CYP74C3 and CYP74A1, plant cytochrome P450 enzymes whose

8th International Symposium on Cytochrome P450 Biodiversity and Biotechnology

CYP74C3 and CYP74A1, plant cytochrome P450 enzymes whose activity is regulated by detergent micelle association, and proposed new rules for the classification of CYP74 enzymes R.K. Hughes1 , E.J. Belfield and R. Casey John Innes Centre, Norwich Research Park, Norwich NR4 7UH, U.K.

Abstract CYP74C3 (cytochrome P450 subfamily 74C3), an HPL (hydroperoxide lyase) from Medicago truncatula (barrel medic), and CYP74A1, an AOS (allene oxide synthase) from Arabidopsis thaliana, are key membraneassociated P450 enzymes in plant oxylipin metabolism. Both recombinant detergent-free enzymes are monomeric proteins with dual specificity and very low enzyme activity that can be massively activated with detergent. This effect is a result of the formation of a complex between the protein monomer and a single detergent micelle and, in the case of CYP74A1, has a major effect on the substrate specificity of the enzyme. Association with a detergent micelle without an effect on protein oligomeric state represents a new mechanism of activation for membrane-associated P450 enzymes. This may represent a second unifying feature of CYP74 enzymes, in addition to their known differences in reaction mechanism, which separates them functionally from more classical P450 enzymes. Highly concentrated and monodispersed samples of detergent-free CYP74C3 and CYP74A1 proteins should be suitable for structural resolution. On the basis of recent evidence for incorrect assignment of CYP74 function, using the current rules for CYP74 classification based on sequence relatedness, we propose an alternative based on substrate and product specificity for debate and discussion.

Introduction Members of the CYP74 (cytochrome P450 subfamily 74) have not been studied extensively. CYP74 enzymes are very different from classical P450 enzymes of microbial [1] or mammalian [2] origin, in that they have an atypical reaction mechanism that requires neither oxygen nor an NADPH-reductase [3], and consequently have extremely high catalytic-centre activities. In this sense, they have more in common with non-classical mammalian P450 enzymes such as thromboxane synthase [4]. CYP74 enzymes all use fatty acid hydroperoxides as substrates and were originally classified on the basis of their product specificity rather than sequence relatedness. Thus CYP74A refers to AOS (allene oxide synthase) and CYP74B to HPL (hydroperoxide lyase) [5]. HPL cleaves hydroperoxides, formed from the oxygenation of PUFAs (polyunsaturated fatty acids) by the action of LOX (lipoxygenase), into an array of volatile and non-volatile products, some of which have antibacterial properties [6] and so are important in plant defence. The volatile aldehydes are Key words: allene oxide synthase, cytochrome P450 subfamily 74 (CYP74), detergent micelle, haem, hydroperoxide lyase, oxylipin. Abbreviations used: AOS, allene oxide synthase; CYP74, cytochrome P450 subfamily 74; DES, divinyl ether synthase; Emulphogene, polyoxyethylene 10 tridecyl ether; HPL, hydroperoxide lyase; 9-HPODE, 9-S-hydroperoxyoctadeca-10E,12Z-dienoic acid; 13-HPODE, 13-Shydroperoxyoctadeca-9Z,11E-dienoic acid; 9-HPOTE, 9-S-hydroperoxyoctadeca-10E,12Z,15Ztrienoic acid; 13-HPOTE, 13-S-hydroperoxyoctadeca-9Z,11E,15Z-trienoic acid; LOX, lipoxygenase; PUFA, polyunsaturated fatty acid; RZ, Reinheitszahl. 1 To whom correspondence should be addressed (email [email protected]).

of great value to the food industry [7] and are generated on an industrial scale using vegetable oils rich in linoleic acid and linolenic acid as a source of PUFA, defatted soya-bean flour as a stable source of LOX and various sources of HPL [8–10]. HPL is present at only very low levels in plants and is relatively unstable [9], so the subsequent conversion of the hydroperoxides into volatile aldehydes is usually carried out by a recombinant HPL present in crude Escherichia coli [8] or yeast [10] extracts; the latter would be more costly to manufacture and the HPL activity in the crude E. coli extract is stable only for 1 month at 4◦ C [8]. The use of a purified HPL would seem to help matters and has been shown to yield 20–85 times more aldehydes per gram of protein in a shorter reaction time than the same protein in chloroplastenriched and crude plant extracts [9]. The stability of HPL activity and protein from any source is evidently a current problem in biocatalysis, but could be overcome to a large extent by the development of a purified recombinant HPL that remains stable and active in the longer term, both on the shelf and during the reaction. All HPLs are membrane-associated and require detergent for extraction and solubilization. It has been difficult to resolve the detergent and protein interactions, and consequently there is some disagreement about the oligomeric state of HPL purified from a number of higher plants including, for example, guava [11], bell pepper [9,12] and tomato [13]. The enzyme has been reported to be either trimeric or C 2006

Biochemical Society

1223

1224

Biochemical Society Transactions (2006) Volume 34, part 6

tetrameric; the oligomeric state of various recombinant HPLs: CYP74B1 [14], CYP74B2 [15], CYP74B3 [16], CYP74B4 [17], CYP74B5 [11], CYP74C1 [18] and CYP74C2 [19], and the effects of detergent removal, are hardly ever reported. The effects of detergent on increasing the activity of HPL are well documented (see [20]) but the molecular mechanism responsible for this activation were unknown. There is no known mammalian equivalent of HPL. HPL has the same substrate specificity as AOS. Unlike HPL, which cleaves hydroperoxides, AOS transforms them into unstable fatty acid epoxides which are then metabolized further by enzymatic processes to jasmonates that are important in the signalling of plant defence responses; the mammalian equivalent of AOS is prostaglandin endoperoxide H synthase [4]. With one exception, that of AOS purified from flax seed [21], the oligomeric state of AOS purified from a number of higher plants, including guayule [4,22] and corn [23], and of recombinant AOSs from barley [24], tomato [5,16] and Arabidopsis [25] has not been reported because, like HPL, it has proved difficult to resolve the detergent and protein interactions. Gel filtration analysis of AOS purified from flax seed in the presence of detergent confirmed that the protein remained as a monomer of molecular mass 55 kDa [21], which suggested that there was no association with detergent micelles and that the protein was entirely watersoluble. In the same work, however, it was reported that the specific activity of the enzyme was enhanced 2–3-fold by detergent, but the molecular mechanism responsible for this activation is unknown. The molecular mechanisms and primary determinants of the differences in CYP74 product specificities are also unknown, primarily because to date there is no published structure of any plant cytochrome P450 enzyme. Detergent micelles have been shown to modify the activities of classical P450 enzymes through an effect on oligomeric state. Thus, in the presence of 10 mM n-octyl glucoside, a complete loss of catalytic activity of CYP2B4 with reductase was observed due to disaggregation of the active pentamer or hexamer into inactive monomers [26]. The effects of the detergent Emulphogene (polyoxyethylene 10 tridecyl ether) on CYP2B4 (and CYP1A2) activity are, however, contradictory (see [27]) with reports of both loss and gain of catalytic activity due to the formation of monomers or dimers respectively. We have expressed in milligram quantities and fully characterized two plant cytochrome P450 enzymes: CYP74C3 (an HPL from Medicago truncatula) [28] and CYP74A1 (an AOS from Arabidopsis thaliana) [29] in the presence and absence of detergent. We first present new and unexpected information on their substrate and product specificities and oligomeric states. Secondly, we describe the ability of both proteins to exhibit activation kinetics through association with a micelle of Emulphogene without an apparent change in oligomeric state. Thirdly, we discuss the production of highly concentrated and monodispersed samples of detergent-free CYP74C3 and CYP74A1 proteins which may be well suited for the purposes of crystallization and structural resolution. C 2006

Biochemical Society

Finally, reasons are outlined for the introduction of a new classification system for CYP74 enzymes based on function rather than sequence relatedness.

CYP74C3 and CYP74A1 are monomeric proteins with dual specificity Analysis of the oligomeric state of CYP74A1 (Figure 1a) and CYP74C3 (Figure 1b) in the absence of detergent has demonstrated for the first time that both CYP74 enzymes are highly water-soluble and monomeric proteins of molecular mass approx. 55 kDa. The enzymes were purified to homogeneity [RZ > 1.3; CYP74C3 [28]; and RZ > 1.1; CYP74A1 [29]; where RZ is Reinheitszahl (or purity index)] with a full complement of haem iron. The catalytic efficiency (kcat /K m ) of the detergent-free proteins (Table 1) compared with the same proteins isolated without detergent removal [28,29] was, however, very low and suggested that the presence of detergent micelles was essential to maintain the most active conformation of both proteins. Phylogenetic analysis of CYP74C3 suggested that it had dual specificity [6] and this was confirmed by GC– MS analysis of the products in reactions of the purified recombinant protein with 9-HPODE (9-S-hydroperoxyoctadeca-10E,12Z-dienoic acid), 9-HPOTE (9-S-hydroperoxyoctadeca-10E,12Z,15Z-trienoic acid), 13-HPODE (13-S-hydroperoxyoctadeca-9Z,11E-dienoic acid) and 13HPOTE (13-S-hydroperoxyoctadeca-9Z,11E,15Z-trienoic acid) [28]. Phylogenetic analysis of CYP74A1 indicated that it was a monospecific 13-AOS [5], and this was apparently confirmed in a previous work using recombinant enzyme [25], although no comments on activities with 9-HPODE or 9-HPOTE were reported in this study. Recent work in our laboratory has revealed, however, that CYP74A1 has major activity with 9-HPODE and 9-HPOTE and exhibits dual specificity [29].

The biological activity of CYP74C3 and CYP74A1 is regulated by monomer–micelle association Analysis by gel filtration in the presence of detergent micelles of the oligomeric states of CYP74A1 and CYP74C3 indicated that CYP74A1 formed almost entirely a complex of approx. 110 kDa (Figure 1c); CYP74C3 formed a number of higher oligomers, but the most active species was a complex of similar size to that formed by CYP74A1 (Figure 1d). Both these complexes were identified as a protein monomer complexed with a single detergent micelle of molecular mass approx. 62 kDa [28,29]. Micelle binding increased the kcat /K m of CYP74C3 and CYP74A1 with the preferred substrate 13HPOTE by approx. 50-fold (Table 1). The increase in kcat /K m of up to 1.6 × 108 and 5.9 × 107 M−1 · s−1 for CYP74C3 and CYP74A1 respectively is especially remarkable for CYP74C3 whose reaction mechanism involves the scission of a C–C bond. Micelle-induced changes in the catalytic efficiencies of both proteins were accompanied by a shift in equilibrium


Figure 1 CYP74C3 and CYP74A1 are monomeric proteins that bind a single detergent micelle Purified CYP74A1 or CYP74C3 was loaded on to calibrated Superdex 200 16/60 or Superdex 75 26/60 columns respectively in the absence (a, b) or presence (c, d) of Emulphogene. Both detergent-free proteins eluted almost exclusively at sizes corresponding to the protein monomers (∼55 kDa). In the presence of detergent, the peak position of CYP74A1 protein shifted to that corresponding to a complex of molecular mass approx. 110 kDa; in addition, the peak position of CYP74C3 protein shifted to a number of species of higher mass. The peak positions of CYP74A1 and CYP74C3 activities corresponded to the size of a monomer–micelle complex.

Table 1 Activation of CYP74C3 and CYP74A1 CYP74C3 Kinetic parameter k cat /K m (µM−1 · s−1 )† 13-HPOTE 13-HPODE 9-HPOTE 9-HPODE Substrate specificity (%) 13-HPOTE 13-HPODE 9-HPOTE

Detergent-free

CYP74A1 Activated*

Detergent-free

Activated*

3.2 0.6

149.3 33.5

1.2 0.9

58.7 8.1

0.1 0.5

0.6 7.2

0.3 0.8

0.2 0.2

100 19 3

100 22 0.4

100 75

100 14

25

0.3

*With 50 µM Emulphogene micelle. †See [28,29] for individual K m and k cat values. Assumes one active site per monomer of molecular mass 56.8 or 55.3 kDa for CYP74C3 and CYP74A1 respectively.

towards low-spin haem iron, which was confirmed by UV– visible and EPR spectroscopy [28,29]. An analysis of micelle binding to both detergent-free proteins indicated that the detergent micelle bound very tightly to both protein monomers with K d values of 6.9 ± 1.1 and 10.7 ± 1.7 µM for

CYP74C3 and CYP74A1 respectively, which are similar or tighter than the preferred substrate 13-HPOTE [28,29]. The substrate specificity of CYP74A1 was greatly modified upon micelle binding to one that had considerably more 13-AOS activity, but the level of 9-AOS activity was not insignificant C 2006

Biochemical Society

1225

1226

Biochemical Society Transactions (2006) Volume 34, part 6

and CYP74A1 would still correctly be described as exhibiting dual specificity even in the micelle-bound form.

Production of CYP74C3 and CYP74A1 for structural resolution To date, no structure has been published for any plant cytochrome P450 or CYP74 enzyme. The structure of the AOS domain of an AOS–LOX chimaeric protein from coral has been solved and shown to be very similar to a catalase [30], but this is highly dissimilar to HPLs, which are neither watersoluble in their most active state nor predicted to have a catalase-fold. CYP74C3 and CYP74A1 have been purified from E. coli to produce detergent-free proteins that are almost entirely monomeric, monodisperse and water-soluble to 50 mg/ml protein, based on haem concentration. Crystallization trials using both oil immersion and hanging drop techniques have been conducted in three different laboratories with relatively little success. No crystals of CYP74C3 have been obtained and the fragile, needle-like crystals of CYP74A1 that were obtained were unsuitable for X-ray diffraction. We have now carried out modelling and mutagenesis experiments to successfully identify mutants of CYP74C3 that are modified in both substrate and micelle binding, the first to be identified for a CYP74 enzyme and which may prove to be more amenable to crystallization.

Proposed new rules for the classification of CYP74 enzymes We suggest a new scheme for CYP74 classification. Since the discovery of the first CYP74 enzyme, an AOS purified from flax seed [3], the classification of CYP74 enzymes has been relatively straightforward and based on sequence relatedness [31]. Fortunately, this also separated them on the basis of function. Thus CYP74A and CYP74B corresponded to AOS and HPL respectively. The substrate specificity of CYP74A and CYP74B enzymes was previously shown to be monospecific and restricted to either 9- or 13-hydroperoxide substrates. The discovery of new HPL [19] and AOS [5] enzymes with dual specificity, turning over both 9- and 13-hydroperoxides, has since led to the emergence of a third group, CYP74C, which contains any CYP74 with dual specificity and is unrelated to product specificity. The recent revelation that recombinant AOS from Arabidopsis (CYP74A1) has dual specificity [29], despite its identification as a 13-AOS based on sequence relatedness, would suggest that CYP74 assignation should be based on evidence of both substrate and product specificity. Moreover, CYP74 designation should only be applied to those sequences that are proven to encode active recombinant proteins and whose substrate and product specificities have been fully determined using purified protein in vitro. The current use of CYP74 classification for unpublished sequences, or where the true function has not been verified, is both misleading and premature. Under the proposed new scheme, CYP74C would include only HPLs with dual specificity: CYP74C1 (cucumber HPL [18]), CYP74C2 (melon HPL [19]) and C 2006

Biochemical Society

CYP74C3 (barrel medic HPL [28]). CYP74D would be assigned to AOS with dual specificity. For the purposes of this discussion, AOS from Arabidopsis remains classified as CYP74A1, but would be reclassified as CYP74D4 in the new scheme: CYP74D1 (tomato AOS [5], previously classified as CYP74C3) and CYP74D2 (barley AOS [24]). A new CYP74C with dual specificity has very recently been described in potato [32] and would be reclassified as CYP74D3 in the new scheme. CYP74E would be assigned to DES (divinyl ether synthase): CYP74E1 (tomato DES [33]). New sequences from rice in the P450 database that are currently classified as CYP74E and CYP74F (http://drnelson.utmem. edu/CytochromeP450.html) are unpublished so would be excluded from this classification scheme. Reclassification based on function is a necessary pre-requisite for the interpretation of the roles of CYP74 enzymes in plant oxylipin metabolism. We respectfully offer our new proposal to the nomenclature committee on P450 classification and the CYP74 scientific community as a basis for debate and discussion.

We thank Professor Mats Hamberg (Karolinska Institute, Stockholm, Sweden) for providing CYP74 substrates. This work was funded by a European Union-funded project ‘NODO (Natural Oxylipins for Defence of Ornamentals)’, project number QLK5-CT-2001-02445, and by the Biotechnology and Biological Sciences Research Council.

