Search

The natural alkaloid camptothecin is the lead compound of a new class of antitumor agents with a unique mechanism of action (i.e. inhibition of DNA topoisomerase I). The pharmacological interest of these agents has generated a large number of derivatives and analogues endowed with potent cytotoxic activity, two of them being in clinical use as antitumor drugs. We have synthesized a new series of camptothecins substituted in position 7 with an alkyl or alkenyl chain bearing cyano and/or carbethoxy groups. These compounds showed potent cytotoxic activity in vitro against the human non-small-cell lung carcinoma H460 cell line, most of them exhibiting IC(50) values in the 0.05-1 microM range, more active than topotecan used as a reference compound. In particular 7-cyano-20S-camptothecin (5a) showed high in vitro cytotoxicity against a topotecan-resistant H460 cell subline (H460/TPT) and a cisplatin-resistant ovarian carcinoma subline (IGROV-1/Pt 1). In an in vivo evaluation of the antitumor activity, 5a appeared significantly more effective than topotecan in the H460 tumor model and comparable with topotecan in a small-cell lung carcinoma model and a colon carcinoma model. The efficacy and good tolerability of this compound increase interest for further preclinical development.

We describe the rational design and synthesis of B- and A, B-ring-modified camptothecins. The key alpha-hydroxy-delta-lactone pharmacophore in 7-tert-butyldimethylsilyl-10-hydroxycamptothecin (DB-67, 14) displays superior stability in human blood when compared with clinically relevant camptothecin analogues. In human blood 14 displayed a t(1/2) of 130 min and a percent lactone at equilibrium value of 30%. The tert-butyldimethylsilyl group renders the new agent 25-times more lipophilic than camptothecin, and 14 is readily incorporated, as its active lactone form, into cellular and liposomal bilayers. In addition, the dual 7-alkylsilyl and 10-hydroxy substitution in 14 enhances drug stability in the presence of human serum albumin. Thus, the net lipophilicity and the altered human serum albumin interactions together function to promote the enhanced blood stability. In vitro cytotoxicity assays using multiple different cell lines derived from eight distinct tumor types indicate that 14 is of comparable potency to camptothecin and 10-hydroxycamptothecin, as well as the FDA-approved camptothecin analogues topotecan and CPT-11. In addition, cell-free cleavage assays reveal that 14 is highly active and forms more stable top1 cleavage complexes than camptothecin or SN-38. The impressive blood stability and cytotoxicity profiles for 14 strongly suggest that it is an excellent candidate for additional in vivo pharmacological and efficacy studies.

Several 5-substituted alkoxy 20(S)-camptothecin analogues having A- and B-ring substituents were prepared via semi-synthesis. Most of these compounds were found to exhibit potent anti-cancer activity based on their in vitro cytotoxicity data obtained against human tumor cell lines.

Homocamptothecins (hCPT) are modified camptothecins (CPT) with a seven-membered beta-hydroxylactone instead of the naturally occurring six-membered alpha-hydroxylactone. This E-ring modification fully conserves the ability to stabilize topo I-DNA single-strand breaks and stimulates high levels of DNA cleavage. A key feature is the irreversibility of E-ring opening, which should give reduced toxicity. Substituted hCPTs have been selected for their high antiproliferative activity on a panel of tumor cell lines, including those with cross resistance, and were found to be active at very low doses in a variety of human tumor xenografts when administered orally. BN 80915, a difluoro-hCPT, has entered clinical trials.

An overview of the cascade radical annulation approach to the camptothecin family of antitumor drugs is presented. This combinatorial synthetic approach involves two key steps: (1) N-propargylation of a lactone/pyridone D/E ring fragment and (2) cascade radical annulation of an A-ring isonitrile to form rings B and C. The synthesis is probably the most flexible and general route to the camptothecin class of molecules. The parallel synthesis of several libraries of silatecan and homosilatecan libraries is summarized. One of the first-generation silatecans, DB-67, is emerging as a serious candidate for cancer chemotherapy.

