The Biosynthesis of Porphyrins Chlorophylls and Vitamin BIZ F. J. Leeper University Chemical Laboratory Lensfield Road Cambridge CB2 7 E W Reviewing the literature between December 1977 and December 1983 1 The Biosynthesis of Haem 1.1 6-Aminolaevulinic Acid Synthase 1.2 The Synthesis of 6-Aminolaevulinic Acid in Plants 1.3 6-Aminolaevulinic Acid Dehydratase (Porphobilinogen Synthase) 1.4 Formation of Uroporphyrinogen I11 Hydroxymethyl-bilane Synthase (Porphobilinogen 1 Deaminase) and Uroporphyrinogen I11 Synthase (Cosynthetase) 1.4.1 Proof that 1-Hydroxymethylbilane is the Intermediate 1.4.2 The Mechanism of Deaminase 1.4.3 The Mechanism of Cosynthetase 1.4.4 Determination of Ratios of Isomers of Uroporp h yrinogen 1.5 Uroporphyrinogen Decarboxylase (Uroporphyrinogen Car boxy -1yase) 1.6 Coproporphyrinogen Oxidase 1.7 Protoporphyrinogen Oxidase 1.8 Ferrochelatase (Haem Synthase Protohaem Ferro- lyase) 1.9 Cytoc hromes 2 The Biosynthesis of Chlorophylls 2.1 Chelation of Magnesium and Methylation 2.2 The Formation of Ring E 2.3 Reduction of the Vinyl Group 2.4 Reduction of Ring D 2.5 Esterification of Chlorophyllide a 2.6 The Formation of Chlorophyll b 2.7 Bac terioc hlorop hylls 2.8 Conclusion 3 The Biosynthesis of Vitamin B12 3.1 The Basic Precursors 3.2 Intermediates on the Pathway 3.2.1 Factor I 3.2.2 Factor 11 Sirohydrochlorin 3.2.3 Factor I11 3.2.4 The Synthesis of Isobacteriochlorins and Chlorins 3.3 Steps beyond Factor I11 3.3.1 Decarboxylation of the Acetate Group at C-12 3.3.2 Methylation Steps 3.3.3 Model Studies of the Methylation Steps 3.3.4 The Extrusion of C-20 3.3.5 Other Steps on the Path to Vitamin B12 3.4 Factor F-430 4 Topics Related to the Biosynthesis of Tetrapyrroles 4.1 The Biosynthesis of the Nucleotide Loop of Vitamin B12 4.2 N.M.R.Spectra and Other Properties of Tetrapyrroles 4.3 The Degradation of Haem 4.4 Medical Aspects 4.5 Evolution of the Biosynthetic Pathways 5 References The biosynthesis of tetrapyrroles has not been reviewed before in this series or in the series of Specialist Periodical Reports on Biosynthesis and so this review will give an overall picture of the field up to the end of 1983 but with particular reference to work published since the last review in Annual Reports on the Progress 0f’Chemistry.l In this time several excellent reviews have appeared in books including that of Akhtar and Jordan in ‘Comprehensive Organic Chemistry’,2 several chapters in Dolphin’s two major series ‘The Porphyrins’ (Vols.I-VII)3 and ‘B1 2’ (Vols. I and 11),4 and ‘Vitamin B1 Proceedings of the Third European Symposium on Vitamin B12 and Intrinsic Factor’ edited by Zagalak and Friedri~h.~ However developments continue apace ;particularly fasci- nating have been the results on the formation of uroporphyrino- gen I11 from four molecules of porphobilinogen and this area alone has been the subject of several reviews.6 Overall pathway.The pathways for the biosynthesis of haem chlorophylls and vitamin B (cyanocobalamin) appear to be essentially the same in all organisms in which they are found. All of them start with 6-aminolaevulinic acid (ALA) two molecules of which are condensed as shown in Scheme 1 to give porphobilinogen (PBG). Four molecules of PBG condense head-to-tail with elimination of ammonia to form a linear tetrapyrrole which is the unrearranged 1-hydroxymethylbi- lane. Cyclization of this unrearranged bilane proceeds enzymatically with a rearrangement of ring D to give the type I11 uroporphyrinogen (uro’gen 111).At this point the pathway to vitamin B12 diverges from that of haem and the chloro- phylls; it will be treated in full in Section 3. The route to haem however involves the decarboxylation of uro’gen I11 at all four acetate side-chains to give coproporphyr- inogen I11 (copro’gen 111). Two of the propionate groups of copro’gen I11 are oxidatively decarboxylated to give the vinyl groups of protoporphyrinogen IX. Oxidation of this porphyr- inogen to protoporphyrin IX (proto IX) and insertion of Fe2+ into the centre of the macrocycle gives haem which is found as such in haemoglobin myoglobin cytochromes b and P-450 peroxi- dases catalase and erythrocruorin. In a modified form haem is present in the a and c cytochromes. Introduction of a Mg2+ ion into the macrocycle on the other hand is the first step on the route to the chlorophylls details of which will be given in Section 2.First we will examine the biosynthesis of haem in more detail. 1 The Biosynthesis of Haem This pathway which is so central in the production of the tetrapyrroles that are vital to all organisms has been intensively studied over many years by chemists and biochem- ists alike so that for every step the enzymes have been isolated and at least partially purified and characterized. Therefore it is appropriate in this case to treat each enzyme in a separate section beginning with the synthesis of ALA. 1.1 6-Aminolaevulinic Acid Synthase (EC 2.3.1.37) The major pathway to ALA (1) in bacteria and animals but not in plants is by the reaction of glycine with succinyl-CoA as shown in Scheme 2.The enzyme that is responsible has been isolated and purified from many sources including Rhodo-pseudomonas sphaeroides,’ Euglena gracilis,s and the livers of rat9 and chicken embryos,1° and the enzyme from soybean callus has been immobilized on a Sepharose support’ (as well as several other enzymes of porphyrin biosynthesis). There appear to be several forms1° of the enzyme in some of these SOU~C~S~.~ and it has been suggested that the chicken embryo liver enzyme that is initially synthesized in the cytosol has a NATURAL PRODUCT REPORTS 1985 CO,H HO 1-Hydroxymethylbilane ALA PBG ICOzH \COzH Protoporphyrin IX Copro'gen II I Uro'gen 111 Vitamin B,? pathway Chlorophyll pathway Scheme I COzH NH NH O,H PLkNH pLpPH (1) (PLP = pyridoxal phosphate CoA = coenzyme A) Scheme 2 molecular mass of 74 000 but a portion is lost when this is trans- ported into the mitochondria to give an active form (of mol.wt 68 000). This transport is inhibited by haemin. 6-Aminolaevulinic acid synthase requires pyridoxal phos- phate which in the absence of substrate forms a Schiff's base with the amino-group of a lysine residue in the enzyme from R. sphaeroides as judged by the detection of pyridoxyl-lysine after hydrolysis of a sample of the enzyme that had been inactivated with NaBH4.1Z Glycine is incorporated into ALA with the loss of C-1 ; in principle the requisite anion of the glycine- pyridoxal phosphate Schiff's base could be formed by this decarboxylation but in fact the loss of 50% of the activity from [2-3H]glycine indicates that it is a deprotonation step which generates the anion.Experiments with [2R-3H]glycine and [2S- 3H]glycine have shown that it is exclusively the pro-R hydrogen which is removed. Reaction with succinyl-CoA is followed by the decarboxylation. It has been shown that the pro-S hydrogen at C-5 of ALA is the one that is derived from glycine thepro-R hydrogen coming from the medium. This was determined by chemical degrada- tion of a sample of porphobilinogen that had been derived from [2-3H]glycine back to [2S-3H]glycine.'4 Thus it appears either that the reaction with succinyl-CoA occurs with inversion at C-2 of glycine and the decarboxylation then occurs with reten- tion of configuration or vice versa.In the absence of succinyl-CoA ALA synthase catalyses the exchange of the pro-R hydrogen of glycine and the 5-pro-R hydrogen of ALA with solvent,I5 as would be expected from the mechanism. It is also reported to cause exchange of the carboxyl group of glycine with NaHITO only in the presence of suc~inyl-CoA.'~ This incorporation is increased by the addition of ALA and coenzyme A and indicates the reversibi- lity of the whole sequence. 1.2 The Synthesis of 6-Aminolaevulinic Acid in Plants It is now clear that plants do not in general use ALA synthase in the production of chlorophyll. Thus I4C-labelled glycine is poorly incorporated (if at all) into ALAI7,'* or chlorophyll;3 where some incorporation is seen both [1-l4C]glycine and [2- 14C]glycine are incorporated.'' It has been shown using I4C-and 13C-labelled precursors that all five carbon atoms of glutamic acid (2) are incorporated into ALA and thence into chlorophylls;*8.19however the order of the likely steps i.e.one reduction and two transaminations is not certain. Enzymes which will cause the transamination of 4,5-dioxovaleric acid (DOVA) (3) and alanine to ALA and pyruvic acid have been isolated from several sou~ces.~ A purified enzyme from cucumber seeds is the first to have been shown to be specific for this transformation only.20 The mammalian liver enzyme with this capability is more active as an alanine :glyoxylate transaminase.2 of cl-ketoglutarate (4)to DOVA has been partially purified from the leaves of Zea mays,22but different workers obtained a NADPH-requiring extract from plastids of the same organism which converted both glutamate (2) and its 1-semialdehyde (5) into ALA ( It seems that either or both of the routes to (3) that are shown in Scheme 3 may be followed.NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER / Ho2cY-co2H 0 Ho2CPco2H (4) NH (2) Scheme 3 The role of DOVA (3) in the biosynthesis of ALA has been questioned by two studies in the first24 it was found that a mutant of R. sphaeroides that requires ALA for its growth could not grow on DOVA instead despite the fact that the transaminase was active. In the second it was found that DOVA inhibited the formation of porphyrins in Clostri-dium tetanomorphum again despite the presence of the transaminase because of inhibition of ALA dehydratase.It has been suggested1 that (3) is an intermediate but that it generally remains enzyme-bound and this could provide an explanation for the above results. It has been found26 that in Euglena gracilis 14C-labelled glutamate but not glycine is incorporated into chlorophyll whereas 14C-labelled glycine is the better precursor of the haems. Under conditions where chloroplasts are absent (in a mutant or when the organism is grown in the dark) [14C]glutamate is not incorporated into porphyrins at all. The indication is that the five-carbon route to ALA (and thence to the tetrapyrroles) is exclusive to the plastids whereas the ALA synthase route only operates in the cytosol and in mitochondria.It is likely that the same situation exists in higher plants also. 1.3 6-Aminolaevulinic Acid Dehydratase (Porphobilinogen Synthase; EC 4.2.1.24) Condensation of two molecules of ALA (1) in a Knorr-type pyrrole synthesis to give porphobilinogen (PBG) (6) is catalysed by ALA dehydratase. The two best studied dehydra- tases are from Rhodopseudomonas sphaeroides and bovine liver. Both are inhibited by NaBH only if ALA or laevulinic acid (7) is present thus indicating that a Schiffs base is formed.27 On hydrolysis of a sample that had been inactivated by NaBH in the presence of [4-14C]ALA a radioactive spot was detected that was chromatographically identical to one for the lysine derivative (8).2s Laevulinic acid alone is not a substrate but if it is mixed with ALA it can be utilized by the enzyme of R.sphaeroides to form the methylpyrrole (9).29Thus it seems that (7) can occupy the A site of the enzyme (the ALA-binding site which provides the acetate side-chain of PBG) and it is assumed that this is the site of formation of a Schiffs base (though it has not been proved). Formation of an imine would facilitate deprotonation on C-3 of ALA to give the enamine and so the likely course of the reaction is as shown in Scheme 4. It is known that it is thepro-R hydrogen which is lost at C-2 of PBG because incorporation of [5S3H]ALA (derived from [2- 3H]glycine -see Section 1.1) occurs without loss of tritium.30 It had been assumed on the basis of the above results that the first molecule of ALA forms a Schiffs base in the A site and then the second molecule occupies the P site.However Seehra and Jordan in an elegant experiment3’ (modelled on experi- ments with PBG deaminase that are described in the next section) have shown that on exposure of the enzyme to 1 equivalent or less of labelled ALA for 0.1 s followed by an excess of unlabelled ALA it is the P site which contains the majority of the labelled ALA. Thus PBG that had been made in this way from [5-13C]ALA contained all of the detectable I3C at c-2. The bovine liver dehydratase is an octamer of molecular weight 285 000,” and it appears (in an electron micrograph) to have the subunits arranged at the corners of a cube.32 It requires four Zn2+ or Cd2+ ions for full activity33 though it can Scheme 4 CO H lo H (7) (9) bind eight Zn2+ ions.34 The ion Hg2+ replaces Zn2+ and inactivates the enzyme.By immobilization of bovine liver dehydratase on a Sepharose support the dissociation of the enzyme into partially active tetramers could be demon-~trated.~~ It is-claimed that a dimer is the minimal complex that retains and this is consistent with the apparent presence of only four active sites per ~ctarner.~’*~’,~~ The N-terminal sequence up to residue 44 has been quoted.39 The subunits are identical except that half of them have the terminal methionine residue acylated and the other half have a free N-terminus.It is thought that the active site of bovine dehydratase involves two thiol groups of cysteine residues and two imidazole moieties of histidine as well as the zinc The thiol groups are easily oxidized in air to a disulphide thus deactivating the enzyme but it can be re-activated by the presence of thiols in the medium. In addition to normal thiol- directed agents the enzyme is also inactivated by 3-and 5-chlorolaevulinic acids.37 These active-site-directed agents alkylate different cysteine residues (only on the reduced enzyme). The results suggest that 5-chlorolaevulinic acid forms a Schiffs base with the enzyme and reacts with a thiol in that NATURAL PRODUCT REPORTS 1985 site whereas the 3-chloro-isomer reacts with a thiol in the alternative site.The role of these thiols is not certain though one or both may be involved in binding the zinc ion -oxidation to the disulphide causes release of the meta1.34,40 There is also speculation about the role of the zinc ion; one possibility is that it acts as a Lewis acid to make the ketone in the P site more electrophilic and the eventual loss of OH-from the same carbon more facile. It has been claimed,40 though that the zinc ion is only necessary to prevent oxidation of the thiol groups. The dehydratase of R. sphaeroides may be a hexameel or an ~ctamer,~~ and unlike the bovine liver enzyme it requires K+ or Mg2+ for acti~ity.~~.~~ In other respects the enzymes from the two sources are similar.6-Aminolaevulinic acid dehydratases from all sources are inhibited by laevulinic acid which has often been used to cause accumulation of ALA in v~uo.~? l 7* *Succinylacetone is also an inhibitor of dehydrata~e~~ and is particularly interesting because it is produced in humans who are suffering from a hereditary disorder of their tyrosine metabolism in which the enzyme that hydrolyses fumarylacetoacetate is defective.44 As a result there is a build up of succinylacetone which inhibits ALA dehydratase and causes excretion of ALA and other symptoms of acute intermittent porphyria. 6-Aminolaevulinic acid dehydratase can be immobilized on a Sepharose support,’ 1935,42 and this produces a very stable system which can be used for long periods of time to produce gram quantities of PBG for studies with the next enzymes in the pathway.4s 1.4 Formation of Uroporphyrinogen 111 Hydroxymethylbilane Synthase (Porphobilinogen Deaminase; EC 4.3.1.8) and Uropor-phyrinogen 111 Synthase (Cosynthetase; EC 4.2.1.75) Until only a very few years ago this step in the biosynthesis of porphyrins remained quite ob~cure.~.~ However as a result of much detailed and painstaking work we now have a considerably clearer overall view of the process.It has been known for some time that two enzymes are involved. The first enzyme PBG deaminase forms a tetramer of PBG with elimination of NH3. The observed product of this enzyme was uroporphyrinogen I (uro’gen I) and so it has been referred to as uro’gen I synthetase but it is now known that uro’gen I is the PBG (6) result of a chemical rather than of enzymic cyclization and so this name for the enzyme is inappropriate.The second enzyme uro’gen I11 synthase (usually described as cosynthetase) in conjunction with deaminase produces uro’gen I11 (instead of the type I isomer) from PBG. It was shown that uro’gen I is not a substrate for cosynthetase and the lack of any other known substrate led to the assumption that cosynthetase binds to and in some way modifies the action of deaminase. The combina- tion of deaminase and cosynthetase is sometimes called porphobilinogenase. It is now generally accepted that the true course of events is as shown in Scheme 5. Four molecules of PBG (6) are initially con- densed head-to-tail by deaminase to form the hydroxymethyl- bilane (10).This intermediate is the substrate for cosynthe- tase which cyclizes it to uro’gen I11 (1 1) with rearrangement of ring D. In the absence of cosynthetase 1-hydroxymethylbilane (10) cyclizes chemically without any rearrangement to give uro’gen I (12). 