The Structure and Mode of Action of the Cofactor of the Oxomolybdoenzymes D. Collison C. D. Garner and J. A. Joule Chemistry Department University of Manchester Manchester M 13 9PL UK 1 Introduction The concept that a small molybdenum-containing unit might act as a cofactor for the molybdoenzymes was first suggested by Pateman et al. 30 years ago; as a result of work on a series of mutant cells lacking both nitrate reductase and xanthine oxidase activity it was proposed that the two enzymes share a common cofactor.’ Support for this idea came from work by Nason et a1.* who showed that a molybdenum-deficient nitrate reductase from a mutant strain of Neurospora crassa Nit-1 could be reactivated by an acid-denatured extract from ‘any molybdoenzyme.’ Subsequent studies principally by Brill et al.achieved the isolation3 of a cofactor from Component I of nitrogenase containing iron molybdenum and acid-labile sulfide. However although this cofactor FeMoco was capable of reconstituting molybdenum- deficient nitrogenase Component I from the Azotobacter vinelandii mutant strain UW45 it did not reconstitute the nitrate reductase mutant Nit-1. On the other hand a cofactor isolated from xanthine oxidase did reactivate the Nit-1 strain -but not the nitrogenase mutant UW45 -this cofactor is now termed Moco and is the subject of this re vie^.^-^ Subsequently activity has been returned to Nit- 1 nitrate reductase using cofactor produced from other oxo- molybdoenzymes aldehyde oxidase sulfite oxidase and nitrate reductase.5 Moco or structural variants thereof is also the cofac- tor for xanthine dehydrogenase pyridoxal oxidase nicotinate oxidase carbon monoxide oxidase formate dehydrogenase tetrathionate reductase chlorate reductase biotin sulfoxide reduc- tase purine hydroxylase and dimethyl sulfoxide (DMSO) reduc-tase all enzymes involved in redox-xygen-transfer processes. In man hepatic sulfite oxidase is an essential enzyme serving a detoxification purpose converting sulfite into sulfate; lack of the enzyme leads in most cases to death soon after birth.6 The major- DK Joule graduated (University of Manchester) in 1958 and after a PhD (Manchester with G.F. Smith} and post-doctoral work (Princeton R. K. Hill Stanford C. Djei-assi) returned to Manchester as a lecturer with subsequent sabbatical periods at the University of Ibadan Nigeria Johns Hopkins Hospital Baltimore and the University of Maryland (UMBC) USA.He is now a Reader. Dr Joule’s reseurch L oncentrates on nitrogen heterocyclic chem- istry and synthesis of heterocyclic natural products Professor Garner graduated (University of Nottingham) in 1963 and after a PhD (Nottingham with Professor C. C. Addison Drs. J. A. Joule C. D. Garner ity of these enzymes are large and complex containing haem Fe- S and/or flavin centres in addition to MOCO.’ 2 Moco Structural Characterisation Moco is unstable when released from its associated protein and has only been characterised by conversion into derivatives all of which lack the molybdenum. Therefore ideas for the exact location of the metal have been developed using a combination of the evidence for the structure of the organic moiety termed molybdopterin spectro-scopic information as to the oxidation state of the metal and its primary coordination sphere in the enzyme (Section 2.2) and indications from the structures of pteridine-containing model com- pounds (Section 5).2.1 The Organic Ligand From experiments on denatured enzyme extracts Rajagopalan et al.x isolated two pteridines,g ‘Form A 1 (after exposure to H+/I,/KI) and ‘Form B’ 2a (K,[Fe(CN),]/HO-). They noted that the fluorescence of these compounds was not present in the whole enzyme and concluded that molybdopterin in Moco does not contain a fully aromatic pteridine and that 1 and 2a therefore must be formed by oxidative degradations. The phosphate in Form B could be removed by treatment with phosphatase the resulting material 2b being shown to contain a 1,2-diol by its positive reaction with periodic acid.The absolute configuration of the alcohol-bearing side-chain carbon in 1 was later established by comparison of the CD spectrum of the corresponding diol with that of synthetic material (the synthesis is summarised in Section 4).1° D. Sutton and S. C. Wallwork) and post-doctoral research (Caltech H. B. Gray; Nottingham ICI Research Fellow) was appointed as a lecturer at the University of Manchester in 1968. He has been vis- iting Professor at the Universities of Lausanne Strasbourg Melbourne and Florence and was awarded the Tilden Medal of the Royal Society of Chemistry in 198516.He is now Professor of Inorganic Chemistry and Head of Department. His research is con- centrated upon coordination chemistry relevant to the biological chemistry of the d-transition metals. Dr. Collison graduated (University of Manchester) in 1976 and after a PhD (Manchester with C. D. Garner and I. H. Hillier} and post-doctoral work (Manchester F E. Mabbs; Manchester C. D. Garner; UMIST N. J. Blackburn) became an SRC Fellow (Manchester) and subse- quently a Royal Society University Research Fellow (Manchester) and was appointed to his current position of Senior Lecturer at the University of Manchester in 1994.His research concentrates on the electronic structure of d-transition metal compounds.D. Collison 25 The stereochemistry shown for 2 is based on the assumption that it is the same as in 1. 