BIOTROPIA No. 9, 1996: 1-14

NATURAL PRODUCTS IN ORGANIC SYNTHESIS

ROBERT H. BURNELL
Departement de chimie, Universite Laval, Quebec

Quebec, Canada G1K 7P4

ABSTRACT

A resume of four lectures presented by the author during a one month visit to BIOTROP in 1995. The

seminars were held at the University of Indonesia at Depok, the Technical Institute in Bandung, the Bogor

Agricultural Institute and at BIOTROP, Bogor. The talks were grouped under the general title "The Use of Natural

Products in Organic Synthesis".

Key words: Organic compounds/Natural products.

The venerable art of the chemical synthesis of naturally occurring organic compounds

which has inspired and fascinated chemists for roughly one hundred and fifty years

continues to be necessary today for several reasons. First, and traditionally the most

important reason, is that synthesis provides the ultimate proof of structure of a substance by

showing that it can be prepared in an unambiguous fashion from simple building blocks. In

principle the structures of all the hundreds of new compounds isolated from Nature every

year should be synthesized. However, if new substances are very similar in structure they can

be interrelated by conversion to one another or perhaps to products that have already been

synthesized.

A second reason for practicing this demanding art we call synthesis is the challenging

necessity of preparing increasingly complex substances in more efficient ways and this is the

most important source of innovation and invention that assures progress in the organic field.

Of course, synthesis can also provide an alternate to Nature as the source of useful

compounds. For instance, despite the effectiveness of the anti-cancer natural product taxol, it

is a sobering thought that all the Taxus trees in the world could not supply at present enough

of the drug to treat the women actually suffering from breast and cervical cancer. And to

make matters worse, the trees take almost a century to mature. Furthermore it has been

estimated that by the year 2000 there will be a world wide short fall of several tons of the

natural anti-malarial obtained from Artemesia species. The difficulty of shortages can

sometimes be alleviated by synthesizing simpler and cheaper analogs.

BIOTROPIA No. 9, 1996

(Burnell et al. 1985), maytenoquinone (Burnell et al. 1988) and coleon B (Burnell et al. 1983),

while occasionally we were able to correct the published structures, as in the case of the

nellionols (Burnell et al. 1984).

For structure confirmation, the simpler synthesis of the left and right handed mixture (the

racemic substance) is quite adequate because as mentioned, with the exception of the optical

property of rotating the plane of light, the left hand form, the right hand form or the 50:50

mixture of the two, all show the same spectroscopic behavior and these are the

characteristics we use to identify them (ultra-violet, infra-red, nuclear magnetic resonance and

mass spectra).

Strictly for proofs of structure, we have prepared molecules by total synthesis from

smaller symmetrical, easy to buy or make, starting materials (Burnell et al. 1993). The general

strategy is shown in the upper part of Scheme 6. To the rather commonplace aromatic

compound 30 we attached a long chain of carbons derived from natural geraniol 29. This

flexible substituent has two "double" bonds in critical positions enabling it to form two rings

when attacked by a suitable "positively charged" reagent (Ph-S
+
in our case). Both the

Ph-S and the CN groups in the product 32 can be exchanged by oxygen and this completed

a rapid synthesis of methyl nimbinone 33, a diterpenoid from the "Neem" tree, Azadirachta

indica (Ira et al. 1988 (1)).

The final product from the rapid synthesis was "racemic", a 50:50 mixture of the natural

substance and its mirror image and all the physical properties of our sample (except the

rotation of light) were identical with those in the literature. Obviously the proposed structure

is correct! To compare the two approaches we also prepared methyl nimbinone from natural

podocarpic acid (Burnell et al. 1993). This far more arduous route is summarized in the

lower part of Scheme 6. The key step in this sequence was the formation and subsequent

opening of the three-membered ring as a method of installing the two methyl groups on the

first ring of the skeleton. In this case when the synthesized methyl nimbinone 36 was

examined in the polarimeter, it did rotate the plane of polarized light in the same direction

and to exactly the same extent as the product isolated from Nature. This synthesis not only

confirms the general structure forwarded for the diterpene, it also tells us that the substance

elaborated by the Neem tree is precisely and exclusively as represented in formula 36. That is,

if the three rings of the skeleton are in the plane of the paper, the methyl group identified by

the dark wedge between the rings, protrudes towards the reader.

