Walter, Alexander M4; da Silva Pinheiro, Paulo César4; Verhage, Matthijs3; Sørensen, Jakob Balslev5
1 Neuronal Signalling Lab, Department of Neuroscience and Pharmacology, Faculty of Health and Medical Sciences, Københavns Universitet2 Neuropharm and Genetics Lab, Department of Neuroscience and Pharmacology, Faculty of Health and Medical Sciences, Københavns Universitet3 unknown4 Neuropharm and Genetics Lab, Department of Neuroscience and Pharmacology, Faculty of Health and Medical Sciences, Københavns Universitet5 Neuronal Signalling Lab, Department of Neuroscience and Pharmacology, Faculty of Health and Medical Sciences, Københavns Universitet
Neurotransmitter release depends on the fusion of secretory vesicles with the plasma membrane and the release of their contents. The final fusion step displays higher-order Ca(2+) dependence, but also upstream steps depend on Ca(2+). After deletion of the Ca(2+) sensor for fast release - synaptotagmin-1 - slower Ca(2+)-dependent release components persist. These findings have provoked working models involving parallel releasable vesicle pools (Parallel Pool Models, PPM) driven by alternative Ca(2+) sensors for release, but no slow release sensor acting on a parallel vesicle pool has been identified. We here propose a Sequential Pool Model (SPM), assuming a novel Ca(2+)-dependent action: a Ca(2+)-dependent catalyst that accelerates both forward and reverse priming reactions. While both models account for fast fusion from the Readily-Releasable Pool (RRP) under control of synaptotagmin-1, the origins of slow release differ. In the SPM the slow release component is attributed to the Ca(2+)-dependent refilling of the RRP from a Non-Releasable upstream Pool (NRP), whereas the PPM attributes slow release to a separate slowly-releasable vesicle pool. Using numerical integration we compared model predictions to data from mouse chromaffin cells. Like the PPM, the SPM explains biphasic release, Ca(2+)-dependence and pool sizes in mouse chromaffin cells. In addition, the SPM accounts for the rapid recovery of the fast component after strong stimulation, where the PPM fails. The SPM also predicts the simultaneous changes in release rate and amplitude seen when mutating the SNARE-complex. Finally, it can account for the loss of fast- and the persistence of slow release in the synaptotagmin-1 knockout by assuming that the RRP is depleted, leading to slow and Ca(2+)-dependent fusion from the NRP. We conclude that the elusive 'alternative Ca(2+) sensor' for slow release might be the upstream priming catalyst, and that a sequential model effectively explains Ca(2+)-dependent properties of secretion without assuming parallel pools or sensors.
P L O S Computational Biology (online), 2013, Vol 9, Issue 12, p. 1-15