WO2001083482A1 - Dna alkylating agent and activation thereof - Google Patents

Abstract

N2-derivatives of methyl 1,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate (CPyl) were synthesized and shown to have DNA alkylation activity and cytotoxic activity that is susceptible to catalysis by metal ions, including Zn2+. This activation promotes DNA minor groove adenine N3 alkylation in a manner analogous to that of CC-1065 and the duocarmycins, and represents a new means of in situ activation for this class of DNA alkylating agent.

Description

DNA ALKYLATING AGENT AND ACTIVATION THEREOF

Description Technical Field:

The invention relates to DNA alkylating agents and to processes for their activation and use in situ. More particularly, the invention relates to DNA alkylating agents having a methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido

[3,2-e]indol-4-one-7-carboxylate (CPyl) head group and to catalytic processes for their metal cation complexation, activation, and use as DNA alkylating agents and cytotoxic agents.

Background:

Several well established methods have been discovered or developed for selective activation of DNA binding agents which initiate reactivity toward DNA including reductive activation (mitomycins), oxidative activation (aflatoxin), disulfide or trisulfide cleavage (calicheamicin), photochemical activation (psoralen), and oxidant activation of metal complexes (bleomycin), e.g., see D. S. Johnson, et al., In Comprehensive Supramolecular Chemistry; Lehn, J. M., Ed., Pergamon: Oxford, 1996, Vol. 4, Chapter 3, pp 73-176; and, more generally, Molecular Aspects of Anticancer Drug-DNA Interactions; S. Neidle, et al., Eds.; CRC: Boca Raton, 1993 and 1994; Vol. 1 and 2. Many of these have been exploited in the development of therapeutics to impart selective activity against tumor cells or to provide a means to effect and control reactivity towards DNA for use as research tools (e.g., synthetic nucleases).

(+)-Duocarmycin SA (DSA, 1) is a remarkably potent antitumor antibiotic (IC₅₀ = 10 pM, L1210) disclosed in 1990 (lchimura, M.; et al. J. Antibiot. 1990, 43, 1037; lchimura, M.; et al. J. Antibiot. 1991, 44, 1045) that has been shown to selectively bind and alkylate DNA (Figure 1) (Boger, D. L; et al. J. Am. Chem. Soc. 1993, 115, 9025). (+)-Duocarmycin SA exhibits enhanced stability and corresponding biological potency compared to its predecessors, (+)-duocarmycin A (DA, 2) (Ohba, K.; et al. J. Antibiot. 1988, 41, 1515; Takahashi, I.; et al. J. Antibiot. 1988, 41, 1915; Yasuzawa, T.; et al. Chem. Pharm. Bull. 1988, 36, 3728; lchimura, M.; et al. J. Antibiot. 1988, 41, 1285; Ishii, S.; et al. J. Antibiot. 1989, 42, 1713) and (+)-CC-1065 (3) (Hanka, L. J.; et al. J. Antibiot. 1978, 31, 1211 ; Chidester, C. G.; et al. J. Am. Chem. Soc. 1981 , 103, 7629), and therefore has been the subject of extensive investigation. Studies have shown that CC-1065 and the duocarmycins tolerate and benefit from structural modifications to the alkylation subunit and that the resulting agents retain their ability to participate in the characteristic sequence selective DNA alkylation reaction (For mechanistic aspects see: Boger, D. L; Johnson, D. S. Angew. Chem., Int. Ed.

Engl. 1996, 35, 1439; For synthetic aspects see: Boger, D. L; et al. Chem. Rev. 1997, 97, 787). Such structural modifications and the definition of their effects have served to advance the understanding of the origin of sequence selectivity (For mechanistic aspects see: Boger, D. L; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439; Sun, D.; et al. Biochemistry 1993, 32, 4487 and references cited therein) and catalysis (Boger, D. L; Garbaccio, R. M. Ace. Chem. Res. 1999, 32, 1043; Boger, D. L; Garbaccio, R. M. Bioorg. Med. Chem. 1997, 5, 263; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 4977; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 4987; Warpehoski, M. A.; Hurley, L. H. Chem. Res. Toxicol. 1988, 1, 315) of the DNA alkylation reaction by 1-3.

The substitution of the fused pyrrole C ring of CPI, the alkylating subunit of CC-1065 (3), with the six-membered benzene ring in CBI was shown to increase relative stability (4*) and biological potency (4*) without affecting DNA alkylation selectivity (Boger, D. L; et al. J. Org. Chem., 1990, 55, 5823; Boger, D. L; et al. J. Am. Chem. Soc. 1989, 111, 6461 ). The preparation of methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate (CPyl) is disclosed herein. CPyl contains a similar net one carbon expansion of the C ring pyrrole found in the DSA alkylating subunit but with incorporation of a pyridine (Figure 2).

Although there are several well established methods available for selective activation of DNA binding agents including reductive activation (mitomycins), oxidative activation (aflatoxin), disulfide or trisulfide cleavage (calicheamicin), photochemical activation (psoralen), and oxidant activation of metal complexes (bleomycin), examples of tunable metal cation Lewis acid activation of a DNA alkylating agent are unknown.

Comparative trace metal analysis of cancerous and noncancerous human tissue have revealed significant distinctions (Mulay, I. L.; et al. J. Natl. Cancer Inst. 1971, 47, 1). Although no generalizations were possible across all tumor types, within a given tumor type these were significant and potentially exploitable differences. For example, Zn was found in breast carcinoma at levels 700% higher than normal breast cells of the same type white lung carcinoma exhibited a revised and even larger 10-fold difference. Thus, chemotherapeutic agents subject to Zn activation can exhibit an enhanced breast carcinoma attributable to this difference in Zn levels (Mulay, I. L.; et al. J. Natl. Cancer Inst. 1971 , 47, 1).

What was needed was a class of DNA alkylating agents employable as cytotoxic agents capable of tunable metal cation Lewis acid activation with Zn and other metals.

Summary:

A new means of in situ activation for a class of DNA alkylating agents is disclosed herein. Methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2-e]indol -4-one-7-carboxylate (CPyl, 17) contains a unique 8-ketoquinoline structure which provides a tunable means to effect activation via selective metal cation complexation (Figure 2). This activation promotes a DNA minor groove adenine N3 alkylation in a manner analogous to that of CC-1065 and the duocarmycins, upon which CPyl was based.

N ²- derivatives of methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2- e]indol-4-one-7-carboxylate (CPyl) are synthesized and characterized. The unique 8-ketoquinoline structure of CPyl is disclosed to provide a tunable means to affect activation via selective metal cation complexation. The synthetic approach was based on a modified Skraup quinoline synthesis followed by a 5- exo-trig aryl radical cyclization onto an unactivated alkene with subsequent TEMPO trap or 5-exo-trig aryl radical cyclization onto a vinyl chloride for synthesis of the immediate precursor. Closure of the activated cyclopropane, accomplished by an Ar-3' spirocyclization, provided the CPyl nucleus in 10 steps and excellent overall conversion (29%). The evaluation of the CPyl-based agents revealed an intrinsic stability comparable to that of CC-1065 and duocarmycin A, but that it is more reactive than duocarmycin SA and the CBI-based agents (3-4^'). A pH rate profile of the addition of nucleophiles to CPyl demonstrated that an acid-catalyzed reaction is observed below pH 4 and that an un-catalyzed reaction predominates above pH 4. The expected predictable activation of CPyl by metal cations toward nucleophilic addition was found to directly correspond to established stabilities of the metal complexes with the addition product (Cu²⁺ > Ni²⁺ > Zn²⁺ > Mn²⁺ > Mg²⁺), and provides the opportunity to selectively activate the agents upon addition of the appropriate Lewis acid. This tunable metal cation activation of CPyl constitutes the first example of a new approach to in situ activation of a DNA binding agent complementary to the well-recognized methods of reductive, oxidative, or photochemical activation. This activation promotes a DNA minor groove adenine N3 alkylation in a manner analogous to that of CC-1065 and the duocarmycins upon which CPyl is based, and represents a new means of in situ activation of a novel class of DNA alkylating agents. For the simple alkylation subunit, Λ/-BOC-CPyl, an increase in alkylation efficiency of 1000 x was observed in the presence of Zn²⁺ without altering the inherent DNA alkylation selectivity and this efficiency is within 10-fold of the natural products CC-1065 and duocarmycin SA themselves. Resolution and synthesis of a full set of natural product analogues and subsequent evaluation of their DNA alkylation properties revealed that the CPyl analogues retain identical DNA alkylation sequence selectivity and near identical DNA alkylation efficiencies compared to the natural products. Consistent with past studies and even with the deep-seated structural change in the alkylation subunit, the agents were found to exhibit potent cytotoxic activity that directly correlates with their inherent reactivity. One aspect of the invention is directed to a DNA alkylating agent represented by the following structure:

wherein R is a DNA minor groove binder. In a preferred embodiment, the DNA alkylating agent is represented by the following structure:

and R, the DNA minor groove binder, is represented by one of the following structures:

Preferred examples include DNA alkylating agents represented by the following structures:

Another embodiment of this aspect of the invention is directed to a DNA alkylating agent represented by the following structure:

In the above structure, R represents a DNA minor groove binder, as indicated above. Another aspect of the invention is directed to a DNA alkylating agent represented by the following structure:

In the above structure, R represents a DNA minor groove binder, as indicated above. Preferred embodiments of this aspect of the invention include DNA alkylating agents represented by the following structures :

In the above structures, R is a DNA minor groove binder represented by any of the following structures:

Preferred examples of this aspect of the invention include DNA alkylating agent represented by the following structures:

Another aspect of the invention is directed to a process for catalyzing a solvolysis of a cyclopropyl ring of an N ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate. The process employs the step of contacting the N ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate under aqueous conditions having a pH greater than 4 with a catalytic concentration of a metal ion sufficient to catalyze the solvolysis of the cyclopropyl ring of the N ²- derivative of methyl 1 ,2,9,9a-tetrahydrocyclopropa[ yrido[3,2-e]indol-4-one-7-carboxylate. The metal ion is selected from Cu²⁺, Ni²⁺, Zn²⁺, Cr²⁺, Fe²⁺, Mn²⁺, and Mg²⁺. A preferred metal ion is Zn²⁺. Another aspect of the invention is directed to a process for catalyzing the production of a DNA alkylation product. The process employs the step of contacting DNA with an N ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate under aqueous conditions having a pH greater than 4 in the presence of a catalytic concentration of metal ion sufficient to catalyze the alkylation of the DNA by the N ²- derivative of methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate for producing the DNA alkylation product. The metal ion is selected from Cu²⁺, Ni²⁺, Zn²\ Cr²⁺, Fe²⁺, Cr³*, Fe³⁺, Mn²⁺, and Mg²⁺. A preferred metal ion is Zn²⁺.

Another aspect of the invention is a DNA alkylation product produced according to the above method.

Another aspect of the invention is directed to a process for catalyzing cell death by DNA alkylation. The process employs the step of contacting a cell, under aqueous conditions having a pH greater than 4, with a concentration of an N ²- derivative of methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one- 7-carboxylate sufficient, in the presence of a catalytic concentration of metal ion, to catalyze cell death by DNA alkylation. The metal ion is selected from Cu²⁺, Ni²⁺, Zn²⁺, Cr²"^", Fe²⁺, Cr³⁺, Fe³⁺, Mn²\ and Mg²⁺. A preferred metal ion is Zn²⁺.

