ABSTRACT
UNLABELLED: Quantum mechanical calculations using the Marcus equation are applied to compare the electron-transfer probability for two distinct crystal structures of the Escherichia coli protein WrbA, an FMN-dependent NAD(P)H: quinone oxidoreductase, with the bound substrate benzoquinone. The calculations indicate that the position of benzoquinone in a new structure reported here and solved at 1.33 Å resolution is more likely to be relevant for the physiological reaction of WrbA than a previously reported crystal structure in which benzoquinone is shifted by â¼5 Å. Because the true electron-acceptor substrate for WrbA is not yet known, the present results can serve to constrain computational docking attempts with potential substrates that may aid in identifying the natural substrate(s) and physiological role(s) of this enzyme. The approach used here highlights a role for quantum mechanical calculations in the interpretation of protein crystal structures.
Subject(s)
Benzoquinones/chemistry , Escherichia coli Proteins/chemistry , Escherichia coli/metabolism , Flavin Mononucleotide/chemistry , Quantum Theory , Repressor Proteins/chemistry , Catalytic Domain , Crystallography, X-Ray , Electron Transport , Escherichia coli Proteins/metabolism , Protein Structure, Tertiary , Repressor Proteins/metabolismABSTRACT
WrbA is a novel multimeric flavodoxin-like protein of unknown function. A recent high-resolution X-ray crystal structure of E. coli WrbA holoprotein revealed a methionine sulfoxide residue with full occupancy in the FMN-binding site, a finding that was confirmed by mass spectrometry. In an effort to evaluate whether methionine sulfoxide may have a role in WrbA function, the present analyses were undertaken using molecular dynamics simulations in combination with further mass spectrometry of the protein. Methionine sulfoxide formation upon reconstitution of purified apoWrbA with oxidized FMN is fast as judged by kinetic mass spectrometry, being complete in â¼5 h and resulting in complete conversion at the active-site methionine with minor extents of conversion at heterogeneous second sites. Analysis of methionine oxidation states during purification of holoWrbA from bacterial cells reveals that methionine is not oxidized prior to reconstitution, indicating that methionine sulfoxide is unlikely to be relevant to the function of WrbA in vivo. Although the simulation results, the first reported for WrbA, led to no hypotheses about the role of methionine sulfoxide that could be tested experimentally, they elucidated the origins of the two major differences between apo- and holoWrbA crystal structures, an alteration of inter-subunit distance and a rotational shift within the tetrameric assembly.
Subject(s)
Apoproteins/chemistry , Escherichia coli Proteins/chemistry , Methionine/analogs & derivatives , Molecular Dynamics Simulation , Repressor Proteins/chemistry , Apoproteins/isolation & purification , Apoproteins/metabolism , Binding Sites , Escherichia coli Proteins/isolation & purification , Escherichia coli Proteins/metabolism , Flavin Mononucleotide/chemistry , Flavin Mononucleotide/metabolism , Kinetics , Methionine/chemistry , Methionine/metabolism , Oxidation-Reduction , Protein Binding , Protein Conformation , Protein Stability , Repressor Proteins/isolation & purification , Repressor Proteins/metabolism , Spectrometry, Mass, Electrospray Ionization , Structure-Activity RelationshipABSTRACT
The Escherichia coli protein WrbA, an FMN-dependent NAD(P)H:quinone oxidoreductase, was crystallized under new conditions in the presence of FAD or the native cofactor FMN. Slow-growing deep yellow crystals formed with FAD display the tetragonal bipyramidal shape typical for WrbA and diffract to 1.2â Å resolution, the highest yet reported. Faster-growing deep yellow crystals formed with FMN display an atypical shape, but diffract to only â¼1.6â Å resolution and are not analysed further here. The 1.2â Å resolution structure detailed here revealed only FMN in the active site and no electron density that can accommodate the missing parts of FAD. The very high resolution supports the modelling of the FMN isoalloxazine with a small but distinct propeller twist, apparently the first experimental observation of this predicted conformation, which appears to be enforced by the protein through a network of hydrogen bonds. Comparison of the electron density of the twisted isoalloxazine ring with the results of QM/MM simulations is compatible with the oxidized redox state. The very high resolution also supports the unique refinement of Met10 as the sulfoxide, confirmed by mass spectrometry. Bond lengths, intramolecular distances, and the pattern of hydrogen-bond donors and acceptors suggest the cofactor may interact with Met10. Slow incorporation of FMN, which is present as a trace contaminant in stocks of FAD, into growing crystals may be responsible for the near-atomic resolution, but a direct effect of the conformation of FMN and/or Met10 sulfoxide cannot be ruled out.
Subject(s)
Escherichia coli Proteins/chemistry , Repressor Proteins/chemistry , Crystallization , Crystallography, X-Ray , Escherichia coli Proteins/metabolism , Flavin Mononucleotide/chemistry , Flavin Mononucleotide/metabolism , Flavin-Adenine Dinucleotide/chemistry , Flavin-Adenine Dinucleotide/metabolism , NAD(P)H Dehydrogenase (Quinone)/chemistry , NAD(P)H Dehydrogenase (Quinone)/metabolism , Oxidation-Reduction , Protein Binding , Repressor Proteins/metabolism , X-Ray DiffractionABSTRACT
Raman microscopy permits structural analysis of protein crystals in situ in hanging drops, allowing for comparison with Raman measurements in solution. Nevertheless, the two methods sometimes reveal subtle differences in structure that are often ascribed to the water layer surrounding the protein. The novel method of drop-coating deposition Raman spectropscopy (DCDR) exploits an intermediate phase that, although nominally "dry," has been shown to preserve protein structural features present in solution. The potential of this new approach to bridge the structural gap between proteins in solution and in crystals is explored here with extrinsic protein PsbP of photosystem II from Spinacia oleracea. In the high-resolution (1.98 Å) x-ray crystal structure of PsbP reported here, several segments of the protein chain are present but unresolved. Analysis of the three kinds of Raman spectra of PsbP suggests that most of the subtle differences can indeed be attributed to the water envelope, which is shown here to have a similar Raman intensity in glassy and crystal states. Using molecular dynamics simulations cross-validated by Raman solution data, two unresolved segments of the PsbP crystal structure were modeled as loops, and the amino terminus was inferred to contain an additional beta segment. The complete PsbP structure was compared with that of the PsbP-like protein CyanoP, which plays a more peripheral role in photosystem II function. The comparison suggests possible interaction surfaces of PsbP with higher-plant photosystem II. This work provides the first complete structural picture of this key protein, and it represents the first systematic comparison of Raman data from solution, glassy, and crystalline states of a protein.
Subject(s)
Photosystem II Protein Complex/chemistry , Plant Proteins/chemistry , Spinacia oleracea/chemistry , Amino Acid Motifs , Amino Acid Sequence , Crystallography, X-Ray , Hydrogen Bonding , Models, Molecular , Molecular Sequence Data , Protein Structure, Secondary , Protein Structure, Tertiary , Spectrum Analysis, RamanABSTRACT
The nucleotide-binding domain of the alpha subunit of mouse brain Na+/K+-ATPase was expressed and isolated from Escherichia coli cells. A model structure was constructed by comparative modeling with and without docked ATP. This was compared with the secondary structure determination from UV circular dichroism and Raman spectroscopy. Thus, we support the quality of the model and the correct folding of the recombinant protein. ATP binding was followed by Raman difference spectroscopy, and its influence on the secondary structure of the N domain seems to not be significant.