We have collected SAXS curves for two multiprotein complexes (C1 and
C2) where each complex is a specific combination of two kinds of
Protomers (P1 and P2). P1 and P2 are well-folded structures and their
atomic coordinates can be treated as rigid bodies. Both C1 and C2 can
be thought of as a string of protomers from N- to C-terminus, where
adjacent protomers are connected by flexible linkers of 15 amino
acids. The sequence of protomers in C1 and C2 are P1P1P2P1P1
and P1P2P1 respectively. C2 is a subset of C1 in solution.
Additionally we have collected a third SAXS curve of multiprotein
complex3 (C3) that is composed of two sets of C2 (C2 forms stable
dimers under certain conditions). There are no linker restrains
between the two sets of C2 in C3 as the interaction between the sets
is purely electrostatic in solution.
Taken together, C2 is a subset of both C1 and C3. The structural
arrangement of the protomers in C2 is preserved in C1 and C3. Since
there is a presumed C2 commonality between the three multiprotein
complexes, we consider it best to refine C1, C2 and C3 simultaneously
as it provides cross validation across the incremental steps of the
refinement process.
Our program of choice so far is CORAL. We were successful in refining
C1 and C2 simultaneously. The PDB of C1 (P1P1P2P1P1) was
represented as a single chain and the middle constituents (P1P2P1)
between the flanking P1 protomers were defined as C2. In this
instance, the input files were a single PDB and two SAXS curves of C1
and C2. We are at a loss for a method to incorporate the SAXS curve of
C3. One possibility is to somehow define C3 as two sets of C2 and note
the absence of linker constrains between the sets, in which case the
input for refinement will consist of the original PDB and 3 SAXS
curves. Another possibility is to represent C3 in a separate PDB where
each set of C2 is specified in a discrete chain, in which case the
input for refinement will consist of two PDBs and 3 SAXS curves. In
any case we lack the CORAL-proficiency in testing these possibilities.
All help is welcome and most appreciated.
Amongst other things, we have no knowledge of the symmetries in the
above complexes. Kratky analysis along with a variety of biophysical
and biochemical data assures us of the structural stability and the
stoichiometry of the multiprotein complexes.
--
Balendu Avvaru
Postdoctoral Fellow
Cytoskeletal dynamics and Motility
Laboratory of Structural Enzymology and Biochemistry (LEBS)
CNRS Bâtiment 34, Avenue de la Terrasse
FRANCE 91198