1.4.1. Proof that 1-Hydroxymethylbilane is the Intermediate Since the discovery that PBG is the precursor of uro’gen 111 many different hypotheses for the steps that are involved have been proposed.46 These can be divided into categories depending on the stage at which the rearrangement occurs mono- di- tri- or tetra-pyrrole. Some of the hypotheses could be discounted by the results of experiments with 13C-labelled PBG. For example,47 (2,ll- 3C2)PBG(13) diluted with unlabelled PBG was incubated with the deaminase-cosynthetase system to give a sample of uro’gen I11 in which eight atoms were labelled.In the 13C n.m.r. spectrum of the labelled protoporphyrin IX (14) that was obtained from this uro’gen 111 the signals of three of the meso carbons were fine doublets (J = 5.5 Hz) owing to long-range coupling across a pyrrolic ring (C-20 to C-4 C-5 to C-9 and C-10 to C-14) but the signal of one meso carbon (C-15) was a widely spaced doublet (J = 72 Hz) owing to coupling with a directly adjacent 13C (at C-16) -see Scheme 6. The implication of this experiment is that only one rearrangement occurs and it must involve an intramolecular inversion of the pyrrole ring that is destined to become ring D of uro’gen 111,in the course of which C-11 of PBG migrates from C-5 to C-2.COzH HO NH HN2 1 W ‘zH H cosynthetase f 602H Uro’gen 111 (1 1) k02H Uro’gen I (12) Scheme 5 NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER In experiments to find out at what stage the rearrangement of ring D occurs no significant incorporations were obtained with isoporphobilinogen (1 5Y8 or with di- or tri-pyrrolic com-pounds. The results with the di- and tri-pyrrolic compounds were confused by non-enzymic condensations and fragmenta- tions which could give bilanes and porphyrinogens but in general the overall rate of formation of a porphyrinogen was not significantly increased by the presence of the enzyme^.^^^^ The first result pointing in the right direction was the discovery of the unrearranged bilanes (16)-( 18) as by-products that are formed from PBG by deaminase in the presence of the appropriate nucleophiles (NH3 NH20H or NH20Me).49*50 The aminomethylbilane (16) was chemically synthesized5 and was found to be converted into uro’gen I faster in the presence of deaminase than purely chemically; furthermore with added cosynthetase uro’gen I11 was formed.The rearrangement to form uro’gen I11 was shown to be intramolecular by double- labelling with l3C. However it was soon realised that (16) could not be a true intermediate52 because (i) it is converted into uro’gens by deaminase (or deaminase-cosynthetase) 15 times slower than PBG; (ii) it is not a substrate for cosynthetase alone but is converted by deaminase into a compound which is a substrate for cosynthetase; (iii) it is not detected as a normal product of deaminase except in the presence of a large concentration of ammonia.The presence of an enzyme-free intermediate between PBG and the uro’gens was deduced from the fact that the initial production of porphyrinogens by deaminase alone was much slower than the consumption of PBG.52 The intermediate that is involved was first observed by 3C n.m.r. spectroscopy of the HO? (,, 11 &2 H H2N (13) I3C at CO2H k02H HOg2H (14) Scbeme 6l 3C is at either 0 or 0 or~orl c o NH2 H (10) X = OH (16) X = NH2 (15) (17) X = NHOH (18) X = NHOMe 23 product of incubation of [1 1-13C]PBG with deamina~e.~~.~~ In addition to the expected signals around 6 24.4p.p.m.for the pyrrole-CH,-pyrrole methylenes a signal was seen at 6 57.1* p.~.m.~~ This compound was stable in alkali but at pH 8.5 and 37 “C it gave uro’gen I chemically with a half-life of 4 minutes. This intermediate proved to be an excellent substrate for cosynthetase which very rapidly converted it into uro’gen III.52-54 (See later for methods of analysis of porphyrin isomers.) Scott and co-~orkers~~ considered several possible structures for the intermediate and to help distinguish among them they incubated deaminase with [l l-13C,1-15N]PBG.55 In the 13C n.m.r. spectrum of the resulting intermediate they observed that the signal at 6 57.1 p.p.m.was broadened; this was interpreted as being a doublet for which J = 6 Hz arising from a one-bond sN-l 3C coupling. They therefore decided on structure (19) for the intermediate. Battersby and co-~orkers~~ also obtained a 13C n.m.r. spectrum of the intermediate that was derived from [11-I3C 1-l5N]PBG but observed a doublet at 6 57.1 p.p.m. with a coupling constant of only 2.4 Hz -consistent with a two-bond 5N-13C coupling. As a result they proposed the open-chain hydroxymethylbilane structure (10). Further proof has been obtained that (10) is indeed the true intermediate. First the 3C n.m.r. spectrum of the model compound (20) showed a doublet (J = 2.3 2 0.2 Hz) at 6 57.3 p.p.m. for the hydroxymethyl carbon,52 very similar to that of the enzymic intermediate.Secondly the 3C n.m.r. spectra of the model compounds (21)57 and (22)58 showed a much more widely spaced doublet of coupling constant 10-10.7 Hz at 6 41.5-42.0 p.p.m. for the 1-pyrrolylmethyl carbon -very different from that of the enzymic intermediate. Thirdly the hydroxymethylbilane (lo) synthe- sized as shown in Scheme 7 was shown to be identical to the enzymic intermediate in all respects including its reaction with c~synthetase.~~ Also very importantly the rate of cyclization of (10) by cosynthetase when deaminase was present was identical with that when deaminase was nor present.63 The only remaining objection to the role of the hydroxy- methylbilane as the true intermediate is that it may be in equili- brium withacompound which is theactualproductofdeaminase and the substrate of cosynthetase such as the azafulvene (23).However the extremely rapid reaction of cosynthetase with (10) makes involvement of such a chemical step seem unlikely. * These values are taken from ref. 52; the values in ref. 53 are 2.3 p.p.m. lower only because they are measured from a different standard. 2H HOzC (19) NATURAL PRODUCT REPORTS 1985 Mg MeOgco'Me+ ___+ MeO& several-steps MeowozMe But02C C02Bn Bu'02C C0,Bn OHC H AcO H H H H H // I (Bn = CH,Ph) I ji Reagents i TsOH CH2C12; ii NaBH,; iii AcOH CH,C12; iv NaOH Scheme 7 H H H H (23) Production of (23) at the active site of deaminase followed by its trapping by water or other available nucleophiles would account for the formation of bilanes (10) and (1 6)-( 18) but the non-enzymic reaction of (10) with NH3 or NH20H is reported to be very and so the formation of (23) is ruled out except when it is bound to an enzyme.In view of the above results the systematic name for PBG deaminase has been officially changed to hydroxymethylbilane synthase [EC 4.3.1.8].60 Similarly as cosynthetase does not in fact act in co-operation with deaminase but can function quite independently its name is also inappropriate and it should properly be uroporphyrinogen III synthase [EC 4.2.1.751. 1.4.2 The Mechanism of' Deaminase The order of assembly of the four PBG molecules to give the hydroxymethylbilane has been studied using pulse-labelling experiments (similar to that described earlier for 2 ALA - PBG).Unlabelled PBG (0.5 pmol) was added to deaminase (-0.25 pmol) that had been isolated from Euglena gracilis; this quantity of PBG was insufficient to allow formation of the biiane which was completed by addition of excess [ll-13C]PBG.59,61The resultant uro'gen I11 was chemically converted into coproporphyrin I11 tetramethyl ester and the relative proportions of 12C and 13C at the four meso positions were determined from the * H n.m.r. of the attached hydrogens in the presence of a shift reagent. It was quite clear that C-20 had the most 12C followed by C-5 C-10 and C-15 respectively. Thus the order of binding of the four pyrrolic rings is first ring A then ring B ring C and finally ring D.The ratios of '*C to 13C at the four sites were very close to the statistical distribution that would result if the rates of binding of each of the four successive PBG molecules were identical. The same order of assembly was found for deaminase that was isolated from Rhodopseudomonas sphaeroides. This ex per imen t em- ployed 14C-labelled PBG and the labelled sites were located by degradation of protoporphyrin IX.* Serious inconsistencies appear in the radioactivity measurements in this work; e.g.,degradation of samples of protoporphyrin IX showed that virtually all of the activity was in the two moles of ethylmethylmaleimide that were formed from rings A and B despite the fact that halfof the activity from [2,1 l-l4C2]PBG should be at the meso carbons which are lost in the degradation.Kinetic experiments on deaminase have shown that the first PBG molecule is covalently bound to the en~yrne~~.~~ with release of 1 mole of ammonia.63 Thus the necessary steps for the formation of the hydroxymethylbilane are as shown in Scheme 8. In this scheme the di- and tri-pyrrolic intermediates remain bound to the enzyme which explains why these types of intermediate cannot in general be detected or incorporated during conversion of PBG into Anderson and De~nick,~~ working with human deaminase observed four different forms of deaminase to which PBG was bound. These separated forms had 1 2 3 and 4 molecules of PBG bound per molecule of deaminase. Subsequently Jordan and co-workers66 have separated by gel electrophoresis three forms of deaminase with [3,5-l4C2]PBG bound to them.All three forms were able to complete the synthesis of 14C-labelled hydroxymethylbilane when excess unlabelled PBG was sup- plied. By subsequent conversion of the hydroxymethylbilane into protoporphyrin IX and degradation they were able to show that the three forms had respectively 1 2 and 3 molecules of PBG bound. There was no sign of a form with four PBG molecules bound in this experiment with deaminase from R. sphaeroides. In contrast Battersby et have found that the tetrapyrrolic hydroxymethylbilane (1 0) forms a relatively stable complex with deaminase from Euglena gracilis. The nature of the enzymic group X (see Scheme 8) to which the first PBG molecule becomes attached has not been rigorously established but there are indications that it might be an amino-group of a lysine residue.First,63 the enzyme having a molecule of [ 1 1-l 3C]PBG covalently bound to it was digested with proteolytic enzymes to obtain small peptide fragments with the pyrrole unit(s) attached. In the I3C n.m.r. spectrum signals appeared at 6 24.5 p.p.m. (attributed to pyrrole-CH,- pyrrole) and others were just visible at 6 42-43 p.p.m. attributed to peptide-NH-CH,-pyrrole. Se~ondly,~' deamin-ase is inhibited by pyridoxal phosphate. With NaBH this inhibition becomes irreversible indicating that a Schiffs base is formed between the pyridoxal phosphate and an amino- group at the active site. On hydrolysisof a sample of the enzyme that had been inactivated with pyridoxal phosphate and [3H]NaBH, E-pyridoxyl-lysine was positively identified by dilution analysis.It is conceivable however that the lysine residue that is thus modified is involved in some other role in the active site. One further interesting observation is that while 14C-labelled PBG remains covalently bound to deaminase even after denaturation treatment of the denatured complex with hydroxylamine does release the radi~activity.~~ As a denatured enzyme cannot catalyse the release of the substrate it may be that hydroxylamine causes SN2-type displacement of the enzymic group. The experiments that have been described in this section have been largely on deaminase that was isolated and purified NATURAL PRODUCT REPORTS 1985 -F.J. LEEPER 25 Enz-X’ \ CO?H C02H HOzC HecozH Enz-Xl HN+cozH HOzC Ho2:wcozH HOzC from either Rhodopseudomonas sphaeroides or Euglena gracilis though many other sources have been de~cribed.~ The enzyme has usually been found to have a molecular weight of about 40 00049.66.69 (but one report68 quotes 20 000). For the enzyme from E. gracilis the values of KMfor PBG depend on whether hydroxymethylbilane is present or absent. Under the best conditions a value of 105pmoldm-3 was obtained6j and a value for k,, of 0.27 s-l has been ~alculated.~~ As yet no sub- strate for deaminase other than PBG [and the aminomethyl- bilane (1 6) and some of its isomers5*] has been found. Similar pyrroles such as isoporphobilinogen (1 5) opsopyrroledicar-boxylic acid (24),and 2-aminomethylpyrrole-3,4-diaceticacid (25) are competitive inhibitors but are not used as substrates by the en~yme.~ Several questions about the action of deaminase remain unanswered.The stereochemistry of all of the steps is unknown and so is the mechanism by which the tetrapyrrole (but not the di- or tri-pyrrole) is released. It is however known that for the enzyme from Euglena gracilis release of the hydroxymethyl- bilane is slow except in the presence of PBG.63 In the biosynthesis of fatty acids and polyketides the condensations are thought to occur repeatedly at the same active site with the growing chain being attached to a long flexible pantotheine chain; the same principle is used in the enzymic synthesis of some oligopeptides e.g.gramicidin S. It is possible that deaminase uses a single catalytic site in a similar fashion though the alternative that there are four separate catalytic sites one for each PBG unit is just as likely. I .4.3 The Mechunism of’ Cosyntherase There have been very few studies on cosynthetase itself partly because a convenient was not available until its substrate was identified as (10). One plausible mechanism for HOzC C02H COzH (23) CO,H H2N rq n (24) the rearrangement,47 based on a proposal first put forward in 1961,73 is shown in Scheme 9. In this mechanism loss of water from the hydroxymethylbilane to give (23) is followed by cyclization onto the substituted a-position of ring D to give a spiro-compound (26).Fragmentation of (26) could give either (23) again or a different azafulvene (27) which would cyclize in the normal way to give uro’gen 111 (1 1). Another possible route from (26) to (11) would involve a series of [1,5] sigmatropic rearrangements. In a model study to test the feasibility of these rearrangement mechanisms the pyrrolenine (28) was ~ynthesized.’~ Under very mild acid conditions (28) rearranged to give (23) as the major product. The structure of this major product and other minor products shows that the fragmentation-recombination mechanism predominates over the sigmatropic rearrangement in this case. While the spiro mechanism is a highly plausible one another mechanism has been suggesteds7 which also fits the known facts; it is shown in Scheme 10.This scheme uses an enzyme- bound electrophile to hold ring D as it is detached from the other three pyrrole units turned around and re-attached. In principle the electrophile could be a proton but this would not provide covalent attachment to the enzyme. In Scheme 10 the electrophile is shown as an iminium ion. In this context it is interesting that two recent reports suggest that folic acid which is a likely source of such an iminium ion may be involved in the mechanism of cosyntheta~e.’~. 76 NATURAL PRODUCT REPORTS 1985 HO2* ZH HO2 W 2H NHHN 1 NHHN 1 Bu'02C&p -+a H H H CO~BU' .ozcW 2H H O 2 W O2H NH HN NHHN -*co2H \t/ u *co2H HO2C H02CH02C HOzC C02H d02H b02H C02H A* J H O 2 W 02H NHHN I \ H02C C02H COzH C02H Scheme 10 NATURAL PRODUCT REPORTS 1985 -F.J. LEEPER Cosynthetase from a number of different sources has been and the hydroxymethylbilane (10) is a substrate in all cases. The specificity of cosynthetase for various bilanes has been tested several times. In experiments with aminomethylbi- lanes deaminase must also be present to convert the aminomethylbilanes into hydroxymethylbilanes before these can act as substrates for cosyntheta~e.~~ This presents the added complication that the enzymic reaction may fail to give uro’gens faster than the non-enzymic cyclization because the aminomethylbilane does not interact with deaminase. A study of the isomer composition however gives some indication of the extent to which cosynthetase has been involved.Studies on the aminomethylbilanes (16) and (30)-(34) in Cambridge using enzymes of Euglena gracifi~,~~ and on (16) and (33)-(36) in Buenos Aires using enzymes from a variety of source^^^,^^ (Rhodopseudomonas sphaeroides wheat germ rat blood and rat spleen) have been published. Both groups find that the unrearranged bilane (16) is cyclized fastest but the results differ for different organisms on the extent of inversion of ring D that occurs for (16) and the other bilanes. The interpretation of the results is made difficult by the presence of two enzymes and by the high proportion of non-enzymic cyclization that occurs in most cases.More direct studies on the substrate specificity of the cosynthetase of E. gracilis have used hydroxymethylbilanes.80 As mentioned earlier the unrearranged hydroxymethylbilane (10) is a very good substrate and is almost completely rearranged. Bilanes in which the substituents on ring c(38) or on ring D (39) are reversed were cyclized by cosynthetase and underwent considerable inversion -45% for (39) 95% or more for (38). Compounds (37) and (40) were not found to be substrates at all. When bilanes (41) and (42) (from which a carboxyl group on ring D was missing) were incubated with cosynthetase,*l (42) was a good substrate -cyclization occurred at about 25% of the rate of that for (10) -but (41) was a poor substrate. Proton n.m.r. spectroscopy revealed that two isomers were formed from (42) in a ratio which indicated that enzymic cyclization proceeded with 65% inversion of ring D.In conclusion our knowledge about the mechanism of cosynthetase is still incomplete but any detailed hypothesis will have to take into account the fact shown in these studies that a type I bilane is rearranged almost completely to a type I11 porphyrinogen whereas a type I11 bilane (in which substituents on ring D are reversed) is rearranged to a considerable extent to a type I porphyrinogen. Obviously there is no common intermediate from these two bilanes. I .4.4 Determination oj’Ratios oj’Isomers of Uroporphyrinogen Much of the foregoing work on deaminase and cosynthetase was greatly helped by the ability to distinguish which uro’gen isomers are formed and to measure their ratios.This has been made very much easier and more accurate by the use of h.p.1.c. Ratios of uro’gen isomers are not determined directly; instead the uro’gen mixture is invariably oxidized to the much more stable uroporphyrins using iodine or benzoquinone. Uropor- phyrin isomers (see Table 1) have been separated by h.p.1.c. both as the free acids (on reverse-phase and as Y (16) NH2 (30) NH2 (31) NH2 (32) NH2 (33) NH2 (34) NH2 (35) NH2 (36) NH2 (10) OH (37) OH (38) OH (39) OH (40) OH (41) OH (42) OH R’ R2 R3 R4 R5 R6 R7 R8 A P A P A P A P A P A P A P P A A P A P P A A P A P P A A P A P P A A P A P A P P A P A P A A P P A A P P A A P P A P A P A P A A P A P A P A P A P P A A P A P A P A P P A A P A P A P A P P A P A P A P A A P A P A P A P MeP A P A P A P AEt (A = CH2C02H,P = CH2CH,C02H) Table 1 The positions of side-chains in the isomers of uroporp hyrin R’ R3 R’ R4 NH HN- R8 \ R7 Type CH,C02 H I R1 R3 R5 R7 I1 R’,R4 R5 R8 111 RI R3 R5 R8 IV R’ R3 R6 R8 R5 R6 CH2CH2COzH R2 R4 R6 R8 R’ R3 R6 R7 R2 R4 R6 R7 R2 R4 R5 R7 was later develo~ed.