Journal of Medicinal Plant Research Editor - in - Chief Editorial Board Hippokrates Verlag E. Reinhard, Univ. Tubingen H. Ammon, Tubingen Stuttgart Pharmazeutisches lnstitut W. Barz. Mijnster Auf der Morgenstelle E. Reinhard, Tiibingen 7400 Tubingen 0.Sticher, Zurich Vol. 36 H. Wagner, Munchen M. Zenk, Bochum May 1979 NO.1

Biosynthesis of lsoquinoline

J. Staunton

University Chemical Laboratory, Lensfield Road, Cambridge, U.K.

Key Word Index: Biosynthesis; lsoquinoline Alkaloids.

Abstract tial aminoacids tyiosine or phenflala- nine. Decarboxylation gives a phene- The alkaloids and their thylamine building block (C,-C,) which biosynthetic relatives include many then combines with an additional build- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. compounds which show important ing block to form an isoquinoline ring. physiological properties in animals: This stage of the general biosynthetic dopamine, mescaline, , papa- scheme is illustrated by the biosynthesis verine, and narcotine are pertinent of the cactus anhalonidine. examples. Many important classes of alkaloid The biosynthetic routes to the mem- are produced by further transforma- bers of this family start from the essen- tion of simple isoquinoline systems. In 2 Staunton the poppy the benzylisoquino- types are found in cacti; benzylisoqui- line is converted to the two nolines related to (4) and phenethyliso- very different alkaloids morphine and quinolines based on (5) are produced in narcotine. More recently it has been many types of plant. The general struc- estblished that two of the alkaloids of ture of these more simple isoquinoline Stephania. japonica, protostephanine systems is shown in (6). Each diagram andhasubanonine are also produced by is marked in heavy type to indicate the modification of a benzylisoquinoline way in which the structure might be precursor. built up from simpler building blocks. These varied biosynthetic pathways It will be clear that in each case a phe- and the methods used in their elucida- nethylamine unit is converted to an tion are discussed in detail. isoquinoline by combination with a se- cond building block of varied structure. The isoquinoline alkaloids comprise As we. shall see later this analysis of a diverse family of natural products their structural relationship is reflected many of whose members play an im- in their biosynthesis. portant role in medicine. In this con- Scheme 2 shows a selection of more nection one need only mention for complicated isoquinoline alkaloid struc- example morphine and the tures. The skeleton (6) is re- pain-killers, or emetine which is used cognisably an isoquinoline derivative to treat amoebic dysentry. as are those of the morphinans (7), the The range of structures to be found protoberberines (8), the phthalideiso- in this family starts with the relative quinolines, (9), and the erythrina alka- simple tetrahydroisoquinoline deriva: loids (10). At first sight the inclusion tives shown in Scheme I. The first three of (11) the basic skeleton of the alkaloid This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Skeletal Structures of Simple Tetrahydroisoquinoline Alkaloids Scheme 1 Biosynthesis of lsoquinoline Alkaloids 3.

Systems Derived from Benzylisoquinoline Precursors

Scheme 2 protostephanine seems inappropriate isoquinoline structures are elaborated but as we shall see later, this alkaloid to more complicated systems such as is derived via benzylisoquinoline pre- those illustrated in Scheme 2. Our cur- cursors (as are all the other systems rent understanding of the early stages illustrated in Scheme 2) and so it is rea- of isoquinoline alkaloid biosynthesis sonable to list it as a member of this owes much to the pioneering studies car- extensive alkaloid family. ried out on the origin of the three alka- The biosynthesis of isoquinoline al- loids (12), (13), and (14) produced by kaloids is conveniently treated in two the cactus ~bphophorawilliamsii. These parts, starting with the generation of alkaloids have long been of interest on the isoquinoline system and then going account of their hallucinogenic proper- on to those processes by which simple ties. This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

