Biology Reference
In-Depth Information
including endophilin, amphiphysin, intersectin and a-adaptin (McNiven et al.,
2000a). Accordingly, much of the work on Dyn1 and Dyn2 has focused on a
role in vesicle formation from the plasma membrane and mechanochemical
force-generating properties that function to tubulate and sever vesicle necks
(Hinshaw, 2000; Zhang and Hinshaw, 2001). In addition to mediating
endocytosis, Dyn2 also functions during vesicle formation from the Golgi
complex and endosomal sorting centres (Jones et al., 1998; van Dam and
Stoorvogel, 2002). Dyn3 is the least studied of the dynamins, but it has been
localized to the neuronal postsynaptic density (Gray et al., 2003) and is also
hypothesized to function during spermatogenesis in mice (Kamitani et al.,
2002).
More recently, dynamin has been proposed to function during actin-
dependent processes. This hypothesis is supported by data that demonstrates
the dynamin PRD mediates interactions with several actin-binding or actin-
associated regulatory proteins, including profilin, cortactin, Abp1, syndapin
(PACSIN), Grb2 and Nck (Grb4) (reviewed in Orth and McNiven, 2003).
Therefore, it is likely that the membrane modelling activities of dynamin,
coupled with its ability to interact with actin, make this protein an important
player in actin-membrane processes. In this chapter, we will discuss some data
that support a role for dynamin in these processes. The specific cell functions
discussed in this chapter include endocytosis, actin-based vesicle transport,
cytokinesis and dendritic spine morphogenesis.
Dynamin and the actin-binding protein cortactin interact directly
We began analysing the potential role of dynamin at the interface between
actin and membranes as it became clear that dynamin did not localize solely
with endocytic structures. Initial studies examining the localization of Dyn2 in
fixed and living cells demonstrated a strong enrichment at peripheral
membrane ru
es and a co-localization with cortical F-actin. These studies
prompted us to investigate whether dynamin could interact directly with
F-actin binding proteins. In collaboration with others, we discovered that the
SH3 domain of the F-actin binding protein cortactin directly bound to the
PRD of Dyn2 (McNiven et al., 2000b). This direct interaction was confirmed
using co-immunoprecipitations, GST-fusion pull-downs with peptide com-
petition and blot-overlay experiments. These experiments revealed that the
C-terminal half of the Dyn2 PRD was important and specific for binding to
the SH3 domain of cortactin but not to the SH3 domain of PLCg, another
dynamin PRD binding partner. Additional studies recently demonstrated that
point mutation of tryptophan 525 to a lysine residue in the cortactin SH3
domain disrupted the interaction with Dyn2 (Schafer et al., 2002). This
