Structural basis for catalytic and inhibitory divergence between archaeal and bacterial ammonia monooxygenases

Key results

1. ShownArchaeal AMO from Nitrosopumilus maritimus SCM1 (NmAMO) forms a cup-shaped homotrimer with eight subunits per protomer. Beyond a core shared with bacterial AMO, it carries a newly identified single-helix subunit (AmoE) and an unassigned subunit (AmoH) that closes the cytoplasmic base.

   Native-membrane cryo-EM map at 2.88 Å; 51 transmembrane helices, a slightly larger assembly than bacterial AMO; NmAmoB keeps only its N-terminal cupredoxin domain, which flattens the extracellular face; AmoF/AmoG heterodimers clamp neighbouring protomers; trimers pack in hexagonal arrays in isolated membranes. · Overall structure of archaeal AMO; Image processing

2. ShownBacterial AMO (Nitrosomonas halophila ZJB1, NhAMO) is inhibited directly: DMP (dephosphorylated DMPP) and allylthiourea (ATU) bridge the CuC-CuD dicopper center and set off a residue lid closing the hydrophobic channel entrance; the bound inhibitor also splits the channel into two isolated segments.

   Cryo-EM structures at 2.15 Å (DMP) and 2.47 Å (ATU). CuC-CuD distance is 7.9 Å without inhibitor, 3.9 Å with DMP and 4.7 Å with ATU. Channel volume is 3,135 Å^3 vs 1,774 Å^3 with DMP. Mutating lid residues (R209, E218, R145) lowered ammonia oxidation by ~30%; mutating aromatic-cluster residues (F157, F161, F220, F228) also impaired activity. · Inhibition of bacterial AMO; Figure legends

3. SuggestedArchaeal AMO appears to be inactivated indirectly. In the structure from ATU-treated cells no ATU density is seen at the active site; instead copper has moved from the CuD site to the CuC site, which the authors interpret as inhibitor-triggered copper relocation.

   2.75 Å structure of membrane-bound NmAMO from N. maritimus grown with 1000 μM ATU: same overall fold as the active state, no ATU density at the active site, copper absent at CuD and present at CuC. ICP copper content is unchanged vs the active enzyme. Across five inhibitors, AOB were strongly inhibited while AOA were suppressed only at high concentrations. · Inactivated structure of archaeal AMO; Two distinct pathways of AMO inactivation; Inhibitor activity assays; Archaeal membrane fraction extraction and…

4. ShownThe two AMOs capture ammonium differently. NmAMO lacks the C-terminal cupredoxin domain (its structural counterpart, Nmar_1504, is a separate soluble protein) and binds less ammonium but with higher affinity. This matches higher affinity and lower maximal rates for N. maritimus than for AOB.

   Electronegative periplasmic surface ~50% smaller than NhAMO's. NmAMO cupredoxin domain bound ~30% less ammonium but with higher affinity by isothermal titration calorimetry. N. maritimus had significantly lower Km(app); AOB had higher Vmax and ammonia oxidation rates. Altering NhAMO periplasmic domains reduced activity. Seawater microcosms at 1 μM vs 10 μM NH4+ favoured AOA vs AOB. · Divergent ammonium capture in archaeal and…; Ammonium binding capacity of archaeal and…; Calculation of kinetic properties; Microcosm enrichment assay comparing the…

5. SuggestedCalculations favour CuD as the archaeal catalytic copper and predict that a bacterial dicopper center oxidizes ammonia over lower energy barriers. Simulations give NmAMO a narrower hydrophobic channel than NhAMO.

   Active NmAMO has CuB and CuD occupied and CuC empty. Quantum-mechanical (QM) hydroxyl-amino coupling barrier is 13.6 kcal/mol from CuD vs 22.2 from CuC. On the dicopper path, NH3 hydrogen abstraction costs 9.7 and coupling 11.0 kcal/mol. Molecular dynamics minimum channel diameter is ~5.4 Å vs ~7.1 Å. Exit barrier is 34.9 vs 22.7 kcal/mol (computed for DMP per Extended Data Fig. 13). · Active center in archaeal AMO; Channel architecture and catalytic advantage of…; Umbrella sampling; Figure legends