Ancestral Sequence Reconstruction of the Ethylene-Forming Enzyme

Abstract

The ethylene-forming enzyme (EFE) catalyzes two main reactions: the conversion of 2-oxoglutarate (2OG) to ethylene plus CO₂ and the oxidative decarboxylation of 2OG coupled to the C5 hydroxylation of l-arginine (l-Arg). EFE also facilitates two minor reactions: the uncoupled oxidative decarboxylation of 2OG and the generation of 3-hydroxypropionate (3HP) from 2OG. To better understand the evolution of this enzyme's diverse activities, we demonstrated that two distantly related extant enzymes produce trace levels of ethylene and 3HP, and we examined the reactivities of 11 reconstructed ancestors. The structure of one ancestral protein was resolved by X-ray crystallography, while the others were modeled with AlphaFold2. These studies highlight the importance of residues located at the 2OG and l-Arg binding pockets for the varied activities. For example, effective formation of ethylene requires that the 2OG binding pocket be hydrophobic except for interactions with the substrate carboxylates. Newly identified changes near the l-Arg binding site exhibit significant effects on the reactivities of the enzyme's reactions. Analysis of the reconstructed ancestors suggests that the primordial enzyme exhibited both ethylene-forming and l-Arg hydroxylation activities with partition ratios like the extant examples; i.e., an enzyme capable of catalyzing predominantly one of these reactions did not subsequently develop the ability to affect the secondary reaction.

1

Reactions catalyzed by EFE. (A) The major reaction generates ethylene from 2OG in the presence of l-Arg. (B) Oxidative decarboxylation of 2OG drives l-Arg hydroxylation with subsequent spontaneous degradation to guanidine and P5C. (C) Uncoupled oxidative decarboxylation. (D) 3HP production from 2OG in the presence of l-Arg.

2

Views comparing the active sites of PK2 EFE·Mn­(II)·2OG·l-Arg (PDB ID: 5V2Y, yellow carbon atoms) and PaIPNS·Na (PDB ID: 6JYV, magenta carbons). The views emphasize potential interactions with (A) 2OG and (B & C) l-Arg. Both panels show residues in stick view with N atoms in blue, O atoms in red, and with Mn or Na shown as spheres of the respective color. Panel C also depicts protein regions in cartoon mode.

3

Difference absorbance spectra of PaIPNS demonstrates the binding of 2OG to the active site of the protein. Spectra are shown for PaIPNS·Fe­(II)·2OG (blue) and PaIPNS·Fe­(II)·2OG·l-Arg (green) complexes, with the spectrum of PaIPNS·Fe­(II) taken as the blank. The anaerobic samples contained 500 μM PaIPNS, 2 mM sodium dithionite, 1 mM Fe­(NH₄)₂(SO₄)₂, 1 mM 2OG, and (when present) 1 mM l-Arg in 25 mM HEPES buffer, pH 7.5. The pH of l-Arg and 2OG solutions were adjusted to 7.5 prior to degassing.

4

Structural overlap of PK2 EFE·Mn­(II)·2OG·l-Arg (gray) (PDB: 5V2Y) and the Din11 model. 2OG and residues of Din11 presumed to be near this substrate are indicated in green, while l-Arg and residues likely to be present at its binding site are shown in yellow. The side chains are displayed in stick representation whereas 2OG and l-Arg are represented using a ball and stick model. Nitrogen (N) atoms are in blue and oxygen (O) atoms in red, and Mn ion is shown as a magenta sphere.

5

Phylogenetic analysis indicates the sequence relationships among the extant Pd1 EFE, PK2 EFE, PaIPNS, and Din11 sequences, along with the ancestral sequences generated from both MEGA X and AP-LASR. The maximum likelihood phylogenetic tree was constructed using the Jones-Taylor-Thornton model and the CLUSTALW alignment plugin in MEGA11. Gaps and missing data were eliminated using pairwise deletion, and bootstrap analysis used 1000 replicates. A scale bar indicating the number of substitutions per site is shown at the bottom.

6

Analysis of the ethylene-to-P5C partition ratios for PK2 EFE, Pd1 EFE, and selected reconstructed ancestors, based on the studies using concentrated protein samples. The error bars represent standard deviations from at least two technical replicates.

7

Overall fold and metal binding site of the Anc357 protein. (A) Structure of Anc357·Mn­(II) in cartoon view using chain A. Mn­(II) is shown as a green sphere. β-strands are shown in pink, α-helices in cyan, and other regions in salmon color. (B) Metal binding site of the Anc357·Mn­(II) complex.

8

Superposition of Anc357·Mn­(II) and PK2 EFE·Mn­(II)·2OG·l-Arg (PDB: 5V2Y). (A) The Anc357 protein is gray and PK2 EFE with bound 2OG and l-Arg is cyan, both shown in cartoon mode with selected components depicted as sticks. A short loop in the Anc357 protein (yellow) substitutes for a much longer loop in PK2 EFE (red). (B and C) Active site comparisons with 2OG in yellow and l-Arg in salmon, both depicted in ball-and-stick mode, the PK2 EFE Mn­(II) as a teal sphere, and the Anc357 Mn­(II) shown as a gray sphere. Panel C also shows the altered positions of Arg171 in the PK2 EFE apoprotein (PDB: 5V2U, yellow). The flexibility of PK2 EFE Arg171 does not extend to the modeled position of Arg181 in the Anc357 protein.