Choosing between the rather simple straight-forward synthesis that leads to a

mixture of the two image forms of methyl nimbinone or the more difficult and less rewarding

preparation of the real natural diterpene from natural podocarpic acid depends on the real goal

of the synthesis. But it would obviously be an enormous advantage to be able to combine the

simplicity of the former with the specificity of the latter. To investigate this possiblity we

selected another relatively simple compound, methyl nim-

Natural products in organic synthesis - Robert H. Bumell

bionone 44 to experiment with this "enantiomeric" style of synthesis. This diterpene is similar

in name and in structure to methyl nimbinone 36 and it was also extracted from the "Neem"

tree (Ira et al. 1988(2)). Crucial to this approach is the incorporation of an asymmetric unit

(technically a "chiral auxilliary") into the starting material at the outset. In theory, the "left or

right-handedness" of this appendage will just be borrowed temporarily during the synthesis

and it should impose its asymmetry on the reactions to induce the preferred formation of just

one of the mirror image forms of the final product. At a propitious moment this helpful but

perhaps costly auxilliary can be removed and recovered to be used in another synthesis.

BIOTROPIA No. 9,19 96

We chose to attach to the necessary aromatic starting product 43, the asymmetric unit 42

(our auxilliary) invented by Evans (Evans el al. 1985) which we had previously prepared from

the abundant natural amino acid, L-valine 41. As in the earlier synthesis, we added the ten

carbon bromide 29 and immediately we knew the auxilliary was performing its function.

While there are two three-dimensional structures possible for the product 45, only one of the

two was formed in quantity in the reaction !

Natural products in organic synthesis - Robert H. Burnell

Employing roughly the same steps as in the methyl nimbinone sequence (above), the

intermediate 45 was cyclized to compound 46 and the two oxygen functions were introduced

to afford a product which showed all the spectroscopic attributes of methyl nimbionone 47

(Dumont 1993). It was encouraging that examination of the optical properties of the

product in the polarimeter showed that the substance did indeed rotate the plane of light.

However, the magnitude of the rotation was much lower than that recorded for the natural

product which indicated that the high "left hand, right hand" selectivity we had achieved

early in the synthesis (only one form of 45 was isolated), had been scrambled or lost in the

later steps. We suspect that the conditions for the cyclization (45 to 46) are too severe and

our efforts are now aimed at finding milder alternate methods of performing this step. The

improved conditions we are searching for must provide the product in high yield, sufficient

to enable the completion of the synthesis which still involves several steps. Unfortunately, as

yet, we have not found this more gentle procedure.

ACKNOWLEDGEMENTS

The work described was financed partly by NSECC (the Natural Sciences and

Engineering Council of Canada), partly by the Government of Quebec (F.C.A.R.) and more

recently by generous personal funding (Emeritus Professor Charles R. Engel). My sincere

appreciation is extended to the following graduate students whose collaboration and talent

were crucial to the study : Sonia Desfosses, Nathalie Dumont, Nathalie The-berge, Michel Jean

and David Miller.

The author wishes to thank the staff at BIOTROP for their hospitality (in particular Dr.

Hilman Affandi and Dr. Gloria Enriquez) and the Canadian International Development

Agency (CIDA) for making the visit to Indonesia possible.

LITERATURE CITED

BARTON, D.H.R., A.G. BREWSTER, S.V. LEY, C.M. READ and M.N. ROSENFELD. 1981. Oxidation of phenols, pyrocatechols and

hydroquinones to orthoquinones using benzeneseleninic anhydride. J. Chem. Soc., Perkin. Trans. I., 1473-1476.

BREWSTER, J.H. 1959. A useful model of optical activity. J. Amer. Chem. Soc., 81: 5475-5483.

BURNELL, R.H., A. ANDERSEN, M. NERON-DESBENS and S. SAVARD. 1981. Approaches to the synthesis oflyco-xanthol and

similar natural products. Synthesis of coleon U. Can. J. Chem., 59: 2820- 2825.

BURNELL. R.H., M. JEAN and S. SAVARD. 1983. Synthesis of coleon B tetramethyl ether. Can. J. Chem., 61: 2461-2465.

BIOTROPIA No. 9,1996

BURNELL. R.H., M. JEAN, D. POIRIER and S. SAVARD. 1984. The structures of the nellionols. Synthesis of model
abieta-8,11,13- trien-7-ones. Synthesis of 5-dehydronellionol trimethyl ether. Can. J. Chera., 62:

2822-2829.