Brief Description of Figures:

Figure 1 illustrates the structures of naturally occurring DNA alkylating agents, viz.: (+)-duocarmycin SA (1) , (+)-duocarmycin A (2), and of (+)-CC-1065 (3) .

Figure 2 illustrates the structural relationships between CPyl and other DNA alkylating agents, viz.: CPl, CBl, CCBI, MCBl, and DSA. The mechanism of metal chelation and catalysis of solvolysis of the cyclopropyl ring of CPyl is also illustrated. Figure 3 illustrates the kinetics of solvolysis by UV spectra of Λ/-BOC-CPyl (16, top) and CPyl (17, bottom) in 50% CH₃OH-aqueous buffer (pH 2, 4:1 :20 (v:v:v) 1.0 M citric acid, 0.2 M NaH₂PO₄, and H₂O, respectively). The spectra were recorded at regular intervals, and only a few are shown for clarity. Top: (hours) 0, 1 , 2, 3, 4, 5, 7, 9, 12, 16, 41. Bottom: (hours) 0, 5, 10, 20, 26, 35, 47, 66, 93, 144, 283.

Figure 4 illustrates a comparison of the rates of solvolysis for Λ/-BOC-CPyl with that of Λ/-BOC-CPI, A/-BOC-DA, Λ/-BOC-DSA, Λ/-BOC-CBI, Λ/-BOC-MCBI, Λ/-BOC-CCBI, and Λ/-BOC-CI.

Figure 5 illustrates a plot of the log / _obs for solvolysis of /V-BOC-CPyl as a function of pH.

Figure 6 illustrates thermally induced strand cleavage of w794 DNA (SV40

DNA segment, 144 bp, nucleotide nos. 138-5238); DNA-agent incubation for 24 or 48 h, as indicated, at 37 "C, removal of unbound agent and 30 min of thermolysis (100 °C), followed by denaturing 8% PAGE and autoradiography; lanes 1-2, (+)-/V-BOC-DSA (1 x 10^~1 and 1 x 10^~2); lanes 3^, (-)-/V-BOC-DSA (1 x 10^_1 and 1 x 10^"2); lane 5, control DNA; lanes 6-9, Sanger G, C, A and T sequencing reactions; lanes 10-11 , (+)-Λ/-BOC-CPyl (1 x 10^-2 and 1 x 10^~3); lanes 12-13, (-)-Λ/-BOC-CPyl (1 x 10^~2 and 1 x 10^"3); lanes 14-15, (+)-V-BOC-CPyl (1 x 10^~2 and 1 x 10^~3); lanes 16-17, (-)-/V-BOC-CPyl (1 x 10^~2 and 1 x 10^"3)

Figure 7 illustrates thermally induced strand cleavage of w794 DNA (SV40

DNA segment, 144 bp, nucleotide nos. 138-5238); DNA-agent incubation for 24 h at 25 ^°C, removal of unbound agent and 30 min of thermolysis (100 °C), followed by denaturing 8% PAGE and autoradiography; lane 1 , control DNA; lanes 2-3, (+)-duocarmycin SA (1 , 1 x 10^"5 and 1 x 10^-6); lanes 4-6, (+)-CPyl-TMI (25, 1 x 10^-5 to 1 x 10^~7); lanes 7-10, Sanger G, C, A and T sequencing reactions; lanes 11-12, (+)-CC-1065 (3, 1 x 10^"° and 1 x 10^"6); lanes 13-15, (+)-CPyl-indole₂ (31 , 1 x 10^"5 to 1 x 10-⁷); lanes 16-18, (+)-CPyl-CDPI₁ (33, 1 x lO^ to 1 x 10^~7). Figure 8 illustrates thermally induced strand cleavage of w794 DNA (SV40 DNA segment, 144 bp, nucleotide nos. 138-5238); DNA-agent incubation for 72 h at 25 °C, removal of unbound agent and 30 min of thermolysis (100 °C), followed by denaturing 8% PAGE and autoradiography; lane 1 , control DNA; lanes 2-3, (-)-duocarmycin SA (1 , 1 x 10^"5 and 1 x 10^-6); lanes 4-5, (-)-CPyl-TMI (25, 1 x 10^-5 and 1 x 10^-6); lanes 6-9, Sanger G, C, A and T sequencing reactions; lane 10, (+)-CC-1065 (3, 1 x 10^-6); lanes 11-12, (-)-CPyl-indole₂ (31, 1 x 10-⁵ and 1 x 10^"6); lanes 13-14, B-CPyl-CDP^ (33, 1 x lO^""5 and 1 x 10^"6).

Figure 9 illustrates the relative toxicities of various N ²- derivatives of CPyl as compared to CPl, CBl, CCBI, MCBl, and DSA as a function of the particular

D DNNAA mmiinor groove binder employed as the Λ/ ²- derivative, e.g., BOC, TMI, and

CDPI.,.

Figure 10 illustrates the aqueous solvolysis of Λ/-BOC-CPyl and CPyl (pH

2, phosphate buffer).

Figure 11 illustrates the activation of CPyl by metal cations toward nucleophilic addition and the relative reaction rates of such metal cations, viz., Cu²⁺ > Ni²⁺ > Zn²⁺ > Cr³⁺ > Fe³⁺> Mn²⁺ > Mg²⁺.

Figure 12 illustrates the enhanced efficiency of the DNA alkylation reaction of CPyl with w794 DNA as a function of the addition of various metal cations, viz.: Cu²⁺ (100x), Ni²⁺ (100-1000X), and Zn²⁺ (1000x).

Figure 13 illustrates a synthetic scheme for CPyl.

Figure 14 illustrates a synthetic scheme for advanced intermediate 14.

Figure 15 illustrates a synthetic scheme for the synthesis of N ²- derivatives of CPyl using various DNA minor groove binders. Figure 16 illustrates an acid-catalyzed nucleophilic addition of CH₃OH to 16.

Figure 17 illustrates a comparison of the treatment of Λ/-BOC-CPyl (16) with Zn(OTf)₂ in CH₃OH to provide a single product 41 with the inability of

/V-BOC-CBI (37) to similarly exhibit this metal-catalyzed reactivity in the presence of Zn(OTf)₂ .

Figure 18 illustrates solvolysis rates of CPyl and Λ/-BOC-CPyl with phosphate buffer.

Figure 19 illustrates solvolysis rates of Λ/-BOC-CPyl with universal buffer.

Figure 20 illustrates the in vitro cytotoxicity of various N ²- derivatives of CPyl.

Figure 21 illustrates thermally induced strand cleavage of w794 DNA (SV40 DNA segment, 144 bp, nucleotide nos. 138-5238); DNA-agent incubation for 24 h at 25 °C, removal of unbound agent and 30 min of thermolysis (100 ^°C), followed by denaturing 8% PAGE and autoradiography; lanes 1 , control DNA with Zn(acac)₂ (1 x 10° M); lanes 2-5, (+)-Λ/-BOC-CPyl (16, 1 x 10^~2 to 1 x lO^""5 M); lanes 6-7, (+)-duocarmycin SA (1 x 10^~5 and 1 x 10^-6 M); lanes 8-11 , Sanger G, C, A and T sequencing reactions; lane 12, (+)-Λ/-BOC-CPyl (16, 1 x 10^"3 M) with Zn(acac)₂ (1 equiv); lanes 13-15, (+)-Λ/-BOC-CPyl (16, 1 x 10^" M) with Zn(acac)₂ (1 , 10, and 100 equiv); lanes 16-17, (+)-Λ/-BOC-CPyl (16, 1 x 10r⁵ M) with Zn(acac)₂ (1 and 1000 equiv). Detailed Description:

A short and efficient 10 step (29% overall) synthesis of CPyl featuring a modified Skraup quinoline synthesis followed by a 5-exo-trig aryl radical cyclization onto a vinyl chloride is detailed and constitutes a net one carbon expansion of the C ring pyrrole found in the duocarmycin SA alkylation subunit. CPyl was found to be 3-4x less stable than CBl and duocarmycin SA but possesses a superior stability to CC-1065 and duocarmycin A. Nucleophilic addition occurred at the least substituted cyclopropane carbon with a regioselectivity (>20:1) comparable to that of CBl but which exceeds that of the natural products themselves (6-1.5:1). A pH rate profile of the addition of nucleophiles to CPyl demonstrated that it is an acid-catalyzed reaction below pH 4, but an uncatalyzed reaction above pH 4 consistent with the observation that the DNA alkylation reaction at physiological pH is not acid-catalyzed (Boger, D. L.; Garbaccio, R. M. >Acc. Chem. Res. 1999, 32, 1043; Boger, D. L.; Garbaccio, R. M. Bioorg. Med. Chem. 1997, 5, 263; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 4977; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 4987; Boger, D. L; Garbaccio, R. M. J. Org. Chem. 1999, 64, 5666; Boger, D. L.; Tumbull, P. J. Org. Chem. 1998, 63, 8004; Boger, D. L.; Turnbull, P. J. Org. Chem. 1997, 62, 5849; Boger, D. L.; et al. Bioorg. Med. Chem. Lett. 1997, 7, 233). The tunable activation of CPyl by metal cations toward nucleophilic addition, which directly follows established stabilities of the resulting metal complexes with the addition product (Cu²⁺ > Ni²⁺ > Zn²⁺ > Mn²⁺ > Mg²⁺), provides the opportunity to selectively and predictably initiate reactions of the agent simply by addition of the appropriate Lewis acid. This novel activation arises from chelation to the CPyl 8-ketoquinoline core, a unique structural feature that is not found in the natural products or alkylation subunit analogues disclosed to date. Resolution and incorporation of CPyl into a full set of duocarmycin and CC-1065 analogues allowed for examination of their cytotoxic and DNA alkylation properties. The CPyl analogues were potent cytotoxic agents exhibiting picomolar IC₅₀'s which correlated with their relative stability. In addition to smoothly following this correlation, the analogues displayed a smooth trend of increasing cytotoxic potency with the increasing length in the DNA binding subunit. Analogous to the natural products, the (S)-enantiomers possessing the absolute configuration of 1-3, proved to be more potent (3-30x) than the unnatural (R)-enantiomers. DNA alkylation studies revealed that the CPyl analogues exhibited an identical DNA alkylation sequence selectivity and near identical DNA alkylation efficiencies compared to the natural products. Thus, this set of analogues, which contain a unique structural modification in the alkylation subunit, retain full DNA alkylation and cytotoxic properties of the natural products, while possessing a novel capability for predictable and tunable activation by chelation of Lewis acids. Although this Lewis acid activation is of limited use for agents which already display effective DNA alkylation properties, its use is especially effective when applied to CPyl members which are poor at alkylating DNA. This includes Λ/-BOC-CPyl (16) for which the alkylation efficiency is increased 10^3χ and reversed analogs of CPyl in which the DNA binding subunits are attached through the C-terminus methyl ester rather than N-terminus secondary amine.