1 2 R' R2 a H bH PO,H H c SMe H Urothione 2c is present in normal urine and is believed to be a metabolite of Moco;" as such it gives a clue to the locations of two sulfur atoms in molybdopterin. Rajagopalan and Johnson subjected urothione to Raney nickel hydrogenolysis; four compounds were produced one of which presumed to have a dihydrothiophene ring was dehydrogenated with selenium dioxide producing material identical with 'dephospho Form B' 2b. Key information'* on the structure of molybdopterin was gained when Rajagopalan and coworkers working with chicken liver sulfite oxidase and cows' milk xanthine oxidase produced a compound which they believed to represent a trapped molybdopterin.These and other data led to a formulation for Moco as 3 with uncertainties as to (i) the oxidation level of the pyrazine ring; (ii) whether the sulfur atoms are attached only (cf.urothione and Form B) at C-1' and C-2' of the side-chain; (iii) the oxidation levels of the side-chain carbons carrying the sulfur atoms; (iv) the tautomeric form of the (reduced) pteridine moiety and (v) the substituent on the phosphate. Other studies have demonstrated that in carbon monoxide dehy- drogenase the pterin unit is linked via the side-chain phosphate to a guanosine-5'-phosphate and in DMSO reductase to a cytosine-5 '-phosphate in each case via a pyrophosphate link; latterly hypo- xanthine and adenine have been shown to be associated with the molybdopterin in the same way.13 Very recently two X-ray crystal structure determinations have shed new light on molybdopterin-cofactors. A hyperthermophilic tungstopterin enzyme ferredoxin aldehyde oxidoreductase from Pyrococcus furiosus contains14 a cofactor which is fascinatingly similar yet subtly different from that proposed by Rajagopalan and other previous workers.The first difference is that in the tungsten cofactor 4 the metal has two pterin ene-dithiolate ligands. The second significant difference is the presence of a dihydropyran ring formed by the formal cyclisation of a side-chain hydroxy oxygen onto the pteridine 7-position of a 5,6-dihydropteridine. Repre- sentation 4 includes the two 0x0-groups indicated by a tungsten edge EXAFS study;IS the phosphates link to the same Mg2+ ion.I4 The crystal structure determination of the aldehyde oxidase from 4 Desulfovibrio gigas,I6 shows a molybdenum cofactor again in a tri- cyclic form comparable to that in 4 but in this case with a cytosine linked at the terminal phosphate and with only one ene-dithiolate ligating the metal centre 5. Considering the quantities of material available for the earlier chemical degradative and spectroscopic work on Moco the close- ness of the deductions to the structures now determined is CHEMICAL SOCIETY REVIEWS 1996 5 Hd \OH commendable.A caveat must be included regarding the accuracy with which the protein crystal structure determinations can delin- eate the state of oxidation at the pyrazine ring carbon atoms or at the side-chain sulfur-bearing carbon atoms. 2.2 The Metal Centre 2.2.1 Introduction Protein crystallographic studies on oxomolybdoenzymes and tung- sten-containing enzymes (tunzymes) have only recently become available (see above). However these studies do not unambigu- ously define the coordination at the metal. The clearest evidence concerning the coordination at the metal centre has been derived from a variety of spectroscopic studies. The complementary use of X-ray absorption spectroscopy (XAS) notably the extended X-ray absorption fine structure (EXAFS) of the molybdenum K-edge and electron paramagnetic resonance (EPR) spectroscopy have domi- nated these investigations because both are able to probe the metal's environment selectively.Interpretations of the data obtained from the enzyme studies have been significantly strengthened by recording corresponding XAS and EPR information for fully characterised chemical analogues. The prefix 0x0 for this group of enzymes is appropriate; not only does each enzyme catalyse a conversion the net result of which can be represented as oxygen atom transfer but also XAS studies have indicated the presence of at least one terminal 0x0 ligand (Mo=O) in (virtually) every system and state examined. Six-coordination and an octahedral geometry dominate the chemistry of molybde- num(vI) molybdenum(v) and molybdenum(1v).l7 For molybde- num(v1) the cis-dioxo