TIBS 16-APRIL 1991

An unusual yet strongly

THE TRANSFER of reducing equiv- conserved reductase alents between and the elec- tron-carrier cofactors NAD* and NADP÷ is integral to many metabolic reactions. in and mammals While in some instances electrons are transferred directly between the sub- strate and the nicotinamide (as with alcohol dehydrogenase and dihydrofolate reductase), often a bound coenzyme serves as an intermediate in The recent determination of the amino acid sequences of the Bacillus electron transfer. Enzymes utilize a var- megaterium P-450 and the flavoprotein component of iety of such intermediates, including Salmonella typhimurium NADPH-sulfite reductase revealed that these heine, iron-sulfur clusters, flavins, and enzymes contain a flavoptotein moiety remarkably similar to mammalian more rarely, pteridines. Of this group, NADPH-cytochrome P-450 reductase. The presence of this oxidore- fiavins such as ductase in these very different enzymes suggests that this flavoprotein (FMN) and flavin adenine dinucleotide arose early in evolution and was utilized as an enzymological building (FAD) are the most common. block. The multi-domain structure of the reductase further suggests that it Ravoproteins can be divided into at arose through a fusion of genes encoding simple flavin electron-transport least five classes, based upon the . reaction catalysed and the electron acceptor ~,2. Transhydrogenases and endoplasmic reticulum of most eukary- Structure. The sequence of P-450 electron translerases are two classes of otic cells. As an essential component of reductase was first determined for the in which the electron the microsomal cytochrome P-450- raP, and is now known for a variety of acceptor (or donor) is either another dependent system3,4, species, including humans 9-~4. The or a nicotinamide cofac- it catalyses the transfer of electrons mammalian reductases share about tot. Members of these two classes are from NADPH to the P-450, 90?/0 sequence identity, with lesser simi- distinguished from other flavoproteins which are responsible for the oxidation larity to the trout (79%)n and by their limited reactivity with molecu- of innumerable foreign and endogenous (33%) ~2 enzymes. A detailed analysis of lar (05). Transhydrogen~es and compounds including drugs, the reductase amino acid sequence, electron are often associ- metabolites, steroids and prosta- with comparison to flavoproteins of ated with energy-generating electron glandins (for review see Ref. 5). A most known three-dimensional structure, has transport pathways, and the associated notable aspect of cytochrome P-450 facilitated identification of the flavin- redox protein is often a 1-electron reductase is its flavin content: the and cofactor.binding domains of the en- acceptor, such as a cytochrome or a single polypeptide chain binds both an zyme Is. A cartoon representation of the non-heme iron-sulfur protein. Well-char- FMN and an FAD prosthetic group 6. As proposed reductase organization is acterized examples of flavoprotein established by Vermilion et al. 7, the shown in Fig. I. The protein is anchored transhydrogenases include ferredox- pathway of electron transfer proceeds to the endoplasmic reticulum by a in-NADP" reductase and glutathione from NADPH to FAD to FMN to short hydrophobic amino-terminal seg- reductase; electron transferases are cytochrome P-450. ment; the subsequent catalytic portion typified by the . The flavo- The microsomal cytochrome P-450 of the reductase begins with a 150- protein that is the subject of this review system differs from known bacterial residue segment that shows consider- is unusual in that it contains both a and adrenal mitochondrial P-450 able similarity to the bacterial flavo- transhydrogenase domain and an elec- monooxygenase systems in the path- doxins. tron domain on a single way of electron transfer. The bacterial Ravodoxins are small (15-23 kDa), polypeptide, and thereby represents a and mitochondrial systems use two acidic, FMN-containing proteins found fusion of two normally independent separate electron transfer components: in bacteria and algae, but not in higher flavoproteins. This compound flavopro- an FAD-containing reductase, which or animals. Flavodoxins are tein, first