Discovery of a non-canonical prototype long-chain monoacylglycerol lipase through a structure-based endogenous reaction intermediate complex

Abstract

The identification and characterization of enzyme function is largely lacking behind the rapidly increasing availability of large numbers of sequences and associated high-resolution structures. This is often hampered by lack of knowledge on in vivo relevant substrates. Here, we present a case study of a high-resolution structure of an unusual orphan lipase in complex with an endogenous C18 monoacylglycerol ester reaction intermediate from the expression host, which is insoluble under aqueous conditions and thus not accessible for studies in solution. The data allowed its functional characterization as a prototypic long-chain monoacylglycerol lipase, which uses a minimal lid domain to position the substrate through a hydrophobic tunnel directly to the enzyme's active site. Knowledge about the molecular details of the substrate binding site allowed us to modulate the enzymatic activity by adjusting protein/substrate interactions, demonstrating the potential of our findings for future biotechnology applications.

Fig. 1. Biophysical and structural properties of Tth MAG lipase.

a Size exclusion chromatograms indicating dimeric (D) and monomeric (M) fractions and b nano differential scanning fluorimetry profiles of selected Tth MAG lipase variants. For further analysis, see Supplementary Table 1. c topology diagram of the molecular structure of the Tth MAG lipase protomer. CD, yellow; lid domain, pink. Secondary structural elements are in proportion to their length and are labeled. The positions of key active site residues (cf. Supplementary Fig. 5) are highlighted. The active site glycerol interacting residue Glu43 is shown in red. While in the text three-letter codes were used for residue specifications, for the reason of clarity one-letter codes were used in all Figures. d surface presentation of Tth MAG lipase protomer, with the dimeric interface surface highlighted in stronger colors. Polar residues that contribute to specific interactions to the dimeric interface are shown in red (glutamate, aspartate), blue (arginine, lysine), and pale blue (glutamine), and are labeled. Glu72, which was mutated into an arginine for probing the dimeric interface, is highlighted by an asterisk. The surface of the active site residue Ser113 is shown in black, as point of reference for locating the active site. e Tth MAG lipase dimer in surface presentation. The catalytic domains (CD) and lid domains of each protomer are labeled and colored separately. The two Tth MAG lipase protomers form a large joint substrate binding site, as indicated by the bound C18 MAG ligands (Fig. 2a) and two asterisks in red, indicating the two active sites. In this presentation, the proximal part of the C18 MAG ligands is visible (cf. Fig. 2). The orientation of Tth MAG lipase protomer A is rotated by about 90 degrees around a vertical axis with respect to the Tth MAG lipase monomer shown in panel (d), as indicated. Source data are provided as a Source Data file.

Fig. 2. Tth MAG lipase active site and lid domain tunnel for long chain MAG esters.

a Surface presentation of the Tth MAG lipase long-chain MAG binding site, in part generated by a tunnel across the lid domain. Bound C18 MAG intermediate is shown in green. Atom-specific colors are used for oxygen (red) and nitrogen (blue). Every fifth methylene group of C18 MAG is numbered. Those parts of the C18 MAG that are sequestered by the lid domain tunnel (C11–C18) appear to be darker due to the semi-transparent Tth MAG lipase surface. Side chains of key residues contributing to ligand binding and the Tth active site are shown and labeled. Residues used for probing function are labeled with an asterisk. The central part of the active site is zoomed in a box to the right, indicating specific polar interactions of the glycerol product with Tth MAG lipase Glu43. b (Upper inlet), active site of the PMSF-inhibited Tth MAG lipase structure for comparison. c Mesh surface representation of the C18 MAG lid domain tunnel from the top of the lid domain, indicating that it is open at the most distal end (left panel), side view of the lid tunnel (central panel) similar to the orientation in (a). Right panel: for comparison, part of the surface generated by residues Phe37 from the CD and Tyr154 from the lid domain is shown for the distinct conformation of the PMSF-inhibited Tth MAG lipase structure, due to a conformational change of Tyr154, which is labeled in red. This change leads to blockage of the lid tunnel found in the Tth MAG lipase structure without PMSF. d Superposition of Tth MAG lipase lid domain structures, without (violet) and with PMSF (gray). e Surface presentation of Tth MAG lipase lid domain structures, without (violet, left panel) and with PMSF (gray, right panel). Residues involved in the formation of the proximal lid substrate tunnel opening and polar residues forming an outer half-ring wrapping the substrate tunnel exit are labeled. The proximal end of the C18 MAG reaction intermediate bound to the Tth MAG lipase structure without PMSF is shown in green. Comparison of the two conformations shows that the lid tunnel is only opened in the Tth MAG lipase structure without PMSF, when C18 MAG ester ligand is present. For reasons of optimizing illustration of the lid tunnel the orientation is different from the one in panel (b).

Fig. 3. Tth MAG lipase reaction scheme.

Reaction intermediates are indicated. Color codes: acyl group, pink; glycerol group, blue; reactive water, green; side chain of Glu43, red. The ligand density found C18 MAG ester reaction intermediate found in the Tth MAG lipase structure active site (Supplementary Fig. 3) is closest to loosely coordinated tetrahedral transition intermediate 1 (boxed).