A series of C(7)-N-alkylaminoethyl-C(10), C(11)-methylenedioxy- and ethylenedioxy-camptothecin (3a-g, 4a-b) were prepared. Their syntheses and in vitro cytotoxicity were reported. Among 15 derivatives, 3a and 3b showed more potent cytotoxicity than Camptothecin, especially in CAOV-3 cell line.

BN 80927, a novel homocamptothecin derivative, inhibits both topoisomerase I and topoisomerase II mediated DNA relaxation and shows pronounced cytotoxicity against HT29, SKOV-3, DU145 and MCF7 human tumor cell lines.

Recent findings concerning the structure of the covalent binary complex formed by DNA topisomerase I and its DNA substrate, as well as the nature of interactions with inhibitors that bind reversibly to this binary complex, have led to two proposed models for the binding of the prototype inhibitor camptothecin to the DNA-topisomerase I binary complex. While these models differ in many regards, they both suggest the involvement of the 20-OH group of camptothecin in a donor hydrogen bond with an enzyme side chain functional group. Presently, five analogues of camptothecin that differ only at C-20 have been evaluated for their ability to bind to the topoisomerase I-DNA binary complex and thereby inhibit enzyme function. Both 20-chloro- and 20-bromocamptothecin bound as well to the enzyme-DNA binary complex as 20-aminoCPT despite the absence of a substituent at C-20 capable of contributing a donor hydrogen bond.

Homocamptothecin (hCPT), a camptothecin (CPT) analogue with a seven membered beta-hydroxylactone which combines enhanced plasma stability and potent topoisomerase I (Topo I)-mediated activity, is an attractive template for the elaboration of new anticancer agents. Like CPT, hCPT carries an asymmetric tertiary alcohol and displays stereoselective inhibition of Topo I. The preparation and biological screening of racemic hCPT analogues are described. The 10 hCPTs tested were better Topo I inhibitors than CPT. Fluorinated hCPTs 23c, d,f,g were found to have potent cytotoxic activity on A427 and PC-3 tumor cell lines. Their cytotoxicity remained high on the K562adr and MCF7mdr cell lines, which overexpress a functionally active P-glycoprotein. Fluorinated hCPTs were more efficacious in vivo than CPT on HT-29 xenografts. In this model, a tumor growth delay of 25 days was reached with hCPT 23g at a daily dose of 0.32 mg/kg, compared to 4 days with CPT at 0.625 mg/kg. Thus difluorinated hCPT 23g warrants further investigation as a novel Topo I inhibitor with high cytotoxicity toward tumor cells and promising in vivo efficacy.

The key intermediate of a 20(S)-camptothecin 1 synthesis was obtained in a highly enantioselective fashion using an enzyme-catalyzed resolution. A commercially available protease was found to exhibit the highest enantioselectivity with moderate activity, and (S)-ethyl 2-acetoxy-2-[6-(acetoxymethyl)-1,1-(ethylenedioxy)-5-oxo- 1,2,3,5-tetrahydroindolizin-7-yl]butanoate 7c of 98% e.e. was obtained as the remaining substrate.

A series of C(7) and C(20)-substituted camptothecin derivatives (12-14, 16-18) are prepared. Their syntheses and in vitro cytotoxicities are reported.

Nineteen ring A- and F-modified hexacyclic analogues of camptothecin were synthesized by Friedländer condensation of appropriately substituted bicyclic amino ketones with tricyclic ketone and were evaluated for cytotoxicity and topoisomerase I inhibitory activity. Seventeen of the compounds showed cytotoxic effects comparable or superior to those of 7-ethyl-10-hydroxycamptothecin (SN-38) against mouse leukemia P388 and human tumor cell lines HOC-21 and QG-56. Introduction of a compact and inductively electron-withdrawing substituent such as a hydroxy, methoxy, chloro, or fluoro group into position 5 of ring A of the hexacyclic compound remarkably increased the antitumor activity. The potency of topoisomerase I inhibition of these compounds showed good correlation with their cytotoxicity. Among them, the 4-methyl-5-fluoro hexacyclic compound was the most potent of all and was 10 times as active as SN-38 in in vitro antitumor activity.