characterized as microsomal accepts two reducing equivalents from strict electron transferases, reacting cytochrome P-450 reductase, has now a nicotinamide cofactor (NADPH or only with other redox proteins, and are been shown to be a component of two NADH), and a small iron-sulfur protein able to replace in many bacterial enzymes: a Bacillus megaterium (a ), which catalyses two redox pathways (for review see Ref. 16). cytochrome P-450 and sulfite reductase. separate l-electron transfers from The three-dimensional structures of the reductase to cytochrome P-450. several flavodoxins are known. Cytochrome P.450 reductase Although the evolutionary origin of Cytochrome P-450 reductase exhibits NADPH-cytochrome P-450 reductase microsomal cytochrome P-450 re- several regions of strong sequence simi- is a 78 kDa flavoprotein bound to the ductase appears to be distinct from that larity with these fiavoproteins, most of these two-component electron trans- notably with the Desulfovibrio pulgaris T. D. Porter is at the Departmentof Biological port chains, it is related in function, fiavodoxin, which correspond to seg- Chemistry, Medical School, The Universityof and may be derived from a common ments involved in binding the FMN Michigan, Ann Arbor, MI 48109-0606, USA. ancestral pathway. group to these proteins. Moreover, the 154 © 1991,Elsevier Science Publishers Ltd,(UK) 0376-5067/91/$02.00 TIBS 16-APRIL 1991

predicted secondary structure of this reductase is a component of region of the reductase, obtained by the the microsomal fatty acid method of Chou and Fasman, is very desaturase and elongation MADPH similar to the secondary structure of pathways, and also a com- cytosol the flavodoxins, suggesting that this ponent of the methe- FAD region of the reductase binds FMN in an moglobin reductase path- overall configuration like that of flavo- way in red blood cells. The FMN doxin, despite being incorporated into similarity between P-450 a much larger protein TM. reductase and these two Recent studies utilizing site-directed flavoproteins begins approx- mutagenesis have supported this imately 40 residues beyond hypothesis ~7. In flavodoxins there are the FMN-binding domain, two residues, which are usually and extends for approxi- hydrophobic and often aromatic, that mately 60 residues before Rgum 1 are located above and below the FMN being disrupted by a ll7- Domain organization of cytochrome P-450 reductase. isoalloxazine ring. These shielding amino acid segment in P-450 The protein is anchored to the endoplasmic reticulum residues maintain a hydrophobic en- reductase that appears unre- by its hydrophobic amino-terminal segment, which is vironment for the flavin, and may par- lated to these two flavopro- followed by the FMN-, FAD- and NADPH-binding domains. ticipate in pi-pi stacking interactions teins. Following this 'inser- with the isoalloxazine ring ~6. In P-450 tion', the sequence similari- reductase, the proposed flavin-shielding ty resumes, and becomes strongest stitute for ferredoxin in a variety of residues are Tyrl40 and Tyr178 (Ref. in the carboxy-terminal region of the pathways, including the photosynthetic 15). Substitution of aspartate for proteins. By comparison of these reduction of NADP÷ with ferredoxin Tyrl40, the residue proposed to be sequences to , an reductase ~s. It has been suggested that positioned at an angle of 45 ° over the FAD-containing protein whose three- the genes for and ferredoxin interior face of the flavin ring, results in dimensional structure is known, tenta- reductase were arranged in tandem in a fivefold decrease in reductase activity, tive assignments of FAD- and NADPH- an operon of an early organism, and at although it does not significantly alter binding domains have been made ~s. some point fused to give rise to the pri- RVlN binding. In contrast, substitution The initial 60-amino acid segment of mordial P-450 reductase gene~E That of aspartate for Tyr178, the proposed similarity has been proposed as the these two proteins could simply fuse in exterior ring-shielding residue, coplanar FAD pyrophosphate-binding segment, a head-to-tail fashion and yield a func- with the flavin ring in flavodoxin, abol- and the subsequent l l7-amino acid tional protein would be surprising, and