-
The Importance of Utrophin and DystrophinDystrophin is the
protein damaged in many cases of muscular dystrophy. Duchenne
Muscular Dystrophy, the commonest and most severe form, is caused
by the absence of dystrophin and normally leads to death by early
adulthood. It is X-linked.Becker Muscular Dystrophy arises from
point mutations or small deletions in dystrophin and has a range of
severity.Dystrophin is found only in muscle cells, but utrophin, an
autosomal homologue of 69% similarity over more than 3500 residues,
is found in all tissuesIncreasing the amount of utrophin in animal
models of Duchenne Muscular Dystrophy partially corrects for the
defect in dystrophin
-
Cellular function of Utrophin and DystrophinDystrophin and
Utrophin bind to the actin cytoskeleton just under the plasma
membrane with the N terminal endThey bind to an assembly of
proteins at the plasma membrane known as the dystrophin associated
protein complexThis complex of proteins binds to the laminins of
the extracellular matrix forming a link between the actin
cytoskeleton and the extracellular matrix, which is thought to
provide a shock absorber role to the cell maintaining the integrity
of the membrane during muscle contraction
-
Domain OrganisationUtrophin and dystrophin can be divided into
three region, an N terminal region that binds actin, a long middle
section of spectrin like coiled-coil repeats and a C terminal
region which has various motifs involved in protein-protein
interactions. This region interacts with the Dystrophin Associated
Protein Complex
-
Actin Binding RegionThe actin binding region was localised to
the N terminal region in the ~250 amino acids before the
spectrin-like repeats beginThe utrophin actin binding region binds
actin with a stoichiometry of 1:1 in sedimentation assays and with
a dissociation constant of 58M. This is weaker than that reported
for whole dystrophin and one of the spectrin repeats of dystrophin,
but not the equivalent region of utrophin has been shown to bind
actin (1)Biochemical studies have identified three actin binding
sequences (ABS1,2,3). Two of these were peptides that showed
changes in the NMR linewidth on addition of F actin (ABS1 and ABS
3) (2,3). ABS2 was identified as the difference between a
proteolytic fragment that bound actin and one that did not (4).
-
Calponin Homology DomainsThe N Terminal region of about 250
amino acids was identified as being homologous to a region found in
a range of proteins that bind the cytoskeleton including -actinin,
-spectrin and fimbrin (see next page) These actin binding regions
show a weak sequence motif repeat of 120 amino acids. This motif is
also found in a single copy in a number of proteins including
calponin and is known as a calponin homology domain.Although some
of the single calponin homology domain proteins bind actin, for the
family of two domain actin binding regions both domains are
required for full actin binding activity. Isolated CH1 domains have
some affinity for actin alone but not isolated CH2 (5).
-
Actin Binding Region Family
-
Phylogenetic AnalysisThe first calponin homology domains of the
pair (CH1) form one phylogenetic group, the second (CH2) form a
second group and the single domains a third group
-
Structure of the utrophin CH2 domainHuman utrophin 144-261
expressed as a non fusion protein in E.coli.Space group P21 a=63.92
b=32.21 c=65.36 = 116.3o. 2 molecules in asymmetric unit.Data
Daresbury Station 9.5 Wavelength 1.00 Molecular replacement (AMORE)
with Spectrin CH2 domain (Matti Saraste)(6)Resolution 20-2.0
Completeness 99.5% Rmerge 0.032 Rref 0.185, Rfree 0.257 Model Chain
A 147-254, B 151-258 173 WaterPublished (7). PDB 1BHD4 main helices
3 roughly parallel 3, 4 and 6 and one roughly perpendicular 1 as
first seen in spectrin CH2 (6). Smaller helices vary between
domains
-
Structure of the utrophin actin-binding regionHuman Utrophin
28-261expressed without a fusion partner in E.coliSpace Group C2:
a=150.15, b=55.19, c=80.28 =106.0o. 2 molecules in asymmetric unit
SeMet MAD BM14 ESRF. 10 Se in ASU found using SHELXS. Refined
against remote wavelength 0.900 Resolution 24-3.0 Completeness
96.7% Rmerge 0.057 Rref 0.198 Rfree 0.258Model A+B 31-256 12 Water
PDB 1QAG
-
Structure of the dystrophin actin-binding regionHuman dystrophin
1-246 expressed with a C terminal Tag in E.coliSpace Group P1:
a=59.690 b=79.330 c=81.950 =61.08o =78.22o =70.54o. 4 molecules in
AU. Data Trieste 5.2R.Wavelength 1.00 MR (Amore) using utrophin CH1
from chain A/CH2 from chain B.Resolution 40-2.6 Completeness 95.4%
Rmerge 0.051 Rref 0.234 Rfree 0.262Model A+B+C+D 9-246 29 Water
-
Other structures from this actin-binding region familySpectrin
CH2 domain- First calponin homology domain solved. Matti Saraste
and coworkers EMBL (6). PDB 1AA2First actin binding region from
fimbrin. Steven Almo and coworkers (8). PDB 1AOA. This was the
first structure of a CH1 and CH2 together. Monomer in the
asymmetric unit. Fimbrin is different from the other members of the
family in having two actin binding regions (4 CH domains) on the
same chain and is most divergent in sequence. It has an insertion
of 13 amino acids between CH1 and CH2 domains relative to
utrophin/dystrophin. 9 of these are disordered in the crystal
structure but inspection of the distances to the symmetry related
copies confirms that this linker must fold back and the two CH
domains in the tight complex in the crystal come from the same
chain
-Actinin Poster P05.04.006 Uwe Sauer. We have not seen any
details of this
-
Individual CH domains superimposedSuperposition of CH domains.