Synthesis of W-BOC-CPyl. The CPyl synthesis was based on a modified Skraup quinoline synthesis to provide the core structure, followed by the TEMPO trap of an aryl radical-alkene 5-exo-trig cyclization for introduction of the A-ring, and final Ar-3' spirocyclization (Figure 13). Thus, 2-bromoacrolein was treated with 4 in the presence of bromine (1.0 equiv, AcOH, 100 °C, 1 h, 92%) to provide 5 (Baker, R. H.; et al. J. Am. Chem. Soc. 1950, 72, 393; Tinsley, S. W. J. Am. Chem. Soc. 1955, 77, 4175). Protection of phenol 5 (1.2 equiv of BnBr, 1.1 equiv of NaH, DMF, 4 to 25 °C, 24 h, 85%) was followed by nitro reduction with SnCI₂ (5.0 equiv, EtOAc, 0.5 h, 70 °C), and amine protection (4.0 equiv of (BOC)₂O, 2 equiv of Et₃N, dioxane, 70 °C, 1 h, 74% for two steps). The resultant 3-bromoquinoline 7 was subjected to Pd(0)-catalyzed carboxybutylation (Schoenberg, A.; et al. J. Org. Chem. 1974, 39, 3318; Heck, R. F. Palladium Reagents in Organic Synthesis, Academic Press: London, Orlando (Fla), 1985) (0.1 equiv of (PPh₃)₄Pd, CO (g), 1.2 equiv of /7-Bu₃N, n-BuOH, 100 °C, 12 h, 78%) and treatment with LiOMe (1.1 equiv, MeOH, 25 °C, 1.5 h, 91%) to afford 9. As detailed by Heck (Schoenberg, A.; et al. J. Org. Chem. 1974, 39, 3318; Heck, R. F. Palladium Reagents in Organic Synthesis, Academic Press: London, Orlando (Fla), 1985), the use of MeOH as a reaction solvent for direct Pd-mediated carboxymethylation of 7 is limited by the achievable reaction temperatures (65 °C with MeOH versus 115 °C with n-BuOH) and thus requires extended reaction times (≥ 3 d) and generally proceeds in lower yields (< 45%). Following selective, acid-catalyzed C5 iodination (Boger, D. L.; McKie, J. A. J. Org. Chem. 1995, 60, 1271) of 9 with Λ/-iodosuccinimide (1.2 equiv, cat. TsOH, THF-CH₃OH, 0 to 25 °C, 1 d, 88%), Λ/-alkylation of the sodium salt of 10 (1.1 equiv of NaH, DMF, 4 °C, 30 min) with allyl bromide (3 equiv, DMF, 25 °C, 2.5 h, 94%) proceeded smoothly.

In contrast to related substrates (For synthetic aspects see: Boger, D. L.; et al. Chem. Rev. 1997, 97, 787; Boger, D. L; McKie, J. A. J. Org. Chem. 1995, 60, 1271), cyclization of 11 under previously described conditions (6.0 equiv of Bu₃SnH, 6.0 equiv of TEMPO, 70 °C, benzene, 3.5 h) did not proceed in high yield (41-54%). Although tin hydride-mediated reduction has been observed with similar compounds (Ueno, Y.; et al. Tetrahedron Lett. 1982, 23, 2575; Shankaran, K.; et al. Tetrahedron Lett. 1985, 26, 6001), it had not been observed in other CC-1065/duocarmycin systems. Consequently, it was surprising to isolate significant amounts (≥ 30%) of the halogen-reduced product. Characterization of this byproduct (Methyl 8-(benzyloxy)-6-[(Λ/-(te/f- butyloxycarbonyl)- Λ/-(2-propenyl))amino]quinoline-3-carboxylate: FABHRMS (NBA/Csl) m/z 581.1065 (M + Cs⁺, C₂₆H₂₈N₂O₅ requires 581.1053)) was confirmed by direct comparison with an authentic sample prepared by allylation of 9 (3 equiv of allyl bromide, 1.1 equiv of NaH, DMF, 4 to 25 °C, 89%). The radical cyclization was improved utilizing fr/^'s(trimethylsilyl)silane, ((CH₃Si)₃SiH, 5 equiv, 8.0 equiv of TEMPO, toluene, 80 °C, 16 h, 85%), which has a stronger metal-hydride bond than Bu₃SnH (79 vs. 74 kcal/mol) (Giese, B.; Kopping, B. Tetrahedron Lett. 1989, 30, 681 ; Kanabus-Kaminska, J. M.; et al. J. Am. Chem. Soc. 1987, 109, 5267) and therefore a slower rate of aryl radical reduction permitting clean intramolecular 5-exo-trig cyclization. The TEMPO trapped product 12 was reduced with activated Zn (Boger, D. L.; McKie, J. A. J. Org. Chem. 1995, 60, 1271 ; Newman, M. S.; Evans, F. J., Jr. J. Am. Chem. Soc. 1955, 77, 946) (50 equiv, 3:1 HOAc-THF, 60 °C, 11 h, 60%) and the resulting alcohol 13 converted to the primary chloride (Hooz, J.; Gilani, S. S. H. Can. J. Chem. 1968, 46, 86) (3.0 equiv of Ph₃P, 9.0 equiv of CCI₄, CH₂CI₂, 25 °C, 3 h, 73%). Two-phase transfer catalytic hydrogenolysis (Ram, S.; Ehrenkaufer, R. E. Synthesis 1988, 91 ; Bieg, T.; Szeja, W. Synthesis 1985, 76) of the benzyl ether 14 (10% Pd-C, 10 equiv 25% aqueous HCO₂NH₄, 25 °C, 3 h, 99%) and subsequent spirocyclization (Baird, R.; Winstein, S. J. Am. Chem. Soc. 1963, 85, 567; Baird, R.; Winstein, S. J. Am. Chem. Soc. 1962, 84, 788; Winstein, S.; Baird, R. J. Am. Chem. Soc. 1957, 79, 756) by treatment of 15 with DBU (3 equiv, CH₃CN, 25 °C, 3 h, 99%) provided A/-BOC-CPyl (16). Acid-catalyzed deprotection of 16 (3 M HCI-EtOAc, 25 °C, 30 min), which is accompanied by the addition of chloride to the cyclopropane, followed by treatment of the crude hydrochloride salt with K₂CO₃ (10 equiv, acetone, 25 °C, 24 h, 96%) cleanly provided CPyl (17).

In an improved preparation of the advanced (chloromethyl)indoline precursor, adoption of a direct free radical cyclization (Patel, V. F.; et al. J. Org. Chem. 1997, 62, 8868; Boger, D. L.; et al. Tetrahedron Lett. 1998, 39, 2227) of a substrate bearing a vinyl chloride acceptor alkene provided 14 in good conversion (Figure 14). Thus, alkylation of the sodium salt of 10 (1.1 equiv of NaH, DMF, 4 °C, 30 min) with £-1 ,3-dichloropropene (3 equiv, 25 °C, 12 h, 94%) was followed by free radical cyclization of the vinyl chloride (1.5 equiv of Bu₃SnH, cat. AIBN, benzene, 70 °C, 6.5 h, 87%) to provide 14. The incorporation of this improvement provided Λ/-BOC-CPyl (16) in 10 steps and superb overall conversion (29%).

Resolution. In order to assess the properties of both enantiomers of the CPyl based agents, a direct chromatographic resolution on a ChiralCel-OD semi-preparative HPLC column (2 * 25 cm, 50% /-PrOH-hexanes eluant, 7 mL/min, α = 1.43) provided both enantiomers of 16 (>99% ee). Λ/-BOC-CPyl proved to be the only late-stage intermediate that could be effectively resolved on a ChiralCel-OD column, and similar efforts to separate 13-15 were not successful. The slower eluting (+)-enantiomer of 16 (t_R = 38 min) was assigned the natural (3S)-configuration and agents derived from this (+)-enantiomer exhibited the more potent biological activity, the more effective DNA alkylation properties, and a DNA alkylation selectivity identical with the natural products. The faster eluting (- )-enantiomer of 16 (t_R = 27 min) was assigned the unnatural (3r?)-configu ration and agents derived from this (-)-enantiomer exhibited the corresponding less potent biological activity, the less effective DNA alkylation properties, and a DNA alkylation selectivity identical with the unnatural enantiomers of the natural products.

Synthesis of Duocarmycin and CC-1065 Analogues. The CPyl alkylation subunit was incorporated into duocarmycin and CC-1065 analogues as detailed in Figure 15. Deprotection and concurrent ring opening of 16 (3 M HCI-EtOAc) followed by immediate coupling (4 equiv of EDCI, DMF, 25 °C) of the resulting amine hydrochloride salt with 5,6,7-trimethoxyindole-2-carboxylic acid (Boger, D. L; et al. J. Am. Chem. Soc. 1990, 112, 8961 ) (TMI, 19, 10 h, 52%), 5-methoxyindole-2-carboxylic acid (20, 10 h, 59%), indole-2-carboxylic acid (21 , 10 h, 71%), indole₂ (Boger, D. L.; et al. Bioorg. Med. Chem. 1995, 3, 1429) (22, 16 h, 64%), and CDPI., (Boger, D. L; et al. J. Org. Chem. 1987, 52, 1521 ; Boger, D. L; Coleman, R. S. J. Org. Chem. 1984, 49, 2240) (23, 16 h, 41 %) provided 24, 26, 28, 30, and 32, respectively. DBU (3 equiv, 3 h, 25 °C) promoted spirocyclization of 24 (DMF, 86%), 26 (CH₃CN, 90%), 28 (CH₃CN, 93%), 30 (DMF, 97%), and 32 (DMF, 67%) provided 25, 27, 29, 31, and 33, respectively.

Solvolysis: Reactivity and Regioselectivity. Two fundamental characteristics of the alkylation subunits have proven important in past studies (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439). The first is the stereo-electronically-controlled acid-catalyzed ring opening of the activated cyclopropane which dictates preferential addition of a nucleophile to the least substituted cyclopropane carbon. The second is the relative reactivity of the agents as established by their rate of acid-catalyzed solvolysis which has been found to accurately reflect a direct relationship between intrinsic stability and in vitro cytotoxicity (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439). Solvolysis was conducted in 50% CH₃OH-buffer mixtures (pH 3 buffer =

4:1 :20 (v:v:v) 0.1 M citric acid: 0.2 M Na₂HPO₄: H₂O; pH 2 buffer = 4:1:20 (v:v:v) 1.0 M citric acid: 0.2 M Na₂HPO₄: H₂O) and followed spectrophotometrically by UV with the disappearance of the long-wavelength absorption of the CPyl chromophore and with the appearance of a short-wavelength absorption attributable to the solvolysis product (Figures 3 and 18). Λ/-BOC-CPyl (16) proved to be reasonably stable to solvolysis even at pH 2.0 (k = 3.72 * 10^"6 s^' t_V2 = 5.2 h) and pH 3.0 (k = 3.81 x 10^"6 s^"1, t_m= 50.5 h) as compared to /V-BOC-CPI (34, pH 3, t_V2 = 37 h) and Λ/-BOC-DA (35, pH 3, t_m = 11 h) (Figure 4). However, 16 was less stable than Λ/-BOC-DSA (36, pH 3, t_V2 = 177 h) and Λ/-BOC-CBI (37, pH 3, t_V2 = 133 h). The rate of solvolysis was found to be independent of the phosphate buffer countercation (Na⁺, NH₄ ⁺ or K⁺) within experimental error.