(12) mescaline (13) anhalamine (14) anhalonidine

Scheme 3 Alkaloids of Lpphphora williarnsii 4 Staunton

The two isoquinoline alkaloids bear into anhalonidine and the specificity of an obvious structural relationship to the label in the alkaloid was established each other and also to their phenethyla- by a standard Hofmann degradation- mine. congener mescaline. Early feeding oxidation sequence leading to formal- experiments showed that as is so often dehyde. This was produced exclusively the case, tyrosine is the precursor of the from C-3 of the tetrahydroisoquinoline phenethylamine residue [I, 21. In Sche- and as expected the molar activity of me 4, for example, activity from 2-"C this fragment was the same as that of tyrosine was incorporated efficiently the alkaloid.

chemical EH~O HO NH -degradation

Scheme 4 lncorporat~on of Tyrosine do Anhalonidine The sequence of steps leading from convert it to the key intermediate (15). tyrosine to the various alkaloids has At this point the pathway branches. been worked out in great detail by pre- 0-Methylation at one site of (15) leads cursor incorporation experiments (Sche- to (16), which is further methylated to me 5) -[3-81. Firstly tyrosine is decar- give mescaline. Alternatively, O-Me- boxylated to tyramine which is then thylation of (15) at the second hydroxyl subjected to a sequence of hydroxyla- gives (17) which is ultimately converted tion and 0-methylation steps which to the two isoquinoline alkaloids. Sur- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Scheme 5 Pathway from tyrosine to the alkaloids of L. williarnsi~ Biosynthesis of lsoquinoline Alkaloids 5 prisingly perhaps (17) is apparently not this simple expectation. Thus in Scheme converted to mescaline so it would ap-- 6 the phenethylamine condenses with a pear that the fate of material as it flows ' suitable ketoacid to form the carboxy- along the pathway is decided by the derivative (18) of the eventual tetrahy- site of 0-methylation of (15). droisquinoline. The extraneous group This general pattern will be seen is subsequently removed in an intriguing again later when we come to consider oxidative decarboxylation to form an biosynthetic pathways leading to other intermediate imine (19); this is reduced isoquinoline alkaloids: tyrosine is first to the tetrahydroderivative in a subse- elaborated to a suitable .hydroxylated quent step. The evidence in support of and derivatised phenethylamine which this scheme is exceptionally strong in- is then combined with a second building cluding as it does the isolation of the block to give an isoquinoline. It will two ketoacids of structure (18) as well have been clear from Scheme 1 that the as positive 'results from conventional structure of this second building block incorporation experiments. can vary widely. At first sight the task In L. wiilliamsii this pathway opera- of identifying the second unit might tes exclusively but the mode of biosyn- seem straightforward but in fact it has thesis of the related alkaloid lophoceri- often proved to be unexpectedly diffi- ne (21) in L. schottii appears to be less cult. The popular view was until recent- clear cut. Two alternative pathways by ly that the phenethylamine would con- which a phenethylamine could be con- dense with the appropriate aldehyde to verted to the isoquinoline derivative form .a tetrahydroisoquinoline directly; are shown in scheme 7. In the first meva- a persuasive analogy for this hypotheti- lonate is converted to an aldehyde (20) cal scheme can be found in the chemical which could combine with a suitable '. synthesis of tetrahydroisoquinolines as- phenethylamine to give the required sociated with the names of Pictet and skeleton. The second pathway starts Spengler. Nature does not necesssarily with leucine (22) and proceeds via steps follow chemical precedents, however, equivalent to those shown in the pre- and in the case of the peyote cactus al- vious scheme to a tetrahydroisoquino- kaloids the pathway eventually estab- line. Surprisingly tracer experiments lished [9] is markedly different from have provided support for both path- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Scheme 6 Generation of the tsoquinoline System in L. williamsi~ 6 Staunton

Scheme 7 Lophocerine Biosynthes~s In L schottii ways and so it is possible that in this poppy [15, 161 it is firmly established case both routes operate side by side that the second unit is incorporated in [lo-1 21. the form 0f.a ketoacid. Looking to the A number of systems have now been future, the main interest in these early subjected to this type of investigation steps lies i.n the detailed mechanism of and the current situation is summarised such key reactions as the hydroxylation in Scheme 8. In .some instances it would of aromatic rings and the oxidative de- appear that the second building blodr is carboxylation of 1-carb~x~isoquino- incorporated as an, aldehyde [13, 141 lines. whereas in the case of the simpler cactus . Interest in isoquinoline alkaloids bio- alkaloids discussed earlier and the ben- synthesis does not stop at this point zylisoquinoline alkaloids of the opium however because the simple isoquinoli-