~~b Recently a separation of all four isomers as the free acids by reverse-phase h.p.l.c.eluting with MeCN/H20 that is buffered with NH,OAc and HOAc (elution occurred in the order I 111 IV 11) has been described.85 This latter technique effected the separation within 25 minutes; however in general reverse-phase h.p.1.c.has a lower capacity than the normal-phase procedure and so would not be suitable on a semi-preparative scale. The separation of uroporphyrin coproporphyrin and the hepta- hexa- and penta-carboxylated porphyrins that are intermediate between them is of considerable interest in clinical chemistry as these compounds are present in the urine of patients suffering from various porphyrias. Such a separa- tion is relatively easy by h.p.1.c. or t.l.c. but particularly impressive is a simultaneous separation of the types I and I11 isomers of coproporphyrin and uroporphyrin and of the penta- hexa- and hepta-carboxylic intermediates that has been reported recently.86 methods successfully separate uroporphyrins I 11 and 111 but uroporphyrin IV cannot be separated from uroporphyrin 111.The normal sequence of elution is I I11 + IV 11. Uroporphyrins can be decarboxylated to give coproporphyr- ins by heating in dilute HCl (at 180°C) or without solveqt. This technique has been employed by the Cambridge group in their studies on deaminase-cosynthetase because it is possible to separate all four isomers of coproporphyrin. The first separation of the four isomers relied on two successive h.p.1.c. runs. The first run separated types 11 I11 + IV and I in that order and the second run using a different solvent system then separated types I11 and IV after several recycles.84 A more effective separation of all four isomers in a single run as their zinc complexes (when the order of elution was I 11 111 IV) Both 1.5 Uroporphyrinogen Decarboxylase (Uroporphyrinogen Car- their methyl esters .(on normal-phase ~olumns).~~-~~,~~ boxy-lyase; EC 4.1.1.37) The next step on the route to haem is the four-fold decarboxylation of uro’gen I11 (1 1) to copro’gen I11 (46).It was thought that a single enzyme was responsible for all four decarboxylations but until recently the enzyme had not been purified and so this could not be verified. In 1983 four different groups all reported the purification of a single enzyme that is capable of the four decarboxylati~ns.~~-~~ The kinetics of the decarboxylation reactions are complicated by the existence of the hepta- hexa- and penta-carboxylic intermediates and by inhibition of the reactions by the products.It seems that the first decarboxylation to give a heptacarboxylic porphyrinogen is the fastest; this intermediate accumulates initially and then is converted into copro’gen I11 at a lower rate.89-91 The intermed jate hexa- and penta-carboxylic porphyrinogens can be detected but their concentrations are usually low. At high concentrations of uro’gen 111 the heptacarboxylic intermediate accumulates to a greater extent and the overall rate of formation of copro’gen 111 is red~ced.~~-~l It has been s~ggested~~-~’ that there are two (or more) catalytic sites one for the first decarboxylation and the other(s) for the remaining three decarboxylations.The evidence for this is not conclusive but does come from several different experiments. For example the decarboxylation of uro’gen Ill was inhibited by uro’gen I whereas decarboxylation of the intermediates was not inhibited by uro’gen I but was inhibited by intermediate type I porphyrin~gens.~~ Inhibition by polychlorinated biphenyls at low concentration leads to accumulation of the heptacarboxylic intermediate but at higher concentrations uro’gen I11 was hardly decarboxylated at all.89 The decarboxylase is also inhibited by bivalent metal ions such as Cu2+ ZnZ+ and Hg2+ and by thiol-directed reagent~.~~-~O All of the uro’gen isomers are decarboxylated by the enzyme to give the corresponding copro’gen isomers but reports on the relative rates vary.The majority of reports find that the first decarboxylation of uro’gen I11 is faster than for uro’gen I but that subsequent decarboxylations are at similar rates for the two isomeric series except that the pentacarboxylic intermedi- ate accumulates to a greater extent for uro’gen I than for uro’gen 111. For uro’gen 111 the order of the decarboxylations is not random; the first decarboxylation is specifically on ring D to give (43) (known as phyriaporphyrinogen) as the only major heptacarboxylic intermediate. 1,93 The hexa- and penta-car- boxylic porphyrinogens are not usually present in sufficient quantities for structure determination but the corresponding porphyrins are excreted in reasonable amounts by sufferers of porphyria cutanea tarda or by animals in which a porphyria has been induced by prolonged administration of hexachloro- benzene or other halogenated aromatics.Jackson eta/.isolated milligram quantities of the intermediate porphyrins from the faeces of rats that had been poisoned with hexachlorobenzene and showed by total synthe~is,~~.~~ that the porphyrins that were isolated corresponded to the porphyrinogens (43) (44) and (45).Thus it seems likely that the preferred order of decarboxylations is clockwise starting at ring D then A B and finally C. However it was found that the other isomers of the intermediate type I11 porphyrinogens can also be substrates for the decarboxylase and so the order of decarboxylations cannot be considered obligatory.In contrast to uro’gen 111 the decarboxylation of uro’gen I is relatively non-specific ;95 both of the possible hexacarboxylic isomers are formed in roughly equal amounts. Porphyrins of the type I series are excreted by sufferers of certain porphyrias in which uroporphyrinogen I11 synthase is deficient. CO,H \-f \ C02H CO,H R’ R’ R3 RJ (I1)A A A A (43) Me A A A (44) Me Me A A (45) Me Me Me A (46) Me Me Me Me (A =CHZCOZH) NATURAL PRODUCT REPORTS 1985 The most reasonable chemical mechanism for the decarbox- ylation of the acetate side-chains of uro’gen I11 is as shown in Scheme 11. Barnard and Akhta~-~~ have produced evidence in support of this mechanism and shown that the decarboxylation at all four positions proceeds with retention of configuration.Thus (2R)-[2-3H,2-2H]succinate (47) was incorporated into haem by a cell-free preparation from chicken erythrocytes. On degradation of the labelled haem (48) the methyl groups were obtained as acetic acid which was found to be chiral with the S configuration by the malate synthase-fumarase assay. This finding has been confirmed using different organisms in work that will be described later on bacteriochlorophyll aIg2and vitamin B,2.209 1.6 Coproporphyrinogen Oxidase (EC 1.3.3.3) The two oxidative decarboxylations of copro’gen 111 (46) to give proto’gen IX (52) are again catalysed by a single enzyme and in this case the indications are that only a single active site is The propionate on ring A is the first to be converted into a vinyl group ;thus the tricarboxylic porphyrino- gen (50) is the first intermediate which can be detected.This porphyrinogen named hardero’gen because the corresponding porphyrin can be isolated from the harderian gland of rats is a H J Scheme 1 H (47) (46) R’ =P R? =P (49) R’ =HP R’ =P (50) R’ =V R’ =P (51) R’ =P R’ =V (52) R’ =V R’ =V [P =CH2CHICOZH,V =CH=CH2 HP =CH(OH)CHZC02H] NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER much better substrate for the enzyme than its isomer isohardero'gen (51). Kinetic.experiments have shown9' that in the presence of excess copro'gen 111 the proto'gen IX that is produced is all derived directly from copro'gen without release of the intermediate hardero'gen (50).The proportion of (50) which does get released into the medium cannot compete successfully for the active site with the excess of copro'gen I11 that is present.In aerobic organisms oxygen is the electron acceptor in the oxidation but anaerobes have a different enzyme which uses a hydride acceptor as its oxidizing agent.98Rhodopseudomonas sphaeroides has both the aerobic and anaerobic forms depending on the conditions of growth and the existence of two pathways (one dioxygen-dependent one NADP-depen-dent) has been demonstrated in soybean root nodules.99 The mechanism for the aerobic enzyme has been suggestedloOto proceed via a hydroxylation to give (49)as the first intermedi-ate.Synthetic (49)was found to be a good substrate for the enzyme that was purified from bovine liver. However the enzyme did not have the properties that would be expected of a hydroxylation system (there was no requirement for any metal ion or cofactor) and so it is still doubtful whether hydroxylated species are true intermediate^.^^ An alternative mechanism shown in Scheme 12 would account for the involvement of (49).In this mechanism loss of a hydride ion from the p-position of a propionate side-chain would give an intermediate imine of the type (53),also available by dehydration of a hydroxypropionate side-chain. Loss of C02 from (53)would then generate the vinyl group. The stereochemistry of the oxidative decarboxylation has been shown to involve loss of thepro-S hydrogen atom that is p to the carboxyl group in the propionate side-chain in chicken reticulocytes,lO1in Euglena gracilis,'O*and recently in both the aerobic and the anaerobic enzyme of Rhodopseudomonas sphaer~ides.~~ The overall elimination of H and C02 has been shown to be trans in the aerobic enzyme from E.gracilis.loo Thus the stereospecifically labelled PBG (54) gave protopor-phyrin IX that was labelled as shown in structure (55). COZH COzH /-b0-H " \ (53) 1 Scheme 12 H02C-$ H2N H (54) (55) < COZH (56) R' = R' = CH 2CH'COzH (57) R' = CH=CH2 R2 = CHzCHzC02H (58) R' = R2 = CH=CH2 COZH / - (59) R = Me CH=CH, or H 60,H (45) R' = CHZCHICO'H R' = CH'C02H (60) R' = CH=CH2 R' = CH2C02H (50) R' = CH=CHZ R' = CH In eukaryotes coproporphyrinogen oxidase is mainly asso-ciated with the mitochondria and is thought to be located in the intermembrane space.Io3 The purified enzyme from bovine liverloois a monomer of molecular mass 71 600. The substrate specificity of the enzyme from various sources has been studied :97* O4 copro'gens I and I1 do not react but copro'gen IV (56) is a substrate and the product is proto'gen XI11 (2,8,13,17-tetramethyl-1 2,18-divinylporphyrinogen-3,7-dipropionic acid) (58).104.'05 In this conversion the tricarboxylic porphyrin (57) accumulates to a greater extent than from copro'gen III.'04 When all of the results of enzyme specificity are collected it appears that the propionate unit is best decarboxylated if it occurs in the middle of the series of substituents shown in part-structure (59).] O0,] O4 An alternative route to haem may be importantIo6 in porphyria patients in whom the pentacarboxylic porphyrino-gen (45)is not efficiently decarboxylated by uro'gen decarbox-ylase. This porphyrinogen can act as a substrate for copro'gen oxidase to give dehydroisocoproporphyrinogen (60). The propionate on ring B is not oxidatively decarboxylated but the remaining acetate side-chain on ring c can be decarboxylated by uro'gen decarboxylase to give hardero'gen (50),which is the normal intermediate of copro'gen oxidase. In humans who are not suffering from porphyria the concentrations of (45)are so low as to make this alternative pathway insignificant.1.7 Protoporphyrinogen Oxidase (EC 1.3.3.4) The enzyme which oxidizes proto'gen IX (52) to the corre-sponding porphyrin (61) has previously been isolated from R I (61) R = CH=CH, M = H,H (64) R = Et M = H,H (65) R = H M = H,H (66) R = CH=CH2 M = Fe f' \ CO2H CO2H yeast Escherichia coli and rat liver.Io7 As with copro'gen oxidase aerobic organisms use oxygen as the oxidant but it has been shown for E. coli and R. sphaeroides that are growing anaerobically that the enzyme is membrane-bound and the oxidation is linked with the electron-transport chain. lo8 Assay procedures for protoporphyrinogen oxidase have been pub- lishedlo9 but otherwise little work on the enzyme has been published recently.The oxidation involves removal of six hydrogen atoms one from each of the four meso positions and two from nitrogen atoms. It was known that removal of hydrogen from all of the meso positions is stereospecific' lo because the enzymic oxidation of stereorandomly tritiated proto'gen IX occurs with loss of 50% of the label whereas the chemical oxidation occurs with loss of only 4% of the label due to a large isotope effect. As yet it is not known which hydrogen is the one that is removed at any of the sites but the overall stereochemistry from PBG has been studied with interesting results.' [1 1 S3H]PBG was synthesized enzymically from [2-3H2]glycine and was incorpor- ated into protoporphyrin IX with retention of only 25% of the label.Degradation of this labelled protoporphyrin IX to the four isomeric biliverdins e.g. (62) was used to show that the tritium was all at position 10. As the stereochemistry of the various steps in the formation of uro'gen I11 from PBG is not known no conclusion about the stereochemistry of the oxidation of proto'gen can be drawn except that it differs at the various sites. It was suggested that loss of three of the meso hydrogens would be from the same side in the three successive oxidations while loss of the fourth meso hydrogen could be from a different side because it occurs during tautomerism of the resultant product e.g. (63) to the porphyrin. Proto'gen oxidase does not oxidize uro'gens or copro'gens though it is reported to oxidize proto'gen XI11 (58) hardero'gen (50) isohardero'gen (5l) and mesoporphyrinogen IX.lo4 It seems that non-polar groups on rings A and B are an important factor. 1.8 Ferrochelatase (Haem Synthase Protohaem Ferro-lyase; EC 4.99.1.1) The last step in the biosynthesis of haem (66) is the insertion of Fe2+ into the centre of the macrocycle. In eukaryotes the enzyme that catalyses this process is located on the inner face of NATURAL PRODUCT REPORTS 1985 the inner mitochondria1 membrane.' l2 The enzymes from rat liver,' 3 l4 bovine liver,' 5* ' and Rhodopseudomonas sphaer- oides1I7 have recently been solubilized and purified. A sensitive assay for ferrochelatase in tissues using 59Fe has been developed.j9 Kinetic studies' * are claimed to indicate an ordered mechanism in which binding of iron occurs prior to the binding of the porphyrin and then after insertion of the metal haem is released prior to the release of two protons. Ferrochelatase is specific for porphyrins with hydrophobic groups on rings A and B (ie. not uro- or copro-porphyrins) and with the substituents methyl propionate propionate and methyl on rings c and D. Both V,, and KMincrease in the series protoporphyrin IX (61) < mesoporphyrin IX (64) < deutero-porphyrin IX (65),' 14-' and the porphyrin with no substi- tuents on ring A or B is the best substrate that has been studied so far. One explanation that has been put forward' l8 is that release of the haem is the rate-determining step and protoporphyrin IX (which is bound most strongly) is also released most slowly.Other bivalent metals than iron can be inserted into porphyrins. In particular Zn2+ is reported' to be a better substrate even than Fez+; Co2+and Ni2+ can also be substrates. The ions Cuz+ Mnz+ Pb2+ and Hgz+ however are not substrates but inhibitors,' l4 as are thiol-directed reagents. Other inhibitors of ferrochelatase are N-methylated proto- porphyrins which occur in animals to which drugs such as 3,5- dicarbethoxy-l,4-dihydrocollidine or griseofulvin have been administered. The N-methyl derivatives arise by occasional methylation of the haem moiety of cytochrome P-450during metabolism of the drugs in the liver. l 2o The compounds 6v that are methylated on each of the four nitrogen atoms are almost equally good inhibitors' 2o (K = 7 nmol dm-3; ref.116). The N-methylated protoporphyrins that are produced in the liver are optically active but it has not been determined which face of the porphyrin has been methylated.120 1.9 Cytochromes As mentioned earlier unmodified haem (66) is the prosthetic group of several enzymes including the cytochromes b and P-450. The prosthetic group haem a (67) has been isolated from cytochrome oxidase.' There are very few studies on the biosynthesis of haem a and so the order of the steps is not known. It seems likely that protoporphyrin IX (61) is the precursor and that either the 18-formyl compound (68) or the 3- alkylated compound (69) is formed next.Both (68) and (69) have been synthesized. Results suggesting that mevalonate is the precursor of the side-chain have been obtained122 but otherwise no descriptions of incorporation experiments have been published. The likely mechanism for formation of the side-chain at C-3 using farnesyl pyrophosphate is shown in Scheme 13. The stereochemistry of the secondary alcohol of haem a is not known. 