Applications of direct methods in protein crystallography for dealing with diffraction data down to 5 Å resolution

Direct methods applying to protein diffraction data below atomic resolution are described. Typical examples for SAD phasing, model completion and phase extension are given in detail.

Improved direct-method aided iterative dual-space model completion

It has been proved that direct methods are efficient in providing phase constraints within the dual-space phase-model iterative framework [1]. The program OASIS is used for the direct-method implementation. Two kinds of iterative direct methods are performed by OASIS. One involves the use of SAD/SIR information [2], while the other dose not [3]. Improvements have been made on both kinds of iterative direct methods. First, the Srinivasan's weighting function replaces previously used Sim's weighting function in the direct-method phase derivation leading to better estimation of phases. This affects both kinds of direct-method iteration. Second, the process of the second kind direct-method iteration has been made multi-threaded and parallel. This significantly speeds up the iteration in multi-CPU systems. Finally, a new figure of merit combining the R factor and Srinivasan's weighting function is used to pick up the best partial structure instead of using the R factor alone. The improved procedure has been tested using a set of SIRAS data from the protein LegC3N with Hg-derivative at 5.0Å resolution and native at 2.1Å resolution. With the previous version of OASIS, SAD phasing at 5.0Å resolution was successful but phases were failed to extend to the 2.1Å resolution native data as is described in §3.2 of [1]. However the new version of OASIS has succeeded in extending 5.0Å resolution SAD phases to 2.1Å resolution native phases leading to a nearly complete structure model.

An unusual metal ion configuration in a viral DNA-packaging nuclease active site

Nucleic acid metabolism is fundamental to many biological processes. A large class of enzymes such as RNase H, reverse transcriptase, retroviral integrase, topoisomerase, DNA and RNA polymerase, transposase, Holliday-junction resolvase, RNAi slicer Argonaute, and viral DNA-packaging terminase, utilize a common two-metal-ion catalytic mechanism for cleavage or synthesis of nucleic acid chains. Here we report an unusual metal-ion cluster in the active site of the nuclease domain of a viral DNA-packaging terminase unveiled by X-ray structures up to 1.38 Angstrom resolution. Two Mg2+ ions are situated in a coupled octahedral coordination system with liganding oxygen atoms from aspartic acid residues as well as water molecules. The two Mg2+ ions are located within a strikingly short distance of ~2.5 Å, which is unusual given the 1.6 Å atomic radius of Mg2+ and is shorter than previously observed metal-metal distances in metallocluster-containing enzymes or other biological systems. This provides the structural basis for distinguishing Mg2+ from other metal ions such as Ca2+ which are well known to support binding of the nucleic acid substrate but not support catalysis. Such an ultra-short distance between two metal-ions may be essential for generation of a highly positive niche, leading to nucleophilic attack at the phosphodiester bond of DNA. These results have defined the precise chemical configuration of the active site in nucleases using two-metal-ion catalytic mechanism. Moreover, assembly of this two-metal-ion cluster in the viral DNA-packaging terminase is mediated by an adjacent Lys residue, likely serving as a regulatory mechanism for activation of the nuclease activity of the terminase during packaging of viral genome.

Visualizing ATP hydrolysis in a viral DNA-packaging molecular motor

Many DNA viruses encode powerful molecular machines to package viral genome into preformed protein shells. These DNA-packaging motors contain an ATPase module that converts the chemical reaction of ATP hydrolysis to physical motion of DNA. We previously determined the structures of the DNA-packaging motor gp2 of Shigella phage Sf6 in the apo form and in complex with ADP and ATP-gamma-S (Zhao et al, 2013, PNAS, 110, 8075). Here we report the structure of gp2 in complex with its substrate ATP at 2.0 Angstrom resolution. To our knowledge, this is the first time to capture, at high resolution, a precatalytic state for ASCE-superfamily ATPases, which include a large group of nucleic acid helicases and translocases involved in a broad range of cellular and viral processes. The structure reveals the precise architecture of the ATP-bound state of the motor immediately prior to catalysis. Comparison with structures of the apo and ADP-complexed forms unveils motions of the Walker A motif coupled with ATP and Mg2+ binding and ATP hydrolysis. In the Walker B motif, residue E118 undergoes a side chain conformational switching coupled with the ATP hydrolysis, whereas residue E119 locks residue R51 side chain to a conformation that is readily reachable to residue E118 side chain. Residue E121 in the Walker B motif deprotonates a water molecule, which acts as a nucleophile to attack the gamma-phosphorous, leading to ATP hydrolysis. The alpha-helix (residue G182-R194) in the linker domain undergoes a translational motion against the ATPase domain triggered by ATP hydrolysis, serving as a mechanism for translating the energy from the chemical reaction into physical movement of DNA. We further observed the time course of ATP hydrolysis by gp2 by determining structures of gp2:ATP complexes captured at various incubation time. These structures have made it possible to delineate, at atomic detail, the complete cycle of ATP hydrolysis of this viral DNA-packaging molecular motor.