The acid-catalyzed nucleophilic addition of CH₃OH to 16 was conducted on a preparative scale to establish the regioselectivity of addition and was confirmed by synthesis of the expected product 41 derived from nucleophilic addition to the least substituted cyclopropane carbon. Treatment of Λ/-BOC-CPyl (16) with catalytic CF₃SO₃H (0.3 equiv, CH₃OH, 25 °C, 20 h, 91%) resulted in the clean addition to provide a single product 41 (Figure 16). Similarly, treatment of 16 with HCI-EtOAc (2 equiv, THF, -78 °C, 2 min, 96%) provided 15 as a single product. Clean cleavage of the C8b-C9 bond with S_N2 addition of CH₃OH or HCI to the least substituted C9 cyclopropane carbon was observed, and no cleavage of the C8b-C9a bond with ring expansion was detected (>20:1 ). This is in sharp contrast to the natural products where substantial amounts of the alternative ring expansion addition products have been observed (6-1.5:1) (Warpehoski, M. A.;

Harper, D. E. J. Am. Chem. Soc. 1994, 116, 7573; Warpehoski, M. A.; Harper, D. E. J. Am. Chem. Soc. 1995, 117, 2951 ; Boger, D. L; et al. Bioorg. Med. Chem. Lett. 1996, 6, 1955; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 311 ; Boger, D. L.; et al. J. Org. Chem. 1996, 61, 1710; Boger, D. L; et al. J. Org. Chem. 1996, 61, 4894). Nonetheless, the observations are consistent with prior studies with Λ/-BOC-CBI (37), Λ/-BOC-MCBI (38), and N -BOC-CCBI (39) where no (>20:1) ring expansion solvolysis product was detected (Figure 4) (Boger, D. L.; et al. J. Am. Chem. Soc. 1990, 112, 8961 ; Boger, D. L; et al. J. Org. Chem. 1996, 61, 1710; Boger, D. L; et al. J. Org. Chem. 1996, 61, 4894).

Solvolysis pH Dependence. Duocarmycin SA (1) is known to be exceptionally stable at neutral conditions and it was interesting to observe that Λ/-BOC-CPyl (16) possessed measurably solvolytic reactivity in 50% aqueous CH₃OH (pH 7). A full pH rate profile in 50% CH₃OH-universal buffer (pH 2-12, B(OH)₃-citric acid- Na₃PO₄) (Perrin, D. D.; Dempsey, B. Buffers forpH and Metal Ion Control; Chapman and Hall: London, 1979; p 156) demonstrated a near first-order rate dependence on acid at pH 2-4 (Figures 5 and 19). Above pH 4, the dependence on acid concentration disappeared indicating a change in mechanism from acid-catalyzed solvolysis to one which is uncatalyzed. A loss of isobestic behavior in the UV trace at pH 11 indicated a second reaction event at this pH, presumably arising from hydrolysis of either the BOC or methyl ester. From a regression analysis best fit plot of the k_obs versus pH, rate constants of 3.37 * 10^"3 M^"V¹ and 8.36 * 10^"7 s^"1 for the acid-catalyzed and uncatalyzed reactions, respectively, were established. This, plus the demonstration that related reactions above pH 4 are not only not specific acid-catalyzed, but also not general acid-catalyzed (Boger, D. L.; Garbaccio, R. M. J. Org. Chem. 1999, 64, 5666; Boger, D. L.; Turnbull, P. J. Org. Chem. 1998, 63, 8004; Boger, D. L; Tumbull, P. J. Org. Chem. 1997, 62, 5849; Boger, D. L; et al. Bioorg. Med. Chem. Lett. 1997, 7, 233), suggest this is simply an uncatalyzed S_N2 nucleophilic addition. The demonstration of an uncatalyzed solvolysis above pH 4 is consistent with the observation that the DNA alkylation reaction at the physiological pH of 7.6 is not acid-catalyzed and that catalysis must come from an alternative source (Boger, D. L.; Garbaccio, R. M. Ace. Chem. Res. 1999, 32, 1043; Boger, D. L.; Garbaccio, R. M. Bioorg. Med. Chem. 1997, 5, 263; Boger, D. L.; et al. J. Am. Chem. Soc. 1997, 119, 4977; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 4987; Boger, D. L; Garbaccio, R. . J. Org. Chem. 1999, 64, 5666; Boger, D. L; Turnbull, P. J. Org. Chem. 1998, 63, 8004; Boger, D. L; Turnbull, P. J. Org. Chem. 1997, 62, 5849; Boger, D. L.; et al. Bioorg. Med. Chem. Lett. 1997, 7, 233).

Solvolysis: Metal Catalysis. Central to the projected use of the CPyl agents and their potential of metal cation activation was their reactivity in the presence of metal cations (Boger, D. L.; et al. J. Am. Chem. Soc. 1993, 115, 10733; Yoshida, K.; et al. Bull. Chem. Soc. Jpn. 1988, 61, 4335; Yoshida, K.; et al. Chem. Lett. 1986, 1059; Pratt, Y. T. J. Org. Chem. 1962, 27, 3905). Treatment of Λ/-BOC-CPyl (16) with Zn(OTf)₂ (1.1 equiv) in CH₃OH (25 °C, 4 h, 92%) on a preparative scale resulted in clean addition to provide a single product 41 (Figure 17). In contrast, /V-BOC-CBI (37) did not exhibit this metal-catalyzed reactivity in the presence of Zn(OTf)₂ (1.5 equiv, CH₃OH, 27 h, 25 °C, 100% recovery), confirming that the CPyl 8-ketoquinoline structure is key to metal chelation, activation, and Lewis acid-catalyzed reaction.

The rates of metal-catalyzed (Cu²⁺, Ni²⁺, Zn²⁺, Mn²⁺, Mg²⁺, Mg²⁺, Fe³⁺, Cr³*, and Ti⁴⁺) addition to Λ/-BOC-CPyl (16) in CH₃OH were measured spectrophotometrically by UV with the disappearance of the long-wavelength absorption of the CPyl chromophore and with the appearance of a short-wavelength absorption (Figure 11). Within a series of divalent metals with acetylacetonate (acac) ligands, the relative rates of solvolysis (rate: Cu²⁺ > Ni²⁺ > Zn²⁺ > Mn²⁺ > Mg²⁺) corresponded directly with established stability constants of the metal complexes with 8-hydroxyquinoline (stability: Cu²⁺ > Ni²⁺ > Zn²⁺ > Mn²⁺ > Mg²⁺) (Phillips, J. P. Chem. Rev. 1956, 56, 271 ; Irving, H.; Williams, R. J. P. J. Chem. Soc. 1953, 3192). Higher valence metals including Fe³⁺, Cr³⁺, and Ti⁴⁺ also demonstrated an analogous activation of CPyl for nucleophilic addition. Notably, no apparent catalysis was observed for Mg(acac)₂ over and beyond the background rate indicating that this endogenous metal cation, like Na⁺, does not activate CPyl effectively. Stronger Lewis acids were found to provide at a faster reaction rate (Zn(OTf)₂: t_V2 = 0.8 h versus Zn(acac)₂: t_V2 = 11.5 h) and incorporation of H₂O in the solvent slowed the reaction (Zn(OTf)₂, 50% aqueous CH₃OH, t_V2 = 33 h). Λ/-BOC-CBI (37) demonstrated no detectable reaction in the Zn(OTf)₂-CH₃OH system after 7 d, further confirming the role of metal cation catalysis for CPyl. Thus, the well-behaved activation of CPyl predictably tunable by choice of the metal cation provides the opportunities to selectively activate the agents.

DNA Alkylation Selectivity and Efficiency. The DNA alkylation properties of the agents were examined within w794 duplex DNA (Boger, D. L.; et al.

Tetrahedron 1991 , 47, 2661 ) for which comparative results are available for related agents (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439). The alkylation site identification and the assessment of the relative selectivity among the available sites were obtained by thermally-induced strand cleavage of the singly 5' end-labeled duplex DNA after exposure to the agents. Following treatment of the end-labeled duplex DNA with a range of agent concentrations and temperatures in the dark, the unbound agent was removed by EtOH precipitation of the DNA. Redissolution of the DNA in aqueous buffer, thermolysis (100 °C, 30 min) to induce strand cleavage at the sites of DNA alkylation, denaturing high-resolution polyacrylamide gel electrophoresis (PAGE) adjacent to Sanger dideoxynucleotide sequencing standards, and autoradiography led to identification of the DNA cleavage and alkylation sites. The full details of this procedure have been disclosed elsewhere (Boger, D. L.; et al. Tetrahedron 1991 , 47, 2661 ). A representative comparison of the DNA alkylation properties of both enantiomers of /V-BOC-CPyl (16) alongside both enantiomers of Λ/-BOC-DSA (36) is presented in Figure 6. Both natural enantiomers exhibited approximately the same efficiencies of DNA alkylation detectable at 10^"3 M (37 °C, 48 h) and prominent at 10^"2 M. (+)-A/-BOC-CPyl was slightly less efficient than the unnatural enantiomer, (-)-/V-BOC-CPyl, at both 24 and 48 h incubations which is also reflected in the relative cytotoxic activities of the two enantiomers. Like the preceding BOC derivatives examined (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439), 16 alkylated DNA much less efficiently than 24-33 (10⁴x), providing detectable alkylation at 10^"2-10^"3 M only under vigorous conditions (37 °C, 24-72 h) and much less selectively than 24-33, exhibiting a two-base pair AT-rich alkylation selectivity (5'-AA > 5'-TA). This unusual behavior of the two enantiomers alkylating the same sites is analogous to past observations (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439). It is a natural consequence of the reversed binding orientation of the two enantiomers and the diastereomeric relationship of the two adducts that result in the two enantiomers covering the exact same binding site surrounding the alkylated adenine. This has been discussed in detail illustrated elsewhere, and Λ/-BOC-CPyl (16) conforms nicely to these past observations and models (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439).

A representative comparison of the DNA alkylation by (+)-CPyl-TMI (25), (+)-CPyl-indole₂ (31), and (+)-CPyl-CDPI₁ (33) alongside that of (+)-duocarmycin SA (1) and (+)-CC-1065 (3) within w794 DNA is illustrated in Figure 7. (+)-CPyl-TMI and (+)-duocarmycin SA alkylate DNA with identical selectivity and near identical efficiency with the latter agent being slightly more effective. This is nicely illustrated in Figure 7 where the two agents detectably alkylate the same high affinity site of 5'-AATTA at 10⁶-10^"7 M (25 °C, 24 h). This is analogous to the observations made in comparisons of duocarmycin SA and CBI-TMI (Boger, D. L.; Yun, W. J. Am. Chem. Soc. 1994, 116, 7996). Like the preceding agents, the CPyl-based agents exhibit AT-rich adenine N3 alkylation selectivities that start at the 3' adenine N3 alkylation site with agent binding in the minor groove in the 3' to 5' direction covering 3.5 or 5 base pairs (data not shown).

A representative comparison of the DNA alkylation by (-)-CPyl-TMI (25), (- )-CPyl-indole₂ (31),and (-)-CPyl-CDPI., (33) alongside the unnatural enantiomer of duocarmycin SA (1) and the natural enantiomer of CC-1065 (3) within w794 DNA is illustrated in Figure 8. The unnatural enantiomer DNA alkylation is considerably slower, and the results shown in Figure 8 were obtained only with their incubation for 72 h (25 °C) versus incubation for 24 h (25 °C, Figure 7) for the natural enantiomers. Despite the longer reaction times, the extent of alkylation by the unnatural enantiomers is lower, requiring higher agent concentrations to detect. The DNA alkylation selectivity and efficiency observed with enf-(-)-CPyl-TMI (25) and enf-(-)-duocarmycin SA (1) were indistinguishable with the latter agent being slightly more effective. The alkylation sites for the unnatural enantiomers proved consistent with adenine N3 alkylation with agent binding in the minor groove in the reverse 5' to 3' direction across a 3.5 or 5 base-pair AT-rich site surrounding the alkylation site (data not shown). This is analogous to the natural enantiomer alkylation selectivity except that it extends in the reverse 5' to 3' direction in the minor groove and, because of the diastereomeric nature of the adducts, is offset by one base pair relative to the natural enantiomers.

