| Δ3-Δ2-Enoyl-CoA isomerase from the yeast Saccharomyces cerevisiae: Molecular and structural characterization | ||
|---|---|---|
| Prev | Next | |
At the beginning of this project, it was suggested that the ECI1 gene product could be involved in the β -oxidation pathway for three reasons: [1] its amino acid sequence is similar to the hydratase/isomerase superfamily members often involved in acyl-CoA ester metabolism, [2] the promoter region of ECI1 contains a sequence resembling OREs (Einerhand et al. 1993, Filipits et al. 1993), which are involved in upregulating the gene expression when S. cerevisiae cells are grown on fatty acids, such as oleic acid, and [3] ECI1 ends with nucleotides encoding a modified PTS1, which could target the gene product into peroxisomes (Gould et al. 1989), where the β -oxidation in yeast exclusively occurs (Kunau et al. 1988). S. cerevisiae genes encoding peroxisomal β -oxidation enzymes, such as POX1 (Dmochowska et al. 1990), FOX2 (Hiltunen et al. 1992), FOX3 (Einerhand et al. 1991) and SPS19 (Gurvitz et al. 1997a, 1997b), are also upregulated when cells are grown on oleic acid or other fatty acids, and this upregulation is mediated by OREs (Einerhand et al. 1993, Filipits et al. 1993). The transcription factors Oaf1p and Pip2p (Luo et al. 1996, Rottensteiner et al. 1997) bind to the OREs and induce the transcription. Transcriptional analysis of ECI1 showed that the amount of ECI1 mRNA was strongly increased in the presence of oleic acid, whereas no transcription could be detected when cells were grown on glucose or ethanol. Likewise, no ECI1 transcripts could be observed in PIP2- and OAF1-deficient cells grown on oleic acid, indicating that the oleic acid induction is mediated by ORE-bound Pip2p and Oaf1p. This was further verified by EMSA experiments.
The ability of the ECI1-deleted strain to utilize various fatty acids was investigated by culturing mutant cells on both saturated and unsaturated fatty acids. Eci1p was found to be dispensable for the metabolism of saturated fatty acids, but it was required for the degradation of double bonds at both odd- (oleic and arachidonic acid) and even-numbered (octadecenoic and arachidonic acid) positions of unsaturated fatty acids. For the metabolism of even-numbered double bonds, an additional auxiliary enzyme, dienoyl-CoA reductase, sps19p (Gurvitz et al. 1997a) is needed, whereas in the degradation of odd-numbered double bonds, Eci1p can act either directly on cis-3-double bonds or via an alternative route also requiring the dienoyl-CoA isomerase, Dci1p, and the dienoyl-CoA reductase (Fig. 3). In yeast, most of the odd-numbered double bonds are probably metabolized via the isomerase-dependent route, leaving the pathway including Dci1p dispensable (Gurvitz et al. 1999, for a review, see Trotter 2001). The fact that the ECI1-deleted strains could not utilize any of the unsaturated fatty acids tested implies that Eci1p is likely to be the only Δ3-Δ2-enoyl-CoA isomerase in S. cerevisiae.
For characterization, Eci1p was expressed in E. coli and purified to homogeneity. On SDS-PAGE, the subunit size was determined to be 32 kDa, and the native molecular mass was found to correspond to that of a hexamer by gel filtration and dynamic light scattering. Hexameric assembly is a common feature in the hydratase/isomerase superfamily. 2-Enoyl-CoA hydratase-1 (Furuta et al. 1980), dienoyl-CoA isomerase (Filppula et al. 1998) and methylmalonyl-CoA decarboxylase (Benning et al. 2000) are also hexamers in solution. In addition, their subunit sizes are comparable to that of Eci1p.
The active sites of hydratase/isomerase proteins commonly display residual catalytic activities that are related to the actual enzyme activity. For example, the rat Δ3,5-Δ2,4-dienoyl-CoA isomerase contains traces of hydratase-1 activity (Filppula et al. 1998) and the rat hydratase-1 contains some isomerase activity (Kiema et al. 1999). This is likely to be due to the common evolutionary origin of these enzymes. Eci1p was, however, found only to have Δ3-Δ2-enoyl-CoA isomerase activity but no detectable hydratase-1 or dienoyl-CoA isomerase activities. Eci1p is thus a monofunctional Δ3-Δ2-enoyl-CoA isomerase with a specific activity of 11.2 µmol/min/mg (kcat 6.0 s-1) and a Km value of 21.5 µM when trans-3-hexenoyl-CoA is used as substrate. The results obtained in this study were also confirmed by another independent study on ECI1 published almost simultaneously (Geisbrecht et al. 1998). Geisbrecht and co-workers (1998) also found that Eci1p is required for the degration of unsaturated fatty acids and that it possesses only Δ3-Δ2-enoyl-CoA isomerase activity. The enzyme activity measurements were done using cis-3-octenoyl-CoA as substrate. The resulting specific isomerase activity was 16 µmol/min/mg (kcat 8.5 sec-1), which is slightly higher than that obtained with the substrate shorter by two carbon atoms in our experiments. In neither of the studies, however, was substrate specificity with different chain lengths determined. Nevertheless, it can be speculated that because Eci1p is most probably the only Δ3-Δ2-enoyl-CoA isomerase in S. cerevisiae, it has to be able to utilize unsaturated fatty acids of all chain lengths. The results of these two studies also showed that Eci1p can use both cis- and trans-3-enoyl-CoAs as substrates, as was expected. The cis-isoform is more abundant in the unsaturated fatty acids occurring in nature, but the trans-isoform is the product of the 2,4-dienoyl-CoA reductase reaction (Dommes & Kunau 1984), which means that both cis- and trans-3-enoyl-CoAs must be acceptable substrates for enoyl-CoA isomerases. The kinetic parameters of the yeast enoyl-CoA isomerase are compared to those of other monofunctional isomerases (Euler-Bertram & Stoffel 1990, Palosaari et al. 1990, Müller-Newen & Stoffel 1993, Geisbrecht et al. 1999b) and peroxisomal MFE-1 (Palosaari & Hiltunen 1990) in table 2. The kcat values of all the monofunctional enoyl-CoA isomerases, with the exception of the bovine mitochondrial one, are more or less comparable with each other, the activity of the yeast enoyl-CoA isomerase, however, being slightly lower. Nevertheless, it should be noted that variable substrates were used in
Table 2. The kinetic parameters of Δ3-Δ2-enoyl-CoA isomerases.
