Body coordination during locomotion

(A) Schematic illustrating three different gaits in mice. The colored rectangle represents the time each limb is in contact with the ground. Each gait is characterized by different coupling of left forelimb (LF), right forelimb (RF), left hindlimb (LH) and right hindlimb (RH).

(B) Schematic illustrating the anatomical connections across half-center modules. Each module is composed of a pool of excitatory and inhibitory neurons controlling extensor (red) and flexor (blue) muscles. Commissural pathways (CIN) at each girdle (green arrows) ensure left and right coordination, ipsilateral (violet) ascending (iLAPN) and descending (iLDPN) propriospinal pathways contribute to the unilateral coordination of forelimb and hindlimb, whereas commissural (orange) ascending (cLAPN) and descending (cLDPN) propriospinal tract connect opposite forelimb and hindlimb.

(C) Schematic illustrating the main classes of premotor neurons forming the central pattern generator (CPG) circuits: the inhibitory ipsilateral neurons (V1 and V2b), the excitatory ipsilateral neurons (V2a), the inhibitory commissural neurons (V0D), the excitatory commissural neurons (V0V) and the dual-projecting excitatory neurons (V3). V1 and V2b neurons preferentially inhibit flexor and extensor motoneurons (MNs), respectively.

Identification of a novel marker of the spinal V2 neuron lineage

(A) Cross-sections of the developing lumbar spinal cord (E9.5 – E13.5) with labeled Hes2 mRNA (red) and counterstained with DAPI (blue). Bottom panels are magnifications of the dotted boxes.

(B) Cross-sections of the developing lumbar spinal cord (E9.5 – E13.5) stained for Hes2 mRNA (red), Hb9 protein (green), and Chx10 protein (blue). Bottom panels are magnifications of the dotted boxes.

Scale bars: 100 μm.

Establishing genetic access to the Hes2 neurons

(A) Whole-mount β-galactosidase staining on E10.5 and E11.5 Hes2iCre;R26LSL-LacZ embryos to characterize the expression of Hes2iCre.

(B) Cross-sections of E12.5 lumbar spinal cord from Hes2iCre;Gata3nlsLacZ;R26LSL-Sun1-GFP embryos stained with antibodies for GFP to label Hes2 neurons (green), β-gal (Gata3nlsLacZ) to label V2b neurons (red), and Chx10 to label V2a neurons (blue). Scale bar: 100 μm.

(C) Bar graph showing the percentages of V2a (Chx10+) and V2b (Gata3nlsLacZ +) neurons within the Hes2 lineage (Hes2+).

(D,E) Bar graphs showing the percentage of V2a (Chx10+) (D) and V2b (Gata3nlsLacZ+) (E) neurons captured by the Hes2iCre.

(F) Cross-sections of E12.5 lumbar spinal cord from Hes2iCre;Gata3nlsLacZ;R26LSL-Sun1-GFP embryos either heterozygous (left) or homozygous (right) for the iCre allele. The sections were stained with antibodies for GFP (green), β-gal (Gata3nlsLacZ) (red) and Chx10 (blue). Scale bars: 100 μm.

(G) Bar graph showing the absence of changes in the number of Hes2 neurons in Hes2 knockout embryos compared to heterozygous littermates, assessed by two-tailed Student’s t-test.

Festinating gait induced by the developmental silencing of the Hes2 neurons

(A) Bottom view of self-paced walking of mice on a runway. Paws are tracked in the indicated colors in control (upper panel) and Hes2 neuron-silenced (bottom panel) mice.

(B) Schematic displaying the calculation of hindlimb and forelimb stride and the color-coding of individual limbs.

(C,E) Bar graphs showing the significant shortening of the stride length of hindlimb (C) and forelimb (E) in Hes2 neuron-silenced mice compared to controls.

(D,F) Bar graphs showing the significant increased stepping cadence of hindlimb (D) and forelimb

(F) in Hes2 neuron-silenced mice compared to controls.

(G) Representative frames showing the alternation of stance and swing phases of the hindlimb in control (upper panels) and Hes2 neuron-silenced (bottom panels) mice during self-paced walking on a wide runaway. Dark and light green arrows point to the hindlimb in stance and swing position, respectively.

(H) Representative schematic showing the reduced duration of the step cycle, stance and swing for the right hindlimb (HR) in Hes2 neuron-silenced mice compared to controls. Scale bar is 70 msec. (I-K) Bar graphs showing the significant shortening of the step cycle (I), swing (J) and stance (K) duration of the hindlimb in Hes2 neuron-silenced mice compared to controls.

(L) Bar graph showing the preserved ratio of stance and swing duration for each step cycle following the developmental silencing of the Hes2 neurons.