R CHO

(0) (0) \ R This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. (0)QNH(0) (0)

C02H R (a) R= 0 Glucose and possibly R= Ar , ArCH2CH2 and Me2CHCH2

Me02C

(b) R = H, Me, PhCH2 and possibly Me2CH CH2

Scheme 8 Alternative Routes to Tetrahydroisoquinoline Alkaloids Biosynthesis of lsoquinoline Alkaloids 7 nes ,we have considered so far can be coupling of phenolic rings. The general transformed into a wide range of dif- concept was first enunciated by early fer&t structures. This is particularly pioneers especially ROBINSON[17] but true of benzylisoquinolines and so the the great breakthrough in our under- rest of the lecture will concentrate on standing of what is involved came when this class of compound. the hypothesis was placed on a sound Some typical transformations are mechanistic basis by BARTON and indicated in Scheme 9 which illustrates COHEN[18]. They pointed out that if the great diversity of skeletal types a phenolate anion is oxidised to the cor- which can be produced. Those of iso- responding radical, the system would boldine and bear an obvious have the potential to combine with a structural relationship to their benzyl- second radical species at any one of isoquinoline precursor, but this rela- four sites: at the oxygen itself or at the tionship is less immediately obvious in ring carbons ortho or para to it. The the case of morphine, and in proto- possible out come of such reactions is stephanine it has been almost comple- shown in Scheme 10. First the phenolate tely obscured. Nevertheless the rela- (25) is oxidised (by loss of an electron tionship has now been firmly establised to a suitable acceptor) to the radical in every case as we shall see in the (26). Reaction of this at the oxygen ensuing account. leads to (27); reaction at ring carbons Before considering detailed case leads initially to the three dienones (28), histories it is desirable to introduce one (29) and (30). The former is capable of of the key processes, that of oxidative surviving as such; the remaining two

Me0 M eO

Me0 OH iso boldine scoulerine

(o)@(0) This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

(0)

NMe

Me0

morphine protostephanine

Scheme 9 Alkaloids Produced from Benzylisoquinoline Precursors 8 Staunton

(27) (I) (32)

Scheme 10 Reactions of Phenoxyradicals would not be expected to do so because known but we will not consider it in they have a hydrogen at the site of detail). The extra C-C bond results in coupling which- can-be readily lost to the fdrmation of a different and more regenerate new aromatic products (30) complicated skeletal structure and we and (32) respectively. Hence the out- will go on to consider how the synthetic come of these radical pairing reactions potential of this reaction is exploited in depends critically on whether or not a some representative biosynthetic path- substituent is present at the site of at- ways. tack. Isoboldine (37) for example has been This type of process is the key to the shown to be formed in papaver somni- biosynthesis of many different classes of ferum by a straightforward intramole- alkaloids. The process is vitally impor- cular oxidative coupling between the tant in the benzylisoquinoline family two aryl rings of reticuline (36) [19]. where there are two linked aryl rings The sites of coupling (marked by dots capable of forming phenoxy radicals. In in Scheme 12) both bear a hydrogen This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. this situation the attacking radical spe- and therefore the two rings can rearo- cies (equivalent to X in Scheme 10) may matise to give the biphenyl system cha- be derived from the second aryl ring as racteristic of the aporphine alkaloids, indicated in formal terms in Scheme 11. though as we shall see the biosynthetic We now have intramolecular oxidative routes leading to members of this group coupling and it can lead to alternative of alkaloids are not always so straight- products such as (34) and (35) (coupl- forward. ing between carbon and oxygen is also It has recently been established that Biosynthesis of lsoquinoline Alkaloids 9