moiety { MoO2l2+ achieves the pseudo-octahedral geometry by binding four donor atoms. Each bond trans to an Mo=O group is generally longer than an equiva- lent bond cis; neutral ligands are often found in the former positions and anionic ligands in the latter. The cis-dioxo geometry maximises the Mo(d,)-O(p,) overlaps. Binuclear complexes of molybdenum(v1) exist but the ten- dency to dimerisation via p2-OH linkages becomes dominant on reduction to molybdenum(v). This aspect of the chemistry means that the synthesis of most monomeric analogue complexes and studies of their spectroscopy and reactivity are performed in non- aqueous media.One terminal 0x0 (or sulfido) group is generally found for monomeric molybdenum(v) complexes and hence there is a single trans-site at which easy substitution chemistry can take place. Thus both five-and six-coordinate geometries are common. Both mono- and di-oxo complexes of molybdenum(1v) are found the latter possessing a mutually trans-dioxo geometry which places the d2 electrons in the same metal (d,) orbital leaving the two remaining d orbitals for Mo(d,)-O(p,) overlaps. 2.2.2 Spectroscopic characterisations 2.2.2.1 X-Ray absorption spectroscopy XAS has played a vital role in defining the chemical nature of molybdenum centres in enzymes and how they respond to changes in the oxidation level of Moco and/or to the presence of substrates substrate analogues or inhibitors of enzymic activity.l8 The molybdenum K-edge EXAFS results achieved19 for chicken liver sulfite oxidase are the clearest such data and the interpretation achieved represents a prototype for other oxomolybdoenzymes. The molybdenum site has been investigated in each of its three accessible oxidation levels [(VI) (v) and (rv)] as a function of pH and chloride concentration. The molybdenum(v1) is coordinated by two 0x0-groups at ca. 1.70 A one oxygen (or nitrogen) and three THE STRUCTURE AND MODE OF ACTION OF THE COFACTOR OF THE OXOMOLYBDOENZYMES-D. COLLISON ET AL. sulfur-donor ligands at ca. 2.06 and 2.42 A respectively; two of these sulfur atoms presumably derived from the molybdopterin.8 The molybdenum(vr) centre is not affected by changing the pH from 6 to 9 or by a variation in the chloride concentration.The molybdenum-(v) and -(Iv) centres each possess a single oxo- ligand at ca. 1.69 A one oxygen (or nitrogen) and three sulfur- donor ligands at ca. 2.00 and 2.37 A respectively. Both of these centres appear to bind chloride at pH 6 in 0.3 mol 1-I KCl. EPR spectroscopy showed that the centre can exist in two different forms which are in a pH- and anion-dependent equilibrium. George et al." concluded that the molybdenum K-edge EXAFS data were consistent with one chloride ligand binding to the low pH form and that the number of 0x0-groups remains the same upon transition from the high-pH to the low-pH molybdenum(v) form. Thus reduc- tion of molybdenum(vr) results in the loss of one 0x0-group pre- sumably due to protonation and the generation of an anion binding site.This behaviour is consistent with the chemistry of molybde- num in its higher oxidation states since a cis-{MO~IO,}~+centre is +generally converted into a { MoVO} or { MoIVO} 2+ centre upon reduction. Xanthine oxidase is the most accessible of the oxomolybdo- enzymes and is readily extracted from cows' milk. This enzyme exists in two forms an active and an inactive form caused by loss of a sulfur atom (desulfo). Molybdenum K-edge EXAFS studies20 have shown that the environments of molybdenum(vr) and molybdenum(rv) in desulfo xanthine oxidase closely resemble that l9 of the corresponding oxidation state for chicken liver sulfite oxidase.The principal differ- ence between the centre of the oxidized active form as compared to the oxidized desulfo form is the presence of one sulfido-group (at ca. 2.18 A) plus one 0x0-group rather than two oxo-groups.21 The molybdenum centre of xanthine oxidase is very reactive and both molybdenum K-edge EXAFS and EPR data indicate that the centre of this reactivity is the Mo=S bond. The terminal sulfido-group is lost upon reduction presumably being protonated to form an Mo-SH moiety. Arsenite is a potent inhibitor of xanthine oxidase clear evi- dence for an Mo- S-As interaction and an interbond angle of ca. 80" has been obtained from combined Mo and As K-edge EXAFS studies. 