Figure 1. Developmental stages of X. laevis and anatomical organization of the spinal
motor columns. A. Relative size differences between stage 50 and 57 larvae (A1) and
associated morphological changes of the hindlimb bud during metamorphosis onset (A2),
from Nieuwkoop and Faber, 1956. B, C. Segmental (B1, C) and rostrocaudal (B2) organization
of the spinal motor column region containing limb MNs at stages 50-52 (B1, B2) and 55 (C).
Inset in B2 shows the location of labeled MNs in the mid-region of the spinal cord. D.
Schematic representation of the segmental organization of the appendicular and axial motor
columns in the larval spinal cord. Ap./Ax. MNs, appendicular/axial motoneurons; Ap./Ax. d.,
Ap./Ax. dendrites; Vr, ventral root; c.c, central canal; SC, spinal cord; D, dorsal; L, lateral.
Scale bars: B1 and C = 20 μm; B2 = 100 μm.
Figure 2. Developmental emergence of the molecular ACh phenotype in appendicular
MNs. A, B. Examples of fluorescence immunolabeling against ChAT (magenta) and VAChT
(red) in appendicular MNs (Ap. MN) labeled with retrograde Alexa Fluor dextran 647 (AD
647, cyan) in stages 53 (A) and 57 (B) tadpoles. Insets (b) in B show magnification of stage
57 appendicular MNs. C. Variation of fluorescence (δF/F) of axial (left plot) and
appendicular (right plot) MNs for ChAT and VAChT immuno-signals at stages 49-51 (n=4),
stages 52-54 (n=7) and stage 55-57 (n=8). Grey dots represent the averaged δF/F values for
ChAT (squares) and VAChT (circles) in each preparation, and the black dots are δF/F grand
means ± SEM for all preparations in a given developmental group. ns non significant, ***
p<0.001, Kruskall-Wallis test. D. Examples of in situ hybridization (ISH) labeling for ChAT
mRNA in the appendicular spinal column at stages 51 and 55. E. Example of fluorescence
immunolabeling against ChAT in stage 62 appendicular MNs that were previously labeled
with AD 647 injected into the hindlimb at stage 51 (see schematic at left ). All scale bars =
50μm; D, dorsal; L, lateral.
Figure 3. Motoneuronal identification of non-cholinergic limb projecting neurons.
A. Example of fluorescence immunolabeling against ChAT (magenta) and Islet1/2 (orange) in
appendicular and axial MNs (Ap. MN, Ax. MN) in a stage 52 tadpole. Appendicular MNs
were previously labeled with retrograde Alexa Fluor dextran 647 (AD 647, cyan). Scale bar =
50μm; D, dorsal; M, medial. B. Protocol for intracellular recordings from appendicular MNs.
Stage 52 MNs were identified by prior rhodamine dextran amine (RDA) retrograde labeling
Figure 3 supplement 1. Presence of spinal Islet1 and ChAT mRNAs at different
developmental stages. In situ hybridization of the two mRNAs was performed with the same
protocol (see Methods). The Islet1 probe was synthesized according to Shi et al. (2009) and
involved amplifying a PCR fragment of 798 pb by the primers 5’-
AAGTGCAACATCGGCTTCAG-3’ and 5’-GCTGTTTGGGGTATCTGGGA-3’. Despite the
occasional appearance of a widespread background signal (due to differences in colorimetric
reaction times), Islet1 mRNAs in darkened clusters were clearly evident in both the axial and
appendicular motor columns at all developmental stages examined (stage 51 (upper left panel)
and 55 (lower left panel) are illustrated), whereas ChAT mRNAs were found in limb MNs
only from stage 55 onwards (c.f., upper and lower right panels).
Figure 4. Switch from non-cholinergic to cholinergic limb muscle activation during axial
swimming at different developmental stages. A. Rhythmic burst discharge (right panel)
recorded from an axial ventral root (VRax) together with a hindlimb bud muscle (HLemg)
and axial myotome (Axemg) during a spontaneous fictive swim episode in a semi-isolated
stage 53 preparation (left panel). B-C. Examples of recordings before (control), during and
after (wash) bath application of d-tubocurarine to a stage 53 (B) and a stage 56 (C)
preparation. Lower plots in B and C show instantaneous l-HLemg (blue) and r-Axemg (green)
discharge rates (spikes/s) averaged ( SEM) over 10-20 l-VRax locomotor cycles (black).
Black arrow in the middle graph of B indicates persistent limb bud EMG activity occurring in
phase with the ipsilateral ventral root during d-tubocurarine application, and which was no
longer present at the later developmental stage (c.f., middle plot in C).
Figure 4 supplement 1. Appendicular MNs are activated centrally by cholinergic inputs.
A. Immuno-labeling of putative cholinergic synapses on limb MNs retrogradely labeled with
Alexa Fluor dextran 647 (AD 647), using antibodies against VAChT and the synaptic marker
synapsin [mouse anti-synapsin (1:200; Synaptic Systems, Germany) revealed by secondary
donkey anti-mouse antibodies coupled to Alexa Fluor 488 (1:500; Thermo Fischer)]. Both
VAChT and synapsin fluorescent signals were found in close apposition surrounding MN cell
bodies, indicating the presence of cholinergic input synapses. Inset: enlargement of synapsin
and VAChT signals in the vicinity of an appendicular MN cell body. Scale bar = 20μm B.
