Limits...
Type II dehydroquinase: molecular replacement with many copies.

Stewart KA, Robinson DA, Lapthorn AJ - Acta Crystallogr. D Biol. Crystallogr. (2007)

Bottom Line: In crystals where this is not the case, multiple copies of the monomer are present, giving rise to significant and often confusing noncrystallographic symmetry in low-symmetry crystal systems.Four commonly used molecular-replacement packages are assessed against two of these examples, one of high symmetry and the other of low symmetry; this study highlights how program performance can vary significantly depending on the given problem.In addition, the final refined structure of the 16-dodecamer triclinic crystal form is analysed and shown not to be a superlattice structure, but rather an F-centred cubic crystal with frustrated crystallographic symmetry.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Chemistry, University of Glasgow, Glasgow G12 8QQ, Scotland.

ABSTRACT
Type II dehydroquinase is a small (150-amino-acid) protein which in solution packs together to form a dodecamer with 23 cubic symmetry. In crystals of this protein the symmetry of the biological unit can be coincident with the crystallographic symmetry, giving rise to cubic crystal forms with a single monomer in the asymmetric unit. In crystals where this is not the case, multiple copies of the monomer are present, giving rise to significant and often confusing noncrystallographic symmetry in low-symmetry crystal systems. These different crystal forms pose a variety of challenges for solution by molecular replacement. Three examples of structure solutions, including a highly unusual triclinic crystal form with 16 dodecamers (192 monomers) in the unit cell, are described. Four commonly used molecular-replacement packages are assessed against two of these examples, one of high symmetry and the other of low symmetry; this study highlights how program performance can vary significantly depending on the given problem. In addition, the final refined structure of the 16-dodecamer triclinic crystal form is analysed and shown not to be a superlattice structure, but rather an F-centred cubic crystal with frustrated crystallographic symmetry.

Show MeSH

Related in: MedlinePlus

A plot of asymmetric unit volume against resolution for the X-ray crystal structures in the PDB. Entries with very small unit cells have been omitted, as have structures determined at higher than 1.5 Å resolution. Four structures are represented by red stars as they do not fit onto the plot; namely (from left to right) entries 1wce (birnavirus, 234 × 106 Å3), 1w8x (bacteriophage Prd1, 192.5 × 106 Å3), 1ohg (Hk97 bacteriophage capsid, 143 × 106 Å3) and 2btv (bluetongue virus core, 123 × 106 Å3). The type II dehydroquinase structure reported here has an asymmetric unit volume of 7.6 × 106 Å3 and is labelled 2bt4.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2394815&req=5

fig3: A plot of asymmetric unit volume against resolution for the X-ray crystal structures in the PDB. Entries with very small unit cells have been omitted, as have structures determined at higher than 1.5 Å resolution. Four structures are represented by red stars as they do not fit onto the plot; namely (from left to right) entries 1wce (birnavirus, 234 × 106 Å3), 1w8x (bacteriophage Prd1, 192.5 × 106 Å3), 1ohg (Hk97 bacteriophage capsid, 143 × 106 Å3) and 2btv (bluetongue virus core, 123 × 106 Å3). The type II dehydroquinase structure reported here has an asymmetric unit volume of 7.6 × 106 Å3 and is labelled 2bt4.