2.2.2.2 Electron paramagnetic resonance spectroscopy EPR spectroscopy selectively probes the molybdenum(v) d1 centres of the oxomolybdoenzymes. The EPR parameters (g-and A-values) of the centre are extremely sensitive to the nature of the coordination sphere.The role of the molybdenum(v) state in the enzymatic reactions has been questioned but nonetheless this state is important since it can be generated within biological samples and it is the only state (other than possibly MolIr) which can be detected by EPR spectroscopy. The work of Bray et ~l.,~,using EPR spectroscopy has profoundly influenced views as to the nature of the active sites of the oxomolybdoenzymes. Indeed the initial experiments of Bray et a1.z3 in 1959 and Meriwether et in 1966 were indicative of sulfur-donor ligands bound to molyb- denum. The presence of nuclei such as IH I3C I4N I7O,31P or 33Sin or near to the coordination sphere of molybdenum(v) can be revealed by EPR spectroscopy as super-hyperfine splittings of resonances.More recently it has become technically feasible to use lH or 31P electron nuclear double resonance (ENDOR) on enzyme samples containing molybden~m(v).2~ 2.2.2.3 Magnetic circular dichroism spectroscopy (MCD) DMSO-reductase from Rhodobacter capsulatus26 and Rhodobacter sphaeroidesZ7 is a soluble periplasmic enzyme (M,= ca. 82000) which contains only Moco as a prosthetic group. This simplicity greatly facilitates spectroscopic studies of the molybdenum centre and this has been exploited in an MCD spectroscopic study of the molybdenum(v) state of these enzymes.,* The spectrum shows six oppositely signed bands ranging in wavelength from 701 to 358 nm which are assigned as dithiolene S to MoVcharge-transfer transitions.2.2.2.4 Resonance Raman spectroscopy Resonance Raman spectroscopy has been used to probe the metal coordination in a variety of metal lop rote in^.^^ However for most pterin-containing molybdenum enzymes other strongly absorbing prosthetic groups dominate the electronic and resonance Raman spectra and to date only DMSO reductase has been studied by this technique. The oxidized and reduced forms of DMSO reductase show vibrations in the 335-385 cm-I region that shift upon enrichment of the enzyme with 34S and therefore have been assigned to Mo-S vibrations.The most prominent feature is the band at 350 cm-' in oxidized DMSO-reductase which shifts to 343 cm-l upon 34S enrichment and has been assigned to a Mo-S (dithiolene) vibration by comparison with the bands at 351 and 348 cm-' in [C,H,MoIV{S,C,[C(O)Me]-quinoxalino}] and [(C,H,),MorV( S,C,[C(O)Me]-pterin J 1 respectively which shift by the same amount.30 3 Synthesis and Properties of Compounds which model the Bioactivities of Moco-containing Enzymes 3.1 Oxygen Atom Transfer 'Molybdenum . . . lies at the epicentre of 0x0 transfer chemistry. More 0x0 compounds have been prepared and characterized more 0x0 transfer reactions are known and more catalytic systems based on these reactions have been devised than for any other and these have been comprehensively re~iewed.~ Molybdenum enzymes catalyse the overall reaction shown by equation 1 where X is the enzyme substrate.The electrons and protons produced by the oxidation of the substrate (or consumed in the reduction of the substrate) may be involved with the molybdenum as the (formal) { MO~IO,}~+/{MoIVO(OH,)J2+ couple. Alternatively the molyb- denum centre may be involved in direct oxygen atom transfer (equation 2.). X +H,O,'XO + 2H+ + 2e (1) Both routes can be employed depending on the enzyme and the operating conditions. Recent work by Holm et al. has demonstrated that DMSO-reductase from R. sphaeroides is an oxotransferase. Thus the overall transfer of an oxygen atom ('*O) from DMSO to 1,3,5-triaza-7-phosphatricyclo[3.3.1.13.7]decane (PTA) is catalysed by the enzyme and the labelling of the substrate demonstrated that the oxygen atom transferred did not arise from the solvent? Mechanistic versatility for oxomolybdenum complexes has been demonstrated by the reaction sequence summarised in Scheme 1.33 This system based on the tris(pyrazoly1)borate ligand (L-N,) models some aspects of the reaction chemistry of sulfite oxidase. [(L-N,)MoV102(SPh)] reacts with PPh to produce [(L-N3)MorVO(SPh)] and is capable of catalysing oxygen atom transfer from Me,SO to PPh,.In the presence of H,O a one-electron reduction takes place to yield [(L-N,)MovO(OH)(SPh)] which can be oxidised by 0 to regenerate the starting material. Oxygen isotope labelling reveals that H,O is the source of the 0x0 ligand not 0 and the oxidation state of the molybdenum is suggested to control the level of protonation of the water-derived ligand.LMovO(OSiMe,)(SPh) 'A [CoCpz][LMovOz(SPh)] COCP2. PY \ LMoJVO(SPh)(py) LMO* 'U2(Sk'h) Scheme 1 Holm et al. have developed some elegant ligand~~~ capable of sustaining molybdenum as a monomeric centre during oxygen atom transfer. The compound 6 synthesised by the reaction of [MoO,(acac),] with the dithiol pro-ligand oxidises phosphines and the resultant molybdenum(1v) form 7 reduces N-and S-oxides. CHEMICAL SOCIETY REVIEWS 1996 4 Synthesis of Degradation Products of Moco Following of deoxyurothione the total synthesis39 of (2)-urothione is a triumph for Taylor's strategy for the regioselec- tive synthesis of 4-substituted pteridines; Scheme 4 summarises the key steps.Scheme 2 Reagents i MeOH CH,Cl room temp. (92%); ii DMF room temp. (65%). The redox pair 8 and 9 oxidise/reduce a variety of substrates such as P- Se- and N-oxides some of which are substrates for oxo-molybdoenzymes. Ar,,Ar Ar ,Ar Ar = 4-But C,H 3.2 Dithiolene Complexes The coordination of molybdenum by sulfur demonstrated by spectroscopic studies of the oxomolybdoenzymes together with the constitution of Moco has usually been interpreted to indicate dithiolene (or ene- 1,2-dithiolate) ligation (see 5). The results of the crystallographic studies of the aldehyde oxidoreductase from D. gigasI6 and P.