Five (20S)-10-hydroxycamptothecin derivatives carrying the long-chain fatty acid esters were prepared for the development of a new class of prodrug-type agents. In vitro experiments using three kinds of purified carboxylesterase isozymes from the liver microsomes of rat, pig, and human demonstrated that these derivatives were efficiently metabolized by enzymes compared with CPT-11.

Eleven camptothecin esters, 6a-e and 7a-f, were prepared by straightforward acylation of camptothecins with the corresponding acylating reagents such as organic anhydrides and carboxylic acid chlorides. The in vitro pharmacokinetic determination of lactone levels of esters 6a and 7b showed that the biological life span of their lactone forms in human and mouse plasma significantly increased when compared with their mother compounds, camptothecin (3) and 9-nitrocamptothecin (4). The differences of lactone levels between human plasma and mouse plasma for 6a and 7b were much smaller than what was observed for their mother compounds. The in vivo antitumor activity and toxicity studies demonstrated that some of these esters were very active against human tumor xenografts in nude mice and had an exceptional lack of toxicity in nude mice, even at enormous doses.

The synthesis, characterization, inhibitory activity against topoisomerase I, and biological evaluation of a series of 14 camptothecin derivatives of polypyrrolecarboxamide (lexitropsin) conjugates of two structural classes: (A) camptothecin-NHCO-lexitropsin 44-51 and (B) camptothecin-CONH-lexitropsin 38-43 are described. All 16 compounds tested, 14 conjugates plus two functionalized camptothecin controls, inhibit topoisomerase I in the concentration range 1.12-16.6 microM that divide into three distinct categories based on activity. The most active enzyme inhibitors belong to structure class A with either cationic dimethylaminium or neutral amide end groups. Generally class B conjugates are less effective in inhibiting topoisomerase I. Cytotoxic potencies of the drugs was tested against four representative human tumor cell lines: SKOV3, SKLVB, HT29, and KB. All 16 drugs gave measurable IC50 values against the KB cell line and fell into two categories with IC50 values of 0.049-0.66 microM (largely structure class B) and 1.0-48 microM (largely class A). Thus the class B conjugates, while less potent against the enzyme, contain two of the most potent drugs, 38 and 39, against KB cell lines. In contrast, in the case of the cell lines SKOV3 and HT29 there was a general correlation between the better topoisomerase inhibitors and their cell cytotoxicities.

African trypanosomes (Trypanosoma brucei species) are parasitic protozoa that cause lethal diseases in humans and cattle. Previous studies showed that camptothecin, a potent and specific inhibitor of DNA topoisomerase I, is cytotoxic to African trypanosomes and related pathogenic hemoflagellates (Bodley AL and Shapiro TA, Proc Natl Acad Sci USA 92: 3726-3730, 1995). In this study, a series of camptothecin analogs was tested against axenically cultured, bloodstream form, T. brucei. Modifications to the pentacyclic nucleus of camptothecin ablated antiparasitic activity. In contrast, activity could be increased by substituents added to the parent ring system (e.g. 10,11-methylenedioxy or ethylenedioxy groups; alkyl additions to carbon 7; or 9-amino or 9-chloro substituents). Cytotoxicity was correlated with the level of cleavable complexes in trypanosomes, implicating topoisomerase I as the intracellular target for these compounds. To obtain some indication of selective toxicity, ten compounds were also tested against L1210 mouse leukemia cells. The 9-substituted-10,11-methylenedioxy analogs caused a disproportionate increase in antiparasitic activity, compared with mammalian cell toxicity. These findings provide a basis for designing further structural modifications and for selecting camptothecin analogs to test in animal models of trypanosomiasis.

