the spinal commissural pathways for crossed reflexes directly involve V3 CINs through a direct excitatory and an indirect inhibitory pathway. The inhibitory pathway is mediated at least with on local inhibitory interneuron (IN) as all V3 CINs are excitatory. V0 CINs although not directly involved in transmitting sensory afferent signals are modulate the inhibitory crossed reflex responses. See also Figure S7.