1.1. PROSTAGLANDINS
It is highly likely that those not themselves involved in scientific research perceive the development of new knowledge within a given area of science as a linear process. The popular view is that the understanding of the specific details of any complex system depends on prior knowledge of the system as a whole. This knowledge is in turn believed to derive from the systematic stepwise study of the particular system in question. The piecemeal, almost haphazard, way in which the details of the existence and later the detailed exposition of the arachidonic acid cascade were put together is much more akin to the assembly of a very complex jigsaw puzzle. This particular puzzle includes the added complication of incorporating many pieces that did not in fact fit the picture that was finally revealed; the pieces that would in the end fit were also found at very different times.
The puzzle had its inception with the independent observation in the early 1930s by Kurzok and Lieb and later von Euler that seminal fluid contained a substance that caused the contraction of isolated guinea pig muscle strips. The latter named this putative compound prostaglandin in the belief that it originated in the prostate gland; the ubiquity of those substances was only uncovered several decades later. The discovery remained an isolated oddity until the mid-1960s, by which time methods for chromatographic separation of complex mixtures of polar compounds and spectroscopic methods for structure determination were sufficiently advanced for the characterization of humoral substances that occur at very low levels. The isolation and structural assignment of the first two natural prostaglandins, [PGE.sub.1] and [PGF.sub.2], were accomplished by Bergstrom and his colleagues at the Karolinska Institute. (The letter that follows PG probably initially referred to the order in which the compounds were isolated: E refers to 9-keto-11-hydroxy compounds and F refers to 9,11-diols; the subscripts refer to the number of double bonds.) The carbon atoms of the hypothetical, fully saturated, but otherwise unsubstituted carbon skeleton, prostanoic acid, are numbered sequentially starting with the carboxylic acid as 1, and then running around the ring and resuming along the other side chain.
The identification of these two prostaglandins in combination with their very high potency in isolated muscle preparations suggested that they might be the first of a large class of new hormonal agents. Extensive research in the laboratories of the pharmaceutical industry had successfully developed a large group of new steroid-based drugs from earlier similar leads in that class of hormones; this encouraged the belief that the prostaglandins provided an avenue that would lead to a broad new class of drugs. As in the case of the steroids, exploration of the pharmacology of the prostaglandins was initially constrained by the scarcity of supplies. The low levels at which the compounds were present, as well as their limited stability, forced the pace toward developing synthetic methods for those compounds. The anticipated need for analogues served as an additional incentive for elaborating routes for their synthesis.
Further work on the isolation of related compounds from mammalian sources, which spanned several decades, led to the identification of a large group of structurally related substances. Investigations on their biosynthesis made it evident that all eventually arise from the oxidation of the endogenous substance, arachidonic acid. The individual products induce a variety of very potent biological responses, with inflammation predominating. Arachidonic acid, once freed from lipids by the enzyme phospholipase [A.sub.2], can enter one of two branches of the arachidonic acid cascade (Scheme 1.1). The first pathway to be identified starts with the addition of two molecules of oxygen by a reaction catalyzed by the enzyme cyclooxygenase to give [PGG.sub.2]. That enzyme, now known to occur in two and possibly three forms, is currently identified by the acronym COX; it is sometimes called prostaglandin synthetase. The reaction comprises the addition of one oxygen across the 9,11 positions to give a cyclic peroxide while the other adds to the 14 position in a reaction reminiscent of that of singlet oxygen to give a hydroperoxide at 14, with the resulting shift of the olefin to the 12 position and with concomitant isomerization to the trans configuration. The initial hydroperoxide is readily reduced to an alcohol to give the key intermediate [PGH.sub.2]. The reductive ring opening of the bridging oxide leads to the PGF series while an internal rearrangement leads to the very potent inflammatory thromboxanes. It was found later that aspirin and indeed virtually all nonsteroid anti-inflammatory drugs (NSAIDs) owe their efficacy to the inhibition of the cylcooxygenase enzymes.
The reaction of arachidonic acid with the enzyme lypoxygenase (LOX), on the other hand, leads to an attack at the 5 position and rearrangement of the double bonds to the 7,9-trans-11-cis array typical of leukotrienes; the initial product closes to an epoxide, thus yielding leukotriene [A.sub.4]. The reactive oxirane in that compound in turn reacts with endogenous glutathione to give leukotriene [C.sub.4]. This compound and some of its metabolites, it turned out, constitute the previously well-known "slow reacting substance of anaphylaxis" (srs-A), involved in allergic reactions and asthma.