Scheme 11 lntramo\ecular Coupling of Phenols

(36) (37) isoboldine (38) bold~ne

Scheme 12 Biosynthesls of Boldine the isomeric alkaloid boldine (8) is pro- sible precursors in order to discover duced in Litsea glutinosa from (37) by which compounds are the .true interme- demethylation followed by remethyla- diates. tion [20]. This functional group inter- A second example of straightforward change, which is of course incidental to phenol coupling takes place in the bio- the process which elaborates the carbon synthesis of the opium alkaloids the- skeleton, can cause considerable pro- baine, codeine, and morphine in Papa- blems for biosynthetic investigators; ver somniferum [21]. As can be seen in 13 whereas the pattern of hydroxyl and Scheme reticuline is again the sub- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. methoxy groups in the aryl rings of strate for the coupling step but this isoboldine provide a reliable clue to the time coupling takes place at different functionality of its benzylisoquinoline sites to give the dienone (38) in which precursor (the new C-C bond is formed only one of the two rings can rearo- ortho to one hydroxyl and para to the matise by enolisation. The carbon other) that of boldine is completely mis- framework of the morphinan alkaloids leading. In a case such as this it may be is now established and since the subse- necessary to test a wide range of pos- quent steps bring about no more than a 10 Staunton

4

Me0

(GO) (41

(42) Codeine (43) Morphine

Scheme 13 Biosynthesis. of Morphine

modification of the peripheral functio- this conrolling influence is brought nal groups they will not be considered home by the results of experiments in in detail [22, 231. which chemists attempt to carry out One point which should be stressed at equivalent oxidative coupling reactions this stage is the importance of the con-. in vitro: the usual result is a complex trol which must be exerted by the enzy- mixture in which the desired product is mes over the mode of oxidative coupl- at best a minor component. ing. Reticuline is the substrate in this In the two examples discussed so far This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. key process in both Schemes 12 and 13. the carbon skeleton of the final alkaloid Therefore the same diradical is presu- is determined by the oxidative coupling mably made to couple in two comple- and remains unchanged in susequent tely different ways. Presumably the steps. This is not always the case, howe- enzymes which catalyse the process in ver, because cyclohexadienones have a each case direct the outcome by forcing strong tendence to undergo rearrange- the benzylisoquinoline to fold in the ment. How this can take place is shown appropriate way. The importance of in Scheme 14. First in the acid-cataly- Biosynthesis of lsoquinoline Alkaloids 11 sed dienonphenol rearrangement the coupling of the benzylisoquinoline parent dienone (44) is converted to the orientaline (52) (an isomer of reticu- phenol (47) by migration of a substi- line) leads to a dienone (53). This is tuent to give a carbonium ion (46) which reduced to the dienol (54) which then can aromatise by loss of a proton. An undergoes a dienol-benzene rearrange- alternative less direct process starts with ment to form the aporphine isothebaine reduction to the dienol (48) which can (55). Note here that there is one less now undergo an equivalent rearrange- oxygen function in the final product ment to give carbonium ion (50) follo- than was present in the benzylisoqiuno- wed by aromatisation to the benzene line and so once again the pattern of derivative (51). Both types of process hydroxyl and methoxy groups is pro- are widely used in alkaloid biosynthe- foundly changed in the course of the sis. Note that the product of the first bio synthesis. aromatisation retains the oxygen of the Migration of a substituent is not the dienone whereas in the second the oxy- only secondary reaction which can lead gen is lost. This possibility has to be to modification of the carbon skeleton borne in mind in speculations concern- of a dienone produced by oxidative ing possible precursors in alkaloid bio- coupling. Thus in the biosynthesis of the synthetic studies [18]. erythrina alkaloid erythraline (60) in An interesting example of the dienol Erythrina crista-galli, the initial die- benzene rearrangement is to be found none (57), produced by oxidative coupl- in the biosynthesis of isothebaine (55) ing of the benzylisoquinoline (56), un- in Papaver orientale [24,2,5]. Oxidative dergoes cleavage as indicated rather

Dienone - Phenol Rearrangement

Dienol- Benzene Rearrangement This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

u2 Q2 - Q + u:qJ -u16

H OH H OH, 0

Scheme 14 Skeletal Rearrangements 12 Staunton

MeO@$?MeHO ::pMe + - Me0

M eO 4'