602H k02H (67) R = CHO M = Fe (69)R = Me M = H,H (70) R = CHO M = H,H NATURAL PRODUCT REPORTS 1985 F. J. LEEPER C02H k02H (61) R = Me (68) R = CHO J &' & Scheme 13 f \ C02H C02H f \ C02H C02H The last step in the biosynthesis is claimed to be the insertion of iron into porphyrin a (8l,82-dihydrocytoporphyrin)(70)'23 because in Candida utilis that has been grown in a medium that is deficient in copper ions an inactive form of the cytochrome is produced which has porphyrin a and not haem a bound to it.The other major class of cytochromes containing a modified haem is the cytochromes c. These have the vinyl groups attached to the thiol group of a cysteine residue of the enzyme. .-,,. I Scheme 14 In most cytochromes c both vinyl groups have reacted with two cysteine residues which are three residues apart in the peptide chain.' X-Ray crystal structures of the enzymes show that in all cases the new chiral centres are of S configuration as in (72). 24 Cytochrome c-558 from Euglena gracilis' 25 and a similar cytochrome from Crithidia oncopelti' 26 have been shown to have the haem attached by only one covalent link to the vinyl group at C-8 as in (71).The mechanism of formation of the thioether link is presumably that shown in Scheme 14. It would be interesting to determine whether the overall addition to the double-bond is cis or trans. Recently an enzyme has been implicated in the formation of the covalent links between haem and the apo- cytochrome c; it has been named cytochrome c synthetase. 27 2 The Biosynthesis of Chlorophylls The path to the chlorophylls diverges from that of the haems at the point of protoporphyrin IX; magnesium instead of iron is inserted in the centre of the porphyrin. There are marked differences in the type of reactions involved in this later part of the biosynthesis of chlorophylls.Whereas in the biosynthetic reactions considered so far the intermediates that have been described have for the most part been free in solution migrating from one enzyme to the next and many of the enzymes themselves have been in solution also the second part of the biosynthesis of the chlorophylls occurs only in chloroplasts and all of the enzymes are bound to membranes and are quite possibly associated with other enzymes of the pathway. The intermediates too become increasingly non-polar and insoluble in water and the later ones are largely protein- bound in the chloroplasts. As a result of these differences the study of the biosynthesis of chlorophylls is not as well advanced as that of the porphyrins.The enzymes that are involved usually cannot be solubilized and sometimes stop functioning when the chloro- plasts are disrupted. Often studies rely on absorption spectra for the identification of intermediates which can be unreliable. For example mono- and di-vinyl compounds cannot easily be distinguished and nor can various esters be distinguished from their parent acids. Furthermore one compound can show a number of different spectroscopic forms depending on its environment (free in solution or bound to different proteins monomeric or dimeric etc.). Lastly it seems that the difficulties in studying the biosynthesis of chlorophylls may be increased by the relative lack of specificity of the enzymes leading to a metabolic network and also by the existence of degradative enzymes that catalyse for example the hydrolysis of esters.These problems will become more apparent when we discuss the steps in more detail. Recent reviews include ones by Jones,' 28 Castelfranco and Beale,' 29 Rebeiz,I3O and Rebeiz and Lascelles. Overall pathway. The production of protoporphyrin IX for the biosynthesis of chlorophylls is thought to follow exactly the same route as for haem but it should be pointed out that the results with Euglenagracilist6 show that the two processes occur in separate compartments and it is quite possible that the enzymes in chloroplasts are substantially different from those in the cytosol or in mitochondria. The basic outline of the route to chlorophyll a is shown in Scheme 15.After insertion of magnesium into protoporphyrin IX the carboxyl group of the propionate that is attached to C-13 is converted into its methyl NATURAL PRODUCT REPORTS 1985 Protoporphyrin IX (61) Protoporphyrin IX Mg complex (73) 1 Divinyl protochlorophyllide (R = CH=CH2) Chlorophyllide a (R = 14) Monovinyl protochlorophyllide (R = Et) Chlorophyll a (R = phytyl) Scheme 15 'H=CH (or Et) d02H (75) R = CH=CHCO,Me (76) R = CH(OH)CH,C02Me (77) R = COCH2C0,Me ester (74). This is followed by an oxidation and cyclization of this side-chain to form ring E. The resulting protochlorophyl- lide (1 7,18-didehydrochlorophy11ide)is reduced in ring D by a light-dependent enzyme to give a chlorophyllide.Reduction of the 8-vinyl group to an ethyl group might it seems occur at any stage along this pathway. The final step to make chlorophyll is esterification of the free carboxyl to give the phytyl ester (phytyl = hexahydrogeranylgeranyl). 2.1 Chelation of Magnesium and Methylation For a long time it was not certain which of these steps came first. The evidence' 28 that chelation preceded methylation was that protoporphyrinatomagnesium(I1)(73)and its methyl ester (74) can be found in some mutants and in organisms that have been treated with inhibitors such as 8-hydroxyquinoline or phenanthroline. However the methyl ester of free protopor- phyrin IX has also been found and until recently the direct enzymic insertion of magnesium into protoporphyrin IX had not been observed.Two groups have now been able to observe the specific chelation of magnesium by preparations of 33 or lysed134 plastids from cucumber. Both groups found that ATP is required for the reaction. The enzyme which performs the methylation of the carboxyl group in the 13-propionate is more well-known;2~128 it uses S-adenosylmethionine (SAM) and protoporphyrinatomagne-sium(I1) is a much better substrate than free protoporphy- rin IX. The enzyme has been called S-adenosyl-L-methio- nine :magnesium protoporphyrin 0-methyltransferase (EC 2.1.1.11); it has been partially purified from barley135 and recently the enzymes from wheat seedlings' 37 and Euglena gracilis have been more highly purified by affinity chromato- graphy.The kinetics of the enzyme of E. gracilis are said to show a random Bi-Bi mechanism with two dead-end ternary complexes.136 The enzyme from wheat on the other hand (78) R = CH=CH2 (79) R = CHZCH3 shows a ping-pong mechanism with binding of protopor- phyrin IX to the enzyme only occurring in the presence of SAM. 37 The observation of magnesium-free protoporphyrin IX methyl ester may be due to non-enzymic demetallation during isolation or to enzymic demetallation which has been observed recently by Crawford and Wang13* in studies on mutants of Chlamydomonus reinhardti. They showed that protoporphyrin IX is not methylated but that the magnesium complex is. If the further metabolism of the methyl ester is blocked removal of the magnesium occurs in a process that is dependent on iron salts and oxygen.2.2 The Formation of Ring E The oxidation and cyclization of the propionate which forms ring E should involve several steps but very little is known about possible intermediates. Ellsworth and Ar~noff,'~~ in the late 1960's reported the isolation of postulated intermediates from mutants of a species of the genus Chorellu; they suggested that these were the divinyl and monovinyl forms of porphyrins with acrylate 0-hydroxypropionate and 0-ketopropionate side-chains which had lost the magnesium from the corre- sponding compounds (79 (76) and (77) during isolation. Since then there has been no confirmation that these are true intermediates on the normal path to the chlorophylls although intermediates from higher plants have been detected by fluorescence spectroscopy,140-14' which are thought to be (79 (76) and (77).131.'40 A likely mechanism2 for the ring-closure is shown in Scheme 16.Cell-free preparations that are capable of transforming (74) into (78) have been and the evidence from these studies and from studies with intact organisms is that the process requires dioxygen and ATP as well as NAD(P)/ NAD(P)H. The preparations also require SAM for maximum NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER NAD+ Scheme 16 efficiency (to reverse an enzymic hydrolysis of the methyl ester) and one report140 finds that coenzyme A (CoA) is important. It is not obvious why either CoA or ATP should be necessary.In whole organisms grown under conditions of iron deficiency (74) accumulates which suggests that an iron-dependent oxygenase may be involved.144 However the same group142 have failed to observe inhibition of the transformation in isolated chloroplasts by iron-sulphur-protein- or haem-directed inhibitors by chelators of copper ion or by peroxide scavengers; inhibition was observed with methyl viologen and similar compounds. 2.3 Reduction of the Vinyl Group The timing of the reduction of the vinyl group has been a matter of controversy for a long time. Earlier the more widely accepted view' 28 was that it occurred at the protochlorophyl- lide stage. Thus it was thought that the divinyl protochlorophyl- lide (78) (also known as magnesium divinylphaeoporphyrin us methyl ester) is reduced to the monovinyl compound (79).The evidence for this rested on the observations128 that (78) can be found in Rhodopseudomonus sphueroides and in cucumber seed coats and that it can be converted into chlorophyllide a by a cell-free system from barley. (The conversion of the protochlor- ophyllide into the chlorophyllide was believed to occur only at the monovinyl level.) However direct transformation of (78) into (79) was not demonstrated. This view is however at variance with the identification of mono- as well as di-vinyl intermediates prior to protochloro- phyllide in Chlorellu mutants that was mentioned earlier. 39 Furthermore Ellsworth and Hsing' 45 prepared a partially purified homogenate from etiolated wheat seedlings that was able to catalyse the reduction of the 8-vinyl group of protoporphyrinatomagnesium(I1) methyl ester by [3H]NADH (but not NADPH).They did not observe significant reduction of the divinyl protochlorophyllide with this preparation. Strong support for the view that reduction can first occur at a protoporphyrinatomagnesium stage is given in numerous papers by Rebeiz and co-workers130,1 31 in which they describe the identification of mono- and di-vinyl analogues of all of the intermediates i.e. protoporphyrinatomagnesium(I1) and its monoester,146 protochlorophyllide,147 chlorophyllide u,148and chlorophyll They also observed in uiuo the conversion of both mono-and di-vinyl protochlorophyllides into their corresponding chlorophyllides and chlorophylls.149Thus they have proposed that the monovinyl and divinyl paths branch at the stage of protoporphyrinatomagnesium(I1)(73) and remain separate until the chlorophyllide stage where reduction of the divinyl to the monovinyl compound can again be ob-served. The reduction of a vinyl group at the protochlor- ophyllide stage as well however has not yet been disproved. It is claimed' 51 that the main biosynthetic route is the divinyl oqe in normal greening tissues while the monovinyl route prevails only in etiolated tissues (of plants that were grown initially in the dark). Rebeiz's two pathways are each proposed152 to be further subdivided into a path having the 17-propionate as a free acid (as described above) and a path with this propionate esterified with a long-chain alkyl group (to be discussed later).In Rebeiz's investigations the various intermediates are separated by t.1.c. on silica gel and then mono- and di-vinyl forms are separated by a reversed-phase type of t.1.c. on polyethylene. The compounds are identified by fluorescence 33 spectra. Because fluorescence emission spectra can be recorded at any number of excitation frequencies (or excitation spectra at any number of emission frequencies) the technique is very useful for analysing mixtures of compounds and this allows Rebeiz and co-workers to identify mono- and di-vinyl species in intact tissue. The intermediates in this work have only been isolated in picomole quantities and this has precluded any characterization other than by spectrofluorometry.Further- more the samples remain contaminated with other non-fluorescent compounds. Full proof of the proposed pathways must await full characterization of the pure intermediates and the investigation of product-precursor relationships prefera- bly using labelled precursors. 2.4 Reduction of Ring D The reduction of a protochlorophyllide to a chlorophyllide involves the overall trans addition of hydrogen across the 17-1 8 double-bond to give the (1 7S,18s)-compound. In higher plants this process requires light; thus etiolated seedlings (grown in the dark) accumulate a form of protochlorophyllide which can be converted into a chlorophyllide by the briefest flash of light.In normal circumstances the level of protochlorophyllide does not continue to build up in the dark because the synthesis of ALA is shut off but if ALA is fed to the seedling the control point is apparently by-passed and protochlorophyllide does accumulate in relatively large quantities. 28 I 29 The course of the reaction in intact tissue has been followed spectroscopically in a number of studies,' 28*129and the results are not easy to reconcile because of all of the different conditions and techniques that have been employed. One reasonable overall view is summarized in Scheme 17. Free protochlorophyllide absorbing at around 635 nm binds to the reduced enzyme (or enzyme-NADPH complex) to give a chromophore absorbing at 650 nm which is the photoredu- cible form.Using a laser with a 5 ns pulse it was foundls3 that this fluorescent protochlorophyll was immediately converted into a non-fluorescent intermediate absorbing at 690 nm which decayed with a time constant of 3 ps to give a fluorescent chlorophyllide absorbing at about 678 nm. The photoconver- sion can occur at very low temperatures l 54 e.g. 4.2 K but the non-fluorescent intermediate is stable at this temperature and starts to decay if the temperature rises above 170K. The absorption spectrum of the chlorophyllide then undergoes a number of changes which can occur in the dark; it is thought that a shift in the absorption to 683 nm which takes about 30 seconds may reflect dissociation from the enzyme and that a later shift to 672 nm may reflect the phytylation step.l 28 Under conditions of lower light intensity where the phototransforma- tion is incomplete other absorption/emission bands appear and this has led to the suggestion that the phototransformable protochlorophyllide (650 nm) may be a dimeric species and that therefore two light reactions are required to reduce both molecules.128 After a flash of light the re-formation of the reduced enzyme-protochlorophyllide complex from oxidized enzyme NADPH and free protochlorophyllide may take up to 30 seconds.The enzyme that is responsible for this photoreduction NADPH :protochlorophyllide oxidoreductase (EC 1.6.99.l) has been isolated several times. 28. 29 Recent reports have hv Pchlide Pchlide. Emr, -Pchlide*- Enzred (635 nm) (690 nm) 'T* (650nm) Enzred 47 EI\ I NADPH Chlorophyll a c--Chlide Chlide -Enz, (672 nm) (683 nm) (678 nm) (Pchlide = protochlorophyllide Chlide = chlorophyllide) Scheme 17 agreed on a minimum molecular weight of 36000-37000.~55 Grifiths' 56 has studied the substrate specificity of the enzyme and found that protochlorophyllide esters' 56 are not substrates and nor are protoporphyrinatomagnesium derivatives; s7 ring E must be present with the correct substituents.However both mono-and di-vinyl protochlorophyllides are reduced. 57 Activity is still observed if magnesium is replaced by zinc but not if it is replaced by other metals or removed entirely.ls6 The purified enzyme-N ADPH-protochlorophyllide com-plex is called the holochrome.The values of molecular weight that have been determined for various holochrome prepara- tions have a very great range indicating that there is a considerable degree of aggregation. The holochrome undergoes similar changes in absorption to those that are observed in intact tissues but the initial absorbance of the protochlorophyl- lide complex is of somewhat shorter wavelength (638 nm). 28 Some algae and lower plants can synthesize chlorophylls in the dark and contain a second enzyme for the reduction of protochlorophyllide which is light-independent. Pine seedlings also have this capability but only for a short time after germination. One recent report' 58 claims that paradoxically barley that has grown in the light has the ability to synthesize chlorophyll when it is put in the dark but that dark-grown barley does not which would explain why the reduction of protochlorophyllide in the dark is not normally observed.2.5 Esterification of Chlorophyllide u Until recently the esterification of chlorophyllide a was thought to be catalysed by an enzyme named chlorophyllase (EC 3.1.1.14) which is found in all green leaves.' 