Methane emission from trees may partially or completely offset the methane sink in upland soils, the only process that has been regularly included in methane budgets for forest ecosystems. Our objective was to analyze multiple biogeochemical processes that influence the production, oxidation and transport of methane in a riparian cottonwood ecosystem and its adjacent river. We combined chamber flux measurements on tree stems, forest soil and the river surface with eddy covariance measurements of methane net ecosystem exchange. In addition, we tested whether methanogens were present in cottonwood stems, shallow soil layers and alluvial groundwater. Average midday peak in net methane emission measured by eddy covariance was c. 12 nmol m −2 s −1. The average uptake of methane by soils (0.87 nmol m −2 s −1) was largely offset by tree stem methane emission (0.75 nmol m −2 s −1). There was evidence of methanogens in tree stems but not in shallow soil. Growing season (May-September) cumulative net methane emission (17.4 mmol CH 4 m −2) included methane produced in cottonwood stems and methane input to the nocturnal boundary layer from the forest and the adjacent river. The multiple processes contributing to methane emission illustrated the linked nature of these adjacent terrestrial and aquatic ecosystems.

Differential mutagenic patterns were recently reported for O-methylated thymine lesions, which indicate that O4methylthymine (O4-Me-T) frequently leads to G misinsertions, whereas O2-methylthymine (O2-Me-T) is primarily nonmutagenic. The reasons for these differences are unclear since both lesions similarly alter the Watson−Crick binding face of T. To rationalize these replication outcomes at a molecular level, this work uses density functional theory calculations and molecular dynamics simulations to probe the lesion base-pairing properties as well as lesion accommodation by human polymerase η (pol η) and postextension DNA duplexes. O4-Me-T forms two strong hydrogen bonds with an opposing G in the active site of pol η, which rationalizes the observed lesion mutagenicity. Nevertheless, dATP insertion opposite O4-Me-T can proceed through water-mediated hydrogen bonding, which is similar to the pathway previously proposed for pol η bypass of abasic sites and other T alkylation lesions. In contrast, the position of O2-Me-T in the pol η active site is dynamic due to the presence of the aberrant methyl group on the minor groove side of DNA. In fact, the experimental replication outcomes can only be rationalized when the syn glycosidic orientation of O2-Me-T is considered, which stabilizes the pre-insertion complex by placing the damage in the polymerase open pocket on the major groove side of DNA. Although dATP insertion can occur opposite syn-O2-Me-T through a water-mediated pathway similar to O4-Me-T replication, rotation about the glycosidic bond precludes a stable pol η ternary complex corresponding to dGTP insertion, which correlates with the reported nonmutagenic bypass of O2-Me-T. In addition to providing structural insights into the differential mutagenicity of methylated T adducts, our data highlight an emerging theme in the literature for the replication of pyrimidine alkylation products in noncanonical glycosidic orientations and sets the stage for future work on the replication of other alkylated lesions by TLS polymerases.

Phosphodiester bond hydrolysis in nucleic acids is a ubiquitous reaction that can be facilitated by enzymes called nucleases, which often use metal ions to achieve catalytic function. While a two-metal-mediated pathway has been well established for many enzymes, there is growing support that some enzymes require only one metal for the catalytic step. Using human apurinic/ apyrimidinic endonuclease (APE1) as a prototypical example and cluster models, this study clarifies the impact of DFT functional, cluster model size, and implicit solvation on single-metal-mediated phosphodiester bond cleavage and provides insight into how to efficiently model this chemistry. Initially, a model containing 69 atoms built from a high-resolution X-ray crystal structure is used to explore the reaction pathway mapped by a range of DFT functionals and basis sets, which provides support for the use of standard functionals (M06-2X and B3LYP-D3) to study this reaction. Subsequently, systematically increasing the model size to 185 atoms by including additional amino acids and altering residue truncation points highlights that small models containing only a few amino acids or β carbon truncation points introduce model strains and lead to incorrect metal coordination. Indeed, a model that contains all key residues (general base and acid, residues that stabilize the substrate, and amino acids that maintain the metal coordination) is required for an accurate structural depiction of the one-metal-mediated phosphodiester bond hydrolysis by APE1, which results in 185 atoms. The additional inclusion of the broader enzyme environment through continuum solvation models has negligible effects. The insights gained in the present work can be used to direct future computational studies of other one-metal-dependent nucleases to provide a greater understanding of how nature achieves this difficult chemistry.