OH (54) (55) lsothebaine

Scheme 15 B~osynthes~sof lsotheba~ne

than migration of a C-C bond. The pro- ring can rearomatise by a simple enoli- duct (58) is then reduced to the dihy- sation. If there is a carbon substituent droderivative (59); we will need to re- at the site of coupling straightforward call these steps later when we consider aromatisation is not possible; the carbon the biosynthesis of the structurally rela- skeleton may then remain unchanged in ted alkaloid protostephanine. In this subsequent steps (morphine) or it may instance the nine-membered ring is sub- rearomatise either by migration of car- jected to further modification via a bon-carbon bond (isothebaine) or by a recyclisation as indicated leading ulti- fragmentation process (erythraline). mately to erythraline [26; 271. The latter two processes greatly extend At this point it is helpful to sum- the range of skeletal types which may This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. marise the processes we have met so far. be formed as a consequence of oxidative Firstly intramolecular oxidative coupl- coupling and this an important factor ing of a benzylisoquinoline leads to the in accounting for the remarkable diver- formation of a new carbon-carbon sity of structure in the benzylisoquino- bond at the ortho or para positions line alkaloid family. with respect to the free hydroxyl groups I would like now to turn to recent in the two aryl rings. When there is a work on the biosynthesis of protoste- hydrogen at the site of coupling the phanine [28, 291. As we shall see this Biosynthesis of lsoauinoline Alkaloids 13

(59) (60) Ery thraline

Scheme 16 Biosynthesis of Erythrina ' Alkaloids pathway is of interest not only because cencentrated on ~ossiblebenzylisoqui- it makes use of both of the key proces- noline precursors. As indicated in struc- ses but also because the.oxidative coupl- ture (61) it seemed likely on mechanistic ing process seems to follow an unusual grounds that the precursor would be course. Moreover, the elucidation of pentaoxygenated (one of the oxygens the biosynthesis proved to be far from being lost subsequently in a dienol-ben- straightforward and so it makes an zene rearrangement). It was assumed in interesting case study to trace in some accord with all the precedents that detail. there would be only one free hydroxyl The structure of protostephanine is group in each aryl ring the rest being shown in Scheme 17 which also shows protected as methoxy groups. These in outline the two most attractive hypo- arguments lead to three candidates (64), theses' for its biosynthesis. Route (a) (65) and (66). In the event all three which was proposed by BARTON [30] gave negative results in precursor in- involves oxidative coupling of a ben- corporation experiments. We were

zylisoquinoline intermediate (61) fol- therefore forced to abandon this intui- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. lowed by skeletal modification. The tive approach (which had proved grati- alternative, (b), is more straighforward fyingly successful in so many earlier in- in that it involves oxidative coupling of vestigations) in favour of patient detec- a so-called open-chain amine (62) and tive work in which the pathway was thus leads directly to the skeleton of the traced forward step by step from the alkaloid [3 11. only known precursor, tyrosine. Of the two hypotheses, we favoured The results of these endeavours are the former and so exploratory work indicated in Scheme 18. At the outset 14 Staunton

,

Me0 O OMe (0) (0) (63) Protostephanine (62)

NMe NMe NMe M eO OMe Me0 HO OMe OH OMe : OMe (64) (65) (66)

SCheme 17 Ideas and Experiments Relating to Pr otostephanine Biosynthesis

and + = '4~ This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Scheme 18 lncorporatlon of Phenethylarnine Precursors into- Protostephan~ne

