AC‐1 and synaptic development

Abstract

Across many developing brain structures, axons initially grow toward, and synapse exuberantly with, target neurons before the later removal of excess inputs. In many cases, this input pruning leads to a remarkable specificity, where a single target is innervated by only one axon. This phenomenon has been observed at the neuromuscular junction, the climbing fibre to Purkinje cell synapse in the cerebellum, and also at relay synapses in the thalamus. The consequence of this one-to-one specificity is quite profound. For example, in both visual and somatosensory thalamus, the summed synaptic strength of a single input can be so large that its activation can be sufficient to drive firing of the postsynaptic cell (Sincich et al. 2007; Wang & Zhang, 2008). Consequentially, this developmental reorganization has the capacity to maintain ‘labelled line’ information from the periphery, since a spike from an individual trigeminal neuron will reliably drive a spike in a synaptically coupled thalamic neuron. How does input competition and reorganization occur during development? A study by Wang et al. (2011) in a recent issue of The Journal of Physiology investigated the mechanisms involved in this process, examining the effect of a Ca2+-activated adenylate cyclase that has been implicated in synaptic plasticity in the developing hippocampus and neocortex (Lu et al. 2003; Yasuda et al. 2003). In young postnatal animals, the majority of thalamic neurons are innervated by multiple inputs (four or more per cell), a number that is winnowed down to one input during the second postnatal week. Concurrent with this process, the ratio of AMPA receptor (AMPAR) mediated to NMDA receptor (NMDAR) currents (A:N ratio) increases over time. Typically, A:N ratios increase because AMPAR currents increase, not because NMDAR currents, become smaller, although this was not directly addressed in the current study. Thus, these data suggest that during normal development, the winning input fibre is associated with stronger AMPAR-mediated transmission (Wang & Zhang, 2008). Because a deficiency in adenylate cyclase type-1 (AC-1) has been implicated in a reduced capacity for both developmental patterning of thalamic inputs into layer 4 of somatosensory cortex and an acute deficit in long-term potentiation at these inputs (Lu et al. 2003), Wang et al. investigated how genetic ablation of this molecule would influence developmental pruning at subcortical synapses, focusing again on trigeminal inputs to thalamic neurons in the ventrobasal nucleus. Interestingly, they found no differences between wild-type and knock-out animals in this property – pruning to a single input per thalamic relay neuron was intact in mutant animals. These data suggest that the developmental mechanisms regulating plasticity may be different in thalamic and cortical neurons. As a further test of the requirement for AC-1 for synaptic change, Wang et al. used sensory deprivation to investigate the effects on developmental maturation of the synapse. By plucking the facial vibrissae whose movement drives activity of trigeminal inputs into thalamic neurons, the investigators could retard synaptic maturation at these inputs in wild-type animals. After whisker deprivation, the one-to-one wiring that was normally achieved during the third postnatal week was impaired, and >40% of thalamic neurons received input from two or more trigeminal neurons. Consistent with this, A:N ratios were lower after sensory deprivation in wild-type mice, suggesting that the normal developmental increase in AMPAR-mediated transmission was impaired. In contrast, synapses from AC-1 knock-out animals were not responsive to sensory deprivation, showing instead the normal pruning mechanism. Eighty per cent of thalamic neurons showed input from a single fibre by the third postnatal week, a number indistinguishable between undeprived wild-type and AC-1 knockout animals. Although AC-1 is expressed both pre- and postsynaptically in this system, the investigators found no effect of AC-1 on presynaptic release properties. Thus, they conclude that the developmental phenotype observed here is associated with postsynaptic AC-1 expression. This study implicates AC-1 in deprivation-dependent synaptic loss at trigemino-thalamic synapses. It complements prior studies that have investigated a role for AC-1 and protein kinase A activation in LTP and LTD generation during early development in acute brain slices (Lu et al. 2003; Yasuda et al. 2003). In contrast to previous work, findings from the current study show that some forms of synaptic strengthening can be preserved in the knock-out, indicating compensatory mechanisms or alternative pathways in thalamic neurons. The results also show that the synaptic depression that occurs during changes in sensory input may be distinct from those that are engaged during normal developmental organization of this circuit. Taken together, these results provide important new insights into the developmental regulation of a fascinating synapse – the thalamic relay synapse – and underscore the remarkable diversity of plasticity processes across the brain.