28 However in uitro,this enzyme could only be shown to catalyse the reverse reaction i.e. hydrolysis to chlorophyllide a and phytol (in methanol or ethanol the methyl or the ethyl ester of chlorophyllide a is obtained). Also chlorophyllase has a broader range of substrates than the esterifying enzyme; in whole leaves phaeophorbides (magnesium-free chlorophyl- lides) are not converted into phaeophytins (magnesium-free chlorophylls) but chlorophyllase will catalyse the reverse reaction.59 Furthermore Akhtar and co-workers using ALA that was labelled with l80in the carboxyl group have shown that in the esterification the 0-phytyl bond of bacteriochloro- phyll a is the one that is formed. 160Thus when bacteriochloro- phyll a that had been biosynthesized from [1-l8O2]ALA was hydrolysed the phytol that was produced contained l80 with at least 90% retention of the label from ALA. If the ester had been formed by chlorophyllase by the reverse of the hydrolysis reaction (which presumably involves cleavage of an 0-acyl bond) then no l80 would be expected in the phytol. If chlorophyllase is not a synthetic enzyme then its role is presumably in the breakdown of chlorophylls that are no longer needed and it is only released when the chloroplast is ruptured.(81) R' = CZOH35 RZ = Et (82) R' = C2OH37 R2 = Et NATURAL PRODUCT REPORTS 1985 In 1977 Riidiger and co-workers showed that geranylgeranyl pyrophosphate rather than the free alcohol is the precursor of the phytyl chain161 and they also demonstrated162 the presence of esters of chlorophyllide a in which the esterifying group was geranylgeranyl (80),dihydrogeranylgeranyl (81) and tetrahy- drogeranylgeranyl(82) as well as chlorophyll a (83) with phytyl (hexahydrogeranylgeranyl) as the esterifying group. It is not known which double-bonds are reduced in the partially reduced intermediates.By following the levels of the various esters they showed that (80)was formed first followed by (81) (82) and then (83). The same conclusion was drawn from a time-course study of the incorporation of 4C-labelled precur- sors into these esters.163 It now seems however that the phytyl ester is produced by at least two different paths because (i) the esterifying enzyme termed chlorophyll synthetase can use phytyl pyrophosphate as well as geranylgeranyl pyrophosphate' 64 (and presumably intermediates between them also) and (ii) an enzyme causing the stepwise reduction of geranylgeranyl pyrophosphate to phytyl pyrophosphate has been detected in chloroplasts.165 The enzymes which reduce the geranylgeranyl-containingchloro-phyll and geranylgeranyl pyrophosphate both use NADPH as hydride donor.66 Chlorophyll synthetase has been located on the thylakoid membrane in the chloroplast.166*167 It appears to be specific for chlorophyllides a and b and simple derivatives of them; little or no reaction was observed with protochlorophyllide phaeophorbide a or bacteriochlorophyllide a.168 As mentioned earlier Rebeiz and co-workers have observed small pools of all of the metabolic intermediates that are esterified with various long-chain alcohols at the propionate group at C-17. 30 l 46-52 Rebeiz has provided evidence that 7 the pathway to these fully esterified intermediates diverges from that of the normal intermediates after protoporphyrinato- magnesiurn(~~)l 69 and that the pools of protochlorophyllide and esterified protochlorophyllide are not interconvertible.70 This is consistent with the observations that protochlorophyllide is not esterified by chlorophyll synthetase168 and nor are protochlorophyllide esters hydrolysed by chlorophyllase. 29 Belanger and Rebeiz observed photoconversion of pools of esters of both monovinyl and divinyl protochlorophyllide into chlorophylls149 but Shioi and Sasa concluded that this did not occur from their measurements of the levels of protochloro- phyllide esters that exist during the synthesis of chloro- phylls. The latter authors have identified the four esters of protochlorophyllide from geranylgeranyl to phytyl and have separated them by h.p.l.c.172 If the protochlorophyllide esters are converted into chloro- phylls then another reductase enzyme must exist because the known protochlorophyllide -reductase was found' 56 not to use the esters as substrates.On the other hand if they are not R2 = Et (84) R' = phytyl RZ = CH=CH2 NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER converted into chlorophylls it is difficult to see what else happens to them. Either way the fully esterified intermediates are much less abundant than the normal free acid interme- diates and constitute (at most) a minor pathway to the chlorophylls. Rebeiz et al. have detected several components in the chlorophyll a fraction from plants by their fluorescence and have partially purified them by h.p.1.~'~~ Chlorophyll a (83) accounts for 80-90% of the fraction and the 81,82-didehydro- chlorophyll a (84) is present in very small amounts (<1%); the remainder is said to consist of three chlorophylls of unknown structure.The divinylated chlorophyll (84) accumulates to a large extent in a mutant of Zea mays and the structure has been well established by n.m.r. and mass spectroscopy. 174 2.6 The Formation of Chlorophyll b Chlorophyll b (86) differs from chlorophyll a (83) only in having the 7-methyl group oxidized to a formyl group. Earlier studies showed that label in (83) is incorporated into (86) and thus (83) was taken to be the immediate precursor of (86)' 28 However more recently a pool of chlorophyllide b (85) has been detected' 75 (along with 8' ,82-didehydrochlorophyllideb) and it was shown that this does not originate by hydrolysis of (86) during isolation.A study of the time course of incorporation of I4C into the various metabolic indicated that labelling of (85) precedes (86) suggesting that (85) is the precursor. It is known that chlorophyll synthetase will also esterify (85),163 and the four esters from chlorophyllide b geranylgeranyl ester to the phytyl ester have all been isolated and separated by h.p.l.~.'~~ Thus it seems that oxidation of chlorophyllide a is one route to chlorophyll b and there is a need for further study to determine whether direct oxidation of chlorophyll a is possible or whether it must first be hydrolysed to the chlorophyllide. 2.7 Bacteriochlorophylls Photosynthetic bacteria contain a number of 'bacteriochloro- phylls' instead of the normal chlorophylls.Bacteriochlorophyll a has the structure (87; R = phytyl). The intermediates of the normal biosynthesis have not been extensively investigated but on the basis of intermediates from mutants' 28v (mostly of Rhodopseudomonas sphaeroides) the most likely path is from chlorophyllide a by hydration of the 3-vinyl group followed by reduction of ring B to give 3-deacetyl-3-( 1-hydroxyethyl)bacter-iochlorophyllide a (88). Oxidation of the hydroxyethyl group then gives (87; R = H) and esterification gives the bacteriochlorophyll. Most of these intermediates have been detected in R. sphaeroides by their fluorescence emission and excitation spectral 78 and by optically detected magnetic resonance at zero field.' 79 Scholz and Ballschmiter180 have investigated minor components of bacteriochlorophyll a fractions by h.p.1.c.and concluded that the six minor compounds that they detected in addition to (87) are diastereoisomers. On the basis that oxidation of the mixture with DDQ gave a single chlorophyll derivative they decided that the components are epimers of (87) at C-7 and C-8 and in ring E; tentative structures have been assigned on the basis of chromatographic mobility. There is more variation in the long-chain ester in bacteria than in higher plants; in Rhodospirillum fulvum the normal phytyl ester of bacteriochlorophyllide a is found but in R. rubrum it is the geranylgeranyl ester. I8O The photosynthetic reaction centres of R.rubrum contain a mixture of this geranylgeranyl bacteriochlorophyllide and bacteriophaeophor- bide a (the magnesium-free compound) as its phytyZ ester. The stereochemistry of the reduction of the 8-vinyl group which occurs at some stage of chlorophyll biosynthesis after protoporphyrin IX has been studied in bacteriochlorophyll a in R. sphaeroides.*82Incorporation of ALA that had 2H and 3H isotopes in the S configuration at C-2 (89) led to the formation of a sample of (88) in which both of the methyl groups that are RO (85) R = H (86) R = phytyl -4Y I (87) XY = 0,R = geranylgeranyl or phytyl (88) X = OH Y = R = H (90) Scheme 18 attached to ring B are in the R configuration as in (90) (established by selective degradation to acetic acid and assay by the usual malate synthase-fumarase method).Thus it is shown that overall the new hydrogen atom at the end of the ethyl group at C-8 is added on the same side as that from which the carboxyl group had been lost. As the stereochemistry of the latter process is known for Euglena gracilis (see Section 1.6) it could be concluded that the addition of hydrogen is from the si-face as shown in Scheme 18. The stereochemistry of the methyl group at C-7 shows that for this organism also the uro'gen decarboxylase step occurs with retention of configuration (see Section 1S). Bacteriochlorophyll b has an exocyclic double-bond2 at C-8 which has now been shown to have the E configuration (91) by n.m.r.spectroscopy using the n.0.e. difference technique. 83 The stereochemistry at C-7 has not yet been determined. Members of the green sulphur bacteria produce a number of closely related pigments known as Chlorobiurn chlorophylls or bacteriochlorophylls c d and e. The structures have been deduced to be (92) for the c series184 and (93) for the d series;185,186 thee series is the same as the d series but with the methyl group at C-7 oxidized to formyl.2 Separation of the magnesium-free methyl ester analogues of these pigments has been achieved using reversed-phase h.p.l.~.,l~~ 87 and partial syntheses of the simplest of them in the c and d series from a derivative of chlorophyll a have been re~0rted.l~~. NATURAL PRODUCT REPORTS.1985 H2N0 H2 H R R' = Et Pr" Bu',or neopentyl R2 = Me or Et (92) R3 = Me (93) R3 = H Noteworthy features of these bacteriochlorophylls are (i) the esterifying alcohol is farnesol (ii) the side-chains at C-8 and at C-12 and the position C-20 bear extra methyl groups which have been found to be derived from methionine and (iii) the configuration of hydroxyethyl groups at C-3 is S when the substituent at C-8 is small (ethyl or propyl) and R when it is large (isobutyl or neopentyl).184,185 2.8 Conclusion The recent results that have been described in this section have shown that the biosynthesis of chlorophylls is not as straightfor- ward as had previously been thought. However much of this work is still very incomplete as many of the intermediates have only been characterized spectroscopically and chromatogra- phically.There is definitely a need for more chemistry to be undertaken in order to isolate purify characterize and synthesize larger quantities of some of the postulated interme- diates and also to perform more definitive experiments using specifically labelled compounds. 3 The Biosynthesis of Vitamin B, The biosynthesis of vitamin B1 (cyanocobalamin) has been covered quite recently in a comprehensive review by Bat- tersby.189 As a result this review will not detail all of the work which has been done but only the important results and those published since the beginning of 1981. 3.1 The Basic Precursors' 7 2 89 Early experiments by Shemin and co-workers established that the cobyrinic acid moiety (94) of vitamin B, (95) is like porphyrins derived from ALA (1) and that the additional methyl groups are provided by methionine.Initial experi- ments with PBG (6) or uro'gen 111 (11) failed to show incorporation of these precursors in intact cells but they are well incorporated in cell-free systems. From uro'gen I11 (1 I) the required steps to cobyrinic acid (94) which is a known precursor of vitamin B1 ,,can be seen to be (i) decarboxylation of the acetate group at C-12; (ii) extrusion of C-20; (iii) introduction of seven methyl groups; (iv) insertion of cobalt. Using 13C n.m.r. spectroscopy it was shown that all of the methyl groups of (94) are labelled by [Me-13C]methionine and that only seven carbons all of them sp2 hybridized are labelled by [5-l3C]ALA.Thus C-20 of uro'gen is lost altogether and does not become the methyl group at C-1 of cobyrinic acid. The possibility that either decarboxylation of the acetate at C-12or methylation at C-1 might occur as the next biosynthetic step after uro'gen I11 (1 1) was investigated using synthetic bilanes and porphyrinogens. The I-methylbilanes that were tested were not incorporated into cobyrinic acid and the incorporation of the decarboxylated uro'gen was very much less than that of uro'gen I11 itself. Therefore it was unlikely that either of these steps could come next. 3.2 intermediates on the Pathway Evidence for the next steps has come from the isolation of compounds that are related to the next three intermediates of the pathway.The structures of the compounds that have been isolated (96)-(98) shown in Scheme 19 are the result of much work by several different groups. Much of the published evidence for the finer details of these structures is biogeneti- cally based. The arrangement of the acetate and propionate side-chains was deduced from the fact that these compounds are derived from uro'gen 111. Once it was known which rings are methylated the position of methylation and its stereochemistry were deduced from the observation that these compounds could be converted into cobyrinic acid. As yet none of these compounds has been synthesized. The compounds (96)-(98) are thought not to be the true intermediates but oxidized forms of them.It has been shown that dihydro-(97) is a better precursor than (97) itself and that a reduced form of (96) is the true intermediate. 3.2.1 Fuctor I Compound (96) was first isolated as its octamethyl ester by and Muller and co-worker~~~~ called Factor I. It was recognized as a chlorin which had one C-methyl group. A sample of methyl-labelled (96) obtained on incorporation of [ Me-'T]SAM was not incorporated into cobyrinic acid unless it was first reduced with sodium amalgam.191 Degradation of the cobyrinic acid showed that the activity was virtually all in rings A + D and knowing the structure of the next intermediate (97) it could be deduced that the site of the methylation was ring A. The methylation of uro'gen I11 at C-2 would naturally lead to a tetrahydro-form of (96) such as (99) or a tautomer of it; see Scheme 19.It is likely that this is the true intermediate and is produced on reduction of (96) with sodium amalgam thus explaining the incorporation results. 3.2.2 Fuctor II Sirohydrochlorin1.2*89 Compound (97) was first isolated in the form of its iron complex as the prosthetic group of bacterial sulphite and nitrite reductases but its structure was not known until the connection with the intermediates of biosynthesis of vitamin B was made. The iron complex was termed sirohaem and the metal- NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER HOz -Methylate C-2 -Me HOz 60,H %-(94) tautomerize CO H C0,H COzH (99) (100) R = H ii 11 (101) iii R11 = Me i i HO?C CO H HOzC C02H (97) R = H (98) R = Me Reagents i air; ii Na analgam; iii H , catalyst or enzyme Scheme 19 ,C02 H HOzC COiH H( fi HOIC COzH C02H ( 102) (103) R = H free compound sirohydrochlorin.Oxidation of (97) can occur (104) R = Me while it is being isolated giving the lactones (102) and (103) Proton n.m.r. spectroscopy of the octamethyl ester of (97) and many of the early structural studies were performed on the showed that the signals of the four meso protons were in the methyl esters of these lactones. They can be reduced with zinc positions that would be expected for an isobacteriochlorin. to give (97) back again. Detailed analysis of the n.m.r. spectrum and biogenetic arguments taken together led to the conclusion that it is rings A and B that are reduced and so the structure is as shown.This was confirmed by incorporation experiments in two ways. First,lg2 sirohydrochlorin that had been labelled with 14Cin the methyl groups by incorporation of [Me-14C]methionine was incorporated into cobyrinic acid (94). Degradation of the heptamethyl ester (cobester) of this labelled compound showed that the methyl groups of ring B and either ring A or ring D were labelled. As it was known that adjacent rings of sirohydro- chlorin are methylated it follows that they must be rings A and B. In the second incorporation experiment lg3 [5-l 3C]ALA was incorporated into sirohydrochlorin. Because of the rearrange- ment of ring D C-14 C-15 and C-16 were all labelled and thus C-15 was immediately recognizable as a triplet in the 3C n.m.r.spectrum. This signal was the most downfield of those from the meso carbons and is due to the meso carbon between the two pyrrolic rings. Thus again it must be rings A and B which are the reduced ones. As with Factor I sirohydrochlorin (97) was thought to be an oxidized form of the true intermediate which on mechanistic grounds should be a dihydro-form (100). By working in rigorously anaerobic conditions Battersby et al. 94 have isolated the dihydrosirohydrochlorin as its methyl ester (101) and shown it to be identical to the product that was obtained by catalytic reduction of sirohydrochlorin ester which was proved to be the ester of (100).Furthermore this dihydro-compound was a better precursor of cobyrinic acid than sirohydrochlorin itself and so it is most likely to be the true intermediate. The incorporation of sirohydrochlorin must be due to a fortuitous enzymic reduction to (100) prior to the next methylation step. The enzyme system that is responsible for the first two methylations of uro’gen I11 to give dihydrosirohydrochlorin has been partially purified to a 50-fold enri~hment.