ishes both reductase activity and FMN insertion in P-450 reductase has been the additional segment (or insertion) binding. Replacement of either, or both, proposed as being involved in orienting discussed earlier may have been necess- tyrosines with phenylalanine, an obvi- the two flavin domains for !nterflavin ary to couple the two proteins effec- ously conservative replacement, does electron transport. Because ferredoxin tively. This segment is predicted to be not affect activity or FMN binding. reductase and cytochrome bs reductase almost entirely a-helical, suggesting Interestingly, FAD binding is reduced by contain only a single flavin, this orient- that it may be located on the surface of the Tyr178~Asp mutation, which may ing segment would be unnecessary in the protein. indicate that FAD incorporation is these enzymes. The proposed nico- The recently determined structure of dependent on formation of an intact tinamide cofactor-binding domain the rat P-450 reductase gene supports FMN domain. These results provide begins at approximately position 480 of a gene fusion origin for this strong support for the hypothesis that P-450 reductase, and extends for flavoprotein 2°. The gene is divided into this region of P-450 reductase binds the approximately 200 residues. The se- fifteen coding exons, of which three FMN prosthetic group. Although the quence similarity among these proteins encode the FMN domain and six encode corresponding mutations have not yet in this region is strong, and is supported the FAD and NADPH domains. The been examined in a flavodoxin molecule by a variety of chemical modification beginning and end of the FMN domain per se, the effects described above are studies that have demonstrated that is clearly coincident with intron pos- consistent with the FMN being bound in bound cofactor in all three flavopro- itions, as is the start of the FAD domain; a manner similar to that found in the teins can protect specific residues in this placement of introns precisely fiavodoxins. this region from modification (for ex- between these functional domains sug- Cytochrome P-450 reductase also amples see Refs 14, 18 and 19). These gests that each was originally a sep- exhibits significant sequence similarity results support the assignment of the arate entity, and that a genetic recombi- to two FAD-containing enzymes, ferre- middle and carboxy-terminal segments nation event brought them together and doxin-NADP÷ reductase and NADH- of P-450 reductase to binding FAD and gave rise to the reductase as a 'fusion ~s. Both of NADPH, respectively. protein'. these enzymes are simple transhy- Evolutionary odglns. The amino-terminal Cytochrome P-450 arose over two bil- drogenase-type flavoproteins, catalysing homology of P-450 reductase with flavo- lion years ago, and members of this electron transfer between a nicotina- doxin, and the carboxy-terminal hom- family of are present mide cofactor and a l-electron ac- ology with ferredoxin-NADP" reductase, in both prokaryotes and eukaryotes ceptor (or donor) protein. Ferredoxin suggest that P-450 reductase arose (for review see Ref. 21). The earliest reductase is the terminal electron car- through a fusion of the ancestral genes P-450s were bacterial, and presumably rier of the photosynthetic electron for these two types of fiavoproteins. It utilized the two-component electron transport chain, and cytochrome b5 is noteworthy that flavodoxin can sub- transport chain (a ferredoxin reductase 155 TIBS 16-APRIL 1991

ferredoxin reductase/ferre- P4SOR .aOS.,OMS['i]T.P,,V*,eVSLFSTTO.VLFS~I,v~V~T,UF|e,,,Kee. P,eS S~ $R NT T~PA~P L T GL L P L N P E 0 LAR LQAAT TO T , E 0 L~A UV , r- Y F UGV L N , R S G A 50 doxin electron transport m chain; the one exception, BM.3 P S T E = $ A K KVRK K AE N A HIM T P L L~]L~G T~]R °~IAD l~., IC['~IF A' ' '2 P450R K| °[TIT A['PlPv~]E $ $ F V''HK K I|G R N ! IMF Y[Y.=~S ° T GT, E_~E F~],RILISKIDA~H R , 6~JNR GN 110 Bacillus megaterium cyto- SR V A VW lill[l-Pv PLPj , ...... iN P R V T L I S AIS 0 T GINIAII~ R V E All.JR DJDiL L A A N L N V T 9& chrome P-450BM~, is unusual & in several respects. V['~TL O SH A G "I.'' ...... E GA~| V T S~H~IP~'-~I{ O~V~'~O~IA ~ A DE V K 100 ,lAID P E Era° L A° L S~]L, E I D K SILVV~FC .L~T 'fOE OJD]P]TO~A]° DIFI,~.