The Organic Chemistry of Drug Synthesis, Volume 7

By Daniel Lednicer

John Wiley & Sons

Copyright © 2008 Daniel Lednicer
All right reserved.

ISBN: 978-0-470-10750-8

Chapter One


Carbocyclic or heterocyclic ring systems comprise the core of chemical structures of the vast majority of therapeutic agents. This finding results in the majority of drugs exerting their effect by their actions at receptor or receptor-like sites on cells, enzymes, or related entities. These interactions depend on the receiving site being presented with a molecule that has a well-defined shape, distribution of electron density, and array of ionic or ionizable sites, which complement features on the receptor. These requirements are readily met by the relatively rigid carbocyclic or heterocyclic molecules. A number of important drugs cannot, however, be assigned to one of those structural categories. Most of these agents act as false substrates for enzymes that handle peptides. The central structural feature of these compounds is an open-chain sequence that mimics a corresponding feature in the normal peptide. Although these drugs often contain carbocyclic or heterocyclic rings in their structures, these features are peripheral to their mode of action. Chapter 1 concludes with a few compounds that act by miscellany and mechanisms and whose structures do not fit other classifications.


A. Antiviral Protease Inhibitors

1. Human Immunodeficiency Virus. The recognition of acquired immune deficiency syndrome (AIDS) in the early 1980s and the subsequent explosion of what had seemed at first to be a relatively rare disease into a major worldwide epidemic, lent renewed emphasis to the study of virus-caused disease. Treatment of viral disease is made particularly difficult by the fact that the causative organism, the virion, does not in the exact meaning of the word, replicate. Instead, it captures the reproductive mechanism of infected cells and causes those to produce more virions. Antiviral therapy thus relies on seeking out processes that are vital for producing those new infective particles. The first drugs for treating human immunodeficiency virus (HIV) infection comprised heterocyclic bases that interfered with viral replication by interrupting the transcription of viral ribonucleic acid (RNA) into the deoxyribonucleic acid (DNA) required by the host cell for production of new virions. The relatively fast development of viral strains resistant to these compounds has proven to be a major drawback to the use of these reverse transcriptase inhibitors. The drugs do, however, still form an important constituent in the so-called cocktails used to treat AIDS patients. Some current reverse transcriptase inhibitors are described in Chapters 4 and 6. The intense focus on the HIV virus revealed yet another point at which the disease may be tackled. Like most viruses, HIV comprises a packet of genetic material, in this case RNA, encased in a protein coat. This protein coat provides not only protection from the environment, but also includes peptides that recognize features on host cells that cause the virion to bind to the cell and a few enzymes crucial for replication. Many normal physiological peptides are often elaborated as a part of a much larger protein. Specialized peptidase enzymes are required to cut out the relevant protein. This proved to be the case with the peptides required for forming the envelopes for newly generated virions. Compounds that inhibit the scission of the protein elaborated by the infected host, the HIV protease inhibitors, have provided a valuable set of drugs for treatment of infected patients. The synthesis of four of those drugs were outlined in Volume 6 of this series. Work on compounds in this class has continued apace as evidenced by the half dozen new protease inhibitors that have been granted nonproprietary names since then.

As noted in Volume 6, the development of these agents was greatly facilitated by a discovery in a seemingly unrelated area. Research aimed at development of renin inhibitors as potential antihypertensive agents had led to the discovery of compounds that blocked the action of this peptide cleaving enzyme. The amino acid sequence cleaved by renin was found to be fortuitously the same as that required to produce the HIV peptide coat. Structure-activity studies on renin inhibitors proved to be of great value for developing HIV protease inhibitors. Incorporation of an amino alcohol moiety proved crucial to inhibitory activity for many of these agents. This unit is closely related to the one found in the statine, an unusual amino acid that forms part of the pepstatin, a fermentation product that inhibits protease enzymes.

This moiety may be viewed as a carbon analogue of the transition state in peptide cleavage. The fragment is apparently close enough in structure to such an intermediate as to fit the cleavage site in peptidase enzymes. Once bound, this inactivates the enzyme as it lacks the scissile carbon-nitrogen bond. All five newer HIV protease inhibitors incorporate this structural unit.

One scheme for preparing a key intermediate for incorporating that fragment begins with the chloromethyl ketone (1) derived from phenylalanine, in which the amine is protected as a carbobenzyloxy (Cbz) group. Reduction of the carbonyl group by means of borohydride affords a mixture of aminoalcohols. The major syn isomer 2 is then isolated. Treatment of 2 with base leads to internal displacement of halogen and formation of the epoxide (3).