~ It has a pH optimum of about 7 a value for the KM for SAM of 217 pmol dm-3 and a molecular weight of about 200 000. Using this enzyme system it was found that the isobacterio- chlorin chromophore is only generated if oxygen is present thus providing further evidence that the true intermediates are reduced forms of the isobacteriochlorins.As mentioned above the iron complex of (97) is well known. In addition a cobalt(m) complex of an isobacteriochlorin has been identified in the sulphite-reducing organisms Desulphoui- brio gigas and D. desulphuri~ans~~~ and it is likely that this is C0”’-(97). As well as the pigments already mentioned epimers at C-3 of (96)Ig6 and (97)’ 97 have been identified and epimers at C-3 and at C-8 were both obtained during the isolation of (100).lg4 It is not known whether these epimers are present in the incubation mixture or are produced in the early stages of isolation. 3.2.3 Factor III The third compound that is related to precursors of vitamin B was identified as a methylsirohydrochlorin. The early struc- tural work was on the bislactone (104) that is produced during isolation (the tentative structure that was proposed had the third methyl group misplaced).Shortly afterwards spectrosco- pic studies on the octamethyl ester of Factor I11 itself provided proof that the structure is (98). lg8 99 From a biogenetic point of view the position of the third methyl group on C-20 was surprising as C-20 is lost on the path NATURAL PRODUCT REPORTS 1985 to cobyrinic acid. Indeed when the trimethylisobacteriochlorin was biosynthetically labelled with 3H in the methyl groups (from [Me-3H]methionine) and 14C in the ring (from [4-I4C]- ALA),200 or vice uersa,lg9 it was found that incorporation into cobyrinic acid resulted in loss of one of the three methyl groups.Thus the possibility that the methyl group at C-20 of (98) migrates to C-1 is discounted. As with sirohydrochlorin it is assumed that (101) i.e. the dihydro-form of (98) is the true biosynthetic intermediate. 3.2.4 The Synthesis of’Isobacteriochlorins and Chlorins While much has been accomplished with the small quantities of Factors 1-111 that are available from natural sources there is still an urgent need for a synthetic route to these compounds for complete confirmation of their structures for further investiga- tion of their chemistry and for the production of isotopically labelled compounds for further biosynthetic studies. Syntheses of isobacteriochlorins with alkyl substituents have been reported which combine a corrinoid A-B half with a porphinoid C-D half either using a metal ion to act as a template20 or using Eschenmoser’s sulphur-contraction proce- dure.202,203 However attempts to repeat these syntheses with the natural acetate and propionate substituents have not been successful.A recent approach which has proved to be applicable to the natural substituents is Battersby’s photochemical cyclization. In this an A-D half is combined with a B-C half to give the seco- compound (105). Irradiation of (105) produces204 the model isobacteriochlorin (106) in 45% yield.205 Using this approach several model isobacteriochlorins204~206 have and chl~rins~~~ been synthesized including the dimethyl ester of ( +)-bonellin (107),208which is a pigment that has been isolated from a marine worm Bonellia viridis.The application of this approach to the synthesis of Factors 1-111 is currently being investigated. 3.3 Steps beyond Factor I11 Despite considerable efforts no intermediates have been detected beyond the trimethylisobacteriochlorin (98). Our knowledge of the later steps therefore relies on (a) some isotopic labelling studies (b) studies on the chemistry of model compounds and (c) mechanistic considerations. 3.3.1 Decarboxylation of’the Acetate Group at C-12 The mechanism of this decarboxylation is most probably the same as that for the decarboxylation of uro’gen I11 to copro’gen I11 which was discussed in Section 1.5. If this is the case it must occur before the methylation at C-12 while this carbon is still sp2hybridized.Apart from this no information is available on the timing of the step. It is known’ that the decarboxylated acetate group provides the pro-S methyl group at C-12 in (94). Recently it has been shown209 that for this methyl group incorporation of ALA having 2H and 3H in the S configuration at C-2 leads to a chiral methyl group of R configuration at C-12 of cobyrinic acid. Thus this decarboxylation occurs with retention of configuration (like that of uro’gen 111). I \ MeO,C C02Me kO,Me (105) NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER 3.3.2 Methylation Steps It has been mentioned earlier that all of the methyl groups that are introduced into (96)(including the one at C-20) are derived from S-adenosylmethionine (SAM).Furthermore using [Me-2H3]-or [Me-2H3,Me-’ 3C]-methionine intact incorporation of the CD3 group at every position has been demonstrated. Using methionine that carries a chiral C1H2H3H group Arigoni210 has shown that the methylations at C-5 C-7 C-12 and C-15 all occur with inversion of configuration of the chiral methyl as expected for a direct SN2 reaction. Information about the order of the methylations after (98) has been obtained2 in a ‘pulse-labelling’ experiment of the same kind as we have seen for PBG deaminase and ALA dehydratase (Sections 1.4.2 and 1.3). A cell-free system from Clostridium tetanomorphum that is capable of synthesizing cobyrinic acid was incubated with sirohydrochlorin and a deficiency of SAM. This was followed by an excess of [Me-’TISAM so that the methyl groups that are added later should contain more 13C than the earlier ones.In the 13C n.m.r. spectrum of the cobyrinic acid heptamethyl ester it could be seen that of the five enriched methyl groups that at C-17 had the least amount of 3C label and was therefore the first of the five to be introduced. In principle this type of experiment could reveal the order of all five methylations and work on this is in progress. Methylation of dihydro-20-methylsirohydrochlorin(101) at C-17 would be expected to furnish a pyrrocorphin (125; R = CH2C02H M =H,H) (Scheme 23) and this (or its 12- decarboxylated derivative) is presumed to be an intermediate on the pathway. Two groups2 have recently reported an efficient method of removing one or both of the methyl groups at C-5 and C-15 of the heptamethyl ester of cobyrinic acid.Nussbaumer and Arigoni used lead tetra-acetate to produce the hydroxymethyl derivatives followed by an acidic ion-exchange resin to remove the hydroxymethyl group. Eschenmoser’s group on the other hand found that heating cobester in the presence of propane- dithiol led to the bis-demethyl compound. These techniques should allow the production of labelled forms of the demethyl compounds in order to test whether they are intermediates in the biosynthesis. 3.3.3 Model Studies of’the Methylation Steps Eschenmoser and co-workers have described some very interesting results on the methylation of 15,23-dihydroisobac- teriochlorins (also called dipyrrocorphins) and pyrrocorphins.For example the pyrrocorphin (109) was obtained213 by tautomerization of octaethylporphyrinogen (108) with the MgI+ salt of triazabicyclo[4.4.0]dec-5-ene(TBD) (see Scheme 20). The magnesium (or zinc) complex of (109) was methylated with Me1 to give the corphin complex (1 lo) which could be further tautomerized to give the more stable pyrrocorphin (1 11) (after demetallation). Even more relevant to the biosynthesis of vitamin BI2 have been the results with 20-methylated tetrapyrroles2 (Scheme 21). When the metal-free dihydroisobacteriochlorin (1 12) reacted with Me1 the major product was the 12-methylated pyrrocorphin (1 13) along with a minor amount of (1 14). The zinc complex of (115) which is a pyrrocorphin that is very similar to (1 13) was methylated at C-17 by Me1 to give the corphin (1 16) (which cannot in this case tautomerize).The regioselectivity of methylations is slightly different from the biosynthetic sequence described in the previous section in which C-17 is methylated before C-12 but it was found that tautomerization of the dihydroisobacteriochlorin (1 12) to the more stable pyrrocorphin (1 17)203 gave a substrate whose magnesium complex is methylated chiefly at C-17 giving the pyrrocorphin (1 18) (after tautomerization of the corphin). Thus Eschenmoser has demonstrated that the required methylations at C-17 and C-12 are chemically possible and this type of bioeenetic svnthesis might also. in the lone run. provide I -E t ’\D N’ Et- Et / Et M Me Et I kt Scheme 20 (1 10) tautomerize - I + I R (113) R =Me (115) R =H4 Scheme 21 compounds with natural substituents for testing as interme- diates on the biosynthetic pathway.3.3.4 The Extrusion ojC-20 Early results on the extrusion of C-20 were claimed to show that this carbon is lost as formaldehyde.’ However this was made doubtful by the finding that C-20 becomes methylated. Re- investigation215 has shown that C-20 and its attached methyl grow are eliminated intact as an acetic acid unit. NATURAL PRODUCT REPORTS 1985 (1 19) \ iii [M= Nil' or CO"~] ii [M = Zn"] or \ TI [M= iv H] iii[M = Nilf] (122) (121) Reagents i heat or hv H+; ii hv; iii heat; iv H+; v OH-Scheme 22 The mechanism for the ring-contraction of a porphinoid to a corrinoid system and the reason for the puzzling methylation of C-20 have been suggested by some more elegant model chemistry from Eschenmoser's group.The nickel or cobalt complexes of the 20-methyl-20-hydroxydihydrocorphin(1 19) gave on melting (at 295°C and 260"C respectively) the corresponding metal complexes of the 19-acetylcorrin (12 (Scheme 22). This reaction may be a concerted migration or may involve ring-opening to a seco-compound followed by a re- closure to the corrin. The seco-compound (1 20; M = Zn") can be produced from (1 19; M = Zn") either by careful heating or photochemically in the presence of traces of acid,217 and it is also formed if (121;M = Zn") is demetallated in the presence of TFA at room temperature.Removal of the acetyl group from the nickel complex of (121) can be effected2I6 by treatment with 2M-KOH at 70 "Cto give the corrin (122) with release of acetic acid. With this precedent it seems most likely that the biosynthesis follows a similar course such as that illustrated in Scheme 23. The need for the methyl group on (2-20 was explained when the model analogous to (1 I9) but lacking this methyl group was synthesized. It was found that this compound tautomerized very readily to give the ketone (123) and no ring-contraction could be effected. 3.3.5 Other Steps on the Path to Vitamin B, The stage in the biosynthesis at which insertion of cobalt occurs has not been determined.On the one hand the identification of cobalt complexes of isobacteriochlorins that was mentioned earlier' 95 and the observation that the intermediate isobacter- iochlorins (97) and (98) are formed when Propionibacterium shermunii is grown in a cobalt-free medium suggest early insertion of the metal. However other organisms that have been grown in cobalt-free medium excrete cobalt-free corrins such as cobalt-free cobyrinic acid amides from Rhodopseudo-monos sphaeroides2 and cobalt-free cobalamin and cobyric acid from Chromatiurn ~inosurn.*'~ While it is not suggested that these compounds are precursors of vitamin B 2 their isolation nevertheless suggests a later rather than an earlier insertion of cobalt. On chemical incorporation of cobalt into the crude metal- free cobyrinic acid c-amide a small amount of the dehydrocobyrinic acid c-amide (124) was isolated,2 as its ester as well as the expected product.It was suggested that this might be an intermediate in the biosynthesis of cobyrinic acid (94). However this view has been contradicated by the finding2'0 that (1 24) is consistently formed as an artefact during handling of the corresponding corrin. Furthermore a very recent result2*' has indicated that the hydrogen atoms at C-18 and C- 19 of cobyrinic acid are derived from the medium rather than from a reduced nicotinamide coenzyme. When cobyrinic acid was biosynthesized from uro'gen I11 by the cell-free enzyme system in D20,deuterium was incorporated at both C-18 and C-19.This deuterium was not detected by deuterium n.m.r. as previous results222 had shown that the n.m.r. signals from deuterium that is attached to the heptamethyl ester of cobyrinic acid are extremely broad; instead the deuterium was detected by the isotopic shift of the n.m.r. signals of adjacent I3Cnuclei at C-19 and the methylene of the acetate group at (2-18 (derived from the appropriately labelled uro'gen 111). These results suggest that the hydrogen atoms at C-18 and C-19 arise through protonation at these positions rather than by reduction of an 18-19 double-bond. As shown in Scheme 23 the biosynthesis of vitamin B could follow a route which has no oxidations or reductions; in view of the fact that vitamin B plays an important role in primitive anaerobic bacteria such a route is extremely attractive.The biosynthesis of cobyrinic acid from the dihydrotrimeth- ylisobacteriochlorin (101) that is shown in Scheme 23 is purely hypothetical but it is as far as possible consistent with current knowledge on the biosynthesis and the chemistry of model systems. However there are many alternative schemes that could be proposed. Nature has provided us with several surprises in the biosynthesis of vitamin B,' and no doubt has one or two more that have still to be discovered. 3.4 Factor F-430 In the past few years a nickel complex of a new reduced porphyrin system has been identified in methanogenic bac- teria. It has been termed factor F-430 and is -apparently involved in the last step in the reduction of C1compounds to methane.The structure of its methyl ester has been eluci- as (126) by a combination of spectroscopic methods including the interpretation of the 3C n.m.r. spectra of samples of (126) that were obtained after incorporation of various 3C-labelled forms of ALA and [Me-13C]methionine. The similarity of structure (1 26) to that of sirohydrochlorin (97) and the mode of incorporation of ALA can leave no doubt that factor F-430 is related to intermediates of the biosynthesis of vitamin B and is probably derived from sirohydrochlorin. NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER k02H (125) R = CH,C0,HorCH3 M = H,H or CO"' HO,k k02H I 30,H y&/ H+ - I H02C COzH HOzd k0,H methylation at C-5 and C-I5 + (94) IHOzC \CO,H Scheme 23 Me02c R Some reports have appeared on the biochemistry of the production of factor F-430225 and on the chemistry of factor F-4302" and model systems.226 Bykhovskii et al.227have isolated several fractions of pigments (as yet unidentified) that are excreted from methane-producing organisms and which ap- pear to be precursors of factor F-430 that are related to sirohydrochlorin.4 Topics Related to the Biosynthesis of Tetrapyrroles The aim of this section is not to give a comprehensive coverage of any topic but rather to give the interested reader some recent leading references to areas which border on the biosynthesis of tetrapyrroles. 4.1 The Biosynthesis of the Nucleotide Loop of Vitamin B, The remainder of the biosynthesis of vitamin B12 after cobyrinic acid has not been covered in this review because it does not affect the tetrapyrrole nucleus.This part of the biosynthesis has been well reviewed re~ently.~,*'~ In the first step cobyrinic acid is amidated at all of the carboxyl groups except the propionate at C-17 giving cobyric acid. This is followed by amidation of the final carboxyl group with aminopropanol phosphorylation of the hydroxyl group attachment of a guanosine phosphate group and then replacement of this by a dimethylbenzimidazole nucleotide. Finally the extra phosphate group on the nucleotide is hydrolysed off to give the vitamin B itself. Forms of vitamin 42 NATURAL PRODUCT REPORTS 1985 (CHOH), I CHiOH Scheme 24 B1 with bases other than dimethylbenzimidazole do occur naturally.The dimethylbenzimidazole has been found to be derived from riboflavin229 as shown in Scheme 24. The biosynthesis of riboflavin is not yet fully understood but a recent paper230 on the incorporation of multiply 3C-labelled precursors has provided some useful information. 4.2 N.M.R.Spectra and Other Properties of Tetrapyrroies Very important in the study of biosynthesis are physical methods for structure determination and for locating isotopic labels. The most important method is undoubtedly n.m.r. spectroscopy and there is always a need for reliable n.m.r. data and assignments. Smith et have reported 13C and lH n.m.r. spectra of six ‘Type IX’ porphyrins as part of a whole series of papers on the n.m.r.spectra of porphyrins. The spectra of most tetrapyrroles are concentration-dependent and the 3C signals for a-pyrrolic positions (and sometimes p-also) can be very much broadened by slow tautomerism of the NH groups. It was found that both of these disadvantages are overcome in the n.m.r. spectra of zinc(I1) complexes of porphyrins in the presence of pyrrolidine. Useful spectroscopic data are also contained in the papers by Clezy and Fooke~,~~~ describing their syntheses of many diacetyl- divinyl- and diformyl- .