|Home | About | Journals | Submit | Contact Us | Français|
Manzamine A represents an important lead structure for the development of novel antimalarial chemotherapies. The synthesis and biological evaluation of a group of simplified analogues of manzamine A, which were designed to examine the roles of the A and D rings and of both the relative stereochemistry and the orientation of the β-carboline heterocycle on the antimalarial activity of manzamine A, are described.
Malaria is a disease that affects 300–500 million people each year and results annually in 1–2 million deaths, mostly among children. Structurally and functionally novel antimalarial agents with new mechanisms of action are needed as monotherapeutic agents or for use in combined chemotherapy with other presently available drugs. The urgent need for new and effective antimalarials escalates as Plasmodium falciparum and other human malaria parasite species have developed resistance to most commercially available antimalarial agents.1
The manzamine alkaloids represent important lead structures for the development of antiinfectives. Manzamine A and related structures represent highly potent, orally bio-available2 antimalarial substances that are more effective than most currently available therapeutics, i.e., chloroquine and artemisinin. In addition, some of the manzamine classes have also demonstrated activity against AIDS-opportunistic infectious diseases including tuberculosis and toxoplasmosis.
The low isolated yield of 1 that was originally reported from a natural source, the Okinawan sponge Haliclona sp.,3 prompted us to examine the SAR of manzamine A in an effort to develop simpler structural analogues with comparable antimalarial properties. We describe herein the synthesis and biological evaluation of a “minilibrary” of analogues of manzamine A that was designed to probe the roles of the A and D rings as well as the relative stereochemistry and β-carboline orientation on the antimalarial pharmacophore. The retention of the β-carboline moiety was deemed critical as the antimalarial activity of ircinol A (2, Figure 1) is ca. 400 times less than that of manzamine A.4
On the basis of the approach that was featured in our total synthesis of manzamine A,5 we reasoned that the readily available photosubstrate 13 could serve as a source of a family of analogues in which both the relative stereochemistry and the orientation of the β-carboline heterocyclic ring could be varied. The synthesis and irradiation of photosubstrate 13 are outlined in Scheme 1. The absolute stereo-chemistry of 13 was established via the alkylation of Myers pseudoephedrine glycinamide6 with allyl bromide to give 14. Hydrolysis of 14, followed by Weinreb amide formation, and LiAlH4 reduction afforded aldehyde 15, which was subjected without purification to Wittig olefination to produce 16 as the desired cis-alkene. Silyl ether deprotection, followed by tosylation, and NaH-mediated ring closure yielded azocine 17. Alloc removal in the presence of Pd(0) followed by alkylation of the resulting secondary amine with 3-butyn-2-one7 afforded photosubstrate 13. Irradiation of 13 through a Pyrex photoreaction vessel (8 mM in acetonitrile) yielded a diastereomeric mixture of intermediate aminals, which in the presence of silica gel afforded tricyclic ketones 11 and 12 in a 1:1.5 ratio, respectively.
The lack of diastereoselectivity in the photocycloaddition of 13 stands in marked contrast to the highly selective formation of the manzamine core that we have described.8 This difference can be understood as a function of the presence or absence of the tetrahydropyridine manzamine A ring in the photocycloaddition reaction. In the case of 18a, the orientation of the A ring is locked as shown in Figure 2. The alternative cyclohexene ring conformation 18c suffers from an internal A(1,2) strain interaction, which accounts for the ca. 3.2 kcal/mol difference in energy between 18a and 18c. The absence of the A ring in 13 significantly decreases the energy difference between the two possible reacting conformers 13a and 13c, and our calculations (MM2) indicate only a ca. 1.0 kcal/mol difference in energy between 13a and 13c, a result that is consistent with the observed 1:1.5 ratio of products.
The syntheses of BCE analogues 3–6 were completed as shown in Schemes 2 and and3.3. Enolization of 11 led to the formation of a mixture of enolates, which afforded an inseparable mixture of ketoesters (not shown). We were delighted to find that conversion of the ketoester mixture to the corresponding enol triflates (NaHMDS, Comins’ reagent) facilitated the separation of the diastereomeric products, which on reduction gave, separately, each of the unsaturated esters 7 and 8 in a ca. a 1:1 ratio. Weinreb aminolysis of each of the unsaturated esters with trimethylaluminum and tryptamine, followed by Bischler–Napieralski cyclization and DDQ-mediated oxidation, afforded the desired BCE analogues 3 and 4. Similar treatment of 12, as outlined in Scheme 3, gave the isomeric β-carboline-containing analogues 5 and 6.
These four novel structures facilitated an interrogation of the role of both the A and D rings and of the relative stereochemistry and β-carboline orientation on the antimalarial activity of 3–6. We note that 3, which embodies both the same relative stereochemistry (cis-exo orientation of the azocine ring relative to the BC ring system) and the same β-carboline orientation (C-10) as that of manzamine A, exhibits the most potent antimalarial activity of the new structures, all of which are significantly less active than manzamine A itself (see Table 1). These results underscore the importance of the A and D rings on the antimalarial activity of manzamine A. We observe a ca. 8-fold decrease in activity with 5 relative to 3, an indication of the significance of the manzamine relative stereochemistry for the antimalarial activity of 3. In contrast, changing the β-carboline orientation, i.e., 4 vs 3, leads to a more modest 2-fold decrease in activity, suggesting a greater role for the manzamine relative stereochemistry in the observed activity of these novel structures. It is more difficult to make direct comparisons with 6, which embodies neither the manzamine relative stereochemistry nor the orientation of the β-carboline found in the natural product.
These simplified analogues (3–6) of manzamine A provide valuable insight into the structural and stereochemical requirements for biological activity in this system. Further studies are currently underway in our laboratory, and our results will be reported in due course.
We are grateful to the NIH, GlaxoSmithKline, Wyeth, Merck, Amgen, Boehringer-Ingelheim, and Novartis for their generous support of this work.