The corresponding analogue (4) in which the amine is protected as a tert-butyloxycarbonyl function (t-BOC) comprises the starting material for the HIV protease inhibitor amprenavir (12). Reaction of 4 with isobutyl amine leads to ring opening of the oxirane and formation of the aminoalcohol (5). The thus-formed secondary amine in the product is converted to the sulfonamide (6) by exposure to p-nitrobenzenesulfonyl chloride. The t-BOC protecting group is then removed by exposure to acid leading to the primary amine (10). In a convergent scheme, chiral 3-hydroxytetrahydrofuran (8) is allowed to react with bis(N-succinimidooxy)carbonate (7). The hydroxyl displaces one of the N-hydroxysuccinimide groups to afford the tetrahydrofuran (THF) derivative (9) equipped with a highly activated leaving group. Reaction of this intermediate with amine 10 leads to displacement of the remaining N-hydroxysuccinimide and incorporation of the tetrahydrofuryl moiety as a urethane (11). Reduction of the nitro group then affords the protease inhibitor (12).

Much the same sequence leads to a protease inhibitor that incorporates a somewhat more complex furyl function-linked oxygen heterocyclic. This fused bis(tetrahydrofuryl) alcohol (16) was designed to better interact with a pocket on the viral protease. The first step in preparing this intermediate consists of reaction of dihydrofuran (13) with propargyl alcohol and iodosuccinimide to afford the iodoether (14). Free radical displacement of the iodine catalyzed by cobaloxime leads to the fused perhydrofuranofuran (15). The exomethylene group in the product is then cleaved by means of ozone; reductive workup of the ozonide leads to racemic 16. The optically pure single entity (17) is then obtained by resolution of the initial mixture of isomers with immobilized lipase.

That product (17) is then converted to the activated N-hydoxysuccinimide derivative 18 as in the case of the monocyclic furan. Reaction with the primary amine 10 used to prepare amprenavir then leads to the urethane (19). Reduction of the nitro group then affords darunavir (20).

The synthesis of the amprenavir derivative, which is equipped with a solubilizing phosphate group, takes a slightly different course from that used for the prototype. The protected intermediate 5 used in the synthesis of 12 is allowed to react with benzyloxycarbonyl chloride to provide the doubly protected derivative 21, a compound that bears a t-BOC group on one nitrogen and a Cbz grouping on the other. Exposure to acid serves to remove the t-BOC group, affording the primary amine 22. This compound is then condensed with the activated intermediate 9 used in the preparation of the prototype to yield the urethane 23. Catalytic hydrogenation then removes the remaining protecting group to give the secondary amine 24. Reaction as before with p-nitrobenzenesulfonyl chloride gives the sulfonamide 25. This intermediate is allowed to react with phosphorus oxychloride under carefully controlled conditions. Treatment with aqueous acid followed by a second catalytic hydrogenation affords the water soluble protease inhibitor fosamprenavir (26).

The preceding three antiviral agents tend to differ form each other by only relatively small structural details. The next protease inhibitor includes some significant structural differences though it shares a similar central aminoalcohol sequence that is presumably responsible for its activity. Construction of one end of the molecule begins with protection of the carbonyl function in p-bromobenzaldehyde (27) as its methyl acetal (28) by treatment with methanol in the presence of acid. Reaction of that intermediate with the Grignard reagent from 4-bromopyridine leads to unusual displacement of bromine from the protected benzaldehyde and formation of the coupling product. Mild aqueous acid restores the aldehyde function to afford 29. This compound is then condensed with carbethoxy hydrazine to form the respective hydrazone; reduction of the imine function leads to the substituted hydrazine (30). Reaction of 30 with the by-now familiar amino-epoxide (4) results in oxirane opening by attack of the basic nitrogen in the hydrazine (30) and consequent formation of the addition product 31. The t-BOC protecting group is then removed by treatment with acid. The final step comprises acylation of the free primary amine in 32 with the acid chloride from the O-methyl urethane (33). This last compound (32) is a protected version of an unnatural [alpha]-aminoacid that can be viewed as valine in with an additional methyl group on what had been the side-chain secondary carbon atom. Thus, the protease inhibitor atazanavir (34) is obtained.