tyrosine is converted to dopamine (69) intermediates gave a significant incor- via alternative intermediates tyrarnine poration which was moreover shown (67) or dopa (68); from then on a linear to be specific as indicated by degrada- . . pathway was traced via (70) as far as tion of the protostephanine. In contrast ; the phenethylamine (71). Each of these with these specific incorporations of Biosynthesis of lsoquinoline Alkaloids 15 phenethylamines, activity was incorpo- The resuls of this mammoth endea- rated from tyrosine into both "halves" vour were very rewarding. Firstly all of the molecule as indicated. the open-chain compounds gave nega- These results are consistent with the tive results. In contrast a number of formation of benzylisoquinoline inter- benzyl gave positive in- mediates in later stages of the biosyn- corporation results which allow us to thesis. Since both mono-methoxy deri- define several more steps on the path- vatives of (71) gave negative incorpora- way leading to protostephanine. As tions it seemed likely that the first iso- shown in Scheme 20 the phenethylamine quinoline derivative to be formed (71) is converted to the benzylisoquino- would have in its phenethylamine unit line (76) presumably by reaction with the same pattern of hydroxyl and the ketoacid (75) (this has not been methoxy groups as this intermediate. tested as a precursor yet). Both the se- Accordingly two classes of coupound condary and tertiary amines depicted represented by structures (72) and (73) by (76) were incorporated and this was were open to consideration. In each case also true of intermediates (77) and (78). the functionality of the aryl ring of the The timing of N-methylation is there- second building block could vary widely fore not clear and it is possible that two as indicated and a further source of interconnected pathways operate in variation lies the nature of the group parallel at these stages of the biosynthe- attached to the nitrogen atom. 1n dl sis. In contrast the order of the hy- sixteen benzylisoquinolines were can- droxylation and methylation by which didates; every one was synthesised in the functionality of the aryl ring is specifically labelled form and tested as modified seems to be precisely defined a precursor. We could not at that stage because the remaining compounds of rule out the possibility that Boit's sche- structure (72) were not incorporated to me might be correct in which case the a significant extent. next intermediates would be compounds The isoquinoline (78) is therefore the of general structure (74). The four se- latest intermediate to be positively condary amines corresponding to this identified. In view of the negative structure were therefore tested as inter- results in our early incorporation ex- mediates along with the benzylisoqui- periments with its two mono-0-methyl nolines. derivatives, (64), and (65), we suggest This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Scheme' 19 Possible Late Precursors of Protostephanine -M:iT (Me) - M::F,&Ie)- HO HO HO Me0 Me0 O OMe (77) (78) (63)

Scheme 20 Benzylisocpinoline Precursors of Protoslephantne that the next step of the ?biosynthesis substrate (that involved directly in the may be intramolecular oxidative coupl- coupling process); the remaining oxygen ing of this intermediate (see Scheme 21). functions have always been protected This suggestion will receive further as methoxy groups. One of the striking comment below but for the moment we features of the biosynthesis of morphine will concentrate on the subsequent hy- in Scheme 13 for instance is the way the pothetical steps by which the inter- two incidental oxygen functions of mediate dienone (80) might be conver- reticuline are protected as methoxyls ted to protostephanine. Note that a mi- prior to oxidative coupling, only to be gration converts (81) to (82) followed demethylated in later stages of the bio- by a fragmentation which is reminiscent synthesis. It is as though Nature in com- of that which occurs in erythraline bio- mon with organic chemists needs to pro- synthesis (Scheme 16). The subsequent tect these extraneous oxygen functions steps which are not specified would be prior to oxidative coupling so as 'to chemically trivial ones of reduction and avoid undesirable side reactions. Indeed methylation. this appears to be so general a feature This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. Hence this biosynthesis would com- of oxidative coupling processes so far bine both types of-skeletal modification investigated in vivo, it has become al- we have seen in earlier schemes. Its real- most a tenet of biosynthetic specula- ly intriguing feature is however the tions. However as was mentioned ear- nature of the proposed oxidative coupl- lier, the two monomethylated deriva- ing step. In every example considered tives of (78) were not incorporated and so far there has been only one free we have therefore to consider seriously hydroxyl group in each aryl ring of the the possibility that the benzylisoquino- Biosynthesis of lsoquinoline Alkaloids 17

