Structure–function relation of the developing calyx of Held synapse in vivo

KEY POINTS: During development the giant, auditory calyx of Held forms a one‐to‐one connection with a principal neuron of the medial nucleus of the trapezoid body. While anatomical studies described that most of the target cells are temporarily contacted by multiple calyces, multi‐calyceal innervation was only sporadically observed in in vivo recordings, suggesting a structure–function discrepancy. We correlated synaptic strength of inputs, identified in in vivo recordings, with post hoc labelling of the recorded neuron and synaptic terminals containing vesicular glutamate transporters (VGluT). During development only one input increased to the level of the calyx of Held synapse, and its strength correlated with the large VGluT cluster contacting the postsynaptic soma. As neither competing strong inputs nor multiple large VGluT clusters on a single cell were observed, our findings did not indicate a structure–function discrepancy. ABSTRACT: In adult rodents, a principal neuron in the medial nucleus of the trapezoid (MNTB) is generally contacted by a single, giant axosomatic terminal called the calyx of Held. How this one‐on‐one relation is established is still unknown, but anatomical evidence suggests that during development principal neurons are innervated by multiple calyces, which may indicate calyceal competition. However, in vivo electrophysiological recordings from principal neurons indicated that only a single strong synaptic connection forms per cell. To test whether a mismatch exists between synaptic strength and terminal size, we compared the strength of synaptic inputs with the morphology of the synaptic terminals. In vivo whole‐cell recordings of the MNTB neurons from newborn Wistar rats of either sex were made while stimulating their afferent axons, allowing us to identify multiple inputs. The strength of the strongest input increased to calyceal levels in a few days across cells, while the strength of the second strongest input was stable. The recorded cells were subsequently immunolabelled for vesicular glutamate transporters (VGluT) to reveal axosomatic terminals with structured‐illumination microscopy. Synaptic strength of the strongest input was correlated with the contact area of the largest VGluT cluster at the soma (r = 0.8), and no indication of a mismatch between structure and strength was observed. Together, our data agree with a developmental scheme in which one input strengthens and becomes the calyx of Held, but not with multi‐calyceal competition.