~~~ porphyrins. Scott et ~1have reported the I3Cand 15Nn.m.r. spectra of uroporphyrinogens. In the chlorophyll field Wray et al.234 have made a comprehensive collection of 3C data of 15-substituted chlorin derivatives and 13Cassignments have also been published for chlorophyll u,235 chlorophyll b,236and bacteriochlorophyll a.237 The lH n.m.r.spectra of chlorophylls have been studied,238 often to obtain information about their aggregati0n.23~ The IH and 13C n.m.r. spectra of cobester have been analysed in and an almost complete assignment of the 13C n.m.r. spectrum of vitamin B12 has also been published recently.241 Also useful in biosynthetic studies may be an account242 of the Fast Atom Bombardment (FAB) mass spectra of several corrins and a collection243 of X-ray crystal data on corrins including cobester. Scott and co-worker~~~~ have used n.m.r. spectroscopy to follow the course of biosynthesis of porphyrins in intact organisms and have been able to observe the conversion of [5-l3C]ALA into PBG and thence into uro’gen and copro’gen.Using this non-invasive technique they have the factors which regulate the biosynthesis of porphyrins in Rhodopseudomonas sphaeroides and Propionibacterium sher- manii. Scott has reviewed his work on n.m.r. studies of metabolism.246 In order to perform isotopic labelling studies a researcher must be able to introduce the isotopic label into the desired precursor. In two recent papers,247 Smith and co-workers have described the introduction of 2H and I3C labels into porphyr- ins including protoporphyrin IX. This may in some cases provide an easier access to labelled precursors than the usual total synthesis. 4.3 The Degradation of Haem248 The pathway for the removal of unwanted haem in mammals involves as the first step the oxidative cleavage of the ring (127) (128) (R = Et Me or H;M = Ni or VO) with loss of C-5.The resulting linear tetrapyrrole biliverdin is then converted into a dihydro-form bilirubin and this is excreted as its conjugates with glucuronic acid. In plants and algae the same degradation of haem is used to make linear tetrapyrroles such as phycocyanobilin (3,3l- dide hydro-2,3-di hydromesobiliverdin) and p hycoerythrobilin (3,3],18l 1 82-tetradehydro-2,3 15,16-tetrahydromesobiliver-din) which when covalently bound to proteins play important functional roles as photoreceptors and photosynthetic light- harvesters. Recently a ‘Symposium-in-Print’ has been devoted to the subject of linear tetrapyrr~les~~~ and it includes papers on the mechanism of degradation of haem250 and the stereochemistry of biliprotein chrom~phores~~~ which both provide useful references.Also a book on the subject has recently been Another form of degradation of tetrapyrroles occurs when organisms die and decompose. The metalloporphyrins which can be found in oil shales marine sediments rocks etc. are for the most part the remnants of the chlorophylls of decayed plants and photosynthetic organisms. Porphyrins of the types (127) and (128) have been identified recently in oil hale^.^^^-^^^ It is most common that all of the side-chains have been reduced to alkyl groups and the magnesium in the centre of the macrocycle has been replaced by nickel or the vanadyl group (VO).Complete structure determination of these porphyrins as indeed of many of the natural products discussed earlier is greatly facilitated by the technique of nuclear Overhauser effect (n.0.e.) difference spectroscopy. 259 The rigid porphyrin macrocycle helps to produce good nuclear Overhauser effects between nearby hydrogen atoms and this allows reliable determination of the position of each of the peripheral substi tuents. 4.4 Medical Aspects One of the incentives for research into the biosynthesis of porphyrins has been the existence of disorders in the human biosynthesis of haem termed ‘porphyrias’. The several different forms of porphyria have now been correlated with biosynthesis to the extent that a defect in each enzyme of the pathway to haem is now known to give rise to a different porphyria.260*26 The effect of chemicals on the enzymes of the porphyrin pathway has been the subject of recent reviews.262 Much interest has been in Pb2+ which is known to inhibit several of the enzymes of the porphyrin pathway (see Section l) especially ALA dehydratase; the assay of this enzyme in the blood can be used as an early indication of lead poisoning.Another indicator which has proved reliable is the level of protoporphyrinatozinc in the blood which rises in parallel with the level of lead. An area of considerable medical interest at present is the use of porphyrins as photosensitizers to combat cancers. ‘Haemato- porphyrin derivative’ which is produced on treatment of haematoporphyrin (129) with acetic and sulphuric acids followed by mild aqueous alkali is absorbed preferentially by cancerous tissue; on irradiation with visible light the tissue is destroyed by the singlet oxygen that is produced by the photosensitizing porphyrin.Although the major components of the ‘haematoporphyrin derivative’ mixture have been identi- NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER OH fied,264 it seems that they are not the active ones. The structure of the active component(s) has yet to be elucidated but they may be dimeric or 01igomeric.~~~ A collection of papers on the subject has been published recently.266 While ‘haematopor- phyrin derivative’ is of considerable promise chlorins have proved better photo sensitizer^^^^ and also have the advantage that they absorb red light which is the colour that is best transmitted through human tissue.The use of chlorins looks particularly interesting especially if one can be found that can be localized in cancerous tissue. 4.5 Evolution of the Biosynthetic Pathways It was shown in the late 1960’s that porphyrins can be formed in a hydrogen-cyanide-producing prebiotic soup,268 and Eschen- moser has speculated that a vitamin-B ,-like corrin could result from the polymerization of hydrogen cyanide. 269 This specula- tion led him to investigate the tautomerization of porphyrino- gens under anaerobic conditions,270 which he has shown can lead to pyrrocorphins;212 this coupled to his experiments on non-enzymic ring-contraction,2 6+ lends a good deal of credence to the idea that corrins may have existed before life began.Vitamin B12 can be found in some of the most primitive of anaerobic organisms,271 which evolved in the reductive atmosphere of this planet before oxygen-producing plants arrived. It is significant then that no oxidation (or reduction) occurs in the biosynthetic pathway. The methylations which occur on the pathway using S-adenosylmethionine are thought to be a more recent type of reaction and so Eschenmoser considers that a ‘protocobyrinic acid’ having protons in place of the methyl groups may have been a predecessor of contemporary corrinoids. In the methylation steps also Eschenmoser has shown that the reactions are chemically very feasible and it would seem possible that Nature has taken pre- existing chemical reactions and improved them by the evolution of enzymes to cataiyse them.The biosynthesis of haem and the chlorophylls is thought to be a later development than that of because it involves oxidative steps that are not possible for primitive anaerobes (some of which still do not make haem) and because their use is in the production and utilization of oxygen. Therefore the development of this pathway is likely to have started only when some organisms began the first primitive form of photosynthesis. A recent review covers this area.273 The universal use of type I11 rather than type I porphyrins may be another case in which Nature has adapted a pre-existing chemical reaction because the tetramerization of four molecules of PBG in random orientation leads statistically (and in practice) to a preponderance of uroporphyrin 111.It may be that the application of evolutionary ideas will lead to fruitful insights into the mechanisms of the enzymes (such as PBG deaminase and cosynthetase) that have been developed to perfect the reactions and will provide the inspiration that is needed to solve the many remaining problems in the biosynthesis of tetrapyrroles which it is hoped this review has served to highlight. Acknowledgements I would like to thank Prof. A. R. Battersby and Dr C. Abell for proof-reading the manuscript and Dr D. G. Buckley for his collection of reprints.5 References 1 D. G. Buckley Annu. Rep. Prog. Chem. Sect. B 1978 74 392. 2 M. Akhtar and P. M. Jordan in ‘Comprehensive Organic Chemistry’ ed. D. H. R.Barton and W. D. Ollis Pergamon Press Oxford 1979 Vol. 5 p. 1121. 3 ‘The Porphyrins’ ed. D. Dolphin Academic Press New York 1979 Vol. 6 Ch. 1-4. 4 ‘B,?’ ed. D. Dolphin John Wiley New York 1982 Vol. 1 Ch. 4 and 5. 5 ‘Vitamin B Proceedings of the Third European Symposium on Vitamin B12 and Intrinsic Factor’ ed. B. J. Zagalak and W. Friedrich Walter de Gruyter and Co. Berlin 1979 Ch. 2. 6 A. R. Battersby C. J. R. Fookes G. W. J. Matcham and E. McDonald Nature (London) 1980,285,17;A. R. Battersby and E. McDonald Ace. Chem. Res. 1979 12 14. 7 R. C. Davies and A. Neuberger Biochem.J. 1979 177 649 and 661 ; Ref. 3 Ch. 4 and references therein. 8 V. Dzelzkalns T. Foley and S. I. Beale Arch. Biochem. Biophys. 1982 216 196. 9 A. Ohashi and G. Kikuchi J. Biochem. 1979 85 239. 10 G. Srivastava I. A. Borthwick J. D. Brooker B. K. May and W. H. Elliott Biochem. Biophys. Res. Commun. 1982 109 305; ibid. 1983 110 23; ibid. 1983 117 344; 1. A. Borthwick G. Srivastava J. D. Brooker B. K. May and W. H. Elliott Eur. J. Biochem. 1983 129 615. 11 A. M. del C. Batlle E. A. Wider de Xifra and A. M. Stella Int. J. Biochem. 1978 9 407 and references therein. 12 D. L. Nandi Arch. Biochem. Biophys. 1978 188 266. 13 Z. Zaman P. M. Jordan and M. Akhtar Biochem. J. 1973 135 257. 14 M. M. Abboud P. M. Jordan and M. Akhtar J. Chem.Soc. Chem. Commun. 1974 643. 15 A. Laghai and P. M. Jordan Biochem. Soc. Trans. 1976 4 52; ibid. 1977 5 299. 16 D. L. Nandi J. Biol. Chem. 1978 253 8872. 17 S. I. Beale and P. A. Castelfranco Plant Physiol. 1974 53 291 297. 18 S. I. Beale S. P. Cough and S. Granick Proc. Natl. Acad. Sci. USA 1975 72 2719; E. Meller S. Belkin and E. Harel Phytochemistry 1975 14 2399. 19 (a) T. Oh-hama H. Seto N. Otake and S. Miyachi Biochem. Biophys. Res. Commun. 1982 105 647; (b)R. J. Porra 0.Klein and P. E. Wright Eur. J. Biochem. 1983 130 509. 20 S. Hayashi and T. Noguchi J. Biol. Chem. 1983 258 13 693. 21 T. Noguchi and R. Mori J. Biol. Chem. 1981 256 10335. 22 J. B. Lohr and H. C. Friedmann Biochem. Biophys. Res. Commun. 1976 69 908. 23 E.Harel and E. Ne’eman Plant Physiol. 1983 72 1057 1064. 24 J. Chen G. W. Miller and J. Y. Takemoto Arch. Biochem. Biophys. 1981 208 22 1. 25 P. J. Brumm G. A. Thomas and H. C. Friedmann Biochem. Biophys. Res. Commun. 1982 104 814. 26 J. D. Winstein and S. I. Beale J. Biol. Chem. 1983 258 6799. 27 D. Shemin Phil. Trans. R. Soc. London Ser. B 1976 273 109. 28 D. L. Nandi Z. Naturjorsch. Sect. C. 1978 33 799. 29 D. L. Nandi K. F. Baker-Cohen and D. Shemin J. Biol. Chem. 1968,243 1224; D. L. Nandi and D. Shemin ibid. pp. 1231 and 1236. 30 M. M. Abboud and M. Akhtar J. Chem. Soc.. Chem. Commun.. 1976 1007; A. G. Chaudry and P. M. Jordan Biochem. Soc. Trans. 1976 4 760. 31 P. M. Jordan and J. S. Seehra FEBS Lett. 1980 114 283; J. Chem.Soc. Chem. Commun. 1980 240. 32 W. H. Wu D. Shemin K. E. Richards and R. C. Williams Proc. Nutl. Acad. Sci. USA 1974 71 1767. 33 D. R. Bevan P. Bodlaender and D. Shemin J. Biol. Chem. 1980 255 2030. 34 P. N. B. Gibbs and P. M. Jordan Biochem. Soc. Trans. 1981 9 232. 35 D. Gurne J. Chen and D. Shemin Proc. Nutl. Acad. Sci. USA 1977 74 1383 36 A. M. del C. Batlle A. M. Stella A. M. Ferramola Y. Sopena E. A. Wider de Xifra and H. A. Sancovich Int. J.Biochem. 1978 9 401. 37 J. S. Seehra and P. M. Jordan Eur. J. Biochem. 1981 113 435. 38 J. S. Seehra M.G. Gore A. G. Chaudry and P. M. Jordan Eur. J. Biochem. 1981 114 263. 39 B. Linger and T. Kleinschmidt 2.Naturforsch. Sect. C. 1983,38 1059. 40 I. Tsukamoto T. Yoshinaga and S.Sano Int. J. Biochem. 1980 12,751;I. Tsukamoto Y. Takeo and S. Sano Biochim. Biophys. Acta 1979 570 167. 41 S. Van Heyningen and D. Shemin Biochemistry 1971 10 4676. 42 D. Shemin J. Biochem. (Suppl.) 1976 79 37P. 43 P. J. Brumm and H. C. Friedmann Bibchem. Biophys. Res. Commun. 1981. 102 854. 44 B. Lindblad S. Lindstedt and G. Steen Proc. Natl. Acad. Sci. USA 1977 74 4641. 45 P. M. Anderson and R. J. Desnick Enzyme 1982 28 146. 46 For references see ref. 6. 47 A. R. Battersby G. L. Hodgson E. Hunt E. McDonald and J. Saunders J. Chem. Soc. Perkin Trans. I 1976 273. 48 A. R. Battersby J. F. Beck K. H. Gibson G. L. Hodgson R. E. Markwell E. McDonald J. Moron and L. N. Nixon J. Chem. Soc.. Perkin Trans. I 1981 2771. 49 R. Radmer and L.Bogorad Biochemistry 1972 11 904. 50 R. C. Davies and A. Neuberger Biochem. J. 1973 133 471. 51 A. R. Battersby C. J. R. Fookes E. McDonald and M. J. Meegan J. Chem. Soc. Chem. Commun. 1978 185; A. R. Battersby C. J. R. Fookes M. J. Meegan E. McDonald and H. K. W. Wurziger J. Chem. Soc.. Perkin Trans. I 1981 2786; P. Kammerer G. Worner and G. Muller Angew. Chem. In?. Ed. Engl. 1979 18 778; H.-0. Dauner G. Gunzer I. Heger and G. Muller Hoppe-Seyler’s 2. Physiol. Chem. 1976 357 147. 52 (a) A. R. Battersby C. J. R. Fookes G. W. J. Matcham E. McDonald and K. E. Gustafson-Potter J. Chem. Soc.. Chem. Commun. 1979 316; (b)A. R. Battersby C. J. R. Fookes K. E. Gustafson-Potter E. McDonald and G. W. J. Matcham J.Chem. Soc. Perkin Trans. I 1982 2413.53 G. Burton P. E. Fagerness S. Hosozawa P. M. Jordan and A. I. Scott J. Chem. Soc. Chem. Commun. 1979 202; A. I. Scott G. Burton P. M. Jordan H. Matsumoto P. E. Fagerness and L. M. Pryde ibid. 1980 384. 54 P. M. Jordan G. Burton H. Nordlov M. M. Schneider L. Pryde and A. I. Scott J. Chem. Soc. Chem. Commun. 1979 204. 55 G. Burton H. Nordlov S. Hosozawa H. Matsumoto P. M. Jordan P. E. Fagerness L. M. Pryde and A. I. Scott J. Am. Chem. Soc. 1979 101 3114. 56 A. R. Battersby R. G. Brereton C. J. R. Fookes E. McDonald and G. W. J. Matcham J. Chem. Soc. Chem. Commun. 1980 1124. 57 A. R. Battersby C. J. R. Fookes K. E. Gustafson-Potter G. W. J. Matcham and E. McDonald J. Chem. Soc. Chem. Commun. 1979 1155; A. R. Battersby C. J. R. Fookes K.E. Gustafson-Potter E. McDonald and G. W. J. Matcham J. Chem. Soc. Perkin Trans. I 1982 2427. 58 A. Gossauer W. Neidhart and A. I. Scott J. Chem. Soc. Chem. Commun. 1983 883. 59 A. R. Battershy C. J. R. Fookes G. W. J. Matcham E. McDonald and R. Hollenstein J. Chem. Soc. Perkin Trans. I 1983 303 1. 60 Biochem. J. 1983 209 p. 11. 61 A. R. Battersby C. J. R. Fookes G. W. J. Matcham and E. McDonald J. Chem. Soc. Chem. Commun. 1979 539. 62 P. M. Jordan and J. S. Seehra FEBS Lett. 1979 104 364; J. S. Seehra and P. M. Jordan J. Am. Chem. Soc. 1980 102 6841. 63 A. R. Battersby C. J. R. Fookes G. Hart G. W. J. Matcham and P. S. Pandey J. Chem. Soc. Perkin Trans. I 1983 3041. 64 P. M. Jordan and A. Berry Biochem. J. 1981 195 177; A. Berry and P.M. Jordan Biochem. Soc. Trans. 1981 9 231. 65 P. M. Anderson and R. J. Desnick J.Biol. Chem. 1980,255 1993. 66 A. Berry P. M. Jordan and J. S. Seehra FEBS Lett. 1981 129 220. 67 G. Hart F. J. Leeper and A. R. Battersby Biochem. J. 1984,222 93. 68 M. V. Rossetti A. A. Juknat de Geralnik M. Kotler S. Fumagalli and A. M. del C. Batlle Int. J. Biochem. 1980 12 761. 69 D. C. Williams G. S. Morgan E. McDonald and A. R. Battersby Biochem. J. 1981 193 301. 70 P. M. Jordan H. Nordlov G. Burton and A. I. Scott FEBS Lett. 1980 115 269; P. M. Jordan J. G. Straka B. F. Burnham and J. P. Kushner Enzyme 1982 28 158. 71 D. J. Wright and C. K. Lim Biochem. J. 1983 213 85. NATURAL PRODUCT REPORTS 1985 72 A. R. Battersby C. J. R. Fookes and G. W. J. Matcham unpublished work Cambridge 1979.73 J. H. Mathewson and A. H. Corwin J. Am. Chem. Soc. 1961,83 135. 74 A. R. Battersby H. A. Broadbent and C. J. R. Fookes J. Chem. Soc.. Chem. Commun. 1983 1240. 75 A. A. Juknat de Geralnik M. V. Rossetti and A. M. del C. Batlle Int. J. Biochem. 1981 13 343. 76 M. Kohashi R. P. Clement J. Tse and W. N. Piper Biochem.J. 1984 220 755. 77 P. M. Jordan and A. Berry FEBS Lett. 1980 112 86. 78 A. Sburlati R. B. Frydman A. Valasinas S. Rose H. A. Priestap and B. Frydman Biochemistry 1983 22 4006. 79 L. Diaz R. B. Frydman A. Valasinas and B. Frydman J. Am. Chem. Soc. 1979 101 2710. 80 A. R. Battersby C. J. R. Fookes G. W. J. Matcham and P. S. Pandey Angew. Chem. In?. Ed. Engl. 1981 20 293. 81 A. R. Battersby C.J. R. Fookes and P. S. Pandey Tetrahedron 1983 39 1919. 82 J. M. Rideout D. J. Wright and C. K. Lim J. Liq.Chromatogr. 1983,6 383; A. W. Wayne R. C. Straight E. E. Wales and E. Englert J. High Resolut. Chromatogr. Chromatogr. Commun. 1979 2 621; I. C. Walker M. T. Gilbert and K. Stubbs J. Chromatogr. 1980 202 491. 83 A. H. Jackson K. R. N. Rao and S. G. Smith Biochem. J.,1982 203,5 15 ;H. Nordlov P. M. Jordan G. Burton and A. I. Scott J. Chromatogr.,1980,190,221 ;J. C. Bommer B. F. Burnham R. E. Carlson and D. Dolphin Anal. Biochem. 1979 95 444. 84 A. R. Battersby D. G. Buckley D. W. Johnson L. N. Mander E. McDonald and D. C. Williams J. Chem. SOC. Perkin Trans. I 1981 2779; A. R. Battersby D. G. Buckley G. L. Hodgson E. McDonald and R.E. Markwell in ‘High Pressure Liquid Chromatography in Clinical Chemistry’ Academic Press Lon- don 1976 p. 63. 85 D. J. Wright J. M. Rideout and C. K. Lim Biochem. J. 1983 209 553. 86 C. K. Lim J. M. Rideout and D. J. Wright Biochem. J. 1983 211 435. 87 G. H. Elder J. A. Tovey and D. M. Sheppard Biochem. J.,1983 215 45. 88 H. de Verneuil S. Sassa and A. Kappas J.Biol. Chem. 1983,258 2454. 89 S. Kawanishi Y. Seki and S. Sano J.Biol.Chem. 1983,258,4285. 90 J. G. Straka and J. P. Kushner Biochemistry 1983 22 4664. 91 A. G. Smith and J. E. Francis Biochem. J. 1981 195 241 ; 1979 183 455. 92 H. de Verneuil B. Grandchamp and Y. Nordmann Biochim. Biophjs. Acta 1980 611 174. 93 A. H. Jackson H. A. Sancovich and A. M. Ferramola de Sancovich Bioorg.Chem. 1980 9 71. 94 M. Chakrabarty S. A. Ali G. Philip and A. H. Jackson Heterocycles 1981 15 1 199. 