Utrophin CH1 yellow CH2 red, Dystrophin CH1 cyan, CH2 black,
spectrin CH2 blue, Fimbrin CH1 purple, CH2 green. As well as the
insertion between domains fimbrin has a large inserted loop in each
domain.
-
Alignment and secondary structureResidues conserved in 6 out of
the 7 species are boxed. The conserved tryptophan is involved in
the inter CH domain interface. The DG is a tight turn between helix
2 and 3. The conserved Asp and Lys are in the interdomain interface
in CH2 but surface exposed in ABS2 presumably for actin binding in
CH1.
-
Dimers SuperimposedThe top picture shows fimbrin in green
superimposed on the dimer of utrophin in red and yellow and
dystrophin in blue and cyan. The CH1/CH2 complex superimposes
closely
The lower picture shows the second CH1/CH2 complex of utrophin
in red and yellow and dystrophin in blue and cyan when the first
are superimposed as above. There is a rotation of about 70 degrees
in the orientation of the domains.
-
Domain swappingSeveral examples of the same or similar proteins
having a conserved interface, which is in one case intrachain and
in another interchain are now known. This is known as 3D-domain
swapping (9).In many cases these are thought to be artefacts of
crystallisation conditions but in other cases, particularly of
homologous rather than identical proteins there is thought to be a
functional and evolutionary link.Gel filtration, NMR and analytical
ultracentrifugation data all indicate that the utrophin actin
binding region is monomeric in solution, so it is probable that the
linker refolds to give a utrophin monomer that resembles the
fimbrin crystal structure. We cannot totally rule out the dimer
purely being an artefact of crystallisation.Eisenberg Model of
Domain swappingClosed MonomersOpenmonomer3D Domain swapped
dimerEvolved Domainswapped dimer
-
Actin Binding ModelsA pseudo atomic model for fimbrin binding to
actin has been proposed based on building the atomic model of
F-actin (10) and the fimbrin actin binding region crystal structure
into a helical EM reconstruction. The simplest model for utrophin
binding to actin would be similar to this (left)An alternative
model would be to have the extended utrophin monomer seen in the
crystal structure binding actin. This does allow ABS1 and ABS3
which are largely buried in the CH domain interface in the fimbrin
like structure to interact with actin directly.
-
EM ReconstructionThe figure below shows two fittings of utrophin
CH domains to the difference map of an EM helical reconstruction of
actin subtracted from a utrophin actin binding domain-actin complex
reconstruction. There is some rearrangement from the crystal
structure to obtain the fit on the right hand side but it is
clearly better than the fimbrin like reconstruction (left).
-
Clinical MutationsDeletions of exon 3 (32-62) in green and exon
5 (89-119) in cyan which each remove large parts of CH1 cause
severe dystrophic symptoms. There are three point mutations that
cause Becker Dystrophy that map to the actin-binding regionL54R
introduces a charged residue into a hydrophobic environment and is
likely to disrupt the actin binding region. Dystrophin is still
found at the plasma membrane but severe symptoms are seen. Mutation
in red surrounding atoms in grey.A168D again introduces a charged
residue in a hydrophobic pocket.and Y231N removes steric bulk from
a hydrophobic core. These lie in CH2 and cause less severe
symptoms.CH1CH2
-
SummaryWe have determined the crystal structures of the CH2
domain of utrophin and the actin binding regions (CH1+CH2) of
utrophin and dystrophin.In contrast to the first fimbrin actin
binding region, which is a monomer, both utrophin and dystrophin
actin-binding regions crystallise as a dimer.The dimer found in
both the dystrophin and utrophin crystals suggests an alternative
model for actin binding from that seen in the EM reconstruction of
fimbrinOur EM helical reconstruction of the utrophin actin-binding
region bound to actin confirms our alternative model for actin
binding
-
Acknowledgements and ReferencesAs well as the authors of the
poster for their various contributions, we would like to thank Dr
John Berriman, Dr Linda Amos and Dr Tony Crowther for their
contribution to the EM reconstruction and the staff of Daresbury,
ESRF and Trieste synchrotrons particularly Dr Andrew Thompson
(ESRF) for assistance with data collectionPictures drawn with
Molscript (Kraulis), Bobscript (Esnouf), Alscript (Barton) and
Raster3d (Merrit)This work was supported by the Muscular Dystrophy
Group (now the Muscular Dystrophy Campaign). Data collection was
supported by the MRC, ESRF and EU.References1) Amann,K.J., Renley,
B.A. & Ervasti,J.M. (1998). J. Biol. Chem. 273, 28419-28423.2)
Levine, B.A., Moir, A.J.G., Patchell, V.B. & Perry, S.V.
(1990). FEBS Lett. 263, 159-162. 3) Levine, B.A., Moir, A.J.G.,
Patchell, V.B. & Perry, S.V. (1992). FEBS Lett. 298, 44-48. 4)
Bresnick, A.R., Warren, V. & Condeelis, J. (1990). J. Biol.
Chem. 265, 9236-9240. 5) Gimona, M. & Winder, S.J. (1998).
Curr. Biol. 8, R674-R675.6) Djinovic-Carugo, K., Banuelos, S. &
Sarraste, M. (1997). Nature Struct. Biol. 4, 175-179. 7) Keep,
N.H., Norwood, F.L.M., Moores, C.A., Winder, S.J. &
Kendrick-Jones, J. (1999). J. Mol. Biol. 285, 1257-1264. 8)
Goldsmith, S.C., Pokala, N., Shen, W., Federov, A.A., Matsudaira,
P. & Almo, S.C. (1997) Nature Struct. Biol. 4, 708-712 9)
Schulunegger, M.P., Bennett, M.J. & Eisenberg, D. (1997). Adv.
Protein Chem. 50, 61-122.10) Hanein, D. et al., & Matsudaira,
P. (1998). Nature Struct. Biol. 5, 787-792