~U~~OSUS'~are consistent with the coordination of one dithiolene to molybdenum and two dithiolenes to tungsten respectively. 2-Scheme 3 Reagents i Na,MoO,. 2H,O H,O pH 6 room temp. (6I %). Dithiolene complexes generally display reversible redox proper- ties. This behaviour is vital for any chemical analogue of the molyb- denum centre of the oxomolybdoenzymes. The maleonitrile dithiolate (mnt) ligand has been to afford [MoO2(mnt),I2- and [M~O(mnt),]~- (Scheme 3) as well as [M~OCl(mnt),]~-; a set of complexes which further reinforce the comparison between these systems and the molybdenum centres of the oxomolybdoenzymes. However the redox potentials reported for oxomolybdoenzymes two couples corresponding to MoV1/MoVand MoV/MoIV processes separated by only ca.200 mV contrast with the observation of two one-electron processes for [Mo(dithiolene),]"- and [MoO(dithio- lene),]"-(n = 0 1 2) complexes which are separated by 3 500 mV.35.36 An intriguing aspect of the properties of Moco is the extent to which the molybdenum and the pterin jointly participate in the redox changes of the centre. Chemical support for this postulate has been demonstrated by studies of [Mo(qdt),I2- (qdt = quinoxaline-2,3-dithiolate) and [(q5-C5H,)Co{S2C,H(quinoxalin-2-y1) }] systems.37 Thus not only may the extent and nature of this cooperation vary from enzyme to enzyme but also it may be meaningless to attempt to separate the metal and ligand redox contributions. Also such behaviour provides an attractive mechanism for modulating the redox potential of Moco by a protein via a control of the state and extent of protonation of the pyrazine ring and the stabilisation of par- ticular tautomers of the partially reduced forms of the pteridine. 0 SMe Scheme 4 Reagents i Et,N EtOH room temp.(92%); ii LiBF aq. MeCN room temp. (98%); iii NaOAc Bu'OH reflux (99%); iv Bu'ONO/CuBr MeCN reflux (56%); v NaSMe THF room temp. (93%); vi NaBH EtOH THF room temp. (90%);vii HC(OMe) p-MeC,H,SO,H room temp. (88%);viii p-MeC6H,S0,H MeOH reflux (86%);ix guanidine hydrochloride NaOMe reflux; x CF,CO,H room temp.; xi 3 mol 1-' H,SO reflux (7996 two steps). If in molybdopterin the sulfur at C-2' is not also linked to the pteridine C-7 then both the formation of urothione as a metabolite and the production of Form B during degradation must involve a cyclisation to produce the thiophene ring this has been inadver- tently m~delled.~~.~' Reaction of the phenyl quinoxalin-2-yl alkyne 10 with [Mo(S4),SI2- (see Section 5 below for fuller discussion of such additions) gave a tris(dithio1ene) complex of molybdenum.Oxidation of this complex produced as well as higher oxidation states of the molybdenum complex a small amount of a metal-free substance shown to be 11 (Scheme 5). 2-11 Scheme 5 Reagents i MeCN reflux (68%);ii e.g.I -50 "C (7%). Reaction of the quinoxaline-dibromoalkene12 with dipotassium trithiocarbonate produced as the main product a tricyclic thiophene then hydrolysed to give 13.It was suggested that the mechanism pro- posed (Scheme 6) to rationalise this unexpected product may well have a bearing on the formation of urothione and of Form B.s Taylor was the first to prove unequivocally the structure of 'Form A' (dephospho) by total synthesis (Scheme 7) in racemic4 and then later in homochiral In the key step a homochiral alkyne 14 obtained from D-mannitol was coupled with a pivaloyl-protected 6-chloropteridine using palladium(0) methodology. Taylor's use of an N-pivaloyl group as a lipophilic protecting group for pteridines considerably facilitates their handling -they are otherwise often very insoluble. THE STRUCTURE AND MODE OF ACTION OF THE COFACTOR OF THE OXOMOLYBDOENZYMES-D. COLLISON ET AL.-HBr then ii l3 f Scheme 6 Reagents i aq. K,CS MeOH room temp. (72%);ii 48% HBr MeOH CH,Cl room temp. (70%). 