WH2 domains and actin variants as multifunctional organizers of the actin cytoskeleton

Actin is one of the most abundant proteins in eukaryotic cells and
regulation of the microfilament system is crucial for a wide range
of cellular functions including cell shape, cell motility, cell
division and membrane dynamics. The aim of this thesis was (1) to
gain a better understanding of the function of distinct actin
binding domains in the regulation of the actin cytoskeleton and (2)
to elucidate the role of actin variants. WH2 domains (WH2,
Wiskott-Aldrich syndrome protein homology 2) are ubiquitous
multifunctional regulators of actin dynamics. The protein Spire
contains four central WH2 domains A-B-C-D with about 20 amino acids
each and the cyclase-associated protein CAP2 contains only one WH2
domain. Under certain conditions, they can (1) nucleate actin
polymerization, (2) disintegrate actin filaments and (3) sequester
actin monomers. Here, the influence of selected Drosophila
melanogaster Spire-WH2 and Mus musculus CAP2-WH2 domain constructs
on actin dynamics was tested in vitro. To act as a filament
nucleator, at least two WH2 domains are required, and nucleation of
actin polymerization was only observed at substoichiometric
concentrations of WH2 domains over actin. At higher concentrations,
the sequestering activity of WH2 domains takes over. Preformed and
purified SpireWH2-actin complexes act as extremely efficient nuclei
for actin polymerization, even at superstoichiometric WH2
concentrations, under which free WH2 domains would sequester actin.
All analyzed constructs, including these with only a single WH2
domain, sequester actin as well as they can disrupt filaments. This
latter and most peculiar behavior of WH2 domains was observed in
fluorometric, viscometric and TIRF assays. The WH2 domains seem to
have such a high affinity for actin that they can forcefully
sequester monomers even from filaments and filament bundles, thus
breaking the whole structures. Taken together, the data clearly
show that SpireWH2-actin complexes are the intermediates that
account for the observed nucleating activity, whereas free WH2
domains can disrupt filaments and filament bundles within seconds,
again underlining the intrinsic versatility of this regulator of
actin dynamics. These data have been confirmed by crystallography
in collaboration with the groups of Prof. Dr. Tad Holak and Prof.
Dr. Robert Huber (Martinsried, Germany). Besides the well-studied
conventional actins many organisms harbor actin variants with
unknown function. The model organism Dictyostelium discoideum
comprises an actinome of a total of 41 actins, actin isoforms and
actin-related proteins. Among them is filactin, a highly conserved
actin with an elongated N-terminus. The 105 kDa protein has a
distinct domain organization and homologs of this protein are
present in other Dictyosteliidae and in some pathogenic Entamoebae.
Here, the functions of filactin were studied in vivo and in vitro.
Immunofluorescence studies in D. discoideum localize endogenous and
GFP-filactin in the cytoplasm at vesicle-like structures and in
cortical regions of the cell. A most peculiar behavior is the
stress-induced appearance of full length filactin in nuclear actin
rods. To perform in vitro analyses recombinant filactin was
expressed in Sf9 cells. Fluorescence studies with the filactin
actin domain suggest that it interferes with actin polymerization
by sequestering G-actin or even capping filaments. Gel filtration
assays propose a tetrameric structure of full length filactin.
Protein interaction studies suggest that filactin is involved in
the ESCRT (endosomal sorting complexes required for transport)
pathway which is responsible for multivesicular body formation. The
data on filactin suggest that only the conventional actins are the
backbone for the microfilamentous system whereas less related actin
isoforms have highly specific and perhaps cytoskeleton-independent
subcellular functions.