Proc. Natl. Acad. Sci. USA Vol. 90, pp. 5881-5888, July 1993 Review Chemistry and biology of natural and designed enediynes K. C. Nicolaou*, A. L. Smitht, and E. W. Yue Department of Chemistry, The Scripps Research Institute, 10666 North Torrey Pines Road, La Jolla, CA 92037; and Department of Chemistry, University of California at San Diego, La Jolