Armand Fulco's studies of fatty acid hydroxylase activity in the soil bacterium Bacillus megaterium revealed the first major P450 redox system deviating from the aforementioned class I and class II types . The hydroxylase activity was found to be associated with a phenobarbital-inducible 119 kDa protein, which was then shown to be a P450 (N-terminal) fused to a CPR module (C-terminal) [68, 69]. The enzyme (P450BM-3 or CYP102A1) is a soluble enzyme and a membrane anchor is absent from both the P450 and reductase domains of the enzyme. Thus, P450BM-3 was the first P450 shown to use a soluble CPR enzyme and the first natural P450/redox partner fusion enzyme [14, 70] (Fig. 5.3a). P450BM-3 has the highest catalytic center activity reported for a P450 monooxygenase system (>15 000 min-1 with arachidonic acid), and a highly efficient intra-reductase domain and reductase-to-heme electron transfer system underlies this rapid fatty acid hydroxylase rate [71, 72]. P450BM-3 catalyzes hydroxylation of a wide range of fatty acids (both saturated and unsaturated), typically hydroxylating close to the ^-terminal at the a-1 to a -3 positions .
As with P450cam, extensive protein engineering studies have revealed various aspects of structure and mechanism in the enzyme. Mutagenesis studies have revealed key roles for the active site residues Phe87 (control of regioselectivity of substrate oxygenation), Arg47 and Tyr51 (interactions with substrate carboxylate group), and Thr268 (oxygen activation and coupling of electron transfer to substrate oxidation) [71, 74, 75]. Key roles for Ala264 (mutants perturb protein conformational equilibria) and Phe393 (interactions with Cys-Fe bond and control of thermodynamic properties of the P450 heme iron) have also been established [76, 77]. In the reductase domain, the "catalytic triad" of residues (conserved across the diflavin reductase family of enzymes) in P450BM-3 comprises residues Cys999, Ser830, and Asp1044. A C999A mutant of P450BM-3 reductase was shown to be dramatically kinetically compromised as a result of a much slower rate of hydride transfer from the NADPH coenzyme and through formation of a stable NADP+ (i.e. product) complex . However, other mutations conducive to the biotechnological application of P450BM-3 have also been reported, and are discussed in more detail in Section 5.4.
Genetic dissection of the P450BM-3 flavocytochrome enabled production of catalytically viable P450 and reductase domains, and led to the production of FAD/NADPH- and FMN-binding domains of the reductase. These experiments validated the proposed domain composition of the flavocytochrome, and were important in confirming that the CPR module resulted from an ancestral fusion of genes encoding ferredoxin reductase-like and flavodoxin-like proteins [79, 80]. Domain dissection studies also enabled determination of reduction potentials for all of the redox centers in the enzyme, and facilitated the demonstration that fatty
Fig. 5.3 Novel types of P450 systems. Several types of P450 systems distinct from those class I and class II systems represented in Fig. 5.1 are now known. (a) P450BM-3 (CYP102A1) type system, with P450 fused to a CPR. The structures shown are the BM-3 heme domain (PDB code 2HPD) and rat CPR (1AMO). (b) CYP116B1/B2 type system with P450 fused to a PDOR enzyme. Representative structures shown are Saccharopolyspora erythrae P450 eryF (CYP107A1, PDB code 1OXA) and the phthalate dioxygenase reductase from Burkholderia cepacia (2PIA). (c) Methylococcus capsulatus CYP51-ferredoxin fusion, represented by M. tuberculosis CYP51 (1X8V) and the Pyrococcusfuriousus 3Fe-4S ferredoxin (1SJ1). (d) Rh. rhodochrous Y-11 XplA enzyme, considered to be a
P450-flavodoxin fusion. The structure is represented by rabbit CYP2B4 (2BDM) and E. coli flavodoxin (1AHN). (e) A P450 system that does not have a protein redox partner (representing the types of P450 driven by peroxide or by direct interaction with NAD(P)H). The protein shown is Fusarium oxysporum P450nor, with bound nitric oxide (1CL6). (f) Likely Ps. fluorescens P450-acyl CoA dehydrogenase fusion represented by Polyangium cellulosum P450 epoK (1Q5E) and Sus scrofa acyl CoA dehydrogenase (3MDE). (g) Non-NAD(P)H-dependent P450 system from Sulfolobus solfataricus, represented by the Desulfovibrio africanus pyruvate ferredoxin oxidoreductase (2C3M), the seven-iron ferredoxin from Azotobacter vinelandii (1FD2) and Sulfolobus solfataricus CYP119A1 (1F4U).
acid substrate binding leads to an increase in heme iron redox potential that triggers NADPH-dependent electron transfer from the redox partner .
