| Δ3-Δ2-Enoyl-CoA isomerase from the yeast Saccharomyces cerevisiae: Molecular and structural characterization | ||
|---|---|---|
| Prev | Chapter 6. Discussion | Next |
The members of the hydratase/isomerase protein family have a common evolutionary origin and they are seen to be both structurally and mechanistically similar, although they are involved in different metabolic pathways. The similar overall structure is seen in all the known structures within the hydratase/isomerase superfamily. The second spiral turn of the core domain and especially helix H2 are the most variable regions, both in the two isomerase structures determined in this study and in all the other members of the hydratase/isomerase superfamily. The marked structural variability of this region correlates with the sequence variability shown by multiple sequence alignments (see article IV, Fig. 1). For example, the yeast enoyl-CoA isomerase and the rat dienoyl-CoA isomerase (FitzPatrick et al. 1995) have a 13-residue insertion when compared to 2-enoyl-CoA hydratase-1 (Minami-Ishii et al. 1989). Since H2 is involved in shaping the substrate-binding pocket, this variability in its structure enables the binding of very different acyl-chains in the active sites of the enzymes and also facilitates the involvement of hydratase/isomerase family members in so many different pathways. In 2-enoyl-CoA hydratase-1 (Engel et al. 1996) and the dienoyl-CoA isomerase (Modis et al. 1998), helix H2 is split into two parts, as in the neutral pH form of the yeast isomerase. As in the yeast enoyl-CoA isomerase, part of this region in hydratase-1 is flexible and has high B-factors. In the octanoyl-CoA complexed structure of hydratase-1 (Engel et al. 1998), this flexible region (residues 114-118) becomes completely disordered and allows the fatty acyl chain to reach towards the intertrimer space. 2-enoyl-CoA hydratase-1 thus does not have a hydrophobic cavity in the core domain for acyl chain binding, as suggested for the yeast enoyl-CoA isomerase.
When the subunits of these enzymes are superimposed, conformational differences are seen, in addition to helix H2, also in H1, H7, H9 and H10. Helix H9 is the most variable helix in the decarboxylase which also has the lowest sequence identity with the yeast enoyl-CoA isomerase. The changes in H2 and H9 influence the mode of assembly into hexamers.
The most notable difference in the yeast isomerase structure when compared to the structures of 2-enoyl-CoA hydratase-1 (Engel et al. 1996), dienoyl-CoA isomerase (Modis et al. 1998) and 4-CBA-dehalogenase (Benning et al. 1996) is the structural switch of the helices H9 and H10 of the C-terminal domain. This is an example of domain swapping, first described for the seminal ribonuclease dimer (Piccoli et al. 1992) and the diphtheria toxin dimer (Bennett et al. 1994). In the yeast enoyl-CoA isomerase, H9 and H10 are positioned so that they fold over the core domain and cover the active site of the same subunit (Fig. 13). In the structures of hydratase-1, dienoyl-CoA isomerase and dehalogenase, however, H9 and H10 protrude away from the core domain and cover the active site of the neighbouring subunit (Fig. 5). The similar structural switch of H9 and H10 as in isomerase is also seen in the structure of methylmalonyl-CoA decarboxylase (Benning et al. 2000). Despite the domain swapping, in the trimers of the hydratase/isomerase enzymes, the positions of H9 and H10 are equivalent, and the active sites are always covered by the C-terminal domain of either the same or the adjacent subunit.
All the enzymes of the hydratase/isomerase superfamily for which the structure is known form crystallographic hexamers, although the 4-CBA-CoA dehalogenase is a trimer in solution (Benning et al. 1996). In addition, all these hexamers are formed by dimerization of two trimeric disks (Engel et al. 1996, Modis et al. 1998, Benning et al. 2000, Kurimoto et al. 2001).
The variability of the helices H2 and H9, which are close to the intertrimer space, correlates with interesting variability in the mode of assembly of trimers into hexamers in the hydratase/isomerase superfamily. In order to compare the hexameric assemblies, the crystallographic packing information was used to construct hexamers of all the known structures, and the assembly was subsequently analyzed by superposition on the neutral pH form of the yeast enoyl-CoA isomerase. In this analysis, the structure of the AUH protein (Kurimoto et al. 2001) was not used, since it was published after the preparation of the original article IV. The difference in assembly can be described by two parameters: a rotation (Κ ) of the trimers with respect to each other and a shift (Δ) along the axis of rotation, indicating the difference in the distance of the two trimers compared to the neutral pH form isomerase hexamer (original article IV, table 2, Fig. 4). It was found that the orientation of the trimeric disks with respect to each other is not conserved, nor is the distance between the disks. The trimers in hydratase-1 (Engel et al. 1996) and the dienoyl-CoA isomerase (Modis et al. 1998) have the same relative orientation with respect to each other, but the distance between the trimers is not the same. This was also found for the low pH form of isomerase and dehalogenase (Benning et al. 1996). A relatively short distance between the trimers is observed for the dehalogenase, although it occurs as a trimer in solution, whereas the longest distance is observed for the decarboxylase, although this enzyme is a hexamer in solution (Benning et al. 2000). The neutral pH form of the yeast isomerase is the most compact structure with the largest number of intertrimer contacts and the shortest distance between the trimeric disks. The low pH form isomerase, however, has the lowest number of contacts. Only a few contacts are also found for the trimeric dehalogenase. Interestingly, the variable helices H2 and H9 are always involved in intertrimer contacts, in many cases also H1. Although, at the monomer level, the overall fold of the enzymes of the hydratase/isomerase superfamily resemble each other and the secondary structure elements are mainly conserved, the mode of assembly of trimer disks into hexamers is clearly not a conserved feature of this enzyme superfamily.