Developing a predictive understanding of Hg transformation in ecosystems requires knowledge of exchange and feedback processes at Critical Interfaces. Task 4 involves molecular structure, dynamics, and mechanisms.
The overarching objective of Task 4 is to understand at the molecular scale how mercury interacts with and is transformed by the species it encounters in natural and contaminated environments. Since our discovery of the mercury methylation genes, we have focused primarily on characterizing the structure and function of HgcA but have also continued work on mercury-ligand interactions and bacterial mercury resistance.
Jointly with Task 3 and Judy Wall's lab (University of Missouri), we have identified several amino acids in HgcA and HgcB that play important roles in mercury methylation. We have also published two papers on mercuric ion reductase (MerA), which is involved in bacterial mercury resistance, and another paper describing computational methods for studying mercury interactions with natural organic matter (NOM) and proteins.
We have made significant progress toward characterizing HgcA. We have developed a protocol to express the cobalamin-binding domain (CBD) of HgcA heterologously in Escherichia coli as an N-terminal maltose binding protein (MBP) fusion (see figure), which greatly improves protein solubility and production yields. Fusion constructs were gifts from Steve Ragsdale (University of Michigan). We can now readily produce sufficient protein for detailed spectroscopic, structural, and functional characterization (Milestone 4b). Reconstitution with hydroxocobalamin (Cbl) yielded ~0.9 to 0.95 mol of Cbl per mol of purified protein. All manipulations from cell lysis to protein purification and cofactor reconstitution are now performed under strictly anaerobic conditions (<1 ppm O2) in an anaerobic chamber. With the availability of sufficient cofactor-loaded HgcACBD, we are now in a position to characterize HgcA by various spectroscopic and biophysical techniques. Experimental characterization is complemented by computational approaches. Confirming the proposed, unprecedented "Cys-on" coordination in HgcA, elucidating its effects on redox chemistry and reactivity, and working toward developing an in vitro enzymatic mercury methylation assay are of particular interest.
Ultraviolet–visible spectra of Aqua-Cbl, GS-Cbl, and recombinantly expressed and reconstituted HgcA.
In addition to ultraviolet–visible (UV/Vis) spectroscopy (see figure), we are working closely with the Ragsdale lab (University of Michigan) to characterize HgcA with electron paramagnetic resonance (EPR), magnetic circular dichroism (MCD), and resonance Raman spectroscopy. All spectroscopic and electrochemical experiments are being performed with complementary quantum chemical approaches.
Through a user proposal with the DOE Environmental Molecular Sciences Laboratory (EMSL; Proposal 48393), we are working to determine the structure of HgcACBD by nuclear magnetic resonance (NMR) spectroscopy. An NMR structure would provide direct evidence for the Rossmann fold, cap helix motif, and "Cys-on" coordination proposed for HgcACBD. We have prepared 15N-labeled protein, and preliminary data from 15N–HSQC NMR (heteronuclear single quantum coherence NMR) revealed a limited number of well-dispersed but broad resonances that indicate sample polydispersity (see figure). HgcA from Desulfovibrio desulfuricans ND132 contains two nonconserved cysteine residues, Cys47 and Cys142, which can form intermolecular disulfides. Dynamic light scattering and SDS-PAGE (sodium dodecyl sulfate–polyacrylamide gel electrophoresis) under reducing and nonreducing conditions revealed disulfide bond formation through Cys crosslinking, which results in high molecular weight oligomers. To overcome this issue, we have obtained a modified expression vector encoding a C47S/C142S double mutant in which only Cys93 is retained. In other work, we have heterologously expressed and purified native (i.e., untagged, full-length) HgcA, which includes its predicted transmembrane domain, and have solubilized it in the detergent n-dodecylmaltoside.
We used DFT calculations to investigate methyl transfer to Hg(II) in a model of HgcA.9 We predicted that a mutant of HgcA with Cys93 replaced by His may still be able to methylate mercury in vivo. This prediction was subsequently verified experimentally. We are now expanding upon that work by computing Co(III/II) and Co(II/I) standard reduction potentials, UV–visible absorption and Co(II) EPR spectra, and other molecular properties that will be used to help interpret experimental data and to make experimentally testable predictions.
In Task 4.2, we investigate mercury-ligand interactions in collaboration with Task 2. We have developed software that facilitates efficient workflows for computational chemistry of these interactions.
Task 4.3 investigates the molecular basis of mercury resistance mechanisms, specifically MerA (Milestone 4f ). To accomplish Hg(II) reduction to Hg(0), MerA transfers Hg(II) from a pair of cysteines at its solvent-exposed C-terminus to another pair of cysteines in its active site. We have combined molecular dynamics simulation on ORNL's Titan supercomputer with neutron scattering experiments to probe the dynamics of a compact state of MerA.26 We have also combined X-ray crystallography, quantum mechanics, and molecular mechanics calculations to study the mechanism of intramolecular Hg(II) transfer in the catalytic core of MerA. This work was performed jointly by ORNL, the University of California–San Francisco (Susan Miller), and the University of Tennessee (UTK) as part of our previous, university-led mercury project funded by the Subsurface Biogeochemical Research (SBR) program within DOE's Office of Biological and Environmental Research.