Computational protocol: Novel protein fold discovered in the PabI family of restriction enzymes

Similar protocols

Protocol publication

[…] Our data collection and refinement statistics are summarized in . X-ray diffraction datasets of native and SeMet variant crystals of R.PabI were collected at the BL-5A and NW-12 beam-lines at the Photon Factory (Tsukuba, Japan), respectively. All the measurements were carried out under cryogenic conditions (95 K) using 20% ethylene glycol (final concentration) as the cryoprotectant. A native crystal of R.PabI diffracted X-rays to a resolution of 3.0 Å. The X-ray diffraction data were integrated and scaled with HKL2000 (). The native R.PabI crystal belonged to the space group P21 with the unit cell dimensions of a = 84.6 Å, b = 114.0 Å, c = 89.2 Å and β = 116.3°. Consideration of the values of VM suggests that this crystal has six protein molecules per asymmetric unit (VM = 2.5 Å3/Da; (). Diffraction data of the SeMet variant of R.PabI were collected and processed in the same way. A SeMet variant crystal diffracted X-rays to a resolution of 2.9 Å and belonged to the same space group with the native crystal, P21 with the unit cell dimensions of a = 84.6 Å, b = 114.2 Å, c = 89.4 Å and β = 116.3°. The crystal structure of R.PabI was determined by the single wavelength anomalous diffraction (SAD) phasing method using the diffraction data set of the SeMet variant of R.PabI. The selenium substructure was solved using the SnB program (). A total of 19 selenium sites were determined in the asymmetric unit. The initial phase was calculated with the program SHARP () using the coordinates of selenium sites solved by SnB program. Phase calculation resulted in an overall figure of merit (FOM) of 0.31 for the resolution range of 20–2.9 Å. After that, density modification and initial model building was performed with the program RESOLVE (). Molecular models of 759 residues (56% of the total) were automatically built with this calculation. The initial model of the SeMet variant of R.PabI was refined and manually rebuilt with the programs CNS () and XtalView (), using 10% (randomly chosen) of the reflections to calculate the Rfree. The partially refined model was transformed into five other subunits using the program MOLREP () in CCP4 (), and refined with the program REFMAC5 () with non-crystallographic symmetry (NCS) restraints. The crystal structure of native R.PabI was determined by a molecular replacement method using the coordinates of the SeMet variant structure of R.PabI with the program MOLREP. The final structure of R.PabI was refined and built using a native crystal diffraction data set (20–3.0 Å) with the program CNS (without NCS restraints) and XtalView. We did not use NCS refinement in the final step of refinement because R-factor and Rfree values became worse when refined with NCS restraint. We could not build coordinates of water molecules in this structure. Though we see some electron density peaks around R.PabI molecules, we could not clearly determine whether they were from water molecules or simply from noises. [...] Structure was analyzed with CCP4 () and APBS () and visualized with PyMol. The DNA-binding region was predicted using the PreDs program (). PreDs classifies each residue on the surface of a protein to the DNA-binding and non-DNA binding regions using a statistical evaluation function (). The evaluation function was developed based on an analysis of the shape and electrostatic properties of DNA-binding regions in structures of 63 protein–DNA complexes. GRAMM 1.03 () was used in the low-resolution docking mode to generate 2000 alternative docking orientations between the idealized symmetrical structure of R.PabI dimer (two identical copies of chain A) and idealized B-DNA 24-mer with the PabI site. The construction of an idealized dimer was necessary for docking because chain B exhibited missing density in the predicted DNA-binding interface. All 2000 orientations were filtered to retain only those matching the DNA-binding site that was predicted using PreDs () and clustered. The variant with the best shape complementarity and minimal number of clashes was refined manually to obtain symmetrical positioning of B-DNA to each monomer of R.PabI. […]

Pipeline specifications

Software tools CNS, Molrep, CCP4, REFMAC5, PyMOL
Applications Small-angle scattering, Protein structure analysis
Organisms Pyrococcus abyssi, Triticum aestivum