An unexpected specialization of the active zone scaffold RIM at high release synapses

The active zone scaffold RIM is canonically viewed as an obligate, universal pillar of the neurotransmitter release machinery. However, whether RIM is strictly required across diverse synapse subtypes with distinct release probabilities has remained an important unresolved question. Here, we report a fundamental revision of this model: RIM is not a constitutive necessity for baseline transmission, but rather a specialized "gain factor" selectively deployed to empower high-release synapses. Utilizing botulinum neurotoxin-based silencing to isolate convergent inputs at the Drosophila neuromuscular junction, we demonstrate that while RIM is essential for high-fidelity transmission at phasic (MN-Is) synapses, it is largely dispensable at low-release tonic (MN-Ib) synapses. This input specificity extends into plasticity: RIM is required for acute presynaptic homeostatic potentiation (PHP) at phasic inputs but is dispensable for the chronic maintenance of PHP at tonic inputs. Mechanistically, super-resolution imaging reveals that RIM is positioned significantly closer to Ca_V2 Ca²⁺ channel nanodomains at phasic synapses. During acute plasticity, RIM coordinates the homeostatic compaction of Ca_V2 channel clusters and drives the rapid expansion of the readily releasable vesicle pool. Our results revise the canonical view of active zone architecture, demonstrating that rather than serving as a static, uniform structural anchor, RIM functions as a tunable, dynamic module deployed to shape the performance and plasticity of high-demand synaptic connections.