Lesions in the Xenopus laevis tadpole hindbrain reveal neural substrates for simple motor decision-making.

Animal survival depends heavily on making the ‘right’ motor decisions. For the hatchling Xenopus laevis tadpole this is the choice between to swim or not swim (1). Tadpoles detect external stimuli through receptors in the skin. Sensory information is then transferred via sensory pathway neurons to the brain where it is processed until the motor decision is made. The tadpole’s sensory pathway as well as the key Central Pattern Generator neurons, which drive swimming, are well-known (2). We have identified long and variable delays (~50-100ms) between firing in the sensory pathway neurons and the first spike in reticulospinal descending interneurons (dINs, 3). The latter signals the start of swimming. Such long and variable delays are key features of decision-making processes within the CNS (e.g. delays to eye movement; 4). However, the underlying neuronal circuitry is not well understood. In the tadpole, we proposed an additional type of neuron (hindbrain extension neurons; hexNs), presynaptic to dINs, which should spike at short latencies and show multiple firing (3,5), thus introducing variability and extending the delay to swim. Here, we aim to determine the location(s) of hexNs and spatial distribution of their sensory signal that allows the tadpole to make motor decisions in a timely manner. In the behaving tadpole (stage 37/38), we monitored the latency to the start of swimming in response to trunk skin touch. Latency data were compared between control tadpoles and those with hindbrain lesions. Removing the rostral part of the brain and/or separating the left and right side of the hindbrain led to an overall significant increase in latency to swim in lesioned animals (p=0.006, Kruskall-Wallis test). Median latency to swim increased from 109 ms in control animals (n=29) to (i) 201 ms in tadpoles lesioned along the midbrain-hindbrain border (n=32), (ii) 163 ms in animals with transverse hindbrain lesion behind the otic capsule (n=32), (iii) 150 ms in animals with lesions along the hindbrain’s midline (n=30). In addition, the above hindbrain lesions led to a change in the tadpole’s side preference for swim initiation. Control animals (n=33) showed 70% preference for the contralateral (contra) vs ipsilateral (ipsi) side (23 contra vs 10 ipsi responses). Tadpoles with rostral hindbrain transverse lesions (n=30) showed enhanced preference (87%) for the contralateral side (26 contra vs 4 ipsi responses). Similar increased preference for the contralateral side (85%) was observed in animals with midline hindbrain lesions (n=30; 24 contra vs 6 ipsi responses). Our novel data indicate that the hexN population must be highly distributed across the hindbrain and essential to making motor decisions in a timely manner. Furthermore, hindbrain lesions tested ‘forced’ the animal to rely on a hard-wired mechanism, which favors the contralateral swim response.