Atomic structures of the P450 (heme) domain have been solved in substratefree and substrate-bound (both palmitoleic acid and N-palmitoylglycine) forms. These reveal considerable conformational alterations [81-83]. It remains unclear whether these major structural rearrangements are a consequence of substrate binding per se, or instead reflect natural conformations of the ligand-free enzyme that may have differing affinities for substrates. Recent structural studies on an A264E mutant of the BM-3 heme domain suggest the latter explanation has some merit .
While there is, as yet, no atomic structure for the CPR domain of the enzyme, the FMN domain has been solved in a non-stoichiometric complex with the heme domain . The FMN domain structure highlights important structural differences between the BM-3 FMN domain and other bacterial flavodoxins that might underlie its tendency to form an anionic (red) as opposed to a neutral (blue) semiquinone species [86, 87]. The BM-3 FMN-heme complex structure located the FMN domain near the proximal face of the heme (the likely docking interface) with its flavin cofactor orientated towards the heme cysteinate ligand . However, a predicted electron transfer pathway through ~50 sigma bonds can be ruled out as a consequence of the discrepancy between the measured and predicted rates of this process [14, 72].
Mobility of the FMN domain (between its electron donor FAD domain and acceptor heme domain) of the enzyme is likely to occur to bring cofactors into close proximity for efficient electron transfer. However, recent data indicate that the flavocytochrome is catalytically functional in fatty acid hydroxylation as a dimer, with electron transfer between FMN in one monomer and heme in the other supporting fatty acid hydroxylation . Thus, a more detailed understanding of the interdomain interactions likely awaits the determination of the atomic structure of the intact flavocytochrome enzyme.
The physiological role of P450BM-3 remains uncertain, but recent genome sequencing efforts have confirmed its presence in several other bacteria - including Ralstonia metallidurans, Bacillus subtilis, and Bradyrhizobium japonicum. In B. subtilis there are two homologs - CYP102A2 and CYP102A3 . Both the A2 and A3 enzymes catalyze fatty acid hydroxylation close to the $-terminal of fatty acid substrates, as does P450BM-3. However, the properties of these enzymes deviate from BM-3 and from one another in terms of substrate preference (Kd values) and turnover rates (kcat values), and with respect to apparent cooperative binding (sigmoidal binding curves) observed for various lipid substrates. The A3 enzyme has been successfully engineered to produce 1-octanol from octane, and it is clear that the catalytic diversity of the CYP102 enzyme class can be exploited for bio-technologically important reactions .
Amino acid sequence alignment of B. megaterium P450BM-3 with its B. subtilis homologs reveals important regions of conservation and deviation between these enzymes (Fig. 5.4). Among the most notable features are the absences of residues corresponding to BM-3 Arg47 and Tyr51 in the A2/A3 enzymes. These residues are implicated in binding fatty acid carboxylate in the BM-3 enzyme, but are clearly not conserved for such a role in A2/A3 . The strong conservation of the I helix region (indicated for the BM-3 enzyme in blue text in Fig. 5.4a) in the A2/A3 enzymes is consistent with the similar substrate selectivity and reactivity of these three enzymes (Fig. 5.4a). In the reductase domain, regions involved in binding coenzyme (NADPH) pyrophosphate and FAD cofactor are cyp102a2 MKETSPIPQPKTFGPLGNLPLIDKDKPTLSLIKLAEEQGPIFQIHTPAg|TTlvlVSGHELW 6C