~L]-QLOJ, T D V 9~IT 164 Cytochrome P-450BM~ is a LV,~ODI~IK~KO|~ISl " - o - EKILI~t VTSI~IOI r-EolEipmeelAIvALs~FItlFsKKAPKILIe ~ 119 kDa fatty acid (o-2 4 i [~R ¥ $~D~'~N U A T T , O K V P A F vI~E~OR T~AE AK-~E H-~A O R['~E A~A ' D ° F~lo T ' E 1,5 hydroxylase that has been I~ VI'[f'~'V 'L' "i"~TM'~]"['r]" ' "[~'~]' oF n A-~QIRl..~F E L L G D DO N L E Of ! :'18 characterized in the labora- ILL #TAIFAVF[ LS~---~JDTSY~.J'IEIFIFICOSIGKIOF SIC AEJLG~GEIRILLORVOA VE--YQAAASE 176 tory of Armand Fulco at '~I°"'s'''R °'v°' ...... F" "AADN 19'~ UCLA2~-25. Like all bacterial o ,MP AMC Is HFIo~],~T r. el~Is~:, °, ,klv v., o. o v A, v Y T G E 14 G R "I'" ,, o, ~ I P-450s, and in contrast to ~...I~~""~I'°M'I'*~I"L°'' ~ w ov L ~I~ A PlVlm~P - - S ~sL.I~ T ~. WV, ...... 0 '. ; ZZS I eukaryotic P-450s, P-450s~ is a cytosolic rather than "-- F O A K N L AJVJT R NQ~J'~T H L N L I S K | E S Y A YP~JA 327 membrane-bound ; unlike all other known P-450 monooxygenases, it requires P K E L VE~JL L~L K G~E P V T V 0 G ...... g T L P L A E~A LiE U H F E LWV IIT A :~26 no additional electron trans- port components for ac- [~QL RAH V P..C~HC K V LEALLEKOAYKEOVF~AKR L ...... T NF~']E L ['~E K AC "4~0 A~IKT ~l tivity. This catalytically self- MIMENYATLTR LLPLVGD AOLOHYAATTPIV° ...... RFSPA 369 sufficient enzyme contains 1 mole of heme and 1 mole each of FMN and FAD on a single polypeptide chain. ~' ~q,~ ° ,FE , ,, ¢, , ,, - - - , =M, TI~' ° ' '['Lu' " v -, =, °~r~q'l~'Fl'FI ',, The molecular weight and prosthetic group content FLI~.I-I~IORVE 6EV VF ! e HNONIFRLPJAapEllTeV!#rlIoPaTT! *P FJRS~F.JsloIoIml 474 suggested that P-450B~ ' °~l'~' '~' "F~IFF~ ,, .~r~Iorr~l. , , ° , ,~l..m-rn~ TF~r~l. , . =, ,,, might represent a fusion of u. ~l'~' ~m_~~ 'I E ' ~m_~' 'I' '.LI,I~ c ,I~ ~ oI ~ o,__~ ,I~ ELE q,~, .Molol,F,l°Itl. VIA F, 'I" E~l' " 6o, ~A e ~ v e~I~_~.I, .... ,~F F~. P. F T~L~.~JF ~I.~LI= ~--].~o~, V V~V ~I~. 0 ,~- ~I~I' e SZ~ genes corresponding to the I mammalian P-450 and its ~ITF~ v" ~ °~l"' 'FIL'[~_L ° ° "'F~rm~m~°F~l,~ ,~l~ .,s ,~o v. o v ,. ,6, microsomal reductase ~, a l¢lVl, v o,_lfl-q¢ "m" e "1' wlx~fl" ~ ~1= * " = v v c ~ =*--q"l" *l¢[b"vl° ,l~lr , o = vl,'~'p~e, ~ss K~JI[Y V O~O K[L[R E O 6 A ELEgit M~.JH ° "[G ,A H | Y V C G D & RJRHErA DL~E K A L L E V IJA E F GiG~LNJO L 578 supposition that has now been confirmed by se- quencing the genezs. T °lily oMv ~ ~1~1. , ~.~ VlS LI° v ~s 6,s The protein is divided into two discrete domains, one of which contains the Rgure 2 heme group, with the sec- Sequence alignment of reductases. The flavoprotein moieties of B. megateriurn cytochrome P-450.f,~l ond domain containing FMN (Ref. 25) (upper line, BM-3) and S. typhirnurium sulfite reductase 2~ (lower line, SR) are aligned with and FAD24. The two domains rat cytochrome P-450 reductase8 (middle line, P450R). Amino acids that match the rat reductase are boxed. Note that the P'450eM-Z sequence begins at position 59 of the rat reductase sequence, and are readily separated by lim- corresponds to amino acid 461 of the complete P'450BM.3sequence. The FMN phosphate-binding seg- ited proteolysis, and each ment is shaded, and the FMN ring-shielding residues are indicated with arrows. The glycine-rich and retains its respective pros- Cys-Gly dipeptide-containing segments in the proposed NADPH domain are also indicated by shading. thetic group(s) when separ- FAD-PPi, proposed FAD pyrophosphate-binding segment. ated. The heme domain of P-450,~ exhibits about 25% sequence identity with and a ferredoxin) discussed earlier. The Bacillus megaterlum P450 reductase mammalian fatty acid hydroxylases, earliest microsomal P-450s arose 1.3 Although evolutionary studies indi- and the flavin domain shows about 33% billion years ago, shortly after the cate that cytochrome P-450 should be identity with mammalian P-450 reduc- prokaryotic--eukaryotic divergence, present in virtually all organisms, tases ~s. The two domains are connected and presumably utilized microsomal including bacteria, its presence in pro- by approximately 30 amino acids, cytochrome P-450 reductase for electron karyotes has only been demonstrated which includes a highly basic segment input, as all present.day microsomai in a limited number of species to date cleaved during