Figure 2.
Figure 2: Comparison of atomic models for wild-type
bacteriorhodopsin and the D96G, F171C, F219L triple mutant.
a,b, Sections of [A]
weighted^27 2F[o]-F[c] density maps for wild-type
bacteriorhodopsin (purple) and the D96G, F171C, F219L triple
mutant (yellow) near the cytoplasmic (a) and extracellular (b)
boundaries. The maps are superimposed on the structure of
wild-type bacteriorhodopsin. Significant rearrangements of
helices F and G occur in the cytoplasmic, but not in the
extracellular region. c, Superposition of C atoms
of wild-type bacteriorhodopsin (purple) and the D96G, F171C,
F219L triple mutant (yellow) illustrating differences in the
cytoplasmic portions of helices F and G. Initially, a minimal
starting model containing only the transmembrane regions of
bacteriorhodopsin was used in a simplified least squares
refinement. Coordinates from six different starting models (PDB
entries 2brd, 1brr, 1ap9, 1brx, 1at9 and 2at9) were tested using
diffraction data sets obtained both from wild-type
bacteriorhodopsin and the triple mutant. The 1brr coordinates^28
were used as a starting model for the next stage of refinement
using the crystallography and NMR system (CNS)^27 suite of
programs, involving simulated annealing followed by temperature
factor refinement. The wild-type structure was refined to a
final R-factor of 23.9% (R[free] 31%), and the triple mutant
structure was refined to a final R-factor of 27.2% (R[ free]
32.1%). The final map was tested by completely omitting from the
starting model the side chains from a series of test residues
such as Phe 42, Trp 86, Trp 189 and Phe 208, or various
combinations of residues at the cytoplasmic ends of helix F or
helix G. In each case, difference maps (F[o]–F[c]) obtained at
the end of the refinement were unambiguous and clear density
peaks were observed for each of the omitted regions in
complete agreement with the atomic models reported here. As the
diffraction data from the triple mutant and wild-type
bacteriorhodopsin are completely independent of the 1brr
starting coordinates, these omit maps constitute stringent and
objective tests of the entire refinement procedure.
The above figure is reprinted
by permission from Macmillan Publishers Ltd:
Nature
(2000,
406,
653-657)
copyright 2000.
[A]
weighted^27 2F[o]-F[c] density maps for wild-type
bacteriorhodopsin (purple) and the D96G, F171C, F219L triple
mutant (yellow) near the cytoplasmic (a) and extracellular (b)
boundaries. The maps are superimposed on the structure of
wild-type bacteriorhodopsin. Significant rearrangements of
helices F and G occur in the cytoplasmic, but not in the
extracellular region. c, Superposition of C 
atoms
of wild-type bacteriorhodopsin (purple) and the D96G, F171C,
F219L triple mutant (yellow) illustrating differences in the
cytoplasmic portions of helices F and G. Initially, a minimal
starting model containing only the transmembrane regions of
bacteriorhodopsin was used in a simplified least squares
refinement. Coordinates from six different starting models (PDB
entries 2brd, 1brr, 1ap9, 1brx, 1at9 and 2at9) were tested using
diffraction data sets obtained both from wild-type
bacteriorhodopsin and the triple mutant. The 1brr coordinates^28
were used as a starting model for the next stage of refinement
using the crystallography and NMR system (CNS)^27 suite of
programs, involving simulated annealing followed by temperature
factor refinement. The wild-type structure was refined to a
final R-factor of 23.9% (R[free] 31%), and the triple mutant
structure was refined to a final R-factor of 27.2% (R[ free]
32.1%). The final map was tested by completely omitting from the
starting model the side chains from a series of test residues
such as Phe 42, Trp 86, Trp 189 and Phe 208, or various
combinations of residues at the cytoplasmic ends of helix F or
helix G. In each case, difference maps (F[o]–F[c]) obtained at
the end of the refinement were unambiguous and clear density
peaks were observed for each of the omitted regions in
complete agreement with the atomic models reported here. As the
diffraction data from the triple mutant and wild-type
bacteriorhodopsin are completely independent of the 1brr
starting coordinates, these omit maps constitute stringent and
objective tests of the entire refinement procedure.