A series of analogs based on a novel template, 11-aza-(20S)-camptothecin, were obtained from total synthesis and tested as potential anticancer drugs in the topoisomerase I enzyme cleavable complex assay. The parent compound 11-aza-(20S)-camptothecin (8) was derived from a Friedlander condensation between the known aminopyridine derivative 3-(3-amino-4-picolylidene)-p-toluidine and optically active tricyclic ketone 7. Compound 8 had activity approximately twice that of (20S)-camptothecin in the calf thymus topoisomerase I cleavable complex assay. Compounds were prepared wherein the 11-aza nitrogen atom was quaternized as either the corresponding N-oxide or methyl iodide. Compounds with quaternized N-11 showed improved water solubility and were equipotent to the clinically investigated camptothecin analog topotecan in the cleavable complex assay. These compounds were evaluated in vivo in nude mice bearing HT-29 human colon carcinoma xenografts. The analog 11-aza-(20S)-camptothecin 11-N-oxide was found to significantly retard tumor growth when compared to untreated controls. Finally, 7,10-disubstituted 11-azacamptothecin analogs were synthesized using Pd(0) coupling reactions of 10-bromo-7-alkyl-11-aza-(20S)-camptothecins 19 and 20, which in turn were available from a Friedlander condensation of the novel bromopyridine derivatives 17a and 17b with 7. Among the 10-substituted series, a number of analogs displayed extremely high in vitro potency against topoisomerase I and improved aqueous solubility. A significant number of the compounds were found to be active in whole cell cytotoxicity assays and several were evaluated in nude mice bearing the HT-29 tumor xenografts. The most effective of these proved to be (S)-11-aza-7-ethyl-10-(aminohydroximinomethyl)camptothecin trifluoracetic acid salt (27), a potent topoisomerase I inhibitor which demonstrated excellent efficacy in both short term and in extended in vivo assays. A comparison between in vitro enzyme data and in vivo data from nude mouse studies in other compounds in this series revealed a poor overall correlation between topoisomerase inhibition in vitro and antitumor efficacy in vivo.

The synthesis and antitumor activities of the novel water soluble camptothecin derivatives 7-[(4-methylpiperazino)methyl]-10,11-(methylenedioxy)-(20S)-campto thecin trifluoroacetate (6) and 7-[(4-methylpiperazino)methyl]-10,11-(ethylenedioxy)-(20S)-camptot hecin trifluoroacetate (7) are described. The solubilities of compounds 6 and 7 were measured to be 4.5 and 5.8 mg/mL, respectively, in pH 5 acetate buffer in contrast to < 0.003 mg/mL for camptothecin in the same buffer. In the purified topoisomerase I cleavable complex enzyme assay, compounds 6 and 7 demonstrated potent inhibition of topoisomerase I with IC50's of 300 and 416 nM, respectively, in comparison to 679 nM for camptothecin and 1028 nM for topotecan. In human tumor cell cytotoxicity assays, compounds 6 and 7 demonstrated potent antitumor activity against ovarian (SKOV3), ovarian with upregulated MDRp-glycoprotein (SKVLB), melanoma (LOX), breast (T47D), and colon (HT29) with IC50's ranging from 0.5 to 102 nM. Compounds 6 and 7 induced tumor regressions in the HT29 human colon tumor xenograft model and demonstrated similar rank order of potency compared to in vitro assay results.

Twenty-six novel A-ring-modified 7-ethylcamptothecins (6) were synthesized by Friedländer's condensation of the chiral tricyclic ketone (5) with aminopropiophenones (4). The compounds substituted with fluorine at the 11 position showed strong cytotoxicity to KB and L1210 cells. The 11-fluoro derivatives also exhibited strong inhibitory activity on DNA topoisomerase I. Nine compounds 6 with four to ten times stronger cytotoxicity than that of camptothecin were selected and converted into water-soluble 17-O-acyl amide derivatives (8). Compounds 8e (10-Me, O-COCH2CH2SCH3) and 8f (11-F, O-COC2H5) showed activity towards Meth A in mice that was comparable to that of CPT-11, at lower doses than CPT-11.