limited proteolysis. P-450s are dependent upon this flavo- (for review see Ref. 22). As noted above, As might be expected, several protein. Thus the evidence from this evolutionary studies would also indi- regions of the Bacillus and mammalian system suggests that the reductase was cate that bacterial I)-450 systems reductases contain highly conserved present in early eukaryotes. The evi- should utilize a two-component elec- segments, corresponding to regions dence presented in the next two sec- tron transport chain, rather than the likely to be involved in flavin binding. tions indicates that this compound compound flavoprotein, cytochrome An alignment of the Bacillus reductase flavoprotein almost certainly had ,°-450 reductase. Indeed, all but one of sequence with the rat microsomal evolved well before the emergence of the bacterial cytochrome P-450 systems reductase is shown in Fig. 2. In the FMN eukaryotes. so far characterized appear to use a domain, a segment proposed to bind 156 TIBS 16-APRIL "1991

the FMN phosphate group is strongly contain an Fe4S4 cluster and a complement (FMN plus FAD), while half conserved (residues 26-34, shaded in group2~. As in cytochrome P-450 re- may lack flavin completely. The actual Fig. 2), and of the two flavin ring-shield- ductase, each flavin has a distinct role distribution of the fiavin groups is not ing residues (indicated by arrows), the in the electron transfer sequence, with yet known. Siegel's group has proposed interior tyrosine (Tyr76) is conserved FAD serving as the entry port for elec- that two a-subunits are required for and the exterior tyrosine is replaced by trons from NADPH, and FMN serving as complete electron transfer, with each tryptophan (Trpll4), an amino acid the exit port to the [5-subuniP. Indeed, binding either FMN or FAD, and that for commonly found in this position in the studies from Siegel's laboratory on each subunit, the binding of one fiavin flavodoxins. The 'insertion' in the FAD the mechanism of electron transfer in precludes the binding of a second flavin domain is maintained in the Bacillus sulfite reductase served as a model for to that subunit. How this is ac- reductase, and, as with ferredoxin the characterization of P-450 reductase complished is not yet understood. reductase and cytochrome b5 re- by Vermilion et al.~; the recent determi- This arrangement raises two ques- ductase, the sequence similarity is nation of the amino acid sequence tions: how is the I:I ratio of FMN- to strongest in the carboxy-terminal of the sulfite reductase fiavoprotein has FAD-containing subunits established in region of the protein, the proposed revealed the basis for the similarity the octameric complex, and how is NADPH-binding domain. Two especially in the mechanism of these two re- intersubunit electron transport accom- conserved segments in this region are ductases 29. plished? Clearly there must be subtle noteworthy: a glycine-rich segment, The flavoprotein moiety (a-subunit) differences between sulfite reductase suggestive of a pyrophosphate-binding of sulfite reductase from either S. and P-450 reductase that facilitate the loop, found in virtually all nicotinamide typhimurium or E. coli shows about 30% octameric assembly and electron trans- cofactor-dependent enzymes (residues sequence identity with rat cytochrome port; one such difference may be the 436--450, shaded); and a segment con- P-450 reductase 29 (see Fig. 2). As with substitution of glutamine for tyrosine at taining a Cys-Gly dipeptide present in the Bacillus megaterium P-450 re- the interior FMN ring-shielding position, all P-450 reductases, and also in ferre- ductase, a segment corresponding to as noted earlier. As demonstrated by doxin reductase and cytochrome bs the fiavodoxin FMN phosphate-binding Shen et al.