Bifunctional architecture enables substrate catalysis and channeling in Paracoccus TMAO demethylase
Figures
Trimethylamine N-oxide demethylase (TDM) catalyzes two sequential reactions.
(A) Steady-state kinetic analysis of TDM-producing formaldehyde (HCHO), in response to varying concentrations of the substrate trimethylamine N-oxide (TMAO) (n=3). (B) Representative high-performance liquid chromatography coupled with mass spectrometry (HPLC-MS) spectrum of dimethylamine (DMA) product from TDM:TMAO reaction. The standard curve for DMA is shown in green, and that for TMAO is shown in orange. The red trace shows the product of the reaction when TMAO is added to TDM, indicating the formation of DMA. Counts in the Y-axis. (C) Linear regression analysis demonstrates a concentration-dependent decrease in HCHO production when tetrahydrofolate (THF) is included in the reaction mixture, suggesting THF-dependent consumption or conversion of HCHO. (D) Representative HPLC-MS confirming the formation of methylenetetrahydrofolate (MTHF) adduct as a product of the TDM:TMAO:THF reaction. The red, gray, and green lines represent the standard curves for THF, MTHF, and the reaction product when TMAO and THF were present with TDM, respectively.
Purification and biochemical characterization of trimethylamine N-oxide demethylase (TDM).
(A) Protein samples were fractionated using a Superdex 200 10/300 column at varying salt concentrations. The peaks below 10 mL were in the void volume and aggregated. At 150 mM NaCl, a peak around 12.5 mL shows the best activity. (B) SDS-PAGE was used to assess sample purity. The gel shows the band under the gray and black fractions (corresponding to elution volumes marked in A) at 150 mM NaCl. Note that all peaks have TDM. The arrow on the right shows a band corresponding to the molecular weight of TDM. (C) TDM catalytic activity at different pH values (20 mM acetate buffer for pH of 4.0–5.0; 20 mM MES buffer for pH of 5.5–6.5; 20 mM Tris-HCl buffer for pH of 7.0–9.0) at 37°C for 30 min and (D) temperature conditions in a buffer of 20 mM MES pH 6.5, 150 mM NaCl. Relative activity was measured in relation to the maximum activity, which was set at 100%. All kinetic experiments were performed in triplicates.
-
Figure 1—figure supplement 1—source data 1
Original, unedited SDS-PAGE gel scan supporting the data presented in Figure 1—figure supplement 1B.
- https://cdn.elifesciences.org/articles/109964/elife-109964-fig1-figsupp1-data1-v1.zip
-
Figure 1—figure supplement 1—source data 2
Original, unedited SDS-PAGE gel scan supporting the data in Figure 1—figure supplement 1B.
Molecular weight standards are indicated in kDa; the blue arrow indicates the target trimethylamine N-oxide demethylase (TDM) protein.
- https://cdn.elifesciences.org/articles/109964/elife-109964-fig1-figsupp1-data2-v1.zip
Overall architecture of trimethylamine N-oxide demethylase (TDM) determined by cryo-electron microscopy (cryo-EM).
(A) Representative cryo-EM maps. (B) Overall structure of the complex, with each subunit colored differently. TDM adopts a 2+2½ complexation. The two half-domains of the complex are the C-terminal domains. (C) Schematic illustration of TDM complexation. TDM monomers with N-terminal, core, and C-terminal domains are colored forest, orange, and blue, respectively. The same coloring is used in panels (A) and (B). (D) Close-up view of the 3Cys:Zn2+ binding motif. (E) Close-up view of the interactions between dimethylamine (DMA) and formaldehyde (HCHO) products with the 3Cys:Zn2+ motif. (F) The trimethylamine N-oxide (TMAO) substrate is trapped in the 3Cys:Zn2+ active site of the double D220A/D367A mutant. The density maps are shown and contoured at 3.0σ with a carve of 2 Å. Bound small molecules are shown as ball-and-stick representations, amino acid residues as sticks, and hydrogen bonds as dashed lines. (G) Relative activity of TDM variants.
The cryo-electron microscopy (cryo-EM) data processing pipeline for trimethylamine N-oxide demethylase (TDM) structures carried out using cryoSPARC.
Following iterative rounds of particle selection, classification, and refinement, a final subset comprising 10–60% of the initially exported particles was used to reconstruct the 3D EM map. The resulting map was subjected to local resolution estimation using the Phenix, which revealed a resolution range spanning from 2.0 to 5.0 Å across different regions of the structure. Both the global and local resolution maps were subsequently visualized and analyzed using UCSF ChimeraX to facilitate model interpretation and validation.
Dimerization of the core domains in the trimethylamine N-oxide demethylase (TDM) complex.