In Vitro Cytotoxic Activity. Past studies with agents in this class have defined a direct correlation between inherent stability and cytotoxic potency (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439). Consistent with their relative reactivity, the CPyl based agents exhibited cytotoxic activity that closely followed this relationship (Figures 9 and 20) (For mechanistic aspects see: Boger, D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1439). The results, which also follow trends established in the DNA alkylation studies, demonstrate that the (+)-enantiomer of the analogs possessing the configuration of the natural products, is the more potent enantiomer by 3-30x. The exceptions to this trend are the simple alkylation subunits 16 and 17 themselves which, like others in the series, typically exhibit comparable activities. The seco precursors, which lack the preformed cyclopropane but possess the capabilities of ring closure, were found to possess cytotoxic activity that was indistinguishable from the final ring-closed agents. Consistent with the unique importance of the C5 methoxy group in the binding subunit of the duocarmycins (Boger, D. L.; Garbaccio, R. M. Ace. Chem. Res. 1999, 32, 1043; Boger, D. L; Garbaccio, R. M. Bioorg. Med. Chem. 1997, 5, 263; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 4977; Boger, D. L.; et al. J. Am. Chem. Soc. 1997, 119, 4987), CPyl-TMI (25) and 27 were found to be equipotent illustrating that the C6 and C7 methoxy groups of 25 are not contributing to its cytotoxic potency. The indole derivative 29 was found to be less potent (10x) than both 25 and 27, further demonstrating the importance of the C5 methoxy which we have suggested is derived from extending the rigid length of the agents and contributing to the alkylation catalysis (Boger, D. L.; Garbaccio, R. M. Ace. Chem. Res. 1999, 32, 1043; Boger, D. L.; Garbaccio, R. M. Bioorg. Med. Chem. 1997, 5, 263; Boger, D. L.; et al. J. Am. Chem. Soc. 1997, 119, 4977; Boger, D. L; et al. J. Am. Chem. Soc. 1997, 119, 4987). Finally, the longer agents, CPyl-indole₂ (31 ) and CPyl-CDP^ (33), displayed the most potent cytotoxic activity reflecting their longer length and greater adduct stability.

Metal Activation of CPyl: A study of CPyl activation by metal cations toward nucleophilic addition

(MeOH, Figure 11) revealed that the relative reaction rates correspond to the established stabilities of the resulting metal complexes (8-hydroxyquinoline, Cu²⁺ > Ni²⁺ > Zn²⁺ > Mn²⁺ > Mg²⁺) (J. P. Phillips, Chem. Rev. 1956, 56, 271; and H. Irving, et al., J. Chem. Soc. 1953, 3192). This provides the opportunity to predict, control, and tune the reactivity over a wide range depending on the application and conditions. Notably, Mg²⁺ provided a rate which was not distinguishable from a background rate of BOC methanolysis, indicating that this prominent endogenous metal cation, like Na⁺, does not appear to activate CPyl effectively.

Consistent with this behavior, the efficiency of the DNA alkylation reaction of 16 (w794 DNA) (see: D. L. Boger, et al., Tetrahedron 1991 , 47, 2661), which occurs at 10^-2 M for 1-3 (24 h, 25 °C for 1 , 37 °C for 2 and 3), was dramatically increased in the presence of Cu²⁺ (1 OOx), Ni²⁺ (100-1 OOOx), and Zn²⁺ (1 OOOx), the three metals selected for study, Figure 12. This enhancement increased with increasing metal cation concentration for 16, but not 35 and 36, and resulted in no change in the DNA alkylation selectivity of 16, which was identical to 35 and 36. Typically studies were conducted with 1 , 10, 100, 1000 equivalents of the metal cation with the former producing significant effects and the latter two concentrations providing the maximal effects. Under the conditions of the measurements reported herein, the enhancement was especially remarkable with Zn²⁺ (1000x) which promoted DNA alkylation of 16 at 10^"5 M. This greater behavior of Zn²⁺ relative to Ni²⁺ and Cu²⁺ (Zn²⁺ > Ni²⁺ > Cu²⁺) is attributed not to an alteration in the relative activations from that expected (Cu²⁺ > Ni²⁺ > Zn²⁺), but rather to an enhanced selectivity under the reaction conditions employed. That is, the greater activation by Cu²⁺ and Ni²⁺ leads to more nonproductive solvolysis relative to Zn²⁺, lowering the apparent efficiencies of DNA alkylation. For any given application, the optimal results are going to depend on the reaction conditions (solvent, buffer, temperature, time) and the optimal catalyst from the range of metal cation catalysts can be established to tune the reactivity. Alkylation at such low concentrations is unprecedented for such simple alkylation subunits and this efficiency is within 10-fold of the natural product (+)-duocarmycin SA (10^"6 M, 25 ^°C), Figure 21. Similar enhancements in the rates of DNA alkylation were also observed. Analogous metal cation enhancements in the rates and efficiency of DNA alkylation for the unnatural enantiomer of 16 were also observed (data not shown). However, similar treatments of 35 and 36 (alkylation at 10^"2 M) or duocarmycin SA and CC-1065 (alkylation at 10^"6 M) did not affect their DNA alkylation rates or efficiencies (data not shown) indicating this behavior is unique to CPyl (16) and its 8-ketoquinoline core structure.

In addition to representing a new and tunable method of in situ activation of a novel class of DNA alkylating agents, the observations have implications on the interpretation of the behavior of CC-1065 and the duocarmycins (D. L. Boger, et al., Angew. Chem., Int. Ed. Engl. 1996, 35, 1439; and D. L. Boger, et al., Chem. Rev. 1997, 97, 787). First, the majority of the efficiency distinctions observed between the simple alkylation subunits such as 16, 35, and 36 and the natural products (10³x of the 10⁴x difference) may be attributed to ineffective catalysis of the DNA alkylation reaction with 16 and related agents and not their intrinsic capabilities or reversibility (kinetic effect) (D. L. Boger, et al., J. Am. Chem. Soc. 1994, 116, 1635; and D. L. Boger, J. Am. Chem. Soc. 1993, 115, 9872). The remaining 10-fold difference may be attributed to differences in the noncovalent binding affinity and/or the minor groove positioning and orientation of the agents consistent with identical conclusions drawn from the results of unrelated studies (D. L. Boger, et al., J. Am. Chem. Soc. 1997, 119, 4977; D. L. Boger, et al., J. Am. Chem. Soc. 1997, 119, 4987). Accordingly, 16, 35, and 36 appear to lack the structural features required for catalysis derived from a DNA binding induced conformational change in the natural products which disrupts the cross conjugated and stabilizing alkylation subunit vinylogous amide activating them for nucleophilic attack (D. L. Boger, et al., Bioorg. Med. Chem. 1997, 5, 263; and D. L. Boger, Ace. Chem. Res. 1999, 32, 1043). Secondly, studies with more advanced CPyl analogues related to the structures of CC-1065 and the duocarmycins, like those with 16 illustrated in Figure 21 , have shown that the DNA alkylation selectivity is unaffected by the metal cation catalysis. This indicates that the source of the alkylation selectivity is not uniquely embedded in the catalysis source which is consistent with proposals that it is derived from the compounds' noncovalent binding selectivity (D. L. Boger, et al., Bioorg. Med. Chem. 1994, 2, 115).

Other examples of metal cation initiation of a DNA alkylation reaction are unknown and, as such, the studies detailed herein appear to constitute the first example. Intriguingly, comparative trace metal analysis of cancerous and noncancerous human tissues have revealed significant distinctions (I. L. Mulay, et al., J. Natl. Cancer Inst. 1971 , 47, 1). Although no generalizations were possible across all tumor types, within a given tumor type these were significant and potentially exploitable differences. For example, Zn was found in breast carcinoma at levels 700% higher than in normal cells of the same type, while lung carcinoma exhibited a reversed and even larger 10-fold difference. Thus, chemotherapeutic agents subject to Zn activation exhibit an enhanced activity against breast carcinoma attributable to this difference in Zn levels. Such therapeutic applications of this class of agents complement their use as research tools and as models to probe the source of the duocarmycin and CC-1065 DNA alkylation selectivity and catalysis.

Synthetic Protocols

3-Bromo-8-hydroxy-6-nitroquinoline (5). A solution of 2-bromoacrolein (5.0 g, 37 mmol, 1.0 equiv) in glacial acetic acid (110 mL) at 25 °C was titrated to the appearance of a faint redish color with bromine (ca. 5.9 g, 37 mmol, 1.0 equiv). 2-Hydroxy-4-nitroaniline (4, 5.7 g, 37 mmol, 1.0 equiv) was added, and the solution was gradually heated to 100 °C. The solution was cooled to 25 °C after 1 h. Filtering and neutralization of the precipitate with sodium phosphate buffer (1 M, pH 7, Na₂HPO₄-NaH₂PO₄) afforded 9.2 g (92%) of 5 as a light yellow solid: mp 240-241 °C; IR (film) v_max3408 (br), 3089, 1587 cm^"1. Anal. Calcd for C₉H₅BrN₂O₃: C, 40.18; H, 1.87; N, 10.41. Found: C, 40.21; H, 1.91; N, 9.98.

8-(Benzyloxy)-3-bromo-6-nitroquinoline (6). A solution of 5 (13.7 g, 51.0 mmol, 1.0 equiv) in anhydrous DMF (150 mL) was cooled to 4 °C under N₂ and treated with Kl (1.70 g, 10.0 mmol, 0.2 equiv) and NaH (60% dispersion in oil, 2.24 g, 56.0 mmol, 1.1 equiv). Benzyl bromide (7.30 mL, 6.10 mmol, 1.2 equiv) was added after 30 min and the reaction was allowed to warm to 25 °C. After 24 h, the reaction volume was reduced by two-thirds in vacuo and EtOAc (200 mL) was added. The reaction mixture was poured on H₂O (200 mL) and extracted with EtOAc (3 x 100 mL). The combined organic extracts were washed with saturated aqueous NaCI (40 mL), dried (Na₂SO₄) and concentrated. Flash chromatography (SiO₂, 5.5 x 20 cm, 50-100% CH₂CI₂-hexane gradient) afforded 6 (15.6 g, 85%) as a yellow solid: mp 170 °C; FABHRMS (NBA/Nal) m/z 359.0040 (M + H⁺,

C, 53.50; H, 3.09; N, 7.80. Found: C, 53.81 ; H, 3.23; N, 7.48.