95 A. H. Jackson K. R. N. Rao D. M. Supphayen and S. G. Smith J. Chem. Soc. Chem. Commun. 1977 696. 96 G. F. Barnard and M. Akhtar J. Chem.Soc.. Perkin Trans. I 1979 2354. 97 G. H. Elder J. 0. Evans J. R. Jackson and A. H. Jackson Biocheni. J. 1978 169 215. 98 J. S. Seehra P. M. Jordan and M. Akhtar Biochem. J.,1983,209 709. 99 J. H. Keithly and K. D. Nadler J. Bacteriol. 1983 154 838. 100 T. Yoshinaga and S. Sano J. Biol. Chem. 1980 255 4722 4727. 101 M. M. Abboud and M. Akhtar Nouv. J. Chim. 1978 2 419; Z. Zaman and M. Akhtar Eur. J. Biochem. 1976 61 215. 102 A. R. Battersby E.uperientia 1978 34 I ; A. R. Battersby E. McDonald H.K. W. Wurziger and K. J. James J. Chem. Soc. Cheni. Commun. 1975 493. 103 G. H. Elder and J. 0. Evans Biochem. J. 1978 172 345. 104 A. H. Jackson G. H. Elder and S. G. Smith In?. J. Biochem. 1978 9 877. 105 A. R. Battersby A. D. Hamilton E. McDonald L. Mombelli and 0.-H. Wong J. Chem. Soc. Perkin Trans. 1 1980 1283; L. Mombelli E. McDonald and A. R. Battersby Tetrahedron Lett. 1976 1037. 106 A. H. Jackson T. D. Lash D. J. Ryder and S. G. Smith Int. J. Biochem. 1980 12 775. 107 R. Poulson and W. J. Polglase J. Bid. Chem. 1975,250 1269; R. Poulson ibid. 1976 251 3730; R. Poulson K. J. Whitlow and W. J. Polglase FEBS Lett. 1976 62 351. NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER 45 108 N. J. Jacobs and J. M. Jacobs Arch.Biochem. Biophys. 1979,197 151 J. X. Duggan and C. A. Rebeiz Plant Sci. Lett. 1982 27 137. 396; ibid. 1981 211 305; Biochim. Biophys. Acta 1977 459 152 S. A. McCarthy F. C. Belanger and C. A. Rebeiz Biochemistry, 141. 198 1 20 5080. 109 N. J. Jacobs and J. M. Jacobs Enzyme 1982 28 206; J.-M. 153 Y. Inoue T. Kobayashi T. Ogawa and K.Shibata Plant Cell Camadro D. Urban-Grimal and P. Labbe Biochem. Biophys. Physiol. 1981 22 197. Res. Commun. 1982 106 724. 154 0. B. Belyaeva E. R. Personova and F. F. Litvin Photosynth. 110 A. H. Jackson D. E. Games P. Couch J. R. Jackson R. B. Res. 1983 4 81. Belcher and S. G. Smith Enzyme 1974 17 81 ;A. R. Battersby 155 M. Ikeuchi and S. Murakami Plant Cell Physiol. 1982 23 1089; E. McDonald J. R. Redfern J. Staunton and R. H. Wightman J.U. Roper P. Bergweiler and C. Lutz Z. Pfunzenphysiol. 1983 Chem. Soc. Perkin Trans. I 1976 266. 112 89 and references therein; K. Apel H.-J. Santel T. E. 111 C. Jones P. M. Jordan A. G. Chaudhry and M. Akhtar J.Chem. Redlinger and H. Falk Eur. J. Biochem. 1980 111 251. Soc. Chem. Commun. 1979 96. 156 W. T. Griffiths Biochem. J. 1980 186 267. 112 M. S. Jones and 0. T. G. Jones Biochem. J. 1970 113 507. 157 W. T. Grifiths N. L. Morgan and R. E. Mapleston in ‘Genetics 113 K. Mailer R. Poulson D. Dolphin and A. D. Hamilton Biochem. and Biogenesis of Chloroplasts and Mitochondria’ ed. T. Bucher Biophys. Res. Commun. 1980 96,777. W. Neupert W. Sebald and S. Werner Elsevier/North-Holland 114 S. Taketani and R. Tokunaga J. Biol. Chem. 1981 256 12 748. Amsterdam 1977 p.1 1 1. 115 S. Taketani and R. Tokunaga Eur. J. Biochem. 1982 127 443. 158 H. Adamson Prog. Clin. Biol. Res. 1982 102 33. 116 H. A. Dailey and J. E. Fleming J. Biol Chem. 1983,258 11 453. 159 A B. Rudoi A. Yu. Vezitskii and A. A. Shlyk Biokhimiya 1982 117 H. A. Dailey J. Biol. Chem. 1982 257 14 714. 47 733. 118 C. L. Honeybourne J. T. Jackson D. J. Simmonds and 0.T. G. 160 A. A. Abid D. Corina and M. Akhtar J. Chem. Soc. Chem. Jones Tetrahedron 1980 36 1833; C. L. Honeybourne J. T. Commun. 1980 51 1. Jackson and 0. T. G. Jones FEBS Lett. 1979 98 207. 161 W. Rudiger P. Hedden H. P. Kost and D. J. Chapman Biochem. 119 J. R. Bloomer and K. 0. Morton Enzyme 1982 28 220. Biophys. Res. Commun. 1977 74 1268. 120 F. de Matteis A. H. Jackson A. H. Gibbs K. R.N. Rao J. Atton 162 S. Schoch U. Lempert and W. Rudiger Z. Pfanzenphysiol. 1977 S. Weerasinghe and C. Hollands FEBS Lett. 1982 142,44 and 83 427. references therein. 163 J. Benz and W. Rudiger Z. Pfanzenphysiol. 1981 102 95. 121 P. S. Clezy and C. J. R. Fookes Ausr. J. Chem. 1981 34 871. 164 W. Rudiger J. Benz and C. Guthoff Eur. J. Biochem. 1980,109 122 J. Keyhani and E. Keyhani FEBS Lett. 1978 93 271. 193. 123 E. Keyhani and J. Keyhani Biochim. Biophys. Acta 1980 633 165 J. Soll and G. Schultz Biochem. Biophys. Res. Commun. 1981 99 211. 907. 124 Ref. 3 Vol. 7 Ch. 4 and 5. 166 J. Soll G. Schultz W. Rudiger and J. Benz Plant Physiol. 1983 125 M. J. Miller and H. Rapoport J. Am. Chem. Soc. 1977,99,3479. 71 849. 126 R. M. Keller D. Picot and K. Wuthrich Biochim.Biophys. Acta 167 M. A. Block J. Joyard and R. Douce Biochim. Biophys. Acta 1979 580 259. 1980 631 210. 127 H. Taniuchi G. Basile M. Taniuchi and D. Veloso J. Biol. 168 J. Benz and W. Rudiger Z. Naturjorsch. Sect. C 1981 36 51. Chem. 1983 258 10 963 and references therein. 169 S. A. McCarthy J. R. Mattheis and C. A. Rebeiz Biochemistry, 128 0. T. G. Jones in ref. 3 Vol. 6 Ch. 3. 1982 21 242. 129 P. A. Castelfranco and S. 1. Beale in ‘The Biochemistry of Plants 170 J. R. Mattheis and C. A. Rebeiz Arch. Biochem. Biophys. 1977 Vol. 8 Photosynthesis’ ed. M. D. Hatch and N. K. Boardman 184 189. Academic Press New York 1981 p. 375. 171 Y. Shioi and T. Sasa Plant Cell Physiol. 1983 24 835. 130 C. A. Rebeiz CHEMTECH 1982 12 52. 172 Y. Shioi and T. Sasa Arch. Biochem.Biophys. 1983 220 286; Y. 131 C. A. Rebeiz and J. Lascelles in ‘Photosynthesis’ ed. Govindjee Shioi R. Fukae and T. Sasa Biochim. Biophys. Acta 1983 722 Academic Press New York 1982 p. 699. 72. 132 A. D. Pardo B. M. Chereskin P. A. Castelfranco V. R. 173 C. A. Rebeiz F. C. Belanger G. Freyssinet and D. B. Saab Franceschi and B. E. Wezelman Plant Physiol. 1980 65 956; Biochim. Biophys. Acra 1980 590 234. T. P. Fuesler L. A. Wright Jr. and P. A. Castelfranco ibid. 1981 174 M. B. Bazzaz C. V. Bradley and R. G. Brereton Tetrahedron 67 246. Lett. 1982 23 121 I ; M. B. Bazzaz and R. G. Brereton FEBS 133 T. P. Fuesler C. M. Hanamoto and P. A. Castelfranco Planf Lett. 1982 138 104. Physiol. 1982 69 421. 175 J. X. Duggan and C. A. Rebeiz Biochim. Biophys. Acta 1982,679 134 M.L. Richter and K. G. Rienits Biochim. Biophys. Acta 1982 248. 717 255. 176 S. Aronoff Biochem. Biophys. Res. Commun. 1981 102 108. 135 J. Shieh G. W. Miller and M. Psenak Plant Cell Physiol. 1978 177 Y. Shioi and T. Sasa Biochim. Biophys. Acta 1983 756 127. 19 1051. I78 J. Beck and G. Drews 2. Naturfbrsch. Sect. C 1982 37 199. 136 S. B. Hinchigeri and W. R. Richards Photosynthetica 1982 16 179 J. Beck J. U. von Schutz and H. C. Wolf Z. Naturjorsch. Sect. C 554. 1983 38 220. 137 S. B. Hinchigeri J. C.-S. Chan and W. R. Richards Phofosynthe-180 B. Scholz and K. Ballschmiter Angew. Chem. Int. Ed. Engl. 1981 tica 1981 15 351. 20 956; J. Chromatogr. 1982 252 269. 138 M. S. Crawford and W.-Y. Wang Plant Physiol. 1983 71 303. 181 E. Walter J. Schreiber E.Zass and A. Eschenmoser Helo. Chim. 139 R. K. Ellsworth and S. Aronoff Arch. Biochem. Biophys. 1969 Acfa 1979 62 899. 130 374; ihid. 1968 125 35 269. 182 A. R. Battersby A. L. Gutman C. J. R. Fookes H. Gunther and 140 R. K. Ellsworth and S. J. Murphy Photosynthetica 1979 13 392. H. Simon J. Chem. SOC.,Chem. Commun. 1981 645. 141 C. A. Rebeiz J. R. Mattheis B. B. Smith C. C. Rebeiz and D. F. 183 N. Risch J. Chem. Res. (S). 1981 116. Dayton Arch. Biochem. Biophys. 1975 166 446. 184 K. M. Smith L. A. Kehres and H. D. Tabba J. Am. Chem. Soc. 142 B. M. Chereskin Y.-S. Wong and P. A. Castelfranco Plant 1980 102 7149. Physiol. 1982 70 987. 185 K. M. Smith D. A. Goff J. Fajer and K. M. Barkigia J. Am. 143 B. M. Chereskin and P. A. Castelfranco Plant Phqsiol.1982 69 Chem. Soc. 1982 104 3747. 112; B. M. Chereskin P. A. Castelfranco J. L. Dallas and K. M. 186 K. M. Smith D. A. Goff J. Fajer and K. M. Barkigia J. Am. Straub Arch. Biochem. Biophyy. 1983 226 10. Chem. Soc. 1983 105 1674. 144 S. C. Spiller A. M. Castelfranco and P. A. Castelfranco Plant 187 K. M. Smith G. M. F. Bisset and M. J. Bushell J. Org. Chem. Phj.siol. 1982 69 107. 1980 45 2218. 145 R. K. Ellsworth and A. S. Hsing Biochim. Biophys. Acra 1973 188 H. Brockmann Jr. U. Jurgens and M. Thomas Tetrahedron 313 119. Lett. 1979 2133. 146 F. C. Belanger and C. A. Rebeiz J. Biol. Chem. 1982 257 1360 189 A. R. Battersby in ref. 4 p. 107. and references therein. 190 R. Deeg H.-P. Kriemler K.-H. Bergmann and G. Muller Hoppe-147 F. C. Belanger and C. A. Rebeiz J.Biol. Chem. 1980 255 1266. Sejder’s 2. Phj’siol. Chem. 1977 358 339. 148 F. C. Belanger J. X. Duggan and C. A. Rebeiz; J. Bid. Chem. 191 K.-H. Bergmann R. Deeg K. D. Gneuss H.-P. Kriemler and G. 1982 257 4849. Muller Hoppe-Seyler’sZ. Physiol. Chem. 1977 358 1315. 149 F. C. Belanger and C. A. Rebeiz Plant Sci. Leff.,1980 18 343. 192 A. R. Battersby E. McDonald M. Thompson and V. Ya. 150 J. X. Duggan and C. A. Rebeiz Plant Sci. Lett. 1982 24 27. Bykhovsky J. Chem. Soc. Chem. Commun. 1978 150. 193 A. I. Scott A. J. Irwin L. M. Siegel and J. N. Shoolery J. Am Chem. Soc. 1978 100 316 7987. 194 A. R. Battersby K. Frobel F. Hammerschmidt and C. Jones J. Chem. Soc.. Chem. Commun. 1982 455. 195 A. R. Battersby and Z.-C. Sheng J. Chem. Soc. Chem. Commun.1982 1393. 196 A. R. Battersby and S. Seo J. Chem. Soc. Perkin Trans. I 1983 3049. 197 A. R. Battersby E. McDonald R. Neier and M. Thompson J. Chem. Soc. Chem. Commun. 1979 960. 198 A. R.Battersby G. W. J. Matcham E. McDonald R. Neier M. Thompson W.-D. Woggon V. Ya. Bykovsky and H. R. Morris J. Chem. Soc. Chem. Commun. 1979 185. 199 G. Muller K. D. Gneuss H.-P. Kriemler A. I. Scott and A. J. Irwin J.Am. Chem. Soc. 1979 101 3655; Tetrahedron 1981 37 81 ;C. Nussbaumer M. Imfeld G. Worner G. Muller and D. Arigoni Proc. Natl. Acad. Sci. USA 1981 78 9. 200 N. G. Lewis R. Neier G. W. J. Matcham E. McDonald and A. R. Battersby J. Chem. SOC.,Chem. Commun. 1979 541. 201 F.-P. Montforts S. Ofner V. Rasetti A. Eschenmoser W.-D. Woggon K. Jones and A.R. Battersby Angew. Chem. Int. Ed. Engl. 1979 18 675; P. Naab R. Lattmann C. Angst and A. Eschenmoser ibid. 1980 19 143. 202 S. Ofner V. Rasetti B. Zehnder and A. Eschenmoser Helu. Chim. Acta 1981 64 1431. 203 K. Hilpert C. Leumann A. P. Davis and A. Eschenmoser J. Chem. SOC.,Chem. Commun. 1983 1401. 204 P. J. Harrison C. J. R. Fookes and A. R. Battersby J.Chem.Soc. Chem. Commun. 1981 797. 205 A. R. Battersby and Z.-C. Sheng unpublished work. 206 D. M. Arnott A. R. Battersby P. J. Harrison G. B. Henderson and Z.-C. Sheng J. Chem. Soc. Chem. Commun. 1984 525. 207 A. R. Battersby C. J. Dutton C. J. R. Fookes and S. P. D. Turner J. Chem. Soc. Chem. Commun. 1983 1235. 208 C. J. Dutton C. J. R. Fookes and A. R. Battersby J.Chem. Soc. Chem.Commun. 1983 1237. 209 A. R. Battersby K. R. Deutscher and B. Martinoni J. Chem. Soc. Chem. Commun. 1983 698. 210 D. Arigoni in ‘Molecular Interactions and Activity in Proteins’ Ciba Foundation Symposium no. 60 Excerpta Medica Amster- dam 1978 p. 243. 21 1 H. C. Uzar and A. R.Battersby J. Chem. Soc. Chem. Commun. 1982 1204; D. M. Doddrell H. C. Uzar and A. R. Battersby J. Org. Chem. 1983 48 3613. 212 C. Nussbaumer and D. Arigoni Angew. Chem. Int. Ed. Engl. 1983 22 737; N. J. Lewis R. Nussberger B. Krautler and A. Eschenmoser ibid. p. 736. 213 R. Wadischatka and A. Eschenmoser Angew. Chem. Int. Ed. Engl. 1983 22 630; R. Wadischatka E. Diener and A. Eschenmoser ibid. p. 631. 214 C. Leumann K. Hilpert J. Schreiber and A. Eschenmoser J.Chem. Soc. Chem. Commun. 1983 1404. 215 L. Mombelli C. Nussbaumer H. Weber G. Muller and D. Arigoni Proc. Natl. Acad. Sci. USA 1981,78 11 ;A. R. Battersby M. J. Bushell C. Jones N. G. Lewis and A. Pfenninger ibid. p. 13. 216 V. Rasetti A. Pfaltz C. Kratky and A. Eschenmoser Proc. Natl. Acad. Sci. USA 1981,78 16. For methods of demetallating these and other metalloporphyrin derivatives see A. R. Battersby K. Jones and R. Snow Angew. Chem. Int. Ed. Engl. 1983,22 734; N. J. Lewis A. Pfaltz and A. Eschenmoser ibid. p. 735. 217 V. Rasetti K. Hilpert A. Fassler A. Pfaltz and A. Eschenmoser Angew. Chem. In?. Ed. Engl. 1981 20 1058. 218 B. Dresow L. Ernst L. Grotjahn and V. B. Koppenhagen Angew. Chem. Int. Ed. Engl. 1981,20 1048; E. Siefert and V.B. Koppenhagen Arch. Microbiol. 1982 132 173 and references therein. 219 R. Bieganowski and W. Friedrich Z. Naturforsch. Sect. C. 1981 36 9. 220 C. Nussbaumer B. Krautler and D. Arigoni unpublished work quoted in ref. 221. 221 A. R. Battersby C. Edington and C. J. R. Fookes J. Chem. Soc. Chem. Commun. 1984 527. 222 A. R. Battersby C. Edington C. J. R. Fookes and J. H. Hook J. Chem. SOC.,Chem. Commun. 1982 181. 223 A. Pfaltz B. Jaun A. Fassler A. Eschenmoser R. Jaenchen H. H. Gilles G. Diekert and R. K. Thauer Helu. Chim. Acta 1982 65 828. 224 H. Gilles and R. K. Thauer Eur. J. Biochem. 1983 135 109; R. Jaenchen G. Diekert and R. K. Thauer FEBSLett. 1981 130 NATURAL PRODUCT REPORTS 1985 133; G. Diekert R. Jaenchen and R. K. Thauer ibid.1980,119 118. 225 J. T. Keltjens C. G. Caerteling A. M. Van Kooten H. F. Van Dijk and G. D. Vogens Arch. Biochem. Biophys. 1983,223,235. 226 A. Fassler A. Pfaltz P. M. Muller S. Farooq C. Kratky B. Krautler and A. Eschenmoser Helv. Chim. Acta 1982 65 812. 227 V. Ya. Bykhovskii N. I. Zaitseva T. N. Zhilina and G. A. Zavarzin Dokl. Akad. Nauk. SSSR 1983 271 470. 228 F. M. Huennekens K. S. Vitols K. Fujii and D. W. Jacobsen in ref. 4 p. 145. 229 J. Horig and P. Renz in ref. 5 p. 323. 230 P. J. Keller Q. Le Van A. Bacher and H. G. Floss Tetrahedron 1983 39 3471. 231 K. M. Smith R. J. Abraham and H. Pearson Tetrahedron 1982 38,2441. 232 P. S. Clezy and C. J. R. Fookes Aust. J.Chem. 1980,33,545,557 and 575. 233 G. Burton P. E. Fagerness P.M. Jordan and A. I. Scott Tetrahedron 1980 36 2721. 234 V. Wray U. Jurgens and H. Brockmann Jr. Tetrahedron 1979 35 2275. 235 S. Lotjonen and P. Hynninen Org. Magn. Reson. 1981 16 304; see also ref. 19a. 236 N. Risch and H. Brockmann Tetrahedron Lett. 1983 24 173. 237 R. G. Brereton and J. K. M. Sanders J.Chem. Soc. Perkin Trans. I 1983 435. 238 J. K. M. Sanders J. C. Waterton and I. S. Denniss J.Chem. Soc. Perkin Trans. I 1978 1150; K. M. Smith D. A. Goff R. J. Abraham and J. E. Plant Org. Magn. Reson. 1983 21 505. 239 R. G. Brereton and J. K. M. Sanders J.Chem. Soc. Perkin Trans. I 1983,423; R.J. Abraham and K. M. Smith J.Am. Chem. Soc. 1983 105 5734; Tetrahedron Lett. 1983 24 2681. 240 A. R. Battersby C. Edington C. J. R. Fookes and J.M. Hook J. Chem. Soc. Perkin Trans. I 1982 2265; L. Ernst Liebigs Ann. Chem. 1981 376. 241 D. L. Anton H. P. C. Hogenkamp T. E. Walker and N. A. Matwiyoff Biochemistry 1982 21 2372. 242 H. Schwarz K. Eckart and L. C. E. Taylor Org. Mass Spectrom. 1982 17 458. 243 K. Kamiya and 0.Kennard J. Chem. Soc. Perkin Trans. I 1982 2279. 244 A. I. Scott G. Burton and P. E. Fagerness J. Chem. SOC.,Chem. Commun. 1979 199. 245 G. Burton P. M. Jordan N. E. Mackenzie P. E. Fagerness and A. I. Scott Biochem. J. 1981 194 627. 246 G. Burton R. L. Baxter J. M.Gunn P. J. Sidebottom P. E. Fagerness K. Shishido J. Y. Lee and A. I. Scott Can J. Chem. 1980,543 1839; A. I. Scott Pure Appl. Chem. 1981,53 1215; A. I. Scott and R. L. Baxter Annu. Rev. Biophys.Bioeng. 1981 10 151. 247 K. M. Smith E. M. Fujinari K. C. Langry D. W. Parish and H.D. Tabba J.Am. Chem. SOC., 1983,105,6638; K. M. Smithand K. C. Langry J. Chem. Soc. Perkin Trans. I 1983 439. 248 Reviewed by R. Schmid and A. F. McDonagh in ref. 3 Vol. 6 p. 258. 249 ‘Linear Tetrapyrroles’ ed. R. Bonnett Tetrahedron 1983 39 1839-1954 250 V. Rajananda and S. B. Brown Tetrahedron 1983 39 1927. 251 A. Gossauer Tetrahedron 1983 39 1933. 252 ‘Bilirubin Metabolism’ ed. K. P. M. Heirwegh and S. B. Brown CRC Press Florida 1982 Vol. 2. 253 C. J. R. Fookes J. Chem. Soc. Chem. Commun. 1983 1472 and 1474. 254 G. A. Wolff M. Murray J. R. Maxwell B. K. Hunter and J. K. M. Sanders J. Chem. Soc. Chem. Commun. 1983 922. 255 A. Ekstrom C. J.R. Fookes T. Hambley H. J. Loeh S. A. Miller and J. C. Taylor Nature (London) 1983 306,173. 256 J. Krane T. Skjetne N. Telnaes M. Bjoroy and H. Solli Tetrahedron 1983 39 4109. 257 J. F. Branthaver C. B. Storm and E. W. Baker Org. Geochem. 1983 4 121. 258 G. Habermehl and G. Springer Naturwissenschaften 1982 69 543. 259 K. N. Ganesh J. K. M. Sanders and J. C. Waterton J. Chem. Soc. Perkin Trans. 1 1982 1617; J. K. M. Sanders and J. D. Mersh Prog. Nucl. Magn. Reson. Spectrosc. 1983 15 353. 260 L. Bogorad in ref. 3 Vol. 6 p. 125; L. Eales ibid. p. 665. 261 S. W. Eubanks J. W. Patterson D. L. May and J. L. Aeling Int. J. Dermatol. 1983 22 337; G. H. Elder Br. J. Dermatol. 1983 108 729. NATURAL PRODUCT REPORTS 1985 -F. J. LEEPER 262 G.S. Marks D. T. Zelt and S. P. C. Cole Can. J. Physwl. Pharmacol. 1982,60 1017; G. S. Marks Trends Pharmacol. Sci. 1981 2 59. 263 See for example A. Fischbein J. C. Thornton L. Sarkozi S. Kon and S. Levin J. Appl. Toxicol. 1982,2,289; M.M. Joselow in 'Zinc in the Environment' ed. J. 0.Nriagu Wiley New York 1980 Vol. 2 p. 171. 264 P. S. Clezy T. T. Hai R.W. Henderson and Le van Thuc Aurt. J. Chem.,1980,33,585; R. Bonnett R. J. Ridge P. A. Scourides and M. C. Berenbaum J. Chem. Soc. Perkin Trans. I 1981 3135. 265 M. C. Berenbaum R. Bonnett and P.A. Scourides Br. J. Cancer 1982 45 571. 266 'Porphyrin Photosensitization' Ado. Exp. Med. Biol. 1983 Vol. 160 ed. D. Kessel and T. J. Dougherty. 267 L. Cariello M. de Nicola Giudici E. Tosti and L.Zanetti Gamete Res. 1982 5 161. 268 G. W. Hodgson and C. A. Ponnamperuma Proc. Natl. Acad. Sci. USA 1968 59 22. 269 A. Eschenmoser lecture given while Alexander Todd Visiting Professor University of Cambridge 20th May 1981. 270 A. Eschenmoser in ref. 5 p 89; Chem. Soc. Rev. 1976 5 377. 271 W. S. Beck in ref. 4 Vol. 2 p. 1. 272 N. H. Georgopapadakou and A. I. Scott J. Theor. Biol. 1977,69 381. 273 G. A. F.Hendry and 0.T. G. Jones J. Med. Genet. 1980,17 1.