Exploring putative enteric methanogenesis inhibitors using molecular simulations and a graph neural network

Abstract

Atmospheric methane (CH₄) acts as a key contributor to global warming. As CH₄ is a short-lived climate forcer (12 years atmospheric lifespan), its mitigation represents the most promising means to address climate change in the short term. Enteric CH₄ (the biosynthesized CH₄ from the rumen of ruminants) represents 5.1% of total global greenhouse gas (GHG) emissions, 23% of emissions from agriculture, and 27.2% of global CH₄ emissions. Therefore, it is imperative to investigate methanogenesis inhibitors and their underlying modes of action. We hereby elucidate the detailed biophysical and thermodynamic interplay between anti-methanogenic molecules and cofactor F₄₃₀ of methyl coenzyme M reductase and interpret the stoichiometric ratios and binding affinities of sixteen inhibitor molecules. We leverage this as prior in a graph neural network to first functionally cluster these sixteen known inhibitors among ~54,000 bovine metabolites. We subsequently demonstrate a protocol to identify precursors to and putative inhibitors for methanogenesis, based on Tanimoto chemical similarity and membrane permeability predictions. This work lays the foundation for computational and de novo design of inhibitor molecules that retain/ reject one or more biochemical properties of known inhibitors discussed in this study.

Keywords: bromoform; climate change; emissions mitigation; enteric methanogenesis; enzyme inhibition.

Figure 1.

A comprehensive schematic illustrating the distribution of greenhouse gas emissions, focusing specifically on methane and detailing its biochemical synthesis and release mechanisms. a) Global representation of GHGs emissions with distributions centered on methane by sector as gathered from literature b) The entire enteric fermentation of carbohydrate (cellulose) feed as a mechanism of methane release. c) Biochemical reaction and the rate-limiting step in enteric methane synthesis catalyzed by MCR enzyme.

Figure 2.

Illustration of the crystal structure of Methyl Coenzyme M Reductase (MCR) (PBD accession ID: 5G0R) from Methanothermobacter marburgensis and the six-chain hexameric complex. a) Each chain of MCR crystal structure has been indicated with six colors. b) Catalytically active chains (A and D) of MCR are shown in green and blue, while other non-catalytic chains are shown in gray. The location of the cofactor F430 in the enzyme structure for both catalytic chains are indicated.

Figure 3.

Representation of all selected anti-methanogenic molecules (inhibitors) structures adopted for this study. a. Bromoform molecule. b. Group of Pterins. c. Group of Nitro- alcohols and esters. d. Group of Coenzyme B analogs. e. Group of Statins or HMG-CoA reductase inhibitors.

Figure 4:

Illustration of all three selected poses of bromoform interacting with Ni(I) of F₄₃₀ in MCR protein and graphical representation of the stoichiometric ratio of individual inhibitors docked to the active site of MCR enzyme in the close vicinity of F430. The dashed lines indicate the distances, in Å, between Ni(I) and bromoform. Cyan: for the distances of the first bromoform molecule. Gold: for the distances of the second bromoform molecule. Magenta: for the distances of the third bromoform molecule. The distances of other inhibitor molecules from Ni(I) are represented in supplementary information (see Figure S2-S17). b. Scatter plot representation of the mean binding affinity values of top three conformations of inhibitor molecules docked to F₄₃₀ of MCR. c. Representation of all positive conformations of inhibitor molecules accurately posed within a 5Å range.

Figure 5.

Two-dimensional t-SNE projection of molecular signatures reveals clustering of methanogenesis inhibitors. a) and b) Visualization of 16 known MCR inhibitors (Red) in relation to their four nearest neighbors (Black) selected from the Milk Composition Database (MCDB). c) and d) Similar visualization with four proximal metabolites (Black) identified in the Bovine Metabolome Database (BMDB).

Figure 5.

Two-dimensional t-SNE projection of molecular signatures reveals clustering of methanogenesis inhibitors. a) and b) Visualization of 16 known MCR inhibitors (Red) in relation to their four nearest neighbors (Black) selected from the Milk Composition Database (MCDB). c) and d) Similar visualization with four proximal metabolites (Black) identified in the Bovine Metabolome Database (BMDB).

Figure 6.

Tanimoto chemical similarity analysis between the LIM and UIMs relative to sixteen MCR inhibitors. (a) The sixteen inhibitors are represented at the periphery of the spider plot. The red-shaded area indicating the similarity of the proximal LIMs while the gray-shaded area represents the farther UIMs. b) Box plots illustrate the similarity of seven LIMs and nine UIMs relative to the 16 known MCR inhibitors. The red boxes represent LIMs, while the gray boxes represent UIMs. A p value of 0.003 indicates that the LIMs exhibit a statistically significant higher similarity to the known sixteen inhibitors compared to UIMs.