15 \ vi\ 0 a 16 Scheme7 Reagents:i,Pd(OAc) (0-Tol),P,CuI,Et,N,MeCN 1OOoC(20%); ii 0.5 mol 1-' HCl reflux; iii K,CO MeOH (42%); iv Et,N/MeSO,Cl,CH,Cl,,O "C (72%); v Ph,P MeCN 90 "C (76%); vi BunLi,THE -78 "C then room temp. in solution (equilibration to all E-alkene) (72%);vii Br CH,Cl 0 "C (67%);viii DBU dioxane 100"C (40%). An alternative route to alkyne-acetal 15 starting from ester-amide 16 itself availableM either viadegradation of folic acid or from syn- thetic 6-hydroxymethylpteri11 is also shown in Scheme 7.45 5 Towards a Total Synthesis of Moco Any strategy aimed at a total synthesis of Moco must have several components; one of these is a means for the generation of a molyb- denum-complexed 1,2-dithiolate (probably an ene- 1,2-dithiolate -a 'dithiolene').Several methods have been described for the pro- duction of such units around molybdenum and other metals. We summarise below those routes which we believe to be of greatest relevance for such synthetic endeavours. Coucouvanis' studies46 of the reactions of the polythiomolybde- num anions generated by reaction of [MoS,I2- or [MoS,O,-,]~- (x = 4 3 or 2) with sulfur may be relevant to the biosynthesis of the dithiolene unit in Moco and could be of value in a laboratory synthesis of molybdenum-dithiolene complexes (see also below). Mechanistic sequences were suggested for the reaction of such anions with dimethyl acetylenedicarboxylate; illustrated in Scheme 8 is the reaction of [(S4)2M~=S]2- under anaerobic conditions giving 17. Shown below are two possible interpretations for the key C-S bonding step the process is (i) initiated by nucleophilic attack by a terminal sulfur on the acetylene with the electron reorganisation shown by the curly arrows on 18,or (ii) viewed as a [2 + 21 cyclo-addition 19.Taylor and Stiefel elegantly utilised similar additions in their syn- theses of pteridinyl- and quinoxalinyl-dithiolenes 20 and 21.47The E 17 E=C02Me Scheme 8 19 initial complexes 22 and 23 could be transformed into the ene-dithi- olate systems believed to exist in molybdopterin by treatment with a phosphine (Scheme 9). Hartzler had demonstrated4* that alkynes carrying at least one electron-withdrawing substituent (ester) react with the betaine (Ph,P+ -CS,-) produced from tributylphosphine and carbon dis- ulfide generating an ylide which can be trapped in a Wittig process.Taylor capitalized on this idea showing that addition to the pteridinyl ketone 25 occurred in the same fashion (Scheme It is important to note that an attempt to utilize (protected) alcohol 24 yielded 'only a minute yield' of addition product -it seems that the electron-withdrawal provided by the heterocycle does not acti- vate the alkyne sufficiently for addition to occur. de Mayo showed that diphenyldithiolene-thione26 reacts with Mo(CO) in the presence of light generating complex 27 though in low yield; a somewhat better yield was obtained in a thermal process using the same metal complex but starting from diphenyl- acetylene and sulfur (Scheme 1 l).so An alternative for the further processing of tri- or di-thiocarbon- ates could rest on Rauchfuss' work on tetrathiapentalenedione 28. 0 n /? \Iqylis-s 22 23 20 21 Scheme 9 Reagents i DMF 70 "C (62%22,36% 23); ii Ph,P DMF 70 "C (95%).CHEMICAL SOCIETY REVIEWS 1996 OTBDMS 0 24 25 Scheme 10 Reagents 1 Bu,NF THF room temp (93%) 11 CrO aq H,SO butan-2-one room temp 111 THF -40 OC iv (EtO,C),C=O -50 "C (69% two steps) Ph Scheme 11 Reagents 1 hv CHCl (9%) 11 PhMe 120 "C (25%) 2-28 L J Scheme 12 This compound served as a precursor to 1,2-dithiolene complexes of molybdenum by reacting with tetrathiomolybdate (Scheme 12) 5' Early studies in Manchester established a route to unsymmetrically substituted trithiolenes hydrolysis of the immediate precursors for the tnthiolenes generated ene-dithiolates in solution which could be trapped by reaction with for example [MoO,(acac),] 52 In the light of remaining uncertainties regarding the oxidation level of sulfur- beanng carbons in molybdopterin the option of producing 1,2-dithiolane precursors was addressed in a model system as shown in Scheme 13 53 It was possible to convert the model dithiolane into the dithiolene vza peracid oxidation Scheme 13 Reagents 1 Br AcOH 90 "C (85%) 11 NaS,CNMe EtOH reflux(67%) 111 conc H,SO room temp (58%) then NaHS aq AcOH room temp (64%),iv [Co(C,H,>(cyclooctadiene)] xylene reflux (51%) v NaBH MeOH 0 "C (87%) vi MeSO,Cl pyndine 0 "C (67%) vii m-ClC,H,CO,H CHCl 0 "C then (CF,CO),O to room temp (84%) It has been known for more than 100 years that ortho-phenylene- diamine reacts with glucose giving 2-(~-urubzno-tetrahydroxy-butyl)quinoxaline 29 Manchester studies41 s4 have utilised this readily available substance as the starting point for studies aimed at modelling methods for the elaboration of the C,-side-chain of molybdopterin (Scheme 14) An extrapolation of the strategy illustrated above into pteridine chemistry required the development of a practical route to a 6-tetrahydroxybutylpterin Although the condensation of 29 Me$' I1 Vlll 1x Me Me Scheme 14 Reagents 1 Me,CO conc H,SO room temp (51%) 11 HC(OEt) p-MeC,H,SO,H CH,Cl room temp (93%) 111 Ac,O reflux (76%) IV Br CH,Cl 0 "C (67%) v NaS,CNMe MeOH reflux (40%) vi Br CH,Cl 0 "C (86%) vii NaOMe MeOH room temp (34%),viii MeSO,Cl pyndine room temp (53% plus 30% dimesylate) ix NaS,CNMe EtOH reflux (73%) x (CF,CO),O pyndine reflux xi H,S room temp (30 9% 31 25% 32 ca 50%) xii