|
Figure 3.
Figure 3. Smc1/3 Dimerization Specificity Is Solely
Conferred by the Hinge Domains(A) The hinge domain is necessary
for Smc1/3 dimerization. Smc1Δhinge or Smc1 were coexpressed
with His[6]Smc3 in insect cells and subjected to a pull-down
assay on Ni^2+-NTA. The presence of Smc1Δhinge or Smc1 in input
(I), unbound (U), and bound (B) fractions was probed by
immunoblotting with an antibody specific to the N terminus of
Smc1 (upper panel) and the efficiency of Smc3 binding to the
resin with anti-His antibody (lower panel).(B) Only molecules
with opposite hinge domains can dimerize. Smc1, HA[3]Smc3, or
Smc1hinge3 were coexpressed in insect cells with either
His[6]Smc3 or His[6]Smc3hinge1, and protein association of each
combination was assayed as in (A).(C) Electron micrographs of
the Smc3hinge1/Smc3 dimer. The His[6]Smc3hinge1/ HA[3]Smc3 dimer
was purified from insect cells over Ni^2+-NTA and gel
filtration. Proteins in the peak fraction from the gel
filtration were rotary shadowed with a 2 nm platinum layer and
visualized in the electron microscope.(D) The hinge domain of
Smc3 is sufficient for binding to Smc1. N-terminal, hinge, and
C-terminal globular domains of Smc3 were coexpressed with Smc1
in insect cells as HA[3]-tagged proteins. The globular domains
were immunoprecipitated and their ability to pull down Smc1 was
tested by immunoblotting for Smc1 (upper panel). Full-length
HA[3]Smc3 was used as a positive control. In addition, the
association of the HA[3]Smc3hinge domain with Smc1hinge3 was
tested. In all experiments, the efficiency of the
HA[3]-immunoprecipitation was tested by blotting against the
HA[3] epitope (lower panel).(E) The Smc3hinge domain binds Smc1
as tightly as the full-length Smc3 protein does. HA[3]Smc3 or
the HA[3]Smc3hinge domain produced in insect cells was bound to
a CM5 sensor chip on the BIAcore system via a monoclonal anti-HA
antibody attached to covalently linked anti-mouse Fc γ-specific
antibody. Insect cell extracts containing defined concentrations
of Smc1 as the ligand (five dilutions, ranging from 20 nM to 200
nM) were floated over the bound analytes, and association and
dissociation kinetics were recorded. For each dilution, the data
were fitted using a 1:1 Langmuir binding model with drifting
baseline and corrected for unspecific binding to uninfected
insect cell extracts. The average association and dissociation
rate constants (k[a] and k[d], respectively) are displayed and
used to calculate the equilibrium binding constant (K[A]).
Low average values of χ^2 indicate the accuracy of the fit and
the suitability of the 1:1 binding model, the variation
coefficients ν for the binding constants show the consistency
of the measurements over the ligand dilution range.(F) Crystal
structure of the hinge domain from Thermotoga maritima SMC
protein (construct HTMC9, residues 473-685). Ribbon drawing of
the hinge domain dimer, showing two stretches of antiparallel
coiled coil (yellow and green). The orientation is essentially
the same as in Figure 1B. The coiled coil segments are formed by
residues from the same chain, resulting in an intramolecular
coiled coil arrangement for SMC proteins. The structure shown
was re-solved in spacegroup C2 by seleno-methionine substitution
and MAD at 3.0 Å resolution.(G) Architecture of SMC
proteins. The intramolecular coiled coil results in the two arms
being formed by separate chains with the hinge domains holding
the two arms together. The coiled coil segments have been
modeled using standard geometry and the crystal structures of
the hinge and head domains have been described here and
elsewhere (Löwe et al., 2001). Figure prepared using
MOLSCRIPT (Kraulis P.J., 1991).
|