BURNELL, R.H., A. ANDERSEN, M. NERON and S. SAVARD. 1985. Synthesis of coleon C tri-O-methyl ether. Can. J. Chem., 63:

2769-2776.

BURNELL, R.H., M. JEAN and D. POIRIER. 1987. Synthesis of taxodione. Can. J. Chem., 65: 775-781.

BURNELL, R.H., M. JEAN and S. MARCEAU. 1988. Synthesis of maytenoquinone. Can, J. Chem., 66: 227-230.

BURNELL, R.H., N. DUMONT and N. THEBERGE. 1993. Synthesis of O- methyl-nimbinone. J. Nat. Prod., 56: 1930-1936.

CARRIERE, Y., J. MILLAR, J.N. MCNEIL, D.J. MILLER and E.W. UNDERBILL. 1988. Identification of the female sex pheromone in

alfalfa blotch leafminer, Agromyzafrontella (Rondani). J. Chem. Ecol., 14: 947-956.

DUMONT, N. 1993. Synthese de diterpenes naturels isoles de 1' Azadirachta indica (Neem). M.Sc. thesis, University

Laval, Canada.

EVANS. D.A., D.J. MATHRE and W.L. SCOTT. 1985. Asymmetric synthesis of the enkephalinase inhibitor Thiorphan. J.

Org. Chem., 50: 1830-1835.

FURIA, T.E. and N. BELLANCA. 1971. Fenaroli's Handbook of Flavor Ingredients. The Chemical Rubber Co., Cleveland,

Ohio.

IRA, B.S., S. SIDDIQUI, S. FAIZI and B.S. SIDDIQUI. 1988 (1). Tricyclic diterpenes from the stem bark of Azadirachta

indica. J. Nat. Prod., 51: 1054-1061.

IRA. B.S., S. SIDDIQUI, S. FAIZI and B.S. SIDDIQUI. 1988 (2). Teipenoids from the stem bark of Azadirachta indica.

Phytochemistry, 27 : 1801-1804.

KING, F.E., T.J. KING and J.G. TOPLISS. 1956. Synthesis of podocarpic acid. Chem. and Ind., 113.

KUPCHAN, S.M., A. KARIM and C. MARCKS. 1969. Tumor inhibitors. XVIII. Taxodione and taxodone, two novel

diterpenoid quinone methide tumor inhibitors from Taxodium distichium. ]. Org. Chem., 34: 3912.

MILLER, D., F. BILODEAU and R.H. BURNELL. 1991. Stereoselective synthesis of isomers of 3,7-dimenthyl-nonadecane, a

sex pheromone of the alfalfa blotch leafminer (Agromyza frontella (Rondani)). Can. J. Chem., 69: 1100-1106.

BIOTROPIA No. 9, 1996

Yet another motivation for synthesis stems from the fact that Nature very often

provides only very small quantities of the substances of the greatest physiological interest.

Modern analytical instruments enable the chemist to elucidate the structures of these

unknown, potentially useful biological, medicinal or phytochemical compounds even if the

elaborate extraction procedures afford only a few milligrams of material. However, for

proper assessment of their physiological activities and their mechanisms of action, much greater

amounts are probably needed and only synthesis could provide the necessary quantities.

Most of the interesting molecules of Nature are asymmetric rather like our hands. Each

hand is made up of the same elements such as fingers, thumbs, palms, nails and so on and while

they are very similar, they are not identical. A right-hand glove does not fit on the left hand ! In

fact our hands are mirror images of one another. Natural molecules made of the same atoms

could be assembled in two fashions, one being the image of the other. Unlike our hands,

however, Nature usually synthesizes only one of the two forms (chemically we call the two

images "enantiomers"). Since all "our human" metabolic intermediates and enzymes are also

asymmetric, the physiological activity of a natural substance may be quite different from that of

its mirror image. The "enantiomeric" form could just be inactive or less active than the natural

product but it could have harmful effects or be highly toxic. The much quoted case of

Thalidomide springs to mind where one enantiomer was a benign sedative recommended for

women suffering nausea during pregnancy whereas the mirror image of exactly the same

molecule was teratogenic producing babies with serious physical defects. So synthesis

must not only aim at preparing the target molecule with the right general structure but the

methodology should also lead to the right three-dimensional structure (or stereochemistry). In

simpler terms, the synthesis must give exclusively the "left" or exclusively the "right" hand

molecule !