A large number of camptothecin (CPT) analogs have been prepared in the 20S, 20RS, and 20R configurations with a number of ring A substituents. Topoisomerase I (T-I) inhibition data (IC50) have been obtained by standard procedures. In general, substitution at the 9 or 10 positions with amino, halogeno, or hydroxyl groups in compounds with 20S configuration results in compounds with enhanced T-I inhibition. Compounds in the 20RS configuration were less active in vitro and in vivo and those in the 20R configuration were inactive. Compounds with 10,11-methylenedioxy substitution on ring A displayed a marked increase in potency in the T-I inhibition assay. The activities of some of the analogs as determined in a variety of in vivo assays including the L-1210 mouse leukemia assay were, in general, in accord with T-I inhibition. A number of water-soluble analogs such as 20-glycinate esters, 9-glycinamides, or hydrolyzed lactone salts were prepared and tested in in vitro and in vivo assays. In general, these compounds were less active than CPT both in terms of T-I inhibition and life prolongation in the L-1210 assay. However, certain 20-glycinate esters showed good in vivo activity after iv administration.

The structure of the N-amino pyridone (4a) obtained by the reaction of camptothecin (1a) with hydrazine was determined by X-ray crystallography. A mixture of 7-ethylcamptothecin (1b) and hydrazine hydrate was stirred at room temperature, and the hydrazide (2b) was isolated as its diacetate 2c. Treatment of the 17-O-acetyl amide (5a) with hydrazine gave 1b (74% yield) and the N-amino lactam 6 (11% yield). Compounds with bulky acyl groups, 5c--e, gave 6 in modest yields. The N-amino lactam 6 was smoothly dehydrated into the pyridone 4d by refluxing in hydrazine hydrate.

7-Ethylcamptothecin (1d), a model which does not have any site on the A-ring for further modification was converted into water-soluble derivatives by opening the E-ring lactone. 1d was heated in N,N-dimethylenediamine to yield amide 2a, and this was then acylated to furnish 3a-q, which were soluble in water as their HCl salts. The propionyl (3b), butyryl (3c) and methylthiopropionyl (3h) derivatives showed higher activity than the sodium salt of 1d. The acyl group makes the derivatives more lipophilic, and ease of hydrolysis of amide 2a to 1d is thought to be necessary for significant activity.

Camptothecin (1), a potent antitumor alkaloid, is known to inhibit topoisomerase I, an enzyme that relaxes supercoiled DNA. Modifications have been made to the B, D, and E rings of this natural product. Specifically, compounds 2-10 either have an ester moiety in place of the E ring lactone, a methyl ester attached to position 14, a saturated (or nonexistent) deaza B ring, or contain a combination of these permutations. We have conducted in vitro assays against the topoisomerase I relaxation reaction which verify the necessity for a lactone in the E ring. Furthermore, steric requirements at position 14 are shown to be crucial for activity, and planarity of the A and B rings of camptothecin is also implicated in the ability of the drug to inhibit topoisomerase I. Speculation on the nature of the drug binding pocket is presented.

Several E-ring-modified analogues of (RS)-camptothecin were synthesized by total synthesis via Friedländer condensation and evaluated for cytotoxicity and antitumor activity against P388 mouse leukemia cells. Among them, (RS)-20-deoxyamino-7-ethyl-10-methoxycamptothecin (25c) was found to be more active than (RS)-camptothecin (1) in the in vivo assay.

Water-soluble derivatives having the lactone ring intact were synthesized starting from 7-ethyl-10-hydroxycamptothecin (1). Glycosides (2) of the phenolic hydroxyl group of 1 were obtained by reaction with acetylated alpha-bromosugars in acetone or aqueous acetone in the presence of potassium carbonate, followed by deprotection. Phosphates (3) were prepared by reaction of 1 with phosphoryl chloride in pyridine or with dibenzylchlorophosphoridate. Sulfates (4) were obtained by reaction of 1 with sulfur trioxide-pyridine complex in the presence of a tertiary amine. The organic ammonium salts of monophosphate (3p) and sulfates (4a and 4b) showed significant activity against L1210 in vivo.