BnO,CCl NaB(CN)H MeOH room temp (65%) 5,6-diaminopyrimidines with 1,2-dicarbonyl compounds leading to pteridines is a well-known approach to pteridines -the Isay syn- thesis -it usually produces mixtures of 6- and 7-substituted pterins -early reports of the regioselective reactions of sugars with 4,5-diaminopyrimidines had been shown to be in error However a very careful examination of reaction conditions produced a recipe for the synthesis of (D-aruhzno)-tetrahydroxybutylpterin,as its tetraacetate-acetamide 33 in acceptable yields on a multigram scale as shown in Scheme 15 5s The pterin 33 has been processed following the quinoxaline model sequence to afford alkene 34 HO HH OH % 33 Me 34 Scheme 15 Reagents 1 N,H H,O AcOH 100 "C (38%) 11 Ac,O pyn dine 100 "C then recrystallise (70%) 111 K,CO MeOH room temp (55%) iv Me,CO p-MeC,H,SO,H room temp v HC(OEt) p MeC,H,SO,H pyndme CH,Cl room temp (44% two steps) vi 160 "C (67%) The oxidation level and the precise tautomeric form adopted by the pteridine in Moco are still not certain because of this any synthetic strategy must allow vanation in the oxidation level of the pyrazine nng 56 To this end it was shown that reductions of quinoxalinyl-trithiolene35 and quinoxaline-trithiolane 36 produced tetrahydr0-N- protected derivatives 37 and 38," in each of which importantly the sulfur-containing units were untouched It has subsequently been shown that 37 can be selectively oxidized to the mono-protected- dihydroquinoxaline 39 which can then be transferred to a metal centre to give a dithiolene complex 40 (Scheme 16) 58 THE STRUCTURE AND MODE OF ACTION OF THE COFACTOR OF THE OXOMOLYBDOENZYMES-D COLLISON ET AL 10 11 12 13 36 Scheme 16 Reagents I BnO,CCl NaB(CN)H (97%) 11 MnO CH,CI 14 room temp (72%) 111 [Co(C,H,)cyclooctadiene)] (13%) iv BnO,CCl NaB(CN)H (42%) 15 A comparable reduction of 6-substituted-pteridines of which the 16 most relevant to molybdopterin synthesis is 33 produced interest- ingly tetrahydro-derivatives (40) but with the protecting group on 17 N-5 adjacent to C-6-substituents (Scheme 17) 55) 18 BnOC 19 33 41 ti 20Scheme 17 Reagents 1 BnO,CCl NaB(CN)H (65%) 21 22 236 Conclusion This is an exciting time in the development of the understanding of 24 the nature and mode of action of the molybdenum cofactors of the oxomolybdoenzymes and their tungsten counterparts Thus protein 25 cystallography is now providing vital information which comple- ments earlier spectroscopic and chemical studies the total synthesis of Moco is in prospect and with this and the preparation of variants 26 and close analogues studies of the natural systems and the mode of action will be considerably augmented 27 Acknowledgements Work in Manchester has been supported by the 28 SERC and now by the EPSRC and the Royal Society (D C ) the assistance from which we gratefully acknowledge 29 307 References 1 J A Pateman,D J Cove,B M ReverandD B Roberts,Nature 1964 31 201,58 2 P A Ketchum H Y Cambrier W A Frazier C H Madansky and A Nason Proc Natl Acad Sci USA 1970,66,1016,A Nason K Y Lee S S Pan P A Ketchum A Lamberti and J Devnes Proc Natl Acad 32 Sci USA 1971,68,3242 33 3 V K Shah and W J Brill Proc Natl Acad Sci USA 1977,74,3249 P T Pienkos V K Shah and W Brill Proc Natl Acad Sci USA 1977 34 74,5468 4 In the light of several previous reviews on Moco and related topics (refs 5,6a and S Goswami Heterocycles 1993,35 155 1) we reference here 35 only key original papers concerned with the development of the subject 5 J L Johnson in Molybdenum and Molybdenum Containing Enzymes 36 ed M P Coughlan Pergamon Press Oxford 1980 pp 345-383 J C Wootton R E Nicholson J M Cock D E Walters J F Burke W A Doyle and R C Bray Biochem Biophys Acta 199 1,1057 I57 37 6 (a)J L Johnson in The Metabolic Basis of Inherited Disease ed C R Scriver A L Beaudet W S Sly and D Valle McCraw Hill New York 1989 p 1463 (b) G K Brown R D Scholeum H B Croll J E Wraith and J J McGill Neurology 1989,39,252 7 Molybdenum Enzymes ed T G Spiro Wiley New York 1985 38 Molybdenum Enzymes Cofactors and Model Systems ed E I Stiefel D Coucouvanis and W E Newton ACS Symposium Series 535 1993 39 J H Enernark and C S Young Ad\ Inorg Chem 1994,40 1 40 8 J L Johnson B E Hainline and K V Rajagopalan J Biol Chem 1980 255 1783 J L Johnson B E Hainline K V Rajagopalan and 41 B H Arison .I Bid Chem 1984,259 5414 9 The name ‘pteridine’ for pyrazino[2,3-d]pynmidine was coined by 42 Wieland and refers to the first isolations of such compounds in nature -43 the pigments of butterfly wings The term ‘ptenn’ is used for 2-amino 4 hydroxyptendine For a review of pteridine chemistry see W 44 Pfleiderer I Heterocycl Chem 1992,29 583 E C Taylor P S Ray I S Darwish J L Johnson and K V Rajagopalan J Am Chem Soc 1989,111,7664 J L Johnson and K V Rajagopalan Proc Natl Acad Sci USA 1982 79,6856 S P Kramer J L Johnson A A Ribeiro D S Millington and K V Rajagopalan