Fig. 1. The reaction catalyzed by MECR/ETR1. MECR/ETR1 reduces the (2E) double bond of the substrate. During the reaction cycle, the H⁻ and H⁺ are transferred to the β- and α-carbons, respectively. MECR/ETR1 accepts fatty acyl groups that are covalently bonded to either CoA or ACP.

Fig. 2. Sequence alignment of 2-enoyl thioester reductases from Homo sapiens (HsEtr), Rattus norvegicus (RnEtr), Mus musculus (MmEtr), Bos taurus (BtEtr), C. tropicalis (CtEtr1p and CtEtr2p), and Saccharomyces cerevisiae (ScEtr1p). The secondary structures of the crystal structure of the human enzyme are shown above the sequences; the bars indicate α-helices and arrows denote β-strands. The secondary structure elements are colored blue and cyan for the catalytic and cofactor binding domains, respectively. In the alignment, the region of the cofactor binding domain is colored cyan as well. The residues conserved are highlighted in black, and the others in a region with a high degree of sequence identity (85%) are highlighted in dark gray. Ser85, Tyr94, Thr170, and Trp311 are the residues in the active site that were mutated in this study; they are highlighted in red. The residues forming a hydrogen bond with the modeled NADPH are colored green, and the residues in blue are putative interaction sites with the recognition helix of the substrate–ACP molecule. The orange bars above the human sequence indicate the residues that line the fatty acyl thioester binding pocket; three of these residues (Ile129, Gly165, and Phe324, colored pink) were mutated to study the chain length preference. The asterisk stands for the N-terminus of the recombinant HsEtr (MECR/ETR1). The X and Y regions are not present in mammalian enzymes. Every 10th residue has been indicated by a black rectangle. The alignment was prepared with the ClustalX program.²⁰

Fig. 3. The structure of human recombinant MECR/ETR1. (a) Quaternary structure of MECR/ETR1. The structure consists of two identical monomers and they are colored differently. The arrow and αd′ indicate the putative interaction site of the recognition helix of a substrate–ACP molecule (Fig. 7). (b) Tertiary structure of the MECR/ETR1 monomer. The catalytic domain is colored blue and the cofactor binding domain is colored cyan. The color code for the domains is the same as that used in Fig. 2. The β-strands and α-helices of the cofactor binding domain are indicated by roman numerals (I–VI) and small letters (a–f), respectively. The β# is not part of the classical Rossmann fold topology. The nomenclature of the catalytic domain is done by numbers (1–9) and capital letters (A–G). NH₂ and COOH label the N-terminus and C-terminus, respectively. In both panels, the asterisk indicates the position of Tyr94 in the active site.

Fig. 4. Comparison of the human sulfate–MECR/ETR1 (this study), the modeled complex of MECR/ETR1 with NADPH (this study), and the NADPH liganded C. tropicalis Etr1p (PDB entry code 1GUF) using the superposition protocol as described in Materials and Methods. (a) Sulfate–MECR/ETR1 (blue) on CtEtr1p (red). The overall fold of the crystal structure of human sulfate–MECR/ETR1 resembles the liganded, closed form structure of the CtEtr1p.¹⁵ In MECR/ETR1, however, some of the loops of the NADPH binding site (βII–αc and βV–#β) are positioned as observed with the apo structure of CtEtr1p. The X and Y are the loop regions that do not exist in mammalian 2-enoyl thioester reductases (Fig. 2). The NADPH of CtEtr1p is shown as atom-colored skeletons. (b) Sulfate–MECR/ETR1 (blue) on the modeled NADPH–MECR/ETR1 complex (pink). The βII–αc loop consists of two arginines, Arg216 and Arg218, which interact with the adenine end of the NADPH. The βV–β# loop contains the TYGGM motif. NH₂ and COOH label the N-terminus and C-terminus, respectively. (c) The zoomed-in view of the active site of sulfate–MECR/ETR1 together with the modeled NADPH. The omit (F_o − F_c)α_c map, calculated after omit refinement (leaving out the three sulfates of the model), is contoured at 3 σ.