20(S)-Camptothecin derivatives having nitro, amino, chloro, bromo, hydroxyl and methoxyl groups in the A-ring were synthesized. B-Ring hydrogenated camptothecin (2a) was converted into 10-hydroxycamptothecin (6e) by treatment with lead tetraacetate in trifluoroacetic acid. 10-Substituted derivatives (6) were obtained by a photoreaction of N-oxides (9). The cytotoxicity of the A-ring modified camptothecins was evaluated against KB cells in vitro and leukemia L1210 in mice. 7-Ethyl-10-hydroxycamptothecin (6i) was identified as a potential derivative for further modification.

A radical substitution reaction of 20(S)-camptothecin (1) with methanol furnished 7-hydroxymethylcamptothecin (2). Reaction of 1 with primary alcohols higher than methanol gave 7-alkylcamptothecins (4), of which alkyl groups were one carbon less than the alcohols used and also 7-hydroxyalkylcamptothecins (5). For the preparation of 7-alkylcamptothecin (4), aldehydes were used as a radical source and several alkylated derivatives were synthesized. 7-Acyloxymethyl derivatives (6), 7-carbaldehyde (7), iminomethyl derivatives (10), acid (11), esters (12) and amides (13) were synthesized starting from 2. 7-Ethyl- (4b) and 7-propylcamptothecin (4c), acyloxymethyl compounds 6a, 6c and ethyl ester (12b) exhibited higher antitumor activity than 1 against L1210 in mice.

Novel 36 derivatives (6), bonding the phenolic hydroxyl group of 7-ethyl-10-hydroxycamptothecin (4) with diamines through a monocarbamate linkage, were synthesized and their antitumor activity was evaluated in vivo. The derivatives were soluble in water as their HCl salts with the E lactone ring intact and exhibited significant antitumor activity. One of the derivatives, 6-27 showed excellent activity against L1210 leukemia and other murine tumors. The structure of its hydrochloride trihydrate (CPT-11) was determined by spectroscopic and crystallographic methods.

Water-soluble analogues of the antitumor alkaloid camptothecin (1) were prepared in which aminoalkyl groups were introduced into ring A or B. Most of the analogues were prepared by oxidation of camptothecin to 10-hydroxycamptothecin (2) followed by a Mannich reaction to give N-substituted 9-(aminomethyl)-10-hydroxycamptothecins (4-12) or by subsequent modification of Mannich product 4 (13, 15, 17, 19, 21). Others were obtained by modification of the hydroxyl group of 2 (25,26) or by total synthesis (35,42,43). These analogues, as well as some of their synthetic precursors, were evaluated for inhibition of topoisomerase I, cytotoxicity, and antitumor activity. Although there was not a quantitative correlation between these assays, compounds that inhibited topoisomerase I were also cytotoxic and demonstrated antitumor activity in vivo. Further evaluation of the most active water-soluble analogue led to the selection of 9-[(dimethylamino)methyl]-10-hydroxycamptothecin (4, SK&F 104864) for development as an antitumor agent. In addition to its water solubility, ease of synthesis from natural camptothecin, and high potency, 4 demonstrated broad-spectrum activity in preclinical tumor models and is currently undergoing Phase I clinical trials in cancer patients.

The total synthesis of the pentacyclic camptothecin analogues 3 and 4 in 11 steps from p-tolualdehyde is described. The overall shape of compound 3 is the same as that of potent, naturally occurring camptothecin (1a). Despite the near spatial identity of 3 and 1b (racemic, (20RS)-camptothecin) from a three-dimensional standpoint, the 9KB and 9PS cytotoxicity assays indicate at least a 40-60-fold decrease in activity of 3 compared to that of 1b, and the isomer 4 was inactive. Similarly, studies of the inhibition of topoisomerase I activity indicated only slight activity for 3 and no activity for 4. It is evident that the pyridone ring D is essential for antitumor activity. Three E ring modified analogues of camptothecin, 2d-f, are described in which the net change is replacement of O by N in ring E. Compared to (20S)-camptothecin (1a) or (20RS)-camptothecin (1b), the ring E modified analogues 2d-f display little or no cytotoxic activity, greatly reduced effect on the inhibition of topoisomerase I, and total loss of life prolongation in the in vivo L-1210 mouse leukemia assay, indicative of the highly restricted structural and electronic requirements of ring E for biological activity in camptothecin.