~, mutations at this position reductase (residues 633-643, shaded). residues is strongly conserved, but, in in P-450 reductase inhibit electron Although no function has been assigned contrast to the Bacillus reductase, only transport between the FMN and FAD to this latter segment, a cysteine in cyto- the exterior FMN ring-shielding residue groups without dramatically altering chrome bs reductase located nine is conserved (Tyr158); the interior tyro- FMN binding. Similarly, the presence of residues away from this conserved sine is replaced with glutamine glutamine at this position in sulfite dipeptide has been shown to be pro- (Glull9). The 'insertion' in the FAD reductase may inhibit intramolecular tected from chemical modification by domain, relative to ferredoxin re- electron transport and oblige the NADH19, supporting the assignment of ductase and cytochrome bs reductase, enzyme to shuttle electrons between nicotinamide cofactor binding to this is maintained in sulfite reductase, the fiavins of complementary subunits. region of these reductases. although it is somewhat shorter than in The Bacillus megateriurn P-4508M~ the P-450 reductases. As with the Conclusion thus contradicts two tenets regarding Bacillus protein, the greatest sequence The ability to generate, utilize and cytochrome P-450 monooxygenases: (I) similarity is found in the carboxy-ter- dispose of reducing equivalents in that bacterial systems utilize a two minal portion of the protein, corre- metabolic reactions is an essential life component, ferredoxin reductase/ferre- sponding to the NADPH-binding do- process, and almost certainly was doxln electron transport chain; and (2) main. Notably, the glycine-rich segment present in the earliest life forms. The that cytochrome P-450 reductase is present in the Bacillus and mammalian universality of electron transfer coen- restricted to eukaryotic internal mem- reductases is highly conserved in sul- zymes and cofactors (flavins and branes. The presence of the reductase rite reductase, as is the segment con- nicotinamide dinucleotides) argues that in this bacterium argues that the flavo- taining the Cys-Gly dipeptide. Thus, these molecules were utilized in the protein evolved well before eukaryotes despite the great evolutionary distance, earliest stages of evolution. It is thus and prokaryotes diverged, an argument all three reductases are clearly homolo- perhaps not surprising that proteins supported by the finding of this flavo- gous, and all have been strongly con- that facilitate the inherent activities of protein in sulfite reductase, as dis- served through evolution. these molecules are also evolutionarily cussed below. Despite the sequence similarity ancient. Cytochrome P-450 reductase between sulfite reductase and P-450 represents one such example. As pre- Sullite reductase reductase, the two proteins differ in one sented here, it is a component of three NADPH-sulfite reductase is a multi- important respect: the sulfite reductase evolutionarily remote pathways; meric enzyme that catalyses the six- fiavoprotein component, present as an whether it has been incorporated into electron reduction of sulfite to sulfide2s, octamer of identical a-subunits, binds other electron transport pathways and an obligatory step in the only eight flavin molecules rather than enzymes, in other organisms, remains of cysteine from sulfate in Escherichia the sixteen that would be expected to be seen. The multi-domain structure coli and Salmonella typhimurium. The based on its homology with P-450 of this enzyme has revealed its evol- enzyme, extensively characterized in reductase. As FMN and FAD are present utionary origins, and raised a wealth of the laboratory of Lewis Siegel at Duke in equal proportions, there are various questions to be addressed by struc- University, has a subunit structure of scenarios for flavin binding in this com- ture--function studies on this unusual a8134, in which the eight a-subunits bind plex: each subunit may bind one flavin flavoprotein. Foremost among these eight flavin molecules, four each of FMN molecule, either FMN or FAD; or half of questions must be the orientation of and FAD, and the four ~-subunits each the subunits may contain a full flavin the two flavins, and the mechanism by 157 TIBS 16-APRIL 1991 which electrons are transferred 7 Vermilion, J. L., Ballou, D. P., Massey, V. and (1985) Arch. Biochem. Biophys. 240, Coon, M. J. (1981) J. BioL Chem. 256, 172-177 between these two prosthetic groups. 