Mentions: As only approximately 1 mg of the CA1 bifunctional inhibitor was available for cocrystallization with S. coelicolor DHQase, extensive crystallization screening was not possible in this case. Cocrystallizations were set up with an in-house PEG/ion screen which had previously been successful in producing a well diffracting orthorhombic crystal form of this enzyme (Roszak et al., 2002 ▶). A dozen crystals were screened for suitably high-resolution diffraction at the Daresbury SRS as the crystal quality was variable and the crystals suffered from anisotropic diffraction and high mosaicity (frequently 2°). One of the crystals diffracted to 1.7 Å resolution with low mosaicity, allowing 250° of oscillation data to be collected. Indexing the diffraction pattern with DENZO gave a large I-­centred tetragonal cell with unit-cell parameters a = b = 196.6, c = 393.6 Å; however, the data did not merge with this symmetry or as I-centred orthorhombic. Indexing the data in lower symmetry such as C-centred monoclinic with unit-cell parameters a = 280.4, b = 280.8, c = 242.7 Å, β = 125.2° was equally unsuccessful; however, the data were not mis-indexed as they merged acceptably in space group P1. This large triclinic crystal form had been encountered previously (Roszak et al., 2002 ▶) and had been abandoned as being impractical. In the absence of any other more amenable crystals of this DHQase–inhibitor complex, a structure solution was attempted using these data. The data were processed in P1 with unit-cell parameters a = 196.61, b = 196.49, c = 240.62 Å, α = 65.9, β = 65.9, γ = 90.1° using the HKL suite of programs dimensioned for viruses. The data were 93% complete with 3 042 000 unique reflections to ∼1.7 Å and an average redundancy of 2.4. The merging R factor was 17% owing to the low redundancy and the weakness of the data beyond 2.0 Å, which was a consequence of the data being collected in one pass. Visual inspection of the merged diffraction data with HKLVIEW (Collaborative Computational Project, Number 4, 1994 ▶) showed no pattern of alternating strong and weak reflections as is observed in superlattice structures where an inexact lattice translation has resulted in a doubling of a cell dimension; instead, a curious striped pattern involving blocks of data was seen (Fig. 2 ▶). Analysis of the self-rotation function showed three perpendicular twofold axes, in addition to other threefold and fourfold rotations at a maximum of 83% of that expected for crystallographic symmetry. The native Patterson map showed a peak half the size of the origin at 0.127, 0.126, 0.240, indicating translational NCS. A Matthews coefficient of 2.4 Å3 Da−1 can be derived for 16 dodecamers in the unit cell; this assumes 50% solvent content for this crystal form as is found in other S. coelicolor DHQase structures. Given such a large number of molecules in the P1 cell, it was presumed that the structure was a superlattice structure of some type. A search of the Protein Data Bank (PDB; Sussman et al., 1998 ▶) based on the size of the asymmetric unit (Fig. 3 ▶) reveals that this structure is not unique in its size, as structures with a complete virus as the asymmetric unit have been solved. However, for its resolution, this structure is a significant outlier some 4–5 times larger in size than anything previously deposited at this or higher resolution.


Type II dehydroquinase: molecular replacement with many copies.

Stewart KA, Robinson DA, Lapthorn AJ - Acta Crystallogr. D Biol. Crystallogr. (2007)

A plot of asymmetric unit volume against resolution for the X-ray crystal structures in the PDB. Entries with very small unit cells have been omitted, as have structures determined at higher than 1.5 Å resolution. Four structures are represented by red stars as they do not fit onto the plot; namely (from left to right) entries 1wce (birnavirus, 234 × 106 Å3), 1w8x (bacteriophage Prd1, 192.5 × 106 Å3), 1ohg (Hk97 bacteriophage capsid, 143 × 106 Å3) and 2btv (bluetongue virus core, 123 × 106 Å3). The type II dehydroquinase structure reported here has an asymmetric unit volume of 7.6 × 106 Å3 and is labelled 2bt4.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2394815&req=5