Chemically producing this "left or right" hand configuration is extremely difficult. The

problem is further compounded by the fact that virtually all the physical properties of the two

image forms are identical which means we cannot easily tell them apart and, furthermore, we

cannot separate one from the other !! However, two characteristics of the enantiomers

(images) are different. First, as we have seen, they can vary greatly in their physiological

activities. Secondly, like all asymmetric substances, they rotate the plane of polarized light but

the images turn the light in opposite directions. Fortunately we can measure this in a

polarimeter and under the conditions defined by the test, a fresh solution of the more prevalent

form of the common sugar glucose rotates light through 112 degrees in a clockwise or

right-handed fashion while its image would also turn light 112 degrees anti-clockwise or to the

left. In fact we often say that a substance such as glucose is dextro-rotatory (from the Latin

dextrus : right) or we just abbreviate this to d-glucose where the small "d" shows that light

passing through a solution of the product

Natural products in organic synthesis - Robert H. Burnell

would be turned to the right. The image form would be levo-rotatory and called 1-glucose.

(Another convention describes rotation to the right as positive (+) and rotation to the left as

negative or (-)).

One way of assuring that the molecule targeted in a synthesis will have the right

stereochemistry (left or right-handedness) is to start the synthesis from an asymmetric

substance available in pure form in Nature. Many potentially useful, commercially available

compounds exist from simple amino acids or sugars to complex steroids, alkaloids and terpenes.

Our work on an insect pheromone is a simple example of such an "enantio-selective"

synthesis (as reaction sequences aimed exclusively at the right or left handed structures are

called).

The insect, the small fly Agromyza frontella was accidentally introduced into Canada

from Europe in the 1960's and its presence was rapidly detected because in its larval stage it

destroyed the leaves of the popular cattle fodder, alfalfa. In their breeding ritual, the female

insects attract the males by emanating a minute quantity of a chemical messenger or

pheromone and although the quantity of this substance available was infinitesimal,

structure 1 (or 3,7-dimethylnonadecane) was advanced for the pheromone. The structure

elucidation was made possible by a skillful study of the mass spectral fragmentation after

rigorous purification by preparative gas chromatography (Carriere et al. 1988).

Potentially this natural chemical could be used to control the population of the insect by

using it to divert the male flies away from the females and thus preventing reproduction !

1 3 7
CH3-CH2-CH-CH2-CH2-CH2-CH -- CH2 11 CH3

CH3 CH3

At first glance the substance is chemically uninteresting. It is just a member of the

stubbornly unreactive class of organic substances known as paraffins or saturated acyclic

hydrocarbons. On reflection, however, the pheromone is a synthetic challenge because the

positions of the two methyl groups (CHs or Me groups on the third and seventh carbons of

the long chain) make the molecule asymmetric so in fact there are four spatial arrangements

possible, as shown in structures 2 to 5. Only one will be the active pheromone but which

one ? Our synthetic strategy should enable us to prepare one (or better all) of them specifically.

BIOTROPIA No. 9, 1996

In designing a synthesis we normally work backwards from the target, hypothetically

disconnecting parts we feel we can graft back onto the remaining core using chemical

methods of proven specificity and geometry. Our first idea for the synthesis of the Agromyza

pheromone 1 arose from the "retro-synthesis" shown in Scheme 1. Disconnections leave a

central portion which contains both of the methyl branches on the long chain. Many naturally

occurring oils (commercially available at low price) contain compounds with this skeleton and

our final choice of a "starting material" would be motivated by the need for elements of

structure permitting the easy grafting of the other units. And, of most importance, the molecule

should be asymmetric with at least one "left" or "right" hand centre. After considering several

possibilities, pulegone 6, the main constituent of the "Oil of European Pennyroyal" from Mentha

pulegioides L. (Furia and Bellanca 1971) seemed ideal for the purpose.

Natural products in organic synthesis — Robert H. Burnell

Without commenting at length on the chemical reactions involved, we constructed the

pheromone as shown in Scheme 2. For those interested in how the transformations were

achieved, the reagents (indicated by the small letters on the arrows) are, of course, published in

detail (Miller et al. 1991). The steps from 6 to 11 not only add the "one carbon unit" but also

induce the stereochemistry (left or right-handedness) at C.3. None of these chemical

transformations could have altered the configuration of the methyl substituent of the original

pulegone so the geometry at the other methyl branch, now further in the chain at C.7, survives

unchanged from beginning to end.