266-277 19 Hackett, C. S., Novoa, W. B., Ozols, J. and 8 Porter, T. D. and Kasper, C. B. (1985) Proc. Strittmatter, P. (1986) J. Biol. Chem. 261, Note added in proof Nat/Acad. Sci. USA 82, 973-977 9854-9857 The recently determined crystal 9 Haniu, M., lyanagi, T., Miller, P., Lee, T. D. and 20 Porter, T. D., Beck, T. W. and Kasper, C. B. Shively, J. E. (1986) Biochemistry 25, (1990) Biochemistry 29, 9814-9818 structure of ferredoxin reductase3° 7906-7911 21 Nelson, D. R. and Strobel, H. W. (1987) MoL strongly supports the predicted domain 10 Katagiri, M., Murakami, H., Yabusaki, Y., Biol. Evol. 4, 572-593 structure of these flavoprotein re- Sugiyama, T., Okamoto, M., Yamano, T. and 22 SUgar, S. G. and Murray, R. I. (1986) in Ohkawa, H. (1986)J. Biochem. 100, 945-954 Cytochrome P-450 Structure, Mechanism and ductases. 11 Urenjak, J., Under, D. and Lumper, L. (1987) Biochemistry (Ortiz de Montellano, P. R., ecl.), J. Chromatogr. 397, 123-136 pp. 429-503, Plenum 12 Yabusaki, Y., Murakami, H. and Ohkawa, H. 23 Narhi, L. O. and Fulco, A. J. (1986) J. BioL References (1988) J. Biochem. 103, 1004-1010 Chem. 261, 7160-7169 1 Walsh, C. (1979) Enzymatic Reaction 13 Yamano, S., Aoyama, T., McBride, O. W., 24 Narhi, L. O. and Fulco, A. J. (1987) J. BioL Mechanisms, pp. 358-420, W. H. Freeman and Hardwick, J. P., Gelboin, H. V. and Gonzalez, Chem. 262, 6683-6690 Company F. J. (1989) MoL Pharmacol. 35, 83-88 25 Ruettinger, R. T., Wen, L-P. and Fulco, A. J. 2 Massey, V. and Ghisla, S. (1983) in Biological 14 Haniu, M., McManus, M. E., Birkett, D. J., Lee, (1989) J. Biol. Chem. 264, 10987-10995 Oxidations (Sund, H. and UIIrich, V., eds), pp. T. D. and Shively, J. E. (1989) Biochemistry28, 26 Siegel, L. M., Davis, P. S. and Kamin, H. (1974) 114-139, Springer-Verlag 8639-8645 J. Biol. Chem. 249,1572-1586 3 Lu, A. Y. H., Junk, K. W. and Coon, M. J. (1969) 15 Porter, T. D. and Kasper, C. B. (1986) 27 Siegel, L. M. and Davis, P. S. (1974) ./. Biol. J. Biol. Chem. 244, 3714-3721 Biochemistry 25, 1682-1687 Chem. 249, 1587-1598 4 Vermilion, J. L. and Coon, M. J. (1974) 16 Mayhew, S. G. and Ludwig, M. L. (1975) in The 28 Faeder, E. J., Davis, P. S. and Siegel, L. M. Biochem. Biophys. Res. Commun. 60, Enzymes (Boyer, P. D., ed.), pp. 57-118, (1974) J. Biol. Chem. 249, 1599-1609 1315-1322 Academic Press 29 Ostrowski, J., Barber, M. J., Rueger, D. C., 5/t.str6m, A. and DePierre, J. W. (1986) Biochem. 17 Shen, A. L., Porter, T. D., Wilson, T. E. and Miller, B. E., Siegel, L. M. and Kredich, N. M. Biophys. Acta 853,1-27 Kasper, C. B. (1989) J. Biol. Chem. 264, (1989) 1 Biol. Chem. 264, 15796-15808 6 lyanagi, T. and Mason, H. S. (1973) 7584-7589 30 Karplus, P. A., Daniels, M. J. and Herriott, J. R. Biochemistry 12, 2297-2308 18 Chan, R. L., Carrillo, N. and VaUejos, R. H. (1991) Science 251, 60-66

TIBS Editorial Policy As the leading review journal in biochemistry and molecular biology, TIBS enables researchers, teachers and their students to keep up with new and recent developments across this broad field.

Reviews form the foundation of each monthly issue. These articles, invited from leading researchers in a specific field, summarize and assess recent and important developments. Short Journal Club articles, which are generally written by a panel of regular correspondents, highlight recent research papers of particular interest. Other features articles provide a forum for discussion: Talking Point articles contribute new perspectives to existing debates, usually based on experimental findings, while Open Question accommodates more hypothetical views. New techniques and their application, or recent developments of established methods, are described in Emerging Techniques articles. Reflections on Biochemistry provides a historical perspective.

Articles for TIBS are generally invited by the Editors but ideas for Reviews and, in particular, Talking Point and Open Question features are welcome. Prospective authors should send a brief summary, citing key references, to a member of the Editorial Board or the Staff Editor in Cambridge, who will supply guidelines on manuscript preparation if the proposal is accepted.

The submission of completed articles without prior consultation is strongly discouraged. As much of the journal content is planned in advance, such manuscripts may be rejected primarily because of lack of space.

Authors should note that all Reviews and Features articles for TIBS are peer reviewed before acceptance and publication cannot be guaranteed.

158