A refined picture of the native amine dehydrogenase family revealed by extensive biodiversity screening

Abstract

Native amine dehydrogenases offer sustainable access to chiral amines, so the search for scaffolds capable of converting more diverse carbonyl compounds is required to reach the full potential of this alternative to conventional synthetic reductive aminations. Here we report a multidisciplinary strategy combining bioinformatics, chemoinformatics and biocatalysis to extensively screen billions of sequences in silico and to efficiently find native amine dehydrogenases features using computational approaches. In this way, we achieve a comprehensive overview of the initial native amine dehydrogenase family, extending it from 2,011 to 17,959 sequences, and identify native amine dehydrogenases with non-reported substrate spectra, including hindered carbonyls and ethyl ketones, and accepting methylamine and cyclopropylamine as amine donor. We also present preliminary model-based structural information to inform the design of potential (R)-selective amine dehydrogenases, as native amine dehydrogenases are mostly (S)-selective. This integrated strategy paves the way for expanding the resource of other enzyme families and in highlighting enzymes with original features.

Fig. 1. Global strategy for discovering AmDHs among the biodiversity.

This includes five main steps: 1) environmental sampling to define the reference AmDH family (ref-AmDHs) and the NAD(P)-dependent enzyme pool, 2) sequence clustering, 3) cluster analysis including the search for distant homologs by HMM-HMM profile comparison, 4) selection of candidate enzymes and 5) production and in vitro tests of selected enzymes.

Fig. 2. Overview of the ref-AmDHs sequence space.

Phylogenetic trees (removing redundancy at 95% identity on 90% coverage) of (A) native AmDHs (nat-AmDHs, from ref. , 3032 sequences) and (B) extended set of nat-AmDHs (ref-AmDHs, this work, 7,620 sequences) with resulting G1–G5 groups. Colored bars indicate the proteins that were successfully modeled and classified using the ASMC method and purple triangles correspond to the 122 AmDHs tested experimentally in this work. The number of sequences in each group is indicated.

Fig. 3. Diversity of active sites from the ref-AmDHs family.

A Residues P1–P21 are considered in this study, including the critical catalytic glutamate (P3) and the residue now at position P5 (green), absent from the Mayol et al. analysis. Top: CfusAmDH active site (PDB ID: 6IAU). For greater clarity, only residues closest to the active site pocket (orange mesh) are shown. Bottom: Active site sequences of CfusAmDH compared to AmDH4. For consistency, coloring refers to the WebLogo3 “chemistry” color scheme as described below. B Hierarchical tree of the 9763 ref-AmDH active sites, made by the ASMC pipeline. Crystallographic structures used in this work are indicated in their respective ASMC groups. Each sequence logo represents the conservation of the P1–P21 residues. Logos were made using WebLogo3 and its “chemistry” color scheme [green: polar, purple: neutral, blue: basic, red: acidic, black: hydrophobic (charges at physiological pH)].

Fig. 4. Phylogenetic tree of representatives of the extended nat-AmDH family experimentally tested and their detected activities towards substrates 1a-5a.

Tested reference enzymes (CfusAmDH, MsmeAmDH, PortiAmDH and AmDH4) are indicated in red. Active nat-AmDHs previously reported but not tested in this screening assay are indicated in blue. Bootstrap values > 80% are indicated with purple circles. Analytical yields in 1b-5b from tested substrates 1a-5a are shown as black triangles with a size gradient that ranges between 0 and 10 mM.

Fig. 5. Substrates and products discussed in this work.

The different ketones, aldehydes (1a-13a) and corresponding amines (1b-13b and 2c-2e) are represented.

Fig. 6. PyMOL visualization of some nat-AmDHs active site.

A Active site cavities of CfusAmDH (left, PDB ID: 6IAU, chain B and METDB-03 (middle, CfusAmDH-based homology model). Their superimposition (right, RMSD = 0.21 Å) highlights the larger cavity of METDB-03, due to the F140/A151 replacement and the W145/W156 displacement, compared to CfusAmDH; B) Positions P5 (italic) and P9 (bold) responsible for pocket enlargement of MGYP000211951848 (right—His135, Ala161) relative to CfusAmDH (left—Phe140, Thr166) and A0A229HGK2 (middle—His145, Leu171). For the sake of clarity, only one correct conformation of docked (3S)-heptan-3-amine (white) for each enzyme is shown.

Fig. 7. Analytical yields in N-alkylamines 2c-2e.

Reactions conditions: 10 mM substrate, 200 mM TRIS.HCl buffer pH 9.0, 250 mM amine donor c-e (or 20 mM), 0.2 mM NADP + , 0.2 mM NAD + , 11 mM glucose, 3 U ml⁻¹ GDH-105, 1.0 mg ml⁻¹ purified enzyme, 24 h, 30 °C. Amounts of amines 2c-2e were obtained after derivatization with BzCl and UHPLC-UV analysis (conditions 2) (see Methods). Bars represent the average of values obtained with two independent experiments (n = 2; dot plots) for the reaction of 2c with 20 mM of c (light blue), 2c with 250 mM of c (dark blue), 2d with 250 mM of d (green) and 2e with 250 mM of e (yellow). Chromatogram and calibration curves are given in Supplementary Fig. 17. Source data are provided as a Source Data file.