criticisms of Bianconi's work were "both accurate and necessary. His reward for carrying on this robust debate, which is absolutely essential to the progress of science, was to be dragged through the courts for three years with claims of defamation and evil motive." But according to Bianconi: "There is also sworn testimony from materials scientists that my science was accurate and correct. Several people testified that the work was valid, that this criticism was unwarranted, and that it was devastating to my career. This was no robust debate. It was a simple breach of contract." Stu Borman
a process of shape-selective recognition. DNA alkylation is then catalyzed by a DNA-binding-induced conformational change that twists a key vinylogous amide group in the agents, disrupting their stability and activating them for the DNA alkylation reaction. Activation occurs preferentially in the narrower, deeper adenine- and thymine-rich minor groove, where the degree of twist in the amide group is greatest—an example of shape-dependent catalysis. The mechanism implies that the drugs' reactions with other nucleophiles will be minimal, says Drugs Boger, "because the agents are preferentially activated for alkylation only once they reach their target." Chemistry professor Marc M. Greenberg of Colorado State University, Fort Collins, whose research focuses on mechanisms of nuChemistry professor Dale L. Boger of cleic acid damage, comScripps Research Institute, La Jolla, Calif., ments that the work by believes he and his coworkers have es- Boger and coworkers reptablished the origin of the sequence se- resents "an impressive comlectivity of CC-1065 and the duocarmy- bination of sophisticated cins—members of a family of antitumor organic synthesis and physantibiotic drugs that alkylate DNA. In ad- ical organic chemistry emdition, he says his group has identified a ployed in concert to adpreviously unrecognized source of catal- dress the mechanism of acysis for the DNA alkylation reaction. tion of the duocarmycins." Boger's interpretation of his results is And chemistry professor not shared, however, by medicinal chem- Peter B. Dervan of California istry professor Laurence H. Hurley of the Institute of Technology, who University of Texas, Austin. Hurley be- specializes in sequence-spelieves the basis for the drugs' selectivity cific recognition of DNA, for specific DNA sequences was estab- says, "Using formidable lished in a set of papers published by his synthetic skills, the Scripps group about 10 years ago, some of team has provided deep which were based on collaborative work mechanistic insight on how with scientists at Upjohn (now Pharma- a DNA binding event in the minor groove cia & Upjohn). activates a covalent reaction." CC-1065 and the duocarmycins are However, Hurley and coworkers have structurally similar polycyclic natural prod- long supported an alternative mechanism: ucts. The discovery of CC-1065, by Up- the alkylation site model [Biochemistry, john researchers, was first reported in 27, 3886 (1988) a n d / Am. Chem. Soc, 1979, and the X-ray crystal structure of the 112, 4633 (1990)]. According to this modcompound was obtained two years later. el, DNA sequences likely to react with the The two classes of duocarmycins (A and drug are "those capable, at the least cost SA) were discovered in 1988 and 1990 by of distortion energy, of adopting an 'acresearchers at Kyowa Hakko Kogyo Co., tive conformation' required for covalent Tokyo. reaction," says Hurley. "In the most obviBoger and coworkers have proposed ous situation, the conformational change a mechanism of DNA recognition of CC- brings the nucleophilic atom on DNA 1065 and the duocarmycins that they re- within bonding distance of the electrofer to as the noncovalent binding model philic center on the drug." \J. Am. Chem. Soc, 119, 4977 and 4987 The reaction is catalyzed when a phos(1997)]. The model holds that the drugs phate in the DNA backbone protonates a bind noncovalently to adenine- and thy- carbonyl in the drugs' alkylation subunit, mine-rich sites in DNA's minor groove by activating it for nucleophilic addition.
"Only selective sequences are predisposed to carry out this catalysis," says Hurley. Boger believes duocarmycin analogs show significant promise as anticancer agents. He points out that a duocarmycin derivative is now in Phase II clinical trials, and that Immunogen, Norwood, Mass., is proposing clinical trials for an antibody-linked conjugate of a CC-1065 analog synthesized by his group. Hurley, on the other hand, believes DNA-DNA cross-linkers derived from CC-
alkylate DNA
Researchers agree to disagree on mechanism of anticancer drugs
1065 are more promising. "Bizelesin, a DNA-DNA interstrand cross-linker designed and synthesized based upon our understanding of how (+)-CC-1065 recognized DNA, is now in Phase I clinical trials in the U.S.," he says. This compound has a unique mechanism of action and molecular mechanism for sequence recognition of DNA that can be predicted from the alkylation site model. Hurley and Boger remain divided on the fundamental issue of how CC-1065 and the duocarmycins alkylate DNA, with no immediate resolution in sight. But there is one point on which they agree. The issue "of the origin of sequence selectivity is a debate that's been going on for 10 years, and it appears it will continue for some time to come," says Boger. Stu Borman NOVEMBER 17, 1997 C&EN 37