Calcium imaging optical recordings (20 fps) from stage 52 limb MNs retrogradely filled with
Calcium Green Dextran Amine 3kD (CGDA, Invitrogen; see schematic at left) during fictive
swimming (n=3). Right panel: individual calcium signals (δF/F) recorded simultaneously
from 10 MN somata (encircled in x40 image at left) during a swim episode recorded from an
axial VR (lower trace). C. Left panel: average fluorescence variation (black trace; SEM,
shaded area; top) for the 10 MNs illustrated in B during fictive swimming (bottom) in control
(black) and in the presence of 50μM d-tubocurarine (d-tubo.; red). Right panel: mean calcium
transient amplitudes in control (black) and under d-tubocurarine (red) expressed as a
percentage of the maximal control amplitude (i.e.,δF/F peak soon after swim episode onset)
for three stage 52 preparations. Unfilled circles represent single acquisitions; filled circles
denote grand means SEM. Long term calcium recordings were performed in control
experiments without drug application to avoid any fluorescence bleaching effect.
Figure 5. Switch from glutamate to acetylcholine receptors in hindlimb muscles. A.
Examples of hindlimb innervation patterns and distribution of ACh nicotinic receptors in a
whole-mount hindlimb bud, revealed by fluorescence immunolabeling of neurofilament
associated protein (Neurofil., red), glutamate receptor (NR2b, blue) and α-bungarotoxin
labeling (α-bungaroTx, yellow) respectively, at stages 52 (upper panels) and 57 (lower
panels). Inset drawings at bottom left of each panel show bud morphology at the two
representative larval stages. Scale bars: 200μm for stage 52, 100μm for stage 57. B. Examples
of fluorescence immunolabeling against neurofilament associated protein (red), synaptophysin
(Synaptoph., green) and NMDA glutamate receptor subunit 1 (NR1, blue) in whole-mount
limb bud at stage 52. White arrowheads indicate sites of apposition of all 3 markers. Scale
bar: 20μm. C. Examples of fluorescence immunolabeling against neurofilament associated
transporter 1 (VGluT1, yellow) and the glutamate
receptor subunit NR2b (blue) in whole-mount limb at stage 52. Scale bar: 50μm.
Figure 5 supplement 1. VGluT1 expression appendicular MNs. A. Fluorescence
immunolabeling against ChAT (red) and VGluT1 (yellow) in limb MNs retrogradely labeled
with Alexa Fluor dextran 647 (AD 647, cyan) at stage 52. Whereas ChAT is absent from
retrogradely labeled MNs, VGluT1 is expressed throughout almost the entire ventral horn,
including in limb MN somata. Bottom images are magnifications of the area (a) delineated by
the dotted line square in the merged image at top right. Punctiform VGluT1 expression
occurred around MN somata (indicating labeling of glutamatergic presynaptic terminals) and
a more diffuse VGluT1 signal also appeared in MN cytoplasm surrounding cell nuclei. B.
Fluorescence immunolabeling against synaptophysin (Synaptoph., green), NMDA glutamate
receptor subunit 1 (NR1, blue), VGluT1 (red) and neurofilament (Neurofil., grey) in hindlimb
muscle tissue at stage 52 (B1) and stage 55 (B2). VGluT1 expression colocalized with
synaptophysin and postsynaptic glutamate receptors were detected at stage 52, but not at stage
54. The bottom images in B1 are magnifications of the area (a) indicated in the merged image
at top right. Scale bars = 20μm.
Figure 6. Functional switch in hindlimb neuromuscular transmission. A. EMG recordings
(HLemg) from an isolated hindlimb preparation (schematic at left) in response to single pulse
(100μs) electrical stimulation (Stim) of a limb motor nerve (HL nerve) in stages 53 and 57
larvae. In each case, the black arrow indicates the beginning of the EMG response and the
black trace represents the mean profile of 6 superimposed responses. Note that the trace
illustration for stage 53 was taken from a 1 mM Mg2+ saline experiment. B. Integrated motor
nerve-evoked HLemg responses in control (black), and under d-tubocurarine (d-tubo.; red) or
CNQX+AP5 (purple) bath application to stage 53 and 57 preparations. Thick lines represent
the mean response profile (SEM); the area under each curve (grey) was used to measure the
response size. C. Mean (SEM) EMG response area as percentage of control response during
d-tubocurarine (black) or CNQX/AP5 (grey) bath application at stages 52-54 and 55-57,
respectively. ** p<0.01 and *** p<0.001, Mann–Whitney U-test.
Figure 7. Schematic representation of neurotransmitter phenotype switching associated
with the establishment of functional limb muscle innervation at different stages of
Xenopus metamorphic development. See text for further explanation. AChRs: nicotinic
ACh receptors; GluRs: glutamate 1035 receptors; Ap. MN: appendicular MNs; Ax. MN: axial
MNs; c.c: central canal