8-(Benzyloxy)-3-bromo-6-(tø-(tert-butyloxycarbonyl))aminoquinoline (7). A solution of 6 (0.20 g, 0.56 mmol, 1.0 equiv) in EtOAc (1.1 mL) at 25 °C was treated with SnCI₂-2H₂O (0.63 g, 2.8 mmol, 5.0 equiv). The reaction mixture was heated to 70 °C under N₂ until an orange slurry formed (ca. 0.5 h). After cooling to 25 °C, the reaction mixture was poured on ice and neutralized with 1 N aqueous NaOH. The aqueous layer was extracted with EtOAc (3 x 15 mL) and the combined organic layers were filtered, washed with saturated aqueous NaCI (10 mL), dried (Na₂SO₄) and concentrated. The yellow solid was placed under vacuum for 0.5 h and then dissolved in anhydrous dioxane (5.0 mL) and treated with di-fetf-butyl dicarbonate (0.49 g, 2.3 mmol, 4.0 equiv) and Et₃N (0.16 mL, 1.1 mmol, 2.0 equiv). The reaction mixture was warmed to 70 °C under Ar for 1 d. After cooling to 25 °C, the solvent was removed in vacuo. Chromatography (SiO₂, 3 13 cm, 25% EtOAc-hexane) afforded 7 (0.18 g, 74%) as a light yellow solid: mp 162 °C; IR (film) v_max 3354, 2971 , 2919, 1807, 1766, 1724 cm^'1; FABHRMS (NBA/Csl) m/z 429.0825 (M + H⁺, C₂₁H₂₁BrN₂O₃ requires 429.0814). n-Butyl 8-(Benzyloxy)-6-(A/-(ferf-butyloxycarbonyl)amino)quinoline- 3-carboxylate (8). A solution of 7 (4.4 g, 10 mmol, 1.0 equiv) in n-BuOH (85 mL) was degassed with N₂. Pd(PPh₃)₄ (1.2 g, 1.0 mmol, 0.1 equiv) and t?-Bu₃N (2.9 mL, 12 mmol, 1.2 equiv) were added and the solution was again purged with N₂. The reaction mixture was flushed with CO and then slowly heated to 100 °C under a CO atmosphere. Upon complete reaction (ca. 12 h), H₂O (50 mL) and saturated aqueous NH₄CI (50 mL) were added. The organic layer was separated and the aqueous layer was extracted with EtOAc (3 50 mL); The combined organic layers were washed with saturated aqueous NaCI (40 mL), dried (Na₂SO₄) and concentrated. Chromatography (SiO₂, 5.5 20 cm, 25% EtOAc- hexane) afforded 8 (3.6 g, 78%) as a yellow solid: mp 135-136 °C; FABHRMS (NBA/Csl) m/z 451.2249 (M + H⁺, C₂₆H₃₀N₂O₅ requires 451.2233). Methyl 8-(Benzyloxy)-6-(Λ^(ferf-butyloxycarbonyl)amino)quinoline- 3-carboxylate (9). A solution of 8 (2.9 g, 6.4 mmol, 1.0 equiv) in CH₃OH (70 mL) was cooled to 4 °C under N₂ and treated with LiOMe (0.28 g, 7.1 mmol, 1.1 equiv). The reaction mixture was allowed to warm to 25 °C after 20 min. Upon complete reaction (ca. 1.5 h), H₂O (100 mL) was added. The organic layer was separated and the aqueous layer was extracted with EtOAc (3 x 30 mL). The organic layers were combined, washed with saturated aqueous NaCI (30 mL), dried (Na₂SO₄) and concentrated. Chromatography (SiO₂, 5 x 19 cm, 25-30% EtOAc-hexane gradient) afforded 9 (2.4 g, 91 %) as a yellow solid: mp 173-174 °C; FABHRMS (NBA/Csl) m/z 409.1773 (M + H⁺, C₂₃H₂₄N₂O₅ requires 409.1763), Anal. Calcd for C₂₃H₂₄N₂O₅: C, 67.63; H, 5.92; N, 6.86. Found: C, 68.00; H, 5.98; N, 6.75.

Methyl 8-(Benzyloxy)-6-(Λ/-(terf-butyloxycarbonyl)amino)-5-iodoquinoline-3- carboxylate (10). A solution of 9 (2.1 g, 5.2 mmol, 1.0 equiv) in a 1 :1 mixture of THF-CH₃OH (85 mL) was cooled to 4 °C and treated with catalytic TsOH (40 mg) in THF (0.5 mL). Λ/-lodosuccinimide (1.4 g, 6.2 mmol, 1.2 equiv) in THF (10 mL) was slowly added over 10 min. After 1.5 h, the reaction mixture was warmed to 25 °C and stirred 45 h. Upon complete reaction, saturated aqueous NaHCO₃ (100 mL), Et₂0 (100 mL), and 100 mL H₂O (100 mL) were added. The organic layer was separated and the aqueous layer was extracted with Et₂O (3 x 50 mL) and EtOAc (50 mL). The organic layers were combined, washed with saturated aqueous NaHCO₃ (50 mL) and saturated aqueous NaCI (50 mL), dried (Na₂SO₄) and concentrated. Chromatography (SiO₂, 5 19 cm, hexanes then 30% EtOAc- hexane) provided 10 (2.3 g, 84%, typically 80-88%) as a yellow solid: mp 182- 183 °C; FABHRMS (NBA Csl) m/z 535.0743 (M + H⁺, C₂₃H₂₃IN₂O₅ requires 535.0730).

Methyl 8-(Benzyloxy)-6-[(W-(tert-butyloxycarbonyl)-W-(2-propenyl)) amino]-5-iodo- quinoline-3-carboxylate (11). A solution of 10 (0.40 g, 0.75 mmol, 1.0 equiv) in anhydrous DMF (6.2 mL) at 4 °C in a flamed dried round bottom flask was treated with NaH (60% dispersion in oil, 33 mg, 0.82 mmol, 1.1 equiv) and stirred under Ar. After 30 min, allyl bromide (94 μL, 2.3 mmol, 3.0 equiv) was added and the reaction mixture was warmed to 25 °C and stirred 2.5 h. Saturated aqueous NaHCO₃ (1.5 mL) and EtOAc (5 mL) were added and the reaction was then poured on H₂O (100 mL). The aqueous layer was extracted with EtOAc (4 x 20 mL) and the combined organic extract was washed with H₂O (2 40 mL) and saturated aqueous NaCI (2 x 40 mL), dried (Na₂SO₄), and concentrated in vacuo. Chromatography (SiO₂, 2.5 x 16 cm, 20-25% EtOAc- hexane gradient) afforded 11 (0.40 g, 94%) as a gold solid (mixture of amide rotamers in CDCI₃): mp 128-129 °C; FABHRMS (NBA/Csl) m/z 575.1036 (M + H⁺, C₂₆H₂₇IN₂O₅ requires 575.1043).

Methyl 5-(Benzyloxy)-3-(fe^butyloxycarbonyl)-1-[[^2\6 6^,-tetramethyl piperidino)-oxy]methyl]-1,2-dihydro-3H-pyrido[3,2-e]indole-8-carboxylate (12). A solution of 11 (20 mg, 35 μmmol, 1.0 equiv) in anhydrous toluene (1.2 mL) was treated with a solution of TEMPO (16 mg, 0.11 mmol, 3.0 equiv) in toluene (0.11 mL) and (TMS)₃SiH (11 μL, 37 μmol, 1.05 equiv). The solution was warmed to 80 °C and 5 equiv of TEMPO (2 x 14 mg in 0.29 mL toluene) and 4 equiv (TMS)₃SiH (4 x 11 μL) were added in portions over the next 4 h. After 16 h, the reaction mixture was cooled to 25 °C and the volatiles removed in vacuo. Chromatography (SiO₂, 1.5 x 12 cm, 20% EtOAc-hexane) provided 12 (18 mg, 85%) as a gold solid: mp 157-158 °C; FABHRMS (NBA/Csl) m/z 726.2383 (M + Cs⁺, C₃₅H₄₅N₃O₆ requires 736.2363). Anal. Calcd. for C₃₅H₄₅N₃O₆Η₂O: C, 67.61 ; H, 7.62; N, 6.76. Found: C, 67.70; H, 7.11 ; N, 6.27. Methyl 5-(Benzyloxy)-3-(tert-butyloxycarbonyl)-1-(hydroxymethyl)- 1 ,2-dihydro-3H- pyrido[3,2-e]indole-8-carboxylate (13). A solution of 12 (20 mg, 33 μmol, 1.0 equiv) in a 3:1 mixture of THF-H₂O (0.90 mL) was treated with activated zinc powder (54 mg, 0.80 mmol, 25 equiv) and HOAc (0.20 mL) and the resulting suspension was warmed to 60 °C with vigorous stirring. After 7 h, additional Zn (54 mg) was added and the reaction was stirred 4 h. The Zn powder was removed by filtration through Celite with a CH₂CI₂ wash (10 mL), and the mixture was concentrated in vacuo. The resulting residue was dissolved in EtOAc (10 mL), filtered through Celite, and the solution was concentrated in vacuo. Chromatography (SiO₂, 1 x 12 cm, 20-50% EtOAc-hexane) provided 13 (9.2 mg, 60%): mp 186 °C; FABHRMS (NBA/Csl) m/z 465.2032 (M + H⁺, C₂₆H₂₈N₂O₆ requires 465.2026). Methyl

5-(Benzyloxy)-3-(fert-butyloxycarbonyl)-1-(chloromethyl)-1,2-dihydro-3H -pyrido[3,2-e]indole-8-carboxylate (14). Method A: A solution of 13 (6.7 mg, 14 μmol, 1.0 equiv) in anhydrous CH₂CI₂ (0.14 mL) under Ar was treated sequentially with Ph₃P (13 mg, 48 μmol, 3.0 equiv) and CCI₄ (14 μL, 0.15 mmol, 9.0 equiv). The reaction mixture was stirred at 25 °C for 3 h. The solvent was then evaporated under a stream of N₂. Radial chromatography (SiO₂, 1.0 mm, 20% EtOAc-hexane) afforded 14 (5.1 mg, 73%) as a gold solid: mp 194 °C (EtOAc- hexane); FABHRMS (NBA/Nal) m/z 483.1675 (M + H⁺, C₂₆H₂₇CIN₂O₅ requires 483.1687). Anal. Calcd for C₂₆H₂₇CIN₂O₅: C, 64.66; H, 5.63; N, 5.80. Found: C, 64.78; H, 5.73; N, 5.66. Methyl 3-(ferf-Butyloxycarbonyl)-1 -(chloromethyl)-5-hydroxy-1 ,2-dihydro-3 V- pyrido[3,2-e]indole-8-carboxylate (15). A slurry of 14 (0.9 g, 1.9 mmol, 1.0 equiv) and 10% Pd-C (0.36 g) in THF (6.2 mL) under N₂ was cooled to -78 °C and degassed under vacuum. The reaction was warmed to 25 °C and treated with 25% aqueous HCO₂NH₄ (1.2 g in 4.7 mL H₂O, 19 mmol, 10 equiv). After 3 h, the catalyst was removed by filtration through Celite (2 x 50 mL Et₂O wash) and the solvent was evaporated under reduced pressure to afford 15 (0.72 g, 99%) as a bright yellow solid: mp 162-163 °C; FABHRMS (NBA/Nal) m/z 393.1228 (M + H⁺, C₁₉H₂₁CIN₂O₅ requires 393.1217).