Four analogues, 10-methoxy (20), 12-aza (29), benz[j] (36), and 18-methoxy (38), of camptothecin were obtained by total synthesis. The two water-soluble analogues, 10-[(carboxymethyl)oxy]- (24) and 10-[2'-(diethylamino)-ethoxy]-20(S)-camptothecin (26), with intact ring E were prepared from natural 10 hydroxycamptothecin (3). In general, there was a good correlation between in vitro 9KB cytotoxicity and activity in the P-388 leukemia system. While the aza analogue 29 was active in P-388 only at a much higher dose level than natural camptothecin (1), the 18-methoxy analogue 38 exhibited activity comparable to that of 1. The water-soluble derivative 24 was inactive. The amine hydrochloride 26 showed excellent activity at a high dose level. This could be due to its hydrolysis to 3. dl-Camptothecin (17) was roughly half as active as 1, indicating that the l isomer is inactive.

Several camptothecin derivatives containing a modified hydroxy lactone ring have been synthesized and evaluated for inhibition of topoisomerase I and cytotoxicity to mammalian cells. Each of the groups of the hydroxy lactone moiety, the carbonyl oxygen, the ring lactone oxygen, and the 20-hydroxy group, were shown to be critical for enzyme inhibition. For example the lactol, lactam, thiolactone, and 20-deoxy derivatives did not stabilize the covalent DNA-topoisomerase I complex. With a few exceptions, those compounds that did not inhibit topoisomerase I were not cytotoxic to mammalian cells. Two cytotoxic derivatives that did not inhibit topoisomerase I were shown to produce non-protein-associated DNA single-strand breaks and are likely to have a different mechanism of action. One of these compounds was tested for antitumor activity and was found to be inactive. The present findings, as well as other reports that the hydroxy lactone ring of camptothecin is critical for antitumor activity in vivo, correlate with the structure-activity relationships at the level of topoisomerase I and support the hypothesis that antitumor activity is related to inhibition of this target enzyme.

The camptothecin analogues (+-)-7-ethyl-10-methoxy-20-deoxyaminocamptothecin (2) and (+-)-7-ethyl-10-hydroxy-20-deoxyaminocamptothecin.HCl (3) were synthesized from indolizine compound 4 via Friedländer condensation to construct a pentacyclic ring system, and were tested in a P388 mouse antileukemia assay. Compounds 2 and 3 were more active and less toxic than (+)-camptothecin (1), and therefore had higher therapeutic ratios.

Several compounds having portions of the camptothecin ring system were prepared. These compounds were screened against L1210 lymphoid leukemia with negative results. Two of the analogs which contained the pyridine and hydroxylactone D and E rings were also screened for inhibition of DNA and RNA syntheses in HeLa cells. Each of these analogs had decreased activity as compared with comptothecin and there was no degradation of DNA in the HeLa cells. This suggest that the D and E rings are not a sufficient requirement for camptothecin-like activity.

The resolution of the tricyclic ketone (3a + 3b) by the separation of diastereomeric adducts 4a and 4c of the precursor ketal 5 is described. The regenerated enantiomers 3a and 3b of 100% optical purity represent the key intermediates from which 20(R)-camptothecin (1a) and 20(S)-camptothecin (1b), respectively, have been prepared. The 20R analogue 1a was 10-100 times less active than the natural 20(S)-camptothecin (1b) in 9KB and 9PS cytotoxicity assays and almost inactive in in vivo L-1210 leukemia tests as compared to the highly potent and active natural compound 1b.