Although the core elements of the secretory machine are increasingly becoming understood at a molecular level, our understanding of their integration with the rest of the cellular architecture remains incomplete. This includes important questions such as ‘how do vesicles arrive at the plasma membrane’ and ‘how is this fusion directed to specific sites?’ Specificity of fusion is provided to some extent by the complement of cognate Q-and R-SNAREs found on the vesicular and plasma membranes, but this does not provide an explanation for the existence of exocytic hotspots found on the surface of neuroendocrine cells 
. Equally, while the involvement of the cytoskeleton in secretion is clear, the complexity of that involvement has not been fully elucidated.
There is now extensive evidence that the actin cytoskeleton plays a role in exocytosis, both at fast synapses
and in neuroendocrine cells 
. In electron micrographs of deep-etched freeze-fractured chromaffin cells, a lattice of actin filaments running parallel to the plasma membrane was observed, with clear attachment of the filaments to secretory vesicles 
. This and other studies 
led to the hypothesis that the actin cytoskeleton formed a barrier to secretion located just within the plasma membrane. Vitale and colleagues 
showed that vesicles were excluded from a depth of approximately 50 nm inside the plasma membrane, and that stimulation allowed vesicles to move closer to the plasma membrane. This provided a simple mechanism for two pools of releasable vesicles in these cells; a readily releasable pool comprising vesicles already present at the plasma membrane, and a reserve pool of vesicles which are recruited upon depletion of the first pool. This model has required re-evaluation in the light of more recent findings that actin may influence the exocytic event itself 
Actin dynamics have also been studied in fast synapses, where secretion is mediated by small synaptic vesicles (SSVs). Disruption of the actin cytoskeleton in these systems has been shown to result in an increase in release 
or an impairment in vesicle mobilization 
. As with dense-core vesicles, this may indicate that the cytoskeleton provides both transport and barrier functions.
PC12 cells have been extensively used as a model system for the study of secretion of large dense-cored vesicles (LDCVs), because of their ease of handling and consistency (see for example ref 7). We have used PC12 cells transfected with a releasable fluorescent cargo, ANF-EmGFP 
to study secretion in cells co-expressing fluorescent actin, with and without disruption of the cytoskeleton using the G-actin sequestering toxin Latrunculin A. The goal of this study was to revisit previous findings on the influence of actin sequestering agents on burst and phasic release, using the technique of evanescent wave microscopy to study this process at the level of individual granules. This technique also allowed us to examine spatial aspects of this process, revealing that disruption of the actin cytoskeleton provided more available sites for exocytosis.