(A) The central dimer interface is mediated by residues located within the core domain of each TDM monomer. For clarity, the N-terminal, core, and C-terminal domains are shown in forest green, light orange, and light blue, respectively. This structural arrangement highlights the spatial organization of the domains contributing to dimer formation. (B) The interface between the two core domains is stabilized through a network of interactions, including hydrogen bonds and hydrophobic contacts. Key residues involved in these interactions are positioned symmetrically across the interface, contributing to the overall stability and specificity of the dimeric assembly.
Sequence alignment of trimethylamine N-oxide demethylase (TDM) from Paracoccus sp. and other organisms using multalign.
Green arrows indicate key residues in active site binding to trimethylamine N-oxide (TMAO) substrate. Red, conserved; blue, moderately conserved; black, no conserved residues. Paracoccus sp. DMF (WP_263566861.1), Paracoccus gahaiensis (WP_136886199.1), Tabrizicola flagellates (WP_149588984.1), Rhodobacter sp. SY28-1 (WP_137111587.1), Amaricoccus solimangrovi (WP_140453703.1), Sinirhodobacter sp. WL0062 (WP_233677943.1), Paenirhodobacter sp. CAU 1674 (WP_275732795.1), Pseudodonghicola flavimaris (WP_284481585.1).
Zinc association with the trimethylamine N-oxide demethylase (TDM) protein.
(A) Energy-dispersive X-ray spectroscopy (EDS) analysis was performed to detect the presence and spatial distribution of zinc (Zn) in the purified TDM protein sample. The black-and-white panel represents the overall protein structure, while the green overlay indicates the specific localization of Zn signals, suggesting Zn binding or association with the protein. (B) Quantitative analysis of metal content was conducted using inductively coupled plasma mass spectrometry (ICP-MS). The molar ratios of iron and zinc relative to the TDM protein concentration were determined. Measurements were performed on independently prepared protein samples (n=2).
paraTDM lacks an iron-binding site.
(A) Sequence alignment of M. silvestris trimethylamine N-oxide demethylase (MsTDM, WP_012592556.1) and paraTDM (WP_263566861), highlighting residues proposed to coordinate Fe2+ in MsTDM (arrows). (B) Cryo-electron microscopy (cryo-EM) structure of TDM expressed in LB medium supplemented with 0.5 mM Fe(NH4)2SO4 (Fe-enriched paraTDM). No additional EM density corresponding to a bound metal ion is observed, including in the regions between the proposed coordinating residues D220 and H298 (arrows) or between H298 and D367. Residues are shown in stick representation. The corresponding EM density contoured at 3.0 Å resolution is shown in the three residues.
Interactions of the substrate (trimethylamine N-oxide [TMAO]) and reaction products within the Zn2+-containing active site of trimethylamine N-oxide demethylase (TDM).
TMAO is coordinated within the active site through interactions with Zn2+ and surrounding conserved residues. The proposed product-binding states illustrate interactions of dimethylamine (DMA) and formaldehyde (HCHO) with residues lining the active-site pocket following turnover.
Tetrahydrofolate (THF) binding to trimethylamine N-oxide demethylase (TDM).
(A) The THF-binding domain forms a cloverleaf of three subdomains (I–III), creating a central cavity that accommodates THF. Electron microscopy (EM) density for THF is shown as a surface. (B) Close-up of the binding pocket highlighting hydrophobic and hydrogen-bonding residues that stabilize THF, colored according to their subdomain location in panel A. (C) Isothermal calorimetry (ITC) analysis confirms specific THF binding to TDM (blue), with negligible signal in buffer control (gray).
Structural features of the tetrahydrofolate (THF)-binding domain in the trimethylamine N-oxide demethylase (TDM):dimethylamine (DMA):formaldehyde (HCHO):THF complex.
(A) Dimerization of the THF-binding domains occurs through interactions between adjacent protomers, contributing to the overall oligomeric architecture of the TDM complex. (B) The central dimer interface is primarily formed by residues located on the inward-facing surfaces of the THF-binding domains. Stabilization of the interface involves multiple hydrogen bonds and additional polar contacts, facilitating a tight and specific interaction. (C) Structural comparison revealed a high degree of similarity to a THF-binding protein, as demonstrated by superposition with the T protein (PDB ID: 1WOO).
The binding of tetrahydrofolate (THF) to trimethylamine N-oxide demethylase (TDM) was determined by isothermal calorimetry (ITC).
(A) Representative ITC thermogram showing the heat changes upon titration of THF into a solution containing purified TDM protein. (B, C) Control ITC thermogram showing titration of buffer into TDM (B) and titration of THF into buffer (C).
Formaldehyde (HCHO) channels from the active site to the tetrahydrofolate (THF)-binding site in trimethylamine N-oxide demethylase (TDM) complex.