cyp102a3 mkqasaipqpktygplknlphlekeqlsqslwriadelgpifrfdfpgV|ssVFvsghnlv 6C
cyp102a1 --TIKEMPQPKTFGELKNLPLLNTDKPVQALMKIADELGEIFKFEAPGRVTRB|SSQRLI 5 8
cyp102a2 cyp102a3 cyp102a1
cyp102a2 kevcdeerfdksiegalekvrafsgdglIfTswthepnwrkahnilmptfsqramkdyhek 12C
cyp102a3 aevcdekrfdknlgkglqkvrefggdglIfTswthepnwqkahrillpsfsqkamkgyhsm 12C
cyp102a1 keacdesrfdknlsqalkfvrdfagdglFtswtheknwkkahnillpsfsqqamkgyham 118
cyp102a2 MVDIAVQLIQKWARLNPNEAVDVPGDMTRLTLDTIGLCGFNYRFNSYYRETPHPFINSMV 18C
cyp102a3 MLDIATQLIQKWSRLNPNEEIDVADDMTRLTLDTIGLCGFNYRFNSFYRDSQHPFITSML 18C
cyp102a1 MVDIAVQLVQKWERLNADEHIEVPEDMTRLTLDTIGLCGFNYRFNSFYRDQPHPFITSMV 17 8
RA0DEAMNKLQRANPDDPAYDENKRQFQEDIKYMNDLVDKIIADRKASGEQS-DDLLTHM 2 37
cyp102a2 LNVEDPETGEKLDDENIRFQIITFLIa cyp102a3 LYAKDPVTGETLDDENIRYQIITFLIa cyp102a1 LNGKDPETGEPLddeniryqiitfli
TSGLLSFATYFLLKHPDKLKKAYEEVDRV 3CC TSGLLSFAIYCLLTHPEKLKKAQEEADRV 3CC tsgllsfalyflvkNPHVLQKAAEEAARV 2 97
cyp102a2 LTDAAPTYKQVLELTYIRMILNESLRLWPTAPAFSLYPKEDTVIGGKFPITTNDRISVLI 36C
cyp102a3 LTDDTPEYKQIQQLKYIRMVLNETLRLYPTAPAFSLYAKEDTVLGGEYPISKGQPVTVLI 36C
cyp102a1 LVDPVPSYKQVKQLKYVGMVLNEALRLWPTAPAFSLYAKEDTVLGGEYPLEKGDELMVLI 3 57
CIGMQFALHEATLVLGMI 42C CIGMQFALQEATMVLGLV 42C ClGQQFALHEATLVLGMM 417
cyp102a2 LKYFTLIDHENYELDIKQTLTLKPGDFHISVQSRHQEAIHADVQAAEKAAPDEQKEKTEA 48C
cyp102a3 LKHFELINHTGYELKIKEALTIKPDDFKITVKPRKTAAINVQRKEQADIKAETKPKETKP 48C
cyp102a1 LKHFDFEDHTNYELDIKETLTLKPEGFVVKAKSKKIPLGGIPSPSTEQSAKKVRKKAENA 47 7
Fig. 5.4 Amino acid alignment of Bacillus megaterium P450 BM3 (CYP102A1) and Bacillus subtilis CYP102A2 and CYP102A3.
Various key residues recognized in studies of the P450 BM3 enzyme are highlighted in white text on a black background. Panel A shows an alignment of the heme domains of the three enzymes. Arg 47 and Tyr 51 are considered important for interactions with fatty acid carboxylate groups, but are not retained in CYP102A2 or A3, suggesting a differing substrate interaction mode . Phe 87 is a determinant of regioselectivity of fatty acid hydroxylation in P450 BM3, and is retained in A2/A3 [71, 74]. Mutation of Leu 181 to Arg/Lys led to improved binding of short chain fatty acids in BM3, and this residue is retained in A2/A3 . The I helix residues Ala 264 and Thr 268 are retained in all enzymes, and the entire I helical region (bold text coloured in black and underlined for the BM3 enzyme) is very strongly conserved in these enzymes. In BM3, an A264E mutant produced a novel Cys-Fe-Glu heme iron ligand set . Thr 268 is implicated in oxygen binding/proton delivery . Cys 400 is the thiolate ligand to the heme iron and is retained in all P450s. Phe 393 is retained in A2/A3 and in all P450 monooxygenase enzymes, and is a critical determinant of heme iron reduction potential and oxy complex stability [13, 77].
cyp102a2 cyp102a3 cyp102a1
KGASVIGLNNRPLLVLYGSDTGTAEGVARELADTASLHGVRTKTAPLNDRIGKLPKEGAV 5 4C
K-------HGTPLLVLFGSNLGTAEGIAGELAAQGRQMGFTAETAPLDDYIGKLPEEGAV 5 33
ASTYQYVPRFIDEQL 6CC ASTYQRIPRLIDDMM 593 ATTYQKVPAFIDETL 589
cyp102a2 AEKGATRFSARGEGDVSGDFEGQLDEWKKSMWADAIKAFGLELNENADKE-RSTLSLQFV 6 59
cyp102a3 KAKGASRLTAIGEGDAADDFESHRESWENRFWKETMDAF--DINEIAQKEDRPSLSITFL 6 51
cyp102a1 AAKGAENIADRGEADASDDFEGTYEEWREHMWSDVAAYFNLDIENSEDNK--STLSLQFV 6 47
cyp102a2 RGLGESPLARSYEASHASIAENRELQSADSDRSTRHIEIALPPDVEYQEGDHLGVLPKNS 719