J Bid Chem 1987,262 16357 B Kruger and 0 Meyer Eur J Biochem 1986,157 12 1 J L Johnson K V Rajagopalan and 0 Meyer Arch Biochem Biophys 1990,283 542 J L Johnson N R Bastian and K V Rajagopalan Ptoc Natl Acad Sci USA 1990 87 3190 G Bomer M Karrasch and R K Thauer FEBS Lett 199 1,290,3 1 M K Chan S Mukund,A Kletzin,M W W Adams andD C Rees Science 1995,267 1463 G N George R C Prince S Mukund and M W W Adams I Am Chem Soc 1992,114,3521 M J Romiio M Archer I Moura J J G Moura J LeGall E Engh M Schneider P Hof and R Huber Science 1995,270 1170 C D Gamer and S Bristow in Metal inns in Bioloqy vol 7 ed T G Spiro Wiley New York 1985,pp 343-410 S P Cramer in Adkances in inorganic and Bioinoi yanic Mechanism.ed A G Sykes Academic Press London 1983 p 259 B Hedman K 0 Hodgson and C D Gamer Synchrotron Radiation and Biophlsic v ed S S Hasnain Ellis Horwood Chichester 1990,p 43 C D Garner Ad1 Inorg Chem 199 1,36,703 G N George C A Kipke R C Prince R A Sunde J H Enemark and S P Cramer Biochemistry 1989,28 2075 N A Turner R C Bray and G P Diakun Biochem I 1989,260,563 S P Cramer and R Hille J Am Chem Soc 1985,107 8164 R C Bray Quart Re\ Biophys 1988,21,299 R C Bray B G Malmstrom and T VanngArd Blot hem I 1959 73 193 L S Menwether W F Marzluff and W G Hodgson Name 1966,212 465 B D Howes N M Pinhal N A Turner R C Bray G Anger A Ehrenberg J B Raynor and D J Lowe Bloc hemistry 1990,29,6 120 B D Howes B Bennett A Koppenhofer D J Lowe and R C Bray Biochemistry 1991 30 3969 A G McEwan S J Ferguson and J B Jackson Biochem I 1991 274 305 N R Bastian C J Kay N J Barbara and K V Rajagopalan I Brol Chem 199 1,266,45 N Benson J A Farrar A G McEwan and A J Thomson FEBS Left 1992,307 169 M G Finnegan J Hilton K V Rajagopalan and M K Johnson Inorg Chem 1993,32,2616 Biological Applications of Raman Spectroscopy ed T G Spiro Wiley New York 1988 vol 3 L Kilpatrick K V Rajagopalan J Hilton N R Bastian E I Stiefel R S Pilato and T G Spiro Biochemistry 1995,34 3032 (a) R H Holm and J P Donahue Polyhedron 1993 12 571 R H Holm Coord Chem Re\ 1990,110 183 (b)R H Holm Chem Rer 1987,87 1401 (c)R H HolmandJ M Berg,Acc Chem Res 1986 19,363,Pure Appl Chem 1984,56 1645 B E Schultz R Hille and R H Holm J Am Chem Soc 1995 117,827 Z Xiao C G Young J H Enemark and A G Wedd J Am Chem Soc 1992,114,9194 J A Craig and R H Holm J Am Chem Soc 1989 111,2 1 11 and references therein B E Schultz S F Gheller M C Muettertiey M J Scott and R H Holm J Am Chem Soc 1993,115,2714 S K Das P K Chaudhury D BiswasandS Sarkar J Am Chem Soc 1994,116,9061 E I Stiefel R Eisenberg R C Rosenberg and H B Gray,J Am Chem Soc 1966,88,2956,S L Soong V Chebola S A Koch T O’Sullivan and M Millar Inorg Chem 1986,25,4068 C D Gamer,E M Armstrong,M J Ashcroft M S Austerberry J H Birks D Collison A J Goodwin J A Joule L Larsen D J Rowe and J R Russell in Molybdenum Enzymes Cofactors and Model Systems ed E Stiefel D Coucouvanis and W E Newton ACS Symposium Series 535 1993 pp 98-1 13 A Sakurai and M Goto,J Biochem (Tokyo) 1969,65,755,E C Taylor and L A Reiter J Org Chem 1982,47,528 E C Taylor and L A Reiter .IAm Chem Soc 1989,111,285 C L Soncelli V A Szalai and S J N Burgmayer J Am Chem Soc 1991,113,9877 L Larsen D J Rowe C D Garner and J A Joule J Chem Soc Perkin Trans I 1989 2317 E C Taylor and S Goswami Tetrahedron Lett 1991,32 7357 E C Taylor P S Ray I S Darwish J L Johnson and K V Rajagopalan J Am Chem Soc 1989,111,7664 J R Russell C D Gamer and J A Joule J Chem Soc Peikin Trans I 1992 1245 45 J R Russell C D Garner and J A Joule Tetrahedron Lett 1992,33 337 1 46 D Coucouvanis A Hadjikynacou A Toupadakis S -M Koo 0 Ileperuma M Draganjac and A Salifoglou Znorg Chem 1991 30 754 D Coucouvanis A Toupadakis S -M Koo and A Hadjikyriacou Polyhedron 1989,8 1705 and references cited therein 47 R S Pilato K E Enksen M A Greaney E I Stiefel S Goswami L Kilpatrick T G Spiro E C Taylor and A R Rheingold .I Am Chem Soc 1991,113,9372 48 H D Hartzler J Am Chem Soc 1971,93,4961 49 E C Taylor and R Dotzer J Org Chem 1991,56 1816 50 W Kusters and P de Mayo J Am Chem Soc 1974,96 3502 G N Schrauzer and V P Mayweg Z Nuturforsch Tell B 1964 19 192 51 X Yang G K W Freeman T B Rauchfuss and S R Wilson Znorg Chem 199 1,30,3034 52 S Boyde C D Garner J A Joule and D J Rowe .I Chem Soc Chem Commun 1987,800 CHEMICAL SOCIETY REVIEWS 1996 53 L Larsen C D Garner and J A Joule .I Chem Soc Perkin Trans I 1989,2311 54 L Larsen D J Rowe C D Gamer and J A Joule Tetruhedron Left 1988,29 1453 55 J R Russell C D Gamer and J A Joule Synlett 1992 71 1 R L Beddoes J R Russell C D Gamer and J A Joule Acta Crystulloqr Sect C 1993,49 1649 56 J R Russell C D Gamer and J A Joule J Chem Soc Prrkin Ttuns I 1992 1245 57 R L Beddoes J R Russell C D Gamer and J A Joule Acts Crystalloqr Sect C 1992,48 2075 58 J H Birks PhD Thesis University of Manchester 1994 59 R L Beddoes J R Russell C D Gamer and J A Joule Atru Ctystallogt Sect C 1993,49 1649