Novel A,B,E-ring modified camptothecins displaying high lipophilicity and markedly improved human blood stabilities.

J. Med. Chem. 1999, 42, 3018-3022 Novel A,B,E-Ring-Modified Camptothecins Displaying High Lipophilicity and Markedly Improved Human Blood Stabilities David Bom, † Dennis P. Curran, † Ashok J. Chavan, ‡ Stefan Kruszewski, § Stephen G. Zimmer, |,∇ Kimberly A. Fraley, § and Thomas G. Burke* ,‡,§,∇ Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, Tigen Pharmaceuticals, Inc., Lexington, Kentucky 40506, and Division of Pharmaceutics Sciences, College of Pharmacy, Department of Microbiology and Immunology, College of Medicine, and Markey Cancer Center, University of Kentucky, Lexington, Kentucky 40506 Received May 6, 1999 Camptothecins are DNA topoisomerase I inhibitors that have recently emerged as a prominent class of anticancer agent. 1-3 Topotecan and CPT-11 are water- soluble analogues of the natural product camptothecin and in 1996 were the first two members within the family to gain FDA approval (topotecan as second-line therapy for advanced epithelial ovarian cancer and CPT- 11 as first-line therapy for colon cancer). Other camp- tothecin congeners currently under clinical evaluation include Lurtotecan (GI147211), 9-aminocamptothecin, and DX-8951f, all of which display improved water solubility over camptothecin, and 9-nitrocamptothecin, which is a lipophilic analogue. Although widely used, camptothecins are known to undergo relatively rapid hydrolysis in the bloodstream resulting in a marked loss of therapeutic potential. It is the key R-hydroxy-δ- lactone pharmacophore within the clinically relevant camptothecins that undergoes facile acyl cleavage at physiological pH 4 to yield a biologically inactive 5-7 carboxylate form. In this report we describe the rational design and total synthesis of highly lipophilic A,B,E- ring-modified camptothecins that are the most blood- stable camptothecins displaying intrinsic anticancer activity yet to be identified. Our approach to the design of more blood-stable camptothecin-class topoisomerase I inhibitors was based upon three general considerations. First, structural modifications that eliminated the highly preferential binding of the carboxylate over the lactone form by human serum albumin (HSA) 8-12 were sought. Previous research efforts in our laboratories have shown that 9-aminocamptothecin and camptothecin display ex- tremely poor stabilities in human blood due to the high- affinity, noncovalent binding interactions of their car- boxylate forms with HSA. 11,12 For instance, frequency- domain lifetime fluorometry reveals that HSA preferen- tially binds camptothecin carboxylate with over a 100- * Correspondence to: Thomas G. Burke, Ph.D., University of Kentucky, ASTeCC Building, Lexington, KY 40506. Phone: (606) 257- 2300, ext. 255. Fax: (606) 257-2489. † University of Pittsburgh. ‡ Tigen Pharmaceuticals, Inc. § College of Pharmacy, University of Kentucky. | College of Medicine, University of Kentucky. ∇ Markey Cancer Center, University of Kentucky. fold higher affinity than camptothecin lactone; 11 this selective binding of carboxylate over lactone results in a shifting of the equilibrium in favor of the carboxylate. In this manner, camptothecin opens more rapidly and completely in the presence of HSA than in the absence of the protein. In a solution containing HSA and in human plasma, pH 7.4 and 37 °C, camptothecin and 9-aminocamptothecin open rapidly and essentially com- pletely, such that negligible 0.2% lactone levels remain at equilibrium. 11,12 Time-resolved fluorescence spectro- scopic investigations show that topotecan contains structural features which reduce binding of its carbox- ylate form to HSA; 13 as a result, topotecan displays improved stabilities in human blood and plasma relative to camptothecin and 9-aminocamptothecin. Second, our finding that lactone stabilization is achieved through lipid bilayer partitioning 10,14,15 led us to pursue the design of more lipophilic camptothecin analogues. We have shown previously that lipid bilayer vesicles, 14,15 as well as erythrocytes, 10 promote camp- tothecin drug stability by preferentially binding elec- troneutral lactone over negatively charged carboxylate. Thus, the design of more lipophilic camptothecins would promote the reversible partitioning of the new agents into the lipid bilayers of erythrocytes, thereby protecting the active lactone forms from hydrolysis. Initial concerns about the loss of antitopoisomerase I activity through enhancing compound lipophilicity were lessened by the work of Pommier et al., 16 who in 1990 noted that several more lipophilic camptothecin analogues such as 10,11- (methylenedioxy)camptothecin display superior intrinsic potencies against topoisomerase I. Last, recent studies have shown that expansion of the camptothecin E-ring to a seven-membered system (by insertion of a methylene spacer between the 20-OH functionality and the carboxyl moiety) enhances the solution stability of the agent while maintaining anti- cancer activity. 17,18 Whereas the 20-OH functionality in conventional camptothecins is thought to interact with the carbonyl oxygen and facilitate ring opening, inclu- sion of a methylene spacer decreases the interactions between the β-OH and carbonyl oxygen. This change diminishes hydrogen-bonding interactions between the two groups and is thought to result in a slower rate of lactone hydrolysis. Since our total synthetic approach allows for the E-ring to be readily modified, we included the expanded β-hydroxy-δ-lactone E-ring functionality in our drug design strategy. The new agents, which we refer to as homosilatecan derivatives, were prepared using the cascade radical annulation approach, as summarized in Scheme 1. Enol ether 2, an intermediate in the synthesis of standard E-ring camptothecin analogues, 19,20 was oxidatively cleaved to keto formate 3 by treatment with OsO 4 followed by Pb(OAc) 4 . Chain extension by Reformatsky reaction followed by treatment with TFA provided an expanded lactone, which was then treated with ICl followed by TMSI to generate the pyridone lactone 4. This was then N-propargylated with trimethylsilyl- and tert-butyldimethylsilyl-substituted propargyl bromide to provide 5a,b. In the key cascade radical annulation, 10.1021/jm9902279 CCC: $18.00 © 1999 American Chemical Society Published on Web 07/22/1999