cyp102a3 SEATETPVAKAYGAFEGIVLENRELQTAASTRSTRHIELEIPAGKTYKEGDHIGILPKNS 711
cyp102a1 DSAADMPLAKMHGAFSTNVVASKELQQPGSARSTRHLEIELPKEASYQEGDHLGVIPRNY 7C7
cyp102a2 QTNVSRILHRFGLKGTDQVTLSASGRSAGHLPLGRPVSLHDLLSYSVEVQEAATRAQIRE 779
cyp102a3 RELVQRVLSRFGLQSNHVIKVSGSAHMA-HLPMDRPIKVVDLLSSYVELQEPASRLQLRE 77C
cyp102a1 EGIVNRVTARFGLDASQQIRLEAEEEKLAHLPLAKTVSVEELLQY-VELQDPVTRTQLRA 7 66
cyp102a2 LASFTVCPPHRRELEELS-AEGVYQEQILKKRISMLDLLEKYEACDMPFERFLELLRPLK 838
cyp102a3 LASYTVCPPHQKELEQLVSDDGIYKEQVLAKRLTMLDFLEDYPACEMPFERFLALLPSLK 83C
cyp102a1 MAAKTVCPPHKVELEALL-EKQAYKEQVLAKRLTMLELLEKYPACEMKFSEFIALLPSIR 82 5
cyp102a2 cyp102a3 cyp102a1
PRYyslISSSPRVNPRQASITVGVVRGPAWSGRGEYRGVASNDLAERQAGDDVVMFIRTPE 89 8
pryysIissspkvhanivsmtvgvvkasawsgrgeyrgvasnylaelntgdaaacfirtpq 89C prYysisssprvdekqasitvsvvsgeawsgygeykgiasnylaelqegdtitcfistpq 885
cyp102a2 SRFQLPKDPETPIIMVGPGTGVAPFRGFLQARDVLKREGKTLGEAHLYFGCRN-DRDFIY 9 57
cyp102a3 SGFQMPNDPETPMIMVGPGTGIAPFRGFIQARSVLKKEGSTLGEALLYFGCRRPDHDDLY 9 5C
cyp102a1 SEFTLPKDPETPLimvgpgtgvapfrgfVQARKQLKEQGQSLGEAHLYFGCRSPHEDYLY 945
cyp102a2 RDELERFEKDGIVTVHTAFSRKEGMPKTYVQHLMADQADTLISILDRGGRLYVcpDGSKM 1C17
cyp102a3 REELDQAEQDGLVTIRRCYSRVENEPKGYVQHLLKQDTQKLMTLIEKGAHIYVcpDGSQM 1C1C
cyp102a1 QEELENAQSEGIITLHTAFSRMPNQPKTYVQHVMEQDGKKLIELLDQGAHFYi|cgdgsqm 1Qa5
Fig. 5.4 Continued Panel B shows an alignment of the reductase domains of the three P450 systems. The gap in the A1/A3 alignments at the start indicates a different inter-domain linker region in the A2 enzyme. Residues Gly 570 and Trp 574 are important determinants of FMN binding and are retained in each enzyme . Residues Cys 999, Ser 830 and Asp 1044 comprise a catalytic triad important for efficient electron transfer from NADPH coenzyme. They are retained in each enzyme and across the majority of the diflavin reductases [54, 78]. Ser 830 and Tyr 829 are located in a strongly conserved region involved in binding the FAD cofactor (bold text coloured in black and underlined for the BM3 enzyme), and the side chain of Tyr 829 interacts with the FAD isoalloxazine ring system. The region between amino acids ~890-915 in BM3 encompasses a conserved region involved in interactions with NADP(H) and its pyrophosphate group, and is also coloured black in bold, underlined text) for the BM3 enzyme. Trp 1064 is also conserved in each enzyme, and the side chain of Trp 1064 stacks across the isoalloxazine ring system of the BM3 FAD cofactor and must be displaced to enable hydride ion transfer from NADPH. The W1064A/H mutants of P450 BM3 demonstrate a spectacular switch of coenzyme specificity in favour of NADH .
also strongly conserved in A1-A3, as are the catalytic triad of residues essential for electron transfer from NADPH: Ser830, Cys999, and Asp1044 in BM-3  (Fig. 5.4b).
The CYP102-type fatty acid hydroxylase-CPR type of fusion protein is not restricted to bacteria, and P450foxy (CYP505A1 from the fungus Fusarium oxyspo-rum) was the first eukaryotic representative of this class of enzymes characterized . This protein associates with the cell membrane, but is devoid of any well-defined membrane anchor region. CYP505A1 also hydroxylates fatty acids at the a-1, a-2, and a-3 positions. Similar enzymes are identifiable from sequence analysis of other eukaryotic genomes (e.g. Neurospora crassa), and CYP505B1 from Fusarium verticilloides likely functions as a polyketide hydroxylase in synthesis of the mycotoxin fumonisin .
Was this article helpful?