Fig. 5. The pantetheine binding cleft and the fatty acyl binding pocket in the catalytic domain of sulfate–MECR/ETR1. The view is from the top of the catalytic domain down towards the cofactor binding domain. S3 is the sulfate molecule that is bound at the entrance of the pantetheine binding cleft. The modeled NADPH is shown as atom-colored skeletons. (a) The monomers of MECR/ETR1 are colored blue and red. The regions of the catalytic domain that lines the fatty acyl binding pocket are colored yellow. (b) Color-coded molecular surface of the structure in (a) (atom color coding: oxygen is red, nitrogen is blue, carbon is white, and sulfur is yellow). The fatty acyl binding pocket extends from the catalytic site near the nicotinamide group of NADPH and Tyr94 (shown as 94) towards Z, which labels the end of the pocket. In this pocket, six water molecules (red spheres) were found in the electron density map. The pantetheine binding cleft stretches from the catalytic site to the right, near S3. The βV–β# loop is between the NADPH binding site and the pantetheine binding cleft. (c) Zoomed-in view of the fatty acyl binding pocket. The residues that contribute to the fatty acyl binding pocket, including Ile129, Gly165, and Phe324, which were mutated into methionine, serine, and tyrosine, respectively, are shown (Table 4). Also shown are the six water molecules that are bound in this pocket.

Fig. 6. The active site of the human sulfate–MECR/ETR1 structure. The residues that were mutated in this study (Ser85, Tyr94, Thr170, and Trp311) are shown and labeled. The figure also shows the modeled mode of binding of NADPH (green carbon atoms) and NADPH as bound to CtEtr1p (1GUF, gray carbons). During the reaction cycle, the hydride of C4 of NADPH is transferred to the β-carbon atom of the substrate. The distance between the C4 atom and the OH of the catalytic moiety is 7 Å.

Fig. 7. The substrate–ACP interaction groove. The groove next to the entrance of the pantetheine binding cleft is the putative binding site for the recognition helix of human ACP. In each panel, the same view is used and the arrow points to the same interaction site. (a) The surface presentation of the sulfate–MECR/ETR1 dimer. (b) The molecular surface around the interaction site. Oxygens are marked with red, nitrogens with blue, and carbons with white. The sulfur atoms are colored yellow. (c) Same view as in (b) except that the secondary structures are shown. The fatty acid binding pocket extends from the catalytic site (Tyr94) towards Z, and the pantetheine binding cleft extends from the catalytic site to the right towards the sulfate molecule, S3. The basic residues (Lys248, Lys252, Arg274, Arg278, and Lys303) in the groove are proposed to interact with the recognition helix of ACP. In (a) and (c), monomer A is colored blue and monomer B is colored red. The modeled NADPH is shown as atom-colored skeletons. M identifies the βV–β# loop.

Data collection and refinement statistics

Kinetic parameters of the wild-type MECR/ETR1

The kinetic constants have been measured for a range of (2E)-enoyl–CoA substrate molecules, in the presence of 125 μM NADPH, as described in Materials and Methods. The values expressed in the second and third columns are means ± SD of three determinations.

Kinetic parameters of the wild type and MECR/ETR1 variants, with mutations near the catalytic cavity

The kinetic constants were measured using (2E)-octenoyl–CoA as substrate, in the presence of 125 μM NADPH, as described in Materials and Methods. The values expressed in the second and third columns are means ± SD of three determinations.

Kinetic parameters of the wild type and the pocket variants having mutations lining the substrate binding pocket

The kinetic constants have been measured for a range of (2E)-enoyl–CoA substrate molecules, in the presence of 125 μM NADPH, as described in Materials and Methods. The values expressed in the K_m and k_cat columns are means ± SD of three determinations. The chain length preference is the C14/C8 and the C16/C8 ratio of the k_cat/K_m values of the respective variants.