A terminal cyclic urea derivative of valine is present at one terminus in lopinavir (43). Preparation of this heterocyclic moiety begins with conversion of valine (35) to its phenoxycarbonyl derivative by reaction with the corresponding acid chloride. Alkylation of the amide nitrogen with 3-chloropropylamine in the presence of base under very carefully controlled pH results in displacement of the phenoxide group to give the urea intermediate (37). This compound then spontaneously undergoes internal displacement of chlorine to form the desired derivative (38).

The statine-like aminoalcohol function in this compound differs from previous examples by the presence of an additional pendant benzyl group; the supporting carbon chain is of necessity longer by one member. Condensation of that diamine (39), protected at one end as its N,N-dibenzyl derivative, with 2,6-dimethylphenoxyacetic acid (40) gives the corresponding amide (41). Hydrogenolysis then removes the benzyl protecting groups to afford primary amine 42. Condensation of that with intermediate 34 affords the HIV protease inhibitor 43.

2. Human Rhinovirus. Human rhinoviruses are one of the most frequent causes of that affliction that accompanies cooling weather, the common cold. This virus also consists of a small strand of RNA enveloped in a peptide coat. Expression of fresh virions in this case depends on provision of the proper peptide by the infected host cell. That in turn hinges on excision of that peptide from the larger initially produced protein. Protease inhibitors have thus been investigated as drugs for treating rhinovirus infections. The statine-based HIV drugs act by occupying the scission site of the protease enzyme and consequently preventing access by the HIV-related substrate. That binding is, however, reversible in the absence of the formation of a covalent bond between drug and enzyme. A different strategy was employed in the research that led to the rhinovirus protease inhibitor rupinavir (58). The molecule as a whole is again designed to fit the protease enzyme, as in the case of the anti-HIV compounds. In contrast to the latter compound, however, this agent incorporates a moiety that will form a covalent bond with the enzyme, in effect inactivating it with finality. The evocative term "suicide inhibitor" has sometimes been used for this approach since both the substrate and drug are destroyed.

The main part of the somewhat lengthy convergent synthesis consists of the construction of the fragment that will form the covalent bond with the enzyme. The unsaturated ester in this moiety was designed to act as a Michael acceptor for a thiol group on a cysteine residue known to be present at the active site. The preparation of that key fragment starts with the protected form of chiral 3-amino-4-hydroxybutyric acid (44); note that the oxazolidine protecting group simply comprises a cyclic hemiaminal of the aminoalcohol with acetone. The first step involves incorporation of a chiral auxiliary to guide introduction of an additional carbon atom. The carboxylic acid is thus converted to the corresponding acid chloride and that reacted with the (S)-isomer of the by-now classic oxazolidinone (45) to give derivative 46. Alkylation of the enolate from 46 with allyl iodide gives the corresponding allyl derivative (47) as a single enantiomer. The double bond is then cleaved with ozone; reductive workup of the ozonide affords the aldehyde (48). Reductive amination of the carbonyl group with 2,6-dimethoxybenzylamine in the presence of cyanoborohydride proceeds to the corresponding amine 49. This last step in effect introduced a protected primary amino group at that position. The chiral auxiliary grouping is next removed by mild hydrolysis. The initially formed amino acid (50) then cyclizes to give the five-membered lactam (51). Treatment under stronger hydrolytic conditions subsequently serves to open the cyclic hemiaminal grouping to reveal the free aminoalcohol (52). Swern-type oxidation of the terminal hydroxyl group in this last intermediate affords an intermediate (53) that now incorporates the aldehyde group required for building the Michael acceptor function. Thus reaction of that compound with the ylide from ethyl 2-diethoxyphosphonoacetate adds two carbon atoms and yields the acrylic ester (54).

The remaining portion of the molecule is prepared by the condensation of N-carbobenzyloxyleucine with p-fluorophenylalanine to yield the protected dipeptide (55). Condensation of that intermediate with the Michael acceptor fragment (54) under standard peptide-forming conditions leads to the tripeptide-like compound (53). Reaction of 53 with dichlorodicyanoquinone (DDQ) leads to unmasking of the primary amino group at the end of the chain by oxidative loss of the DMB protecting group. Acylation of that function with isoxazole (55) finally affords the rhinovirus protease inhibitor rupinavir (58).


Excerpted from The Organic Chemistry of Drug Synthesis, Volume 7 by Daniel Lednicer Copyright © 2008 by Daniel Lednicer. 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.