Data are presented as mean ± SEM. Each mouse analyzed is represented with a gray filled circle. Statistical analysis was done using two-tailed Student’s t-test.

Disruption of ipsilateral body coordination by developmental silencing of the Hes2 neurons

(A) Representative frames showing the limb positioning in control (upper panel) and Hes2 neuron-silenced (bottom panel) mice during self-paced walking on a wide runaway. Dark and light green arrows point to the hindlimb in stance and swing position, respectively.

(B) Representative schematics showing the temporal dynamics of interlimb coordination in control (upper panel) and Hes2 neuron-silenced (bottom panel) mice during self-paced walking. Note the extended time spent with four-limb support (grey boxes) in Hes2 neuron-silenced mice compared to controls. HR=hindlimb right, FR=forelimb right, HL=hindlimb left, FL=forelimb left. Scale bar is 100 msec.

(C) Bar graph showing the significant increase in the time spent in four-paw support in Hes2 neuron-silenced mice compared to controls.

(D,E) Bar graphs showing the significant increase in the latency of initiation of swing between diagonal (D) and ipsilateral (E) limbs in Hes2 neuron-silenced mice compared to controls.

(F,G) Representative frames showing the hindpaw trajectory during swing in control (F) and Hes2 neuron-silenced (G) mice during self-paced walking on a wide runaway.

(H,I) Line graphs showing the knee, ankle, and paw kinematics during a normalized step-cycle of control (F) and Hes2 neuron-silenced (G) mice during self-paced walking on a wide runaway. SEM is indicated as shaded lines, and calculated from multiple mice (Control N=4, Silenced N=5). Statistical analysis was done using two-way ANOVA, followed by Tukey’s post hoc test.

(J,K) Bar graphs showing the increased number of slips in Hes2 neuron-silenced mice compared to controls as mice cross an uneven ladder (J) or a narrow beam (K).

Data are presented as mean ± SEM. Each mouse analyzed is represented with a gray filled circle. Statistical analysis, unless otherwise indicated, was done using two-tailed Student’s t-test.

Ablation of spinal V2 neurons in adult induces festination and impairs ipsilateral body coordination

(A) Schematic illustrating the intersectional genetic approach to ablate the spinal V2 neurons in adult mice.

(B) Bottom view of mice during self-paced walking on a runway. Paws are tracked in the indicated colors in control (upper panel) and spinal V2 neuron-ablated (bottom panel) mice. Stride and ipsilateral distance are calculated as indicated.

(C,E) Bar graphs showing the significant shortening of the stride length of hindlimb (C) and forelimb (E) in spinal V2 neuron-ablated mice compared to controls.

(D,F) Bar graphs showing the significant increased cadence of hindlimb (D) and forelimb (F) stepping in spinal V2 neuron-ablated mice compared to controls.

(G) Bar graph showing the increased ipsilateral distance in spinal V2 neuron-ablated mice compared to controls.

(H) Representative frames showing limb positioning in control (upper panel) and spinal V2 neuron-ablated (bottom panel) mice during self-paced walking on a wide runaway. Dark and light green arrows point to the hindlimb in stance and swing position, respectively. Representative schematic showing the temporal dynamics of interlimb coordination during self-paced walking. Note the extended time spent with four-limb support (grey boxes) in V2 neuron-ablated mice compared to controls. HR=hindlimb right, FR=forelimb right, HL=hindlimb left, FL=forelimb left. Scale bar is 100 msec.

(I) Bar graph showing the significant shortening of the step cycle duration of the hindlimb in spinal V2 neuron-ablated mice compared to controls.

(J) Bar graph showing the significant increase in the time spent on four-paw support in spinal V2 neuron-ablated mice compared to controls.

(K,L) Bar graphs showing the significant changes in the latency of initiation of swing between ipsilateral (K) and diagonal (L) limbs in spinal V2 neuron-ablated mice compared to controls.

(M) Line graphs showing the ankle and paw kinematics during a normalized step-cycle of control and spinal V2 neuron-ablated mice during self-paced walking on a wide runaway. SEM is indicated as shaded lines, and calculated from multiple mice (Control N=4, Silenced N=3). Statistical analysis was done using two-way ANOVA, followed by Tukey’s post hoc test.

(N-P) Bar graphs showing the increased number of slips in V2 neuron-ablated mice compared to controls as mice cross an uneven ladder (N), a narrow beam (O) and a circular narrow beam (P). Data are presented as mean ± SEM. Each mouse analyzed is represented with a gray filled circle. Statistical analysis was done, unless otherwise indicated, using two-tailed Student’s t-test.