Methyl 2-(ferf-Butyloxycarbonyl)-1₎2,9,9a-tetrahydrocyclopropa[c]pyrido [3,2-e]indol- 4-one-7-carboxylate (16, Λ/-BOC-CPyl). A solution of 15 (81 mg, 0.21 mmol, 1.0 equiv) in CH₃CN (6.9 mL) at 25 °C under Ar was treated with DBU (0.12 mL, 0.83 mmol, 4.0 equiv) and stirred 3 h. Flash chromatography was applied directly to the reaction mixture (SiO₂, 2.5 x 8 cm, 2% MeOH-CH₂CI₂) and furnished 16 as a light yellow-white solid (68 mg, 93%, typically 90-99%): mp 255 °C (dec); IR (film) v_max2974, 1728 cm^"1; UV (CH₃OH) λ_max 318 (ε = 9000), 240 (shoulder, e = 11000), 218 (e = 15500); FABHRMS (NBA/Nal) m/z 379.1282 (M + Na⁺, C₁₉H₂₀N₂O₅ requires 379.1270).

Resolution of /V-BOC-CPyl. Samples of racemic 16 were resolved by semi-preparative HPLC chromatography on a Daicel ChiralCel OD column (10 μm, 2 x 25 cm) using 50% /-PrOH-hexane eluant (7 mL/min). The enantiomers eluted with retention times of 27.3 min and 37.8 min (α = 1.43). (+)-(8bR, 9aS)-16: [α]²² _D +120 (c 0.46, THF); (-)-(8bS, 9aR)-16: -118 (c 0.47, THF). Methyl 1,2,9,9a-Tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one- 7-carboxylate (17, CPyl). A solution of 16 (4.0 mg, 11 μmol, 1.0 equiv) in 3 M HCI-EtOAc (0.37 mL) was stirred for 30 min at 25 °C. The solvent was removed by a stream of N₂ and the residual salt was dried under vacuum. The residue was taken up in acetone (0.28 mL) and treated with K₂CO₃ (16 mg, 0.11 mmol, 10 equiv). After stirring for 24 h at 25 °C, the reaction mixture was filtered through Celite to provide 17 as a yellow solid film (2.7 mg, 96%): UV (CH₃OH) λ_max 360 (e = 7000), 298 (shoulder, e = 5000), 226 (e = 15000); FABHRMS (NBA/Nal) m/z 257.0927 (M + H⁺, C₁₄H₁₂N₂O₃ requires 257.0926); (+)-(8bR, 9aS)-17: [α]²² _D +41 (c 0.14, CH₂CI₂); (-)-(8bS, 9aR)-17: -45 (c 0.09, CH₂CI₂). Methyl 8-(Benzyloxy)-6-[(/V-(ferf-butyloxycarbonyl)-tø-(£-3-chloro- 2-propenyl))- amino]-5-iodoquinoline-3-carboxylate (18). A solution of 10 (1.2 g, 2.2 mmol, 1.0 equiv) in anhydrous DMF (20 mL) was cooled to 4 °C in a flamed dried round bottom flask under Ar and was treated with NaH (60% dispersion in oil, 98 mg, 2.5 mmol, 1.1 equiv). After 30 min, E-1 ,3-dichloropropene (0.61 mL, 6.7 mmol, 3.0 equiv) was added and the reaction mixture was gradually warmed to 25 °C and stirred 12 h. The reaction was poured on H₂O (100 mL) and saturated aqueous NaHCO₃ (50 mL). The aqueous layer was extracted with EtOAc (4 x 50 mL) and the combined organic extract was washed with saturated aqueous NaHCO₃ (50 mL), H₂O (2 x 40 mL), and saturated aqueous NaCI (2 x 40 mL), dried (Na₂SO₄), and concentrated in vacuo. Chromatography (SiO₂, 5.5 x 14 cm, 20-30% EtOAc-hexane gradient) afforded 18 (1.2 g, 85%, typically 84-94%) as a yellow foam (mixture of amide rotamers in CDCI₃). FABHRMS (NBA/Csl) m/z 740.9646 (M + Cs⁺, C₂₆H₂₆CIN₂O₅l requires 740.9629). Methyl 5-(Benzyloxy)-3-(te/f-butyloxycarbonyl)-1-(chloromethyl)- 1,2-dihydro-3A - pyrido[3,2-e]indole-8-carboxylate (14). Method B: A solution of 18 (0.65 g, 1.1 mmol, 1.0 equiv) in benzene (20 mL) under Ar was treated with π-BuSn₃H (0.15 mL, 0.50 mmol, 0.5 equiv) and catalytic AIBN (18 mg) and stirred at 70 °C. Additional π-BuSn₃H (0.29 mL, 1.1 mmol, 1.0 equiv in 2 portions) was added over the next hour. After 3 h, the reaction mixture was concentrated in vacuo. Chromatography (SiO₂, 4 x 20 cm, 20-30% EtOAc-hexane gradient) provided 14 (0.46 g, 87%).

Methyl 1-(Chloromethyl)-5-hydroxy-3-[(5,6,7-trimethoxyindol-2-yl) carbonyl]-1 ,2- dihydro-3H-pyrido[3,2-e]indole-8-carboxylate (seco-CPyl-TMI, 24). A solution of 16 (10.0 mg, 28.1 μmol, 1.0 equiv) in 3 M HCI-EtOAc (0.935 mL) was stirred for 30 min at 25 °C. The solvent was removed by a stream of N₂ and the residual salt was dried under vacuum. The residue was dissolved in anhydrous DMF (0.300 mL) and treated with 5,6,7-trimethoxyindole-2-carboxylic acid (19, 10.6 mg, 42.1 μmol, 1.5 equiv) and EDCI (27.0 mg, 140 μmol, 5.0 equiv). After stirring for 10 h at 25 °C under Ar, the reaction mixture was concentrated in vacuo and suspended in H₂O. The precipitate was collected by centrifugation and washed with H₂O (4 mL). Flash chromatography (SiO₂, 0.7 x 7 cm, 1-5% MeOH-CHCI₃ gradient) afforded 24 (7.7 mg, 52%) as a light yellow solid: FABHRMS (NBA/Csl) m/z 526.1397 (M + H⁺, C₂₆H₂₄CIN₃O₇CI requires 526.1381); (+)-(1 S)-24: [α]²⁵ _D +6 (c 0.33, CHCI₃); (-)-(1 R)-24: [α]²⁵ _D-6 (c 0.33, CHCI₃).

Methyl 2-[(5,6,7-Trimethoxyindol-2-yl)carbonyl]-1 ,2,9,9a- tetrahydrocyclopropa[c]-pyrido[3,2-e]indol-4-one-7-carboxylate (25, CPyl-TMI). A solution of 24 (1.5 mg, 2.8 μmol, 1.0 equiv) in anhydrous DMF (0.10 mL) at 25 °C was treated with DBU (1.3 μL, 8.6 μmol, 3.0 equiv) and stirred 3 h under Ar. Direct chromatography of the reaction mixture (0.7 x 3 cm, 5% MeOH-CH₂CI₂) furnished 25 as a light yellow solid (1.2 mg, 86%): FABHRMS (NBA/Nal) m/z 490.1629 (M + H⁺, C₂₆H₂₃N₃O₇ requires 490.1614); (+)-(8bR, 9aS)-25: [α]²⁵ _D +42 (c 0.085, CH₂CI₂); (-)-(8bS, 9aR)-25: [α]²⁵ _D -44 (c 0.045, CH₂CI₂).

Methyl 1-(Chloromethyl)-5-hydroxy-3-[(5-methoxyindol-2-yl)carbonyl]- 1,2-dϊhydro- 3 Y-pyrido[3,2-e]indole-8-carboxylate (26). Flash chromatography (SiO₂, 0.7 6 cm, 1-5% MeOH-CHCI₃ gradient) afforded 26 (59%) as a yellow solid: FABHRMS (NBA/Nal) m/z 466.1186 (M + H⁺, C₂₄H₂₀CIN₃O₅ requires 466.1170); (+)-(1 S)-26: [ ]²⁵ _D +13 (c 0.16, CHCI₃); (-)-(1r?)-26: [α]²⁵ _D-12 (_c 0.27, CHCI₃).

Methyl 2-[(5-Methoxyindol-2-yl)carbonyl]-1,2,9,9a-tetrahydrocyclopropa- [c]pyrido[3,2-e]indol-4-one-7-carboxylate (27). Flash chromatography (0.7 x 3 cm, 5% MeOH-CH₂CI₂) furnished 27 as a light yellow solid (90%): FABHRMS (NBA/Nal) m/z 452.1209 (M + Na⁺, C₂₄H₁₉N₃O₅ requires 452.1222); (+)-(8bR, 9aS)-27: [ ]²⁵ _D +49 (c 0.07, CH₂CI₂); (-)-(8bS, 9aR)-27: [α]²⁵ _D -44 (c 0.055, CH₂CI₂). Methyl 1 -(Chloromethyl)-5-hydroxy-3-[(indol-2-yl)carbonyl]-1 ,2-dihydro-3H- pyrido[3,2-e]indole-8-carboxylate (seco-CPyl-indole, 28). Flash chromatography (SiO₂, 0.7 x 7 cm, 1-5% MeOH-CHCI₃ gradient) afforded 28 (71%) as a yellow solid: FABHRMS (NBA/Nal) m/z 436.1052 (M + H⁺, C₂₃H₁₈CIN₃O₄l requires 436.1064); (+)-(1 S)-28: [α]²⁵ _D +3 (c 0.35, CHCI₃); (-)-(1R)-28: [α]²⁵ _D-3 (c 0.29, CHCI₃).

Methyl 2-[(lndol-2-yl)carbonyl]-1,2,9,9a-tetrahydrocyclopropa[c]pyrido [3,2-e]indol- 4-one-7-carboxylate (CPyl-indole, 29). Flash chromatography (0.7 x 3 cm, 5% MeOH-CH₂CI₂) furnished 29 as a light yellow solid (93%): FABHRMS (NBA/Nal) m/z 400.1310 (M + H⁺, C₂₃H₁₇N₃O₄ requires 400.1297); (+)-(8bR, 9aS)-29: [ ] ⁵ _D +48 (c 0.065, CH₂CI₂); (-)-(8bS, 9aR)-29: [α]²⁵ _D -43 (c 0.065, CH₂CI₂).

Methyl 1-(Chloromethyl)-5-hydroxy-3-{[5-[N-(indol-2-yl)carbonyl] aminoindol-2-yl]-carbonyl}-1,2-dihydro-3H-pyrido[3,2-e]indole-8-carboxylate (seco-CPyl-indole₂, 30). Flash chromatography (SiO₂, 0.7 x 6 cm, 10% DMF- CHCI₃) afforded 30 (64%) as a yellow solid: FABHRMS (NBA/Csl) m/z 726.0548 (M + Cs⁺, C₃₂H₂₄CIN₅O₅ requires 726.0520); (+)-(1S)-30: [α]²⁵ _D +26 (c 0.105, DMF); (-)-(1r?)-30: [ ]²⁵ _D-28 (c 0.06, DMF).