.- .-

NMe Me0 Me0

Scheme 21 Hypothesis for the Late Stages of Protostephanlne Blosynthes~s line enters the phenol coupling step stephanine as far as the.benzylisoquino- with 'two free hydroxyl groups in one line (78). Again further methylation of of its aryl rings. That being so the oxi- :the catechol system does not seem to dative coupling may involve nucleo- take place prior,to oxidative coupling. philic addition to an orthoquinone in Clearly the subsequent steps (which are intermediate (79) rather than the usual not specified in Scheme 22) must involve coupling of a biradical. Obviously oxidative coupling between the marked further experiments will be needed to carbons followed by migration of a substantiate this proposal. One cannot C-N bond. Once again the overall rule out the possibility that a protect- transformation of (78) to (83) shows ing group other than methyl is em- several unusual features which warrant ployed in this case. Could the hydroxyl further investigation. be protected as a glycoside derivative Before leaving this topic it is worth

for example? Even so the biosynthesis drawing attention to the profound mo- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. would be exceptional and we are there- dification of the methylation pattern fore keen to investigate the process in the lower aryl ring of (78) which further. occurs in going to (83) : the two oxygens Finally it should be mentioned that which were free in the precursor be- the biosynthesis of hasubanonine (83) come methylated whereas the oxygen in Stephanica japonica shows the same which starts out as a methoxy group

unusual features 128, 281. The pathway ' becomes free. This biosynthesis there- is identical with that leading to proto- fore strikingly demonstrates the danger 18 Staunton - NMe Me0 O OH OH OMe

Scheme 22 B~osynthes~sof Hasubanon~ne

Me0HoxoH L OH

OMe OMe OMe

Scheme 23 Biosynthesis of Scoulerine

of assuming that the pattern of deri- to nitrogen. Experimental supPo& for vatisation in an alkaloid provides a the process is well documented [32, 331. reliable clue to that which obtains in its As was the case with the oxidative precursors. coupling of phenols the initial oxida- Though it provides a very flexible tion product is not necessarily the ter- and widely used means of modifying minus of the biosynthetic 'pathway. the carbon skeleton of benylisoquino- Further modification of the protober- line alkaloids, intramolecular oxidative berine structure may take place leading coupling is not the only device used in to the production of many new systems. Nature for this purpose. Before clos- Some important examples are given in

ing therefore we will briefly examine Scheme 24. The carbon skeletons of This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. an alternative strategy which is widely (86) and narcotine (87) are used and which leads eventually to. a produced by oxidative cleavage of C-N variety of modified structures. The basic bonds (a) and (b) respectively [34, 351. process shown in Scheme 23 leads to the That of chelidonine (88) arises by clea- formation of a protoberberine system. vage of bond (c) followed'by recycli- Again an oxidative cyclisation is used sation at an alternative site [34]. . but this time it takes place between an The biosynthetic pathways leading aryl ring and a methyl group attached to each of these alkaloids in Chelido- Biosynthesis of lsoquinoline Alkaloids 19

(86) Protopine (87) Narcotine (88) Chel~don~ne

Scheme 24 Biosynthet ic Transformations ot Scoulerlne nium majus have been worked out in References considerable detail but they will not be presented in detail here. Suffice it to say 1. Leete, E.: J. Amer. Chern. 'SOC., 88, 4218 that these and other modifications of (1966). the protoberberine skeleton greatly 2. Bartersby, A. R., R.Binks and R. Huxtable: extend the range of structural variants Tetrahedron Lett., 563 (1967). 3. Lundstrom, J. and S. Agurell: Tetrahedron to be found in the isoquinoline family. Lett., 4437 (1968). In conclusion, we can now claim to 4. Battersby, A.R., R.Binks and R. Huxtable: have a detailed knowledge of many of Tetrahedron Lett., 61 11 (1968). the more important biosynthetic path- 5. Khanna, K. L., H. Rosenberg and A. G. ways leading to isoquinoline alkaloids. Paul: J. C. S. Chem. Comm., 315 (1969). 6. Paul, A. G., K. L. Khanna, H. Rosenberg Without exception these pathways are and M. Takido: J. C. S. Chem. Cornrn., remarkable for the way in which a rela- 838 (1969). tively small range of building blocks 7. Lundstrom J. and S. Agurell: Tetrahedron are transformed in one way or another Lett., 3371 (1969). Lundstrom, J.: Acta. Chem. Scand., 25, This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. to produce many diverse structures. The 3489 (1971). pathways involved are often long and 8. Kapadia, G. J., G. Subba Rao, E. Leete, M. complex and require many different B.E. Fayez: enzymes.. Clearly alkaloids are not 9. Vaishnav, Y. N. and H. M. Fales: J. Amer. produced by accident: rather the plants Chem. Soc., 92, 6943, (1970). 10. O'Donovan, D. G. and H. Horan: J. Chem. go to great pains to make them. Why SOC. (C), 2791 (1968). they should do so is for the most part 11. Sdiiitte, H. R. and G. Feelig: Liebig's Ann. still a complete mystery. Chem., 730, 186 (1969). Staunton