Fig. 4. Tth MAG lipase activities for MAG esters.

a Tth MAG lipase activities for MAG esters of variable length. Coloring scheme of esters characterized in vitro assays (b–d). MAG esters found in cell extracts (f) and by structural data (Fig. 2, Supplementary Fig. 3) are also indicated, demonstrating complementarity of different experimental approaches. b Left panel: in vitro turnover of C8 MAG, DAG and TAG esters and C18:1 MAG. b Right panel: C8 MAG turnover of different Tth enzyme variants. c Tth MAG lipase activity for several pNP acyl esters of C2-C18 FAs and ferulate, indicating a broad range of possible substrates. d Steady-state Michaelis-Menten analysis for MAG esters with quantifiable activity profiles. Left panel, apparent substrate binding affinities for C2-C14 MAG esters measured by Michaelis constants K_M, indicating increase of binding with increasing MAG ester length; central panel, catalytic rates (k_cat) for C2-C14 MAG esters, showing an overall profile comparable to that measured for overall substrate turnover (central panel, cf. panel b, left); right panel, catalytic efficiencies k_cat/K_M for the same substrate spectrum, showing only minor variations for most of the MAG esters used (C4-C12). e Left panel: experimental steady-state kinetic data of selected Tth variants (wt, Y154A, E43A) to estimate Michaelis-Menten values for C6 MAG esters. f Fatty Acid Methyl Esters (FAMEs) analysis by high-performance thin-layer chromatography (HPTLC) of E. coli total lipid extracts (TLE) expressing Tth MAG lipase. Left panel, 1, Tth MAG lipase (wt); 2, Tth MAG lipase (S113A); 3, Tth MAG lipase (E43A). Right panel: FAMEs quantification for selected Tth MAG lipase mutants, normalized to Tth MAG lipase (wt). Data statistics (b–e): Data are presented as mean values and individual data points, n = 3–4 (technical replicates). Data statistics (f): Data are presented as means ± SD, and individual data points. Error bars represent the standard deviation from biologically independent replicates (n = 3 for variants S113A, E43A, Y154R; n = 4 for variants E43K, Y154A; n = 6 for WT; and n = 7 for ΔLid). Unpaired t-test was used to evaluate statistical significance applying p values for a two-tailed confidence interval as follows: p < 0.001 (***), p < 0.01 (**), p < 0.05 (*). Definition p values: p = 0.0055 ΔLid, p = <0.0001 S113A, p = 0.0288 E43A, p = 0.0013 E43K, p = 0.4985 Y154A, p = 0.1285 Y154R. Source data and uncropped image are provided as a Source Data file.

Fig. 5. Sequence, structural, and functional relations of esterase/lipases with divergent HBH lid domains.

a Scheme of HBH lid domain sequences related to Tth MAG lipase (Fig. 1c) with secondary structural elements indicated proportional to sequence length. The two-stranded HBH lid β-hairpins are in cyan, sequence, and secondary structural elements inserted between strands βL1 and βL2 are in green. The additional two-stranded β-hairpin structure hovering over the active site in the esterase from L. johnsonii (PDB code: 3S2Z) is shown in dark blue. The related β-hairpin-like structure in the esterase from B. proteoclasticus (PDB code: 2WTM) is shown in the same color with a diagonal stripe pattern. The lid structure of the Osmotically inducible protein C from R. marinus (PDB code: 5CML) comprises an additional third β-strand colored in dark violet, which interacts with the first β-strand of the HBH β-hairpin. Human MAG lipase (3HJU), which does not have a HBH lid motif and presents a canonical MAG lipase, is shown for comparison as well. Further details are summarized in Supplementary Table 8. b HBH lid structures of three representative esterase/lipases with their PDB codes indicated (Supplementary Fig. 12b), demonstrating common elements (N- and C-terminal α-helices, two-stranded β-hairpin) and diversity of the inserted sequences in terms of length and structure. The CDs of these structures are shown by surface presentation in light orange, and the HBH domains are illustrated by cartoon presentation. The surface covered by the catalytic triad serine (Ser113 in Tth MAG lipase) is shown in red, as a point of reference to locate the active site. Active site ligands are shown in sphere presentation in atom-specific colors (carbon, green; oxygen, red; nitrogen, blue). Only in Tth MAG lipase, the C18 ligand crosses the lid domain via a hydrophobic tunnel. The ligands in 3LLC (tetraethylene glycol) and 4KE8 (tetradecyl hydrogen (R)-(3-azidopropyl)phosphonate, chain A) are in similar positions next to the active site but they are too short to cross the respective lid domains (Supplementary Fig. 4a).

Fig. 6. Structural comparison of the ligand bound HBH lid domains and active site area, indicating different substrate specificities.

a Tth MAG lipase in the presence of C18 MAG (left); b esterase from L. johnsonii (PDB code: 3S2Z) in the presence of caffeine (right). Two residues critical for specific hydrogen bond interactions with ferulate in the structure of the esterase from L. johnsonii (Asp138, Tyr169) are substituted by hydrophobic residues in Tth MAG lipase (equivalent residues Leu145, Phe175). Other key residues with critical roles for substrate binding are also shown. In the structure of the esterase from L. johnsonii, an additional N-terminal lid β-sheet is formed (colored in blue) that hovers over the CD active site. At the end of the short turn connecting the two β-strands, there is a highly exposed residue, which is Gln145 in this structure. Residues that are part of the conserved HGF motif (residues 35-37 in Tth MAG lipase, residues 32–34 in the L. johnsonii esterase) and less conserved FSE/D motif (residues 77–79 in Tth MAG lipase, residues 70–72 in in the L. johnsonii esterase) are labeled in red (Supplementary Fig. 2).