Source of the Δ3–Δ2-enoyl-CoA isomerase | Substrate | kcat (sec-1) | Km (µM) | Reference |
|---|---|---|---|---|
Rat mitochondrial | trans-3-C6-CoA | 19 | n.r. a | Palosaari et al. 1990 |
cis-3-C12-CoA | 31 | 37 | Müller-Newen & Stoffel 1993 | |
Bovine mitochondrial | cis-3-C12-CoA | 128 | 32 | Euler-Bertram & Stoffel 1990 |
Yeast peroxisomal | trans-3-C6-CoA | 6.0 | 21.5 | this study |
cis-3-C8-CoA | 8.5 | n.r. | Geisbrecht et al. 1998 | |
Human peroxisomal | cis-3-C8-CoA | 18 | n.r. | Geisbrecht et al. 1999b |
Rat MFE-1 | trans-3-C6-CoA | 0.41 | 23.6 | Kiema, Taskinen, Pirilä, Koivuranta, Wierenga & Hiltunen, unpublished |
a n.r., not reported | ||||
the experiments. Interestingly, the isomerase activity of peroxisomal MFE-1 is considerably lower when compared to the monofunctional enzymes.
Multiple-sequence alignments (see original articles I, III and IV) were used to compare the yeast enoyl-CoA isomerase with the other hydratase/isomerase superfamily members and to elucidate the possible active site amino acid residues. Site-directed mutagenesis was used to test whether the candidates acted as catalytic residues. On the basis of sequence alignments, the catalytic residue of yeast isomerase was suggested to be either Tyr148 or Glu158. It was found that Glu158 is possibly the residue involved in the catalytic mechanism of yeast isomerase, since its mutation to alanine made the enzyme totally inactive, whereas the Tyr148Ala mutation had no effect. Glu158 corresponds to the catalytic residues of the rat Δ3,5-Δ2,4-dienoyl-CoA isomerase and 4-CBA-CoA dehalogenase. Neither of the mutations affected the expression level or solubility of the protein. Although yeast isomerase has only low sequence identity compared to the other hydratase/isomerase proteins, often less than 20 %, it was predicted to have the same elements of secondary structure and thus similar overall structure as 2-enoyl-CoA hydratase-1 (Engel et al. 1996). The structure would thus consist of a spiral core domain for catalysis and an α-helical trimerization domain. Like hydratase-1, yeast isomerase would form disk-like trimers that would dimerize into hexamers.
Proteins are targeted into peroxisomes by either PTS1 (Gould et al. 1989) or PTS2 sequences (Swinkels et al. 1991). ECI1 has both a modified PTS1, HisArgLeuCOOH, in its carboxy terminus and a possible PTS2 in the amino-terminal part (Yang et al. 2001). In our study, only the presence of PTS1 was taken into consideration, and its ability to direct yeast isomerase into peroxisomes was studied by tagging the isomerase with GFP at its N-terminus. The fusion protein was indeed found to be located in peroxisomes. The same result was also obtained by Geisbrecht and co-workers (1998). Recent studies have indicated, however, that the peroxisomal localization process of Eci1p is more complicated than it seemed at first (Yang et al. 2001). Yang and co-workers (2001) also discovered the existence of a PTS2, but found that it is not used in peroxisomal targeting. The PTS1 receptor, pex5, however, is indispensable for the peroxisomal import of Eci1p, although the actual PTS1 of the yeast enoyl-CoA isomerase is not (Yang et al. 2001). This leads to the suggestion that Eci1p forms a complex with another peroxisomal protein and uses the PTS1 of the partner for pex5 binding and peroxisomal import. This partner was found to be Dci1p (Yang et al. 2001), which has 45 % sequence identity to Eci1p and is the yeast dienoyl-CoA isomerase (Gurvitz et al. 1999). If Dci1p is deleted from the yeast strain, Eci1p is able to use, although inefficiently, its own PTS1 for peroxisomal import (Yang et al. 2001). In a normal situation, yeast isomerase possibly forms a hetero-oligomer with Dci1p, which is able to enter the peroxisomes.