Mechanistic and theoretical arguments told us that the final product from our

synthesis was actually structure 3, the drawing of which implies that if the long chain

zig-zags in the plane of the paper then both methyl groups represented by the dotted wedges,

stick out to the rear. Other considerations told us that of the four possible structures this was

actually the active pheromone. When analyzed by polarimetry and using rules formulated by

Brewster for calculating the optical properties of hydrocarbons

(Brewster 1959), we felt the synthetic oil also contained lesser quantities of structure 4 in
which the C.3 methyl has the other configuration.

BIOTROPIA No. 9, 1996

Natural products in organic synthesis - Robert H. Burnell

We had chosen pulegone as the starting material for the synthesis for another reason.

If the disconnetion suggesting possible synthetic approaches is envisaged in another fashion

as shown in Scheme 3, the pheromone could be prepared by just grafting an eleven carbon

fragment onto the pulegone 6.

So, to check both the structure and the purity of the end product from our first

synthesis, we prepared the pheromone a second time by this alternate route which as shown

in Scheme 4, starts from compound 8, an early intermediate in the previous synthesis (Miller et

al. 1991).

This second sequence gave exactly the same pheromone 3 but in higher yield.

Remarkably in this case and as intended, the original asymmetry of pulegone (the methyl

branch) is now at C.3 rather than at C.I as in the first synthesis. This "left hand : right hand"

selection was only possible because the natural pulegone is asymmetric and the Pennyroyal

plant synthesizes only one of the two possible mirror image forms. Interestingly, the image

form of pulegone can be isolated in pure form from other plant species so our synthesis could

also lead to the image of pheromone 3 represented by structure 2.

A second example of synthesis in our laboratory starting from commercially

available, naturally occurring asymmetric substances involves Podocarpic acid 23 which was

itself synthesized as early as 1956 (King et al. 1956). The world source of this acid is the resin

obtained from Podocarpus trees found here in Java and New Zealand. One of the molecules we

hoped to synthesize was Taxodione, an anti-tumoral substance shown to possess structure 28

(Kupchan et al. 1969) and at that time was considered a promising drug for human therapy.

When we started, other syntheses had already been reported by other groups but we felt that

we could improve on existing syntheses in elegance and efficiency. Podocarpic acid was a

natural choice as starting material because even uninformed inspection shows that it already

possesses many of the structural

BIOTROPIA No. 9,1996

elements of taxodione including the three angularly attached six-membered rings. But even

more important, the absolute stereochemistry (the handedness) of the centres identified as C.10

and C.5 is exactly the three dimensional geometry required for "natural" taxodione !

Scheme 5 is a brief outline of the crucial intermediates in our synthesis (Burnell et al.

1987). The white arrows merely draw attention to the part of the molecule that will be

modified in the next transformation. The key step was the use of a selenium reagent (Barton et

al. 1981) to oxidise the phenol 25 to the ortho-quinone 26. Some of the reactions were most

gratifying. For instance, while conventional wisdom would say such a change was unlikely,

compound 26 rearranged spontaneously to 27 and the latter was oxidized to the final taxodione

28 by merely percolating its solution through a column of silica gel. The product was, of course,

pure natural taxodione containing none of the

Natural products in organic synthesis - Robert H. Burnell

mirror image form and this was confirmed by measuring the direction and magnitude of its

ability to rotate the plane of polarized light. If we had started the synthesis from simple

symmetrical chemicals, the final substance would only be half taxodione and half its mirror

image and would show no rotation of the plane of light.

The taxodione preparation was part of a much broader study to corroborate the

structures of several naturally occurring aromatic diterpenes by synthesis. Since in all cases we

started from podocarpic acid (or the very similar equally affordable dehydro-abietic acid) the

final products also possessed the exact stereochemistry (left or right handedness) of the

natural substances. Most of the structures proposed in the literature were shown to be correct

as in the case of coleon U (Burnell et al. 1981), coleon C

1.pdf
10.pdf
11.pdf
12.pdf
13.pdf
14.pdf
2.pdf
3.pdf
4.pdf
5.pdf
6.pdf
7.pdf
8.pdf
9.pdf