12. O'Donovan, D. G. and E. Barry: J. C. S. 26. Barton, D. H. R., R. B. Boar and D. A. Perkin I, 2528 (1974). Widdowson: J. Chem Soc. (C), 1213 (1970). 13.. Battersby, A. R. and B. Gregory: J. C. S. 27. Barton, D. H.R., R. D. Bracho, C. J. Potter Chern. Cornm., 134 (1968). and D A. Widdowson: J. C. S. Perkin I, 14. Battersby, A. R., R. B. Herbert, E. McDo- 2278 (1974). nald, R. Ramage and (the late) J. H. Cle- 28. Battersby, A. R., R. C. F. Jones, R. Kaz- ments: J. C. S. Perkin I, 1741 (1972). lauskas, C. Poupat, C. W. Thornber, S. 15. Battersby, A. R., R. C. F. Jones and R. Ruchirawat and J. Staunton: J. C. S. Chem. Kazlauskas: Tetrahedron Lett., 1873 (1975). Cornrn. 773 (1974). 16. Wilson, M. L. and C. J. Coscia: J. Amer. 29. Battersby, A. R., A. Minta, A. P. Ottridge Chem. Soc., 97, 431 (1975). and J. Staunton: Tetrahedron Letters, 1321 17. Robinson, R.: Structural Relations of Na- (1977). tural Products, Clarendon Press, Oxford, 30. Barton, D. H.R.: Pure and Applied Chem., (1955). 9, 35 (1964). 18. Barton, D. H. R. and T. Cohen: Festschr. 31. Boit, H. G.: Ergebnisse der Alkaloid-Che- A. Stolle, 117 (1957). mie bis 1960, Akademie-Verlag, Berlin, 402 19. Brockmann-Hanssen, J., C.-C. Fu and L. (i96i). Y. Misconi: J. Pharm. Sci., 60, 1880 ('1971). 32. Barton, D. H. R., R. H. Hesse and G. W. 20. Tewari, S., D. S. Bhakuni and R. S. Kapil: Kirby: J. Chem. Soc., 6379 (1965). J. C. S. Perkin I, 706 (1977). 33. Battersby, A. R., R. J. Francis, M. Hirst, 21. Barton, D. H. R., G. W. Kirby, W. Steg- E. A. Ruveda and J. Staunton: J. C. S. Per- lich, G. M. Thomas, A. R. Battersby, T. A. kin I, 1140 (1975). Dobson and H. Ramuz: J. Chem. Soc., 34. Battersby, A. R., J. Staunton, H. R. Wilts- 2423 (1965). hire, R. J. Francis and R. Southgate: J. C. 22. Blaschke, G., H. I. Parker and H. Rapo- S. Perkin I, 1147 (1975). port: J. Amer. Chem. Soc., 89, 1540 (1967). 35. Battersby, A. R., M. Hirst, D. J. McCaldin, 23. Battersby, A. R., J. A. Martin and E.Brock- R. Southgate and J. Staunton: J. Chem. mann-Hanssen: J: Chem. Soc. (C), 1785 SOC.(C), 2163 (1968). (1967). 24. Battersby, A. R., R. T. Brown, J. H. ~le: Address: ments and G. Iverach: C. S. Chem.' Prof. Dr. Staunton G. J. . . 1. Cornrn., 230 (1965). University Chemical Laboratory 25. Battersby, A. R., T. J. Brocksom and R. . . Lensfield Road Ramage: J. C. S. Chem. Cornm., 464 (1969). Cambridge - U.K. This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.