Methyl 2-{[5-[N-(lndol-2-yl)carbonyl]aminoindol-2-yl]carbonyl}-1,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate (CPyl-indole₂, 31). Flash chromatography (SiO₂, 0.7 x 2 cm, 10% DMF-CH₂CI₂) afforded 31 (97%) as a yellow solid: MALDIHRMS (DHB) m/z 558.1757 (M + H⁺, C₃₂H₂₃CIN₅O₅ requires 558.1778); (+)-(1 S)-31 : [α]²⁵ _D +50 (c 0.04, DMF); (-)-(1 R)-31 : [α]²⁵ _D-46 (c 0.05, DMF). Methyl 3-[(3-Carbamoyl-1 ,2-dihydro-3H-pyrrolo[3,2-e]indol-7-yl)carbonyl]-1 - (chloromethyl)-5-hydroxy-1,2-dihydro-3iV-pyrido[3,2-e]indole-8-carboxylate (seco-CPyl-CDPI^ 32). Flash chromatography (SiO₂, 0.7 x 7 cm, 5% MeOH-10% DMF-CHCI₃) afforded 32 (41%) as a yellow solid: MALDIHRMS (DHB) m/z 520.1383 (M + H⁺, C₂₆H₂₂CIN₅O₅ requires 520.1388); (+)-(1 S)-32: [α]²⁵ _D +23 (c 0.04, DMF); (-)-(1R)-32: [ ]²⁵ _D-20 (c 0.06, DMF). Methyl 2-[(3-Carbamoyl-1 ,2-dihydro-3W-pyrrolo[3,2-e]indol-7-yl)carbonyl]- 1,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate (CPyl-CDPI^ 33). Flash chromatography (SiO₂, 0.7 x 4 cm, 5% MeOH-10% DMF-CH₂CI₂) afforded 33 (67%) as a yellow solid: MALDIHRMS (DHB) m/z 484.1613 (M + H⁺, C₂₆H₂₁N₅O₅ requires 484.1621); (+)-(1 S)-33: [α]²⁵ _D +50 (c 0.03, DMF); (-)-(1R)-33: [α]²⁵ _D-53 (c θ.03, DMF).

Aqueous Solvolysis of /V-BOC-CPyl and CPyl (pH 2 and pH 3, phosphate buffer). Samples of 16 (0.15 mg) and 17 (0.05 mg) were dissolved in CH₃OH (1.5 mL) and mixed with pH 3.0 buffer (1.5 mL, 4:1 :20 (v:v:v) 0.1 M citric acid, 0.2 M Na₂HPO₄, and H₂O, respectively). Similarly, samples of 16 (0.1 mg) and 17 (0.05 mg) were dissolved in CH₃OH (1.5 mL) and mixed with pH 2.0 buffer (1.5 mL, 4:1 :20 (v:v:v) 1.0 M citric acid, 0.2 M Na₂HPO₄, and H₂O, respectively). After mixing, the UV spectra of the solution was measured against a reference solution containing CH₃OH (1.5 mL) and the appropriate aqueous buffer (1.5 mL) and these readings were used for the initial absorbance values (/A,). The UV spectrum was measured at regular intervals for 30 d (16 at pH 3), 14 d (16 at pH 2), 40 d (17 at pH 3), and 8 d (17 at pH 2). For 16, the decrease in the long-wavelength absorption at 315 nm and increase in the short-wavelength absorption at 278 nm were monitored. The solvolysis rate constant and half-life (pH 3: k - 3.81 x 10^"6 s^"1, f_1/2 = 51 h, r = 0.99; pH 2: / = 3.72 ^χ 10^"5 s^'1, t_V2 - 5.2 h, r= 0.99) were calculated from the least-squares treatment of the slope of the plot of time versus ln[(/4_f - A J(A_f - A)] (Figure 10). For 17, the decrease in the long-wavelength absorption at 362 nm and increase in the short-wavelength absorption at 290 nm were monitored. The solvolysis rate constant and half-life were calculated by the same treatment providing k = 6.27 x 10^"7 s^"1 (t_υ2 = 310 h, r = 0.99) for pH 3 and k = 4.63 x 10^"6 s ¹ (t_V2 = 42 h, r= 0.99) for pH 2.

Acid-Catalyzed Addition of CH₃OH to W-BOC-CPyl: Methyl 3-(ferf-butyl- oxycarbonyl)-5-hydroxy-1 -(methoxymethyl)-l ,2-dihydro-3H-pyrido[3,2-e]indo le-8- carboxylate (41). A solution of 16 (3.0 mg, 8.4 μmol, 1.0 equiv) in CH₃OH (0.43 mL) was treated with CF₃SO₃H (0.13 μL, 2.5 μmol, 0.3 equiv) at 25 °C. After 20 h, the reaction was quenched by the addition of NaHCO₃ (9 mg), filtered through Celite, and concentrated in vacuo. Chromatography (SiO₂, 0.7 7 cm, 5- 50% EtOAc-CHCI₃ gradient) afforded 41 (3.0 mg, 91 %) as a yellow solid:

FABHRMS (NBA/Nal) m/z 389.1706 (M + H⁺, C₂₀H₂₄N₂O₆ requires 389.1713).

Addition of HCI to N-BOC-CPyl. A solution of 16 (2.3 mg, 6.5 μmol, 1.0 equiv) in THF (0.15 mL) was cooled to -78 °C and treated with 4 M HCI-EtOAc (3.0 μL, 11 μmol, 1.7 equiv). The mixture was stirred for 2 min before the solvent was removed in vacuo. Chromatography (SiO₂, 0.7 x 7 cm, 10% EtOAc-CHCI₃) afforded 15 (2.4 mg, 96%). Aqueous Solvolysis of W-BOC-CPyl (pH 2-11, universal buffer). Samples of 16 (0.025 mg) were dissolved in CH₃OH (1.0 mL), and the resulting solutions were mixed with a universal aqueous buffer (Perrin, D. D.; Dempsey, B. Buffers for H and Metal Ion Control; Chapman and Hall: London, 1979; p 156) (pH 2-11 , 1.0 mL, B(OH)₃-citric acid-Na₃PO₄). After mixing, the UV spectra of the solution were measured against reference solutions and these readings were used for the initial absorbance values ( ). The UV spectrum was measured at regular intervals (pH 2-4: every hour for 1d and then every 24 h; pH 4-10: every 24 h) until no further change in absorbance was observed (>4_f). The decrease in the long-wavelength absorption at 320 nm and increase in the short-wavelength absorption at 278 nm were monitored. The solvolysis rate constants and half-lives were calculated from the least-squares treatment of the slope of the plot of time versus ln[(>4_f - AV(A - A)].

Lewis Acid-Catalyzed Addition of CH₃OH to Λ/-BOC-CPyl. A solution of 16 (1.2 mg, 3.4 μmol, 1.0 equiv) in CH₃OH (0.14 mL) was treated with Zn(OTf)₂ (1.4 mg, 3.9 μmol, 1.1 equiv) at 25 °C. After 4 h, the solvent was removed with a stream of N₂. Flash chromatography (SiO₂, 0.7 x 7 cm, 20-50% EtOAc-CHCI₃ gradient) afforded 41 (1.2 mg, 92%).

Metal-Catalyzed Solvolysis of /v-BOC-CPyl. Samples of 16 (0.025 mg) were dissolved in CH₃OH (1.9 mL) and the resulting solutions were treated with 125 μL (1.0 equiv) or 25 μL (0.2 equiv) of a 0.56 mM solution (CH₃OH) of the desired metal (Cu(acac)₂, Mg (acac)₂, Ni(acac)₂, Zn(acac)₂, Mn(acac)₂, Mg(acac)₂, Fe(acac)₃, Cr(acac)₃, Zn(OTf)₂, Ti(O/-Pr)₄, or Cu(OMe)₂). The solvolysis solution was sealed and kept at 25 °C protected from light. After mixing, the UV spectra of the solution were measured against reference solutions and these readings were used for the initial absorbance values (A . The UV spectrum was measured at regular intervals until no further change in absorbance was observed (>4_f)- The decrease in the long-wavelength absorption at 330 nm and increase in the short-wavelength absorption at 270 nm were monitored. The solvolysis rate constants and half-lives were calculated from the least-squares treatment of the slope of the plot of time versus ln[(>4_f - >4i)/(>4_f - A)].

Claims

What is claimed is:

1. A DNA alkylating agent represented by the following structure:

wherein R is a DNA minor groove binder.

2. A DNA alkylating agent according to claim 1 represented by the following structure:

3. A DNA alkylating agent according to claim 2 wherein R is selected from a group of DNA minor groove binders represented by the following structures:

4. A DNA alkylating agent according to claim 3 represented by the following structure:

5. A DNA alkylating agent according to claim 3 represented by the following structure:

6. A DNA alkylating agent according to claim 3 represented by the following structure:

7. A DNA alkylating agent according to claim 3 represented by the following structure:

8. A DNA alkylating agent according to claim 3 represented by the following structure:

9. A DNA alkylating agent according to claim 1 represented by the following structure:

wherein R is selected from a group of DNA minor groove binders represented by the following structures:

10. A DNA alkylating agent represented by the following structure:

wherein R is a DNA minor groove binder.

11. A DNA alkylating agent according to claim 10 represented by the following structure:

12. A DNA alkylating agent according to claim 11 wherein R is selected from a group of DNA minor groove binders represented by the following structures:

13. A DNA alkylating agent according to claim 12 represented by the following structure:

14. A DNA alkylating agent according to claim 12 represented by the following structure:

15. A DNA alkylating agent according to claim 12 represented by the following structure:

16. A DNA alkylating agent according to claim 12 represented by the following structure:

17. A DNA alkylating agent according to claim 12 represented by the following structure:

18. A DNA alkylating agent according to claim 13 represented by the following structure:

wherein R is selected from a group of DNA minor groove binders represented by the following structures:

19. A process for catalyzing a solvolysis of a cyclopropyl ring of an N - derivative of methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7- carboxylate, the process comprising the following step:

contacting the N ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate under aqueous conditions having a pH greater than 4 with a catalytic concentration of a metal ion sufficient to catalyze the solvolysis of the cyclopropyl ring of the Λ/ ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate, the metal ion being selected from a group consisting of Cu , Ni , Zn , Cr , Fe , Cr³⁺, Fe³⁺, Mn²⁺, and Mg²⁺.

20. The process for catalyzing a solvolysis of the cyclopropyl ring of an N ²- derivative of methyl 1 ,2,9,9a-tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7- carboxylate according to Claim 19 wherein the metal ion is Zn²⁺.

21. A process for catalyzing the production of a DNA alkylation product, the process comprising the following step:

contacting DNA with an N ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate under aqueous conditions having a pH greater than 4 in the presence of a catalytic concentration of metal ion sufficient to catalyze the alkylation of the DNA by the N ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate for producing the DNA alkylation product, the metal ion being selected from a group consisting of Cu²⁺, Ni²⁺, Zn²⁺, Cr²*, Fe²⁺, Cr³⁺, Fe³⁺, Mn²⁺, and Mg²⁺.

22. The process for catalyzing the production of a DNA alkylation product according to Claim 21 wherein the metal ion is Zn²⁺.

23. A DNA alkylation product produced according to the method of claim 21.

24. A DNA alkylation product produced according to the method of claim 22.

22. A process for catalyzing cell death by DNA alkylation, the process comprising the following step:

contacting a cell, under aqueous conditions having a pH greater than 4, with a concentration of an N ²- derivative of methyl 1 ,2,9,9a- tetrahydrocyclopropa[c]pyrido[3,2-e]indol-4-one-7-carboxylate sufficient, in the presence of a catalytic concentration of metal ion, to catalyze cell death by DNA alkylation, the metal ion being selected from a group consisting of Cu²⁺, Ni²⁺, Zn²⁺, Cr²*, Fe²⁺, Cr³⁺, Fe³⁺, Mn²⁺, and Mg²⁺.

23. A process for catalyzing cell death by DNA alkylation according to Claim 22 wherein the metal ion is Zn²⁺.