(A, B) Surface representations of the channels within the TDM complex colored according to either electrostatic potential or radius. (C) Tunnel analysis of a coarse-grained molecular dynamics (CG-MD) simulation suggests that HCHO in each chain can access the THF-binding sites in chains A and B. The five top-ranked paths for chain A are shown in different colors. THF, dimethylamine (DMA), and HCHO are depicted as spheres colored by the atom type. TA, TB, TC, and TD: THF molecules in chains A, B, C, and D, respectively; HA: chain A, HCHO.
Structural stability of the trimethylamine N-oxide demethylase (TDM) complex during coarse-grained molecular dynamics (CG-MD) simulations.
Root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), and radius of gyration (Rg) were analyzed over the course of a 10 μs CG-MD simulation. (A) The RMSD of protein backbone atoms relative to their initial positions stabilized around 0.5 nm, indicating overall structural stability. (B) The RMSF plot shows limited per-residue fluctuations, with elevated flexibility observed primarily in loop regions, consistent with expected dynamic behavior. (C) The Rg of the protein, calculated both overall and along individual spatial dimensions (x, y, and z), remained consistent throughout the simulation, further supporting the maintenance of a compact, folded structure.
The N-terminal domain is involved in the structural stability and assembly of proteins.
(A) The evolutionary conservation of residue positions on the trimethylamine N-oxide demethylase (TDM) structure is estimated based on phylogenetic connections among homologous sequences. (B) N-terminal domain was depicted as forest. Residues ranging from 1 to 20, 150–158 were highlighted. (C) Relative TDM activity of TDM mutants compared to that of wild-type (WT) variant.
Two distinct binding modes between the N-terminal domain (green) and the tetrahydrofolate (THF)-binding domains (blue and light blue).
(A) Interaction interface between N-terminal residues 150–158 and the THF-binding domain. (B) Alternative interaction interface involving N-terminal residues 12–18 with the THF-binding domain.
N-terminal domain facilitates recruitment of the peripheral tetrahydrofolate (THF)-binding domain.
Tables
Cryo-electron microscopy (cryo-EM) data collection, refinement, validation, and statistics.
| PDB EMD | TDM-apo 9Q59 EMD-72223 | TDM:DMA:HCHO:THF 9Q5L EMD-72239 | (DD):TMAO:THF 9Q5A EMD-72224 |
|---|---|---|---|
| Data collection and processing | |||
| Magnification | 105,000 | 105,000 | 105,000 |
| Voltage (kV) | 300 | 300 | 300 |
| Electron exposure (e–/Å2) | 56.8 | 56.8 | 56.8 |
| Defocus range (μm) | 0.8–2.0 | 0.8–2.0 | 0.8–2.0 |
| Raw pixel size (Å) | 0.411 | 0.411 | 0.411 |
| Symmetry imposed | C2 | C2 | C2 |
| Number of initial particle images | 145,414 | 568,305 | 240,010 |
| Number of final particle images | 26,866 | 174,871 | 143,707 |
| Map resolution (Å) | 2.76 | 2.79 | 2.80 |
| FSC threshold | 0.143 | 0.143 | 0.143 |
| Map resolution range (Å) | 2.5–5.0 | 2.5–5.0 | 3.0–5.0 |
| Refinement | |||
| Initial model used | AlphaFold model | Apo model | Apo model |
| Model resolution (Å) | N/A | 2.8 | 2.8 |
| FSC threshold | 0.5 | 0.5 | 0.5 |
| Model resolution range (Å) | N/A | 3.0–50 | 3.0–50 |
| Model composition | |||
| Non-hydrogen atoms | 17,878 | 17,058 | 17,917 |
| Protein residues | 2309 | 2196 | 2306 |
| Ligand: ZN | 2 | 2 | 2 |
| TMO | 2 | ||
| HCHO | 2 | ||
| DMN | 2 | ||
| B factors (Å2) | |||
| Protein | 118.75 | 87.46 | 86.93 |
| Ligand | 144.67 | 89.10 | 90.75 |
| RMSD values | |||
| Bond lengths (Å) | 0.004 | 0.004 | 0.003 |
| Bond angles (°) | 0.586 | 0.608 | 0.501 |
| Validation | |||
| Molprobity score | 2.22 | 2.32 | 2.02 |
| Clash score | 8.74 | 12.12 | 7.78 |
| Poor rotamer (%) | 2.98 | 2.63 | 2.18 |
| Ramachandran plot (%) | |||
| Favored | 94.25 | 93.89 | 95.07 |
| Allowed | 5.49 | 6.01 | 4.80 |
| Outliers | 0.26 | 0.09 | 0.13 |