Much of the early work on this class of compounds focused on developing routes for producing the agents in quantities sufficient for biological investigations. There was some attention paid to elaborating flexible routes as it was expected that there might be some demand for analogues not found in nature. This work was hindered by the relative dearth of methods for elaborating highly substituted five-membered rings that also allowed control of stereochemistry. The unexpected finding of a compound with the prostanoic acid skeleton in a soft coral, the sea whip plexura homomalla, offered an interim source of product. The group at Upjohn, in fact, developed a scheme for converting that compound to the prostagland, which they were investigating in detail. The subsequent development of practical total syntheses in combination with ecological considerations led to the eventual replacement of that marine starting material.
The methodology developed by E. J. Corey and his associates at Harvard provides the most widely used starting material for prostaglandin syntheses. This key intermediate, dubbed the "Corey lactone," depends on rigid bicyclic precursors for controlling stereochemistry at each of the four functionalized positions of the cyclopentane ring. Alkylation of the anion from cyclopentadiene with chloromethylmethyl ether under conditions designed to avoid isomerization to the thermodynamically more stable isomer gives the diene (3-1). In one approach, this is then allowed to react with a-chloroacrylonitrile to give the Diels-Alder adduct (3-2) as a mixture of isomers. Treatment with an aqueous base affords the bicyclic ketone (3-3), possibly by way of the cyanohydrin derived from the displacement of halogen by hydroxide. Bayer-Villiger oxidation of the carbonyl group with peracid gives the lactone (3-4); the net outcome of this reaction establishes the cis relationship of the hydroxyl that will occupy the 11 position in the product and the side chain that will be at 9 in the final product. Simple saponification then gives hydroxyacid (3-5). The presence of the carboxyl group provides the means by which this can be resolved by conventional salt formation with chiral bases. Reaction of the last intermediate with base in the presence of iodine results in the formation of iodolactone; the reaction may be rationalized by positing the formation of a cyclic iodonium salt on the open face of the molecule; attack by the carboxylate anion will give the lactone with the observed stereochemistry. Acetylation of the hydroxyl gives (3-6); halogen is then removed by reduction with tributyltin hydride (3-7). The methyl ether on the substituent at the future 11 position is then removed by treatment with boron tribromide. Oxidation of the primary hydroxyl by means of the chromium trioxide : pyridine complex (Collins reagent) gives Corey lactone (3-9) as its acetate.
A somewhat more direct route to the Corey lactone, developed later, depends on a radical photoaddition/rearrangement reaction as the key step. The scheme starts with the Diels-Alder addition of a-acetoxyacrylonitrile to furan to give the bridged furan (4-1) as a mixture of isomers. Hydrolysis by means of aqueous hydroxide gives the ketone (4-2); this reaction may also proceed through the intermediate cyanohydrin. This cyanohydrin is in fact produced directly by treatment of the mixture of isomers with sodium methoxide in a scheme for producing the ketone in chiral form. The crude intermediate is treated with brucine. Acid hydrolysis of the solid "complex" that separates affords quite pure dextrorotary ketone (4-2); this complex may consist of a ternary imminium salt formed by a sequential reaction with the cyanohydrin function. Irradiation of the ketone in the presence of phenylselenylmalonate leads to the rearranged product (4-5) in quite good yield. The structure can be rationalized by postulating the homolytic cleavage of the C-Se bond in the malonate to give intermediate (4-3) as the first step; the resulting malonate radical would then add to the olefin. Acyl migration would then give the rearranged carbon skeleton of (4-4). Addition of the phenylselenyl radical to that intermediate will then give the observed product. Reduction of the carbonyl group by means of sodium borohydride gives the product of approach of hydride from the more open exo face (4-6). Decarboxylation serves to remove the superfluous carboxyl group to afford (4-7); treatment with tertiary-butyldimethylsilyl chloride in the presence of imidazole gives the protected intermediate (4-8) that contains all the elements of the Corey lactone with the future aldehyde, however, in the wrong a configuration. Saponification of the ester followed by acid hydrolysis, in fact, gives the all cis version of the lactone. The desired trans isomer (4-9) can be obtained by oxidizing the selenide with hydrogen peroxide in the presence of sodium carbonate.
Biological investigations, once supplies of prostaglandins were available, revealed the manifold activities of this class of agents. The very potent effect of [PGF.sub.2[alpha]] on reproductive function was particularly notable. Ovulation in most mammalian species is marked by the formation on the ovary of a corpus luteum that produces high levels of progesterone if a fertile ovum has implanted in the uterus. Administration of even low doses of [PGF.sub.2[alpha]] was found to have a luteolytic effect, with loss of the implanted ovum due to the withdrawal of progestin. This prostaglandin was in fact one of the first compounds in this class to reach the clinic under the United States Adopted Name (USAN) name dinoprost. The development of drugs for use in domestic animals tends to be faster and much less expensive than those that are to be used in humans. This is particularly true if the animals are not used as food, since this dispenses with the need to study tissue residues. It is of interest, consequently, that one of the early prostaglandins that reached the market is fluprostenol (5-8). This compound differs from [PGF.sub.2[alpha]] in that the terminal carbon atoms in the lower side chain are replaced by the trifluromethylphenoxy group; this modification markedly enhances potency as well as stability. This drug is marketed under the name Equimate(r) for controlling fertility in racing mares, a species in which costs are probably of little consequence.