Nucleases catalyze the cleavage of phosphodiester bonds in nucleic acids using a range of metal cofactors. Although it is well accepted that many nucleases rely on two metal ions, the...

Reactive oxygen species damage DNA and result in health issues. The major damage product, 8-oxo-7,8-dihydroguanine (8oG), is repaired by human adenine DNA glycosylase homologue (MUTYH). Although MUTYH misfunction is associated with a genetic disorder called MUTYH-associated polyposis (MAP) and MUTYH is a potential target for cancer drugs, the catalytic mechanism required to develop disease treatments is debated in the literature. This study uses molecular dynamics simulations and quantum mechanics/molecular mechanics techniques initiated from DNA−protein complexes that represent different stages of the repair pathway to map the catalytic mechanism of the wild-type MUTYH bacterial homologue (MutY). This multipronged computational approach characterizes a DNA−protein cross-linking mechanism that is consistent with all previous experimental data and is a distinct pathway across the broad class of monofunctional glycosylase repair enzymes. In addition to clarifying how the cross-link is formed, accommodated by the enzyme, and hydrolyzed for product release, our calculations rationalize why cross-link formation is favored over immediate glycosidic bond hydrolysis, the accepted mechanism for all other monofunctional DNA glycosylases to date. Calculations on the Y126F mutant MutY highlight critical roles for active site residues throughout the reaction, while investigation of the N146S mutant rationalizes the connection between the analogous N224S MUTYH mutation and MAP. In addition to furthering our knowledge of the chemistry associated with a devastating disorder, the structural information gained about the distinctive MutY mechanism compared to other repair enzymes represents an important step for the development of specific and potent small-molecule inhibitors as cancer therapeutics.

The information contained within DNA as a sequence of nucleobases is required for life of most organisms, yet can get altered when the nucleobases are damaged upon exposure to many internal (hormones) and external (ultraviolet sunlight, pollutants) sources. As a result, repair pathways exist to combat the potentially detrimental effects of DNA damage. Nonbulky nucleobase damage (nucleobase oxidation, alkylation and deamination) is commonly removed by the base excision repair (BER) pathway, which involves several enzymes. The first BER enzymes are the DNA glycosylases, which are responsible for identifying the damaged base, flipping the base into the enzyme active site and removing the damaged nucleobase from the sugar–phosphate backbone. Due to the stability of many forms of damaged DNA, the DNA glycosylases must achieve great catalytic power. Understanding the mechanistic details associated with DNA glycosylases is essential for developing detection and treatment strategies for many diseases as abnormal glycosylase function has been associated with cancers, metabolic dysfunctions, neurodegeneration and epigenetic programming during embryo development. Molecular level insight into the function of a wide range of DNA glycosylases has been obtained from computational chemistry, including quantum mechanical cluster calculations, combined quantum mechanics‐molecular mechanics approaches and molecular dynamics simulations. By discussing some of the modeling that has been performed to date on monofunctional DNA glycosylases, the key contributions of the field of computational chemistry to broadening our understanding of the function of this important enzyme family, as well as the critical interplay between traditional biochemical experiments and computer calculations, is highlighted. This article is categorized under: Structure and Mechanism > Reaction Mechanisms and Catalysis Structure and Mechanism > Computational Biochemistry and Biophysics Electronic Structure Theory > Combined QM/MM Methods