fig3: A plot of asymmetric unit volume against resolution for the X-ray crystal structures in the PDB. Entries with very small unit cells have been omitted, as have structures determined at higher than 1.5 Å resolution. Four structures are represented by red stars as they do not fit onto the plot; namely (from left to right) entries 1wce (birnavirus, 234 × 106 Å3), 1w8x (bacteriophage Prd1, 192.5 × 106 Å3), 1ohg (Hk97 bacteriophage capsid, 143 × 106 Å3) and 2btv (bluetongue virus core, 123 × 106 Å3). The type II dehydroquinase structure reported here has an asymmetric unit volume of 7.6 × 106 Å3 and is labelled 2bt4.
Mentions: As only approximately 1 mg of the CA1 bifunctional inhibitor was available for cocrystallization with S. coelicolor DHQase, extensive crystallization screening was not possible in this case. Cocrystallizations were set up with an in-house PEG/ion screen which had previously been successful in producing a well diffracting orthorhombic crystal form of this enzyme (Roszak et al., 2002 ▶). A dozen crystals were screened for suitably high-resolution diffraction at the Daresbury SRS as the crystal quality was variable and the crystals suffered from anisotropic diffraction and high mosaicity (frequently 2°). One of the crystals diffracted to 1.7 Å resolution with low mosaicity, allowing 250° of oscillation data to be collected. Indexing the diffraction pattern with DENZO gave a large I-­centred tetragonal cell with unit-cell parameters a = b = 196.6, c = 393.6 Å; however, the data did not merge with this symmetry or as I-centred orthorhombic. Indexing the data in lower symmetry such as C-centred monoclinic with unit-cell parameters a = 280.4, b = 280.8, c = 242.7 Å, β = 125.2° was equally unsuccessful; however, the data were not mis-indexed as they merged acceptably in space group P1. This large triclinic crystal form had been encountered previously (Roszak et al., 2002 ▶) and had been abandoned as being impractical. In the absence of any other more amenable crystals of this DHQase–inhibitor complex, a structure solution was attempted using these data. The data were processed in P1 with unit-cell parameters a = 196.61, b = 196.49, c = 240.62 Å, α = 65.9, β = 65.9, γ = 90.1° using the HKL suite of programs dimensioned for viruses. The data were 93% complete with 3 042 000 unique reflections to ∼1.7 Å and an average redundancy of 2.4. The merging R factor was 17% owing to the low redundancy and the weakness of the data beyond 2.0 Å, which was a consequence of the data being collected in one pass. Visual inspection of the merged diffraction data with HKLVIEW (Collaborative Computational Project, Number 4, 1994 ▶) showed no pattern of alternating strong and weak reflections as is observed in superlattice structures where an inexact lattice translation has resulted in a doubling of a cell dimension; instead, a curious striped pattern involving blocks of data was seen (Fig. 2 ▶). Analysis of the self-rotation function showed three perpendicular twofold axes, in addition to other threefold and fourfold rotations at a maximum of 83% of that expected for crystallographic symmetry. The native Patterson map showed a peak half the size of the origin at 0.127, 0.126, 0.240, indicating translational NCS. A Matthews coefficient of 2.4 Å3 Da−1 can be derived for 16 dodecamers in the unit cell; this assumes 50% solvent content for this crystal form as is found in other S. coelicolor DHQase structures. Given such a large number of molecules in the P1 cell, it was presumed that the structure was a superlattice structure of some type. A search of the Protein Data Bank (PDB; Sussman et al., 1998 ▶) based on the size of the asymmetric unit (Fig. 3 ▶) reveals that this structure is not unique in its size, as structures with a complete virus as the asymmetric unit have been solved. However, for its resolution, this structure is a significant outlier some 4–5 times larger in size than anything previously deposited at this or higher resolution.

Bottom Line: In crystals where this is not the case, multiple copies of the monomer are present, giving rise to significant and often confusing noncrystallographic symmetry in low-symmetry crystal systems.Four commonly used molecular-replacement packages are assessed against two of these examples, one of high symmetry and the other of low symmetry; this study highlights how program performance can vary significantly depending on the given problem.In addition, the final refined structure of the 16-dodecamer triclinic crystal form is analysed and shown not to be a superlattice structure, but rather an F-centred cubic crystal with frustrated crystallographic symmetry.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Chemistry, University of Glasgow, Glasgow G12 8QQ, Scotland.

ABSTRACT
Type II dehydroquinase is a small (150-amino-acid) protein which in solution packs together to form a dodecamer with 23 cubic symmetry. In crystals of this protein the symmetry of the biological unit can be coincident with the crystallographic symmetry, giving rise to cubic crystal forms with a single monomer in the asymmetric unit. In crystals where this is not the case, multiple copies of the monomer are present, giving rise to significant and often confusing noncrystallographic symmetry in low-symmetry crystal systems. These different crystal forms pose a variety of challenges for solution by molecular replacement. Three examples of structure solutions, including a highly unusual triclinic crystal form with 16 dodecamers (192 monomers) in the unit cell, are described. Four commonly used molecular-replacement packages are assessed against two of these examples, one of high symmetry and the other of low symmetry; this study highlights how program performance can vary significantly depending on the given problem. In addition, the final refined structure of the 16-dodecamer triclinic crystal form is analysed and shown not to be a superlattice structure, but rather an F-centred cubic crystal with frustrated crystallographic symmetry.

Show MeSH
Related in: MedlinePlus