Reaction of the anion from phosphonate (5-1) with ethyl meta-triflurophenoxy-methylacetate results in acylation of the phosphonate by the displacement of ethoxide and the formation of (5-3). Condensation of the ylide from this intermediate with the biphenyl ester at position 11 of Corey lactone (5-4) leads to the enone (5-5) with the usual formation of a trans olefin expected for this reaction. The very bulky biphenyl ester comes into play in the next step. Reduction of the side chain ketone by means of zinc borohydride proceeds to give largely the 15a alcohol as a result of the presence of that bulky group. The ester is then removed by saponification, and the two hydroxyl groups are protected as their tetrahydropyranyl ethers (5-6). The next step in the sequence involves the conversion of the lactone to a lactol; the carbon chain is thus prepared for attachment of the remaining side chain while revealing potential hydroxyl at the 9 position. This transform is affected by treating (5-6) with diisobutylaluminum hydride at -78ºC; over-reduction to a diol occurs at higher temperatures. Wittig reactions can be made to yield cis olefins when carried out under carefully defined, "salt-free" conditions. Condensation of the lactol (5-7) with the ylide from 5-triphenylphosphoniumpentanoic acid under those conditions gives the desired olefin. Treatment with mild aqueous acid serves to remove the protecting groups, thus forming fluprostenol (5-8).
Prostaglandins have been called hormones of injury since their release is often associated with tissue insult. Most of these agents consequently exhibit activities characteristic of tissue damage. Many prostaglandins cause vasoconstriction and a consequent increase in blood pressure as well as the platelet aggregation that precedes blood clot formation. Thromboxane A2 is, in fact, one of the most potent known platelet aggregating substances. Prostacyclin, [PGI.sub.2], one of the last cyclooxygenase products to be discovered, constitutes an exception; the compound causes vasodilation and inhibits platelet aggregation. This agent may be viewed formally as the cyclic enol ether of a prostaglandin that bears a carbonyl group at the 6 position of the upper side chain. This very labile functionality contributes to the short half-life of [PGI.sub.2]. The fact that the lifetime of this compound is measured in single-digit minutes precludes the use of this agent as a vasodilator or as an inhibitor of platelet aggregation.
The analogue in which carbon replaces oxygen in the enol ring should of course avoid the stability problem. The synthesis of this compound initially follows a scheme similar to that pioneered by the Corey group. Thus, acylation of the ester (7-2) with the anion from trimethyl phosphonate yields the activated phosphonate (7-3). Reaction of the ylide from that intermediate with the lactone (7-4) leads to a compound (7-5) that incorporates the lower side chain of natural prostaglandins. This is then taken on to lactone (7-6) by sequential reduction by means of zinc borohydride, removal of the biphenyl ester by saponification, and protection of the hydroxyl groups as tetrahydropyranyl ethers.
The first step in building the carbocyclic ring consists, in effect, of a second acylation on trimethyl phosphonate. Thus, the addition of the anion from that reagent to the lactone carbonyl in (7-6) leads to the product as its cyclic hemiketal (8-1); this last, it should be noted, now incorporates an activated phosphonate group. Oxidation of that compound with Jones' reagent gives the diketone (8-2). The ylide prepared from that compound by means of potassium carbonate in aprotic media adds internally to the ring carbonyl group to give fused cylopentenone (8-3). Conjugate addition of a methyl group to the enone by means of the cuprate reagent from methyl lithium occurs predominantly on the open [beta] face of the molecule to afford (8-4). The counterpart of the upper side chain is then added to the molecule by condensation with the ylide from triphenylphosphoniumpentanoic acid bromide. The product (8-5) is obtained as a mixture of E and Z isomers about the new olefin due to the absence of directing groups. Removal of the tetrahydropyran protecting groups with mild aqueous acid completes the synthesis of ciprostene (8-6). This compound has the same platelet aggregation inhibitory activity as [PGI.sub.2], though with greatly reduced potency.
(Continues...)
Excerpted from Strategies for Organic Drug Synthesis and Design by Daniel Lednicer Copyright © 2009 by John Wiley & Sons, Inc.. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.