References 1 Hasemann, C.A., Kurumbail, R.G., Boddupalli, S.S., Peterson, J.A. and Desienhofer, J. (1995) Structure 2, 41–62 2 Williams, P.A., Cosme, J., Sridhar, V., Johnson, E.F. and McRee, D.E. (2000) Mol. Cell 5, 121–131 3 Song, W.-C., Funk, C.D. and Brash, A.R. (1993) Proc. Natl. Acad. Sci. U.S.A. 90, 8519–8523 4 Pan, Z., Camara, B., Gardner, H.W. and Backhaus, R.A. (1998) J. Biol. Chem. 273, 18139–18145 5 Itoh, A., Schilmiller, A.L., McCaig, B.C. and Howe, G.A. (2002) J. Biol. Chem. 277, 46051–46058 6 Feussner, I. and Wasternack, C. (2002) Annu. Rev. Plant Physiol. Plant Mol. Biol. 53, 275–297 7 Casey, R. and Hughes, R.K. (2004) Food Biotechnol. 18, 135–170 8 Noordermeer, M.A., van der Goot, W., van Kooij, A.J., Veldsink, J.W., Veldink, G.A. and Vliegenthart, J.F.G. (2002) J. Agric. Food Chem. 50, 4270–4274 9 Husson, F. and Belin, J.M. (2002) J. Agric. Food Chem. 50, 1991–1995 10 Hausler, ¨ A., Lerch, K., Muheim, A. and Silke, N. (2001) U.S. Patent 6238898 11 Tijet, N., Waspi, U., Gaskin, D.J., Hunziker, P., Muller, B.L., Vulfson, E.N., Slusarenko, A., Brash, A.R. and Whitehead, I.M. (2000) Lipids 35, 709–720 12 Shibata, Y., Matsui, K., Kajiwara, T. and Hatanaka, A. (1995) Plant Cell Physiol. 36, 147–156 13 Fauconnier, M.-L., Perez, A.G., Sanz, C. and Marlier, M. (1997) J. Agric. Food Chem. 45, 4232–4236 14 Psylinakis, E., Davoras, E.M., Ioannidis, N., Trikeriotis, M., Petrouleas, V. and Ghanotakis, D.F. (2001) Biochim. Biophys. Acta 1533, 119–127 15 Kandzia, R., Stumpe, M., Berndt, E., Szalata, M., Matsui, K. and Feussner, I. (2003) J. Plant Physiol. 160, 803–809 16 Howe, G.A., Lee, G.I., Itoh, A., Li, L. and DeRocher, A.E. (2000) Plant Physiol. 123, 711–724 17 Noordermeer, M.A., van Dijken, A.J.H., Smeekens, S.C.M., Veldink, G.A. and Vliegenthart, J.F.G. (2000) Eur. J. Biochem. 267, 2473–2482


18 Matsui, K., Ujita, C., Fujimoto, S.-H., Wilkinson, J., Hiatt, B., Knauf, V., Kajiwara, T. and Feussner, I. (2000) FEBS Lett. 481, 183–188 19 Tijet, N., Schneider, C., Muller, B.L. and Brash, A.R. (2001) Arch. Biochem. Biophys. 386, 281–289 20 Koeduka, T., Stumpe, M., Matsui, K., Kajiwara, T. and Feussner, I. (2003) Lipids 38, 1167–1172 21 Song, W.-C. and Brash, A.R. (1991) Science 253, 781–784 22 Pan, Z., Durst, F., Werck-Reichhart, D., Gardner, H.W., Camara, B., Cornish, K. and Backhaus, R.A. (1995) J. Biol. Chem. 270, 8487–8494 23 Utsunomiya, Y., Nakayama, T., Oohira, H., Hirota, R., Mori, T., Kawai, F. and Ueda, T. (2000) Phytochemistry 53, 319–323 24 Maucher, H., Hause, B., Feussner, I., Ziegler, J. and Wasternack, C. (2000) Plant J. 21, 199–213 25 Laudert, D.L., Pfannschmidt, U., Lottspeich, F., Hollander-Czytko, H. and Weiler, E.W. (1996) Plant Mol. Biol. 31, 323–335 26 Dean, W.L. and Gray, R.D. (1982) J. Biol. Chem. 257, 14679–14685

27 Viner, R.I., Novikov, K.N., Ritov, V.B., Kagan, V.E. and Alterman, M.A. (1995) Biochem. Biophys. Res. Commun. 217, 886–891 28 Hughes, R.K., Belfield, E.J., Muthusamay, M., Khan, A., Rowe, A., Harding, S.E., Fairhurst, S.A., Bornemann, S., Ashton, R., Thorneley, R.N.F. and Casey, R. (2006) Biochem. J. 395, 641–652 29 Hughes, R.K., Belfield, E.J., Ashton, R., Fairhurst, S.A., Gobel, ¨ C., Feussner, I. and Casey, R. (2006) FEBS Lett. 580, 4189–4195 30 Oldham, M.L., Brash, A.R. and Newcomer, M.E. (2005) Proc. Natl. Acad. Sci. U.S.A. 102, 297–302 31 Nelson, D.R. (2002) Methods Enzymol. 357, 13–15 32 Stumpe, M., Gobel, ¨ C., Demchenko, K., Hoffmann, M., Klosgen, ¨ R.B., Pawlowski, K. and Feussner, I. (2006) Plant J. 47, 883–896 33 Itoh, A. and Howe, G.A. (2001) J. Biol. Chem. 276, 3620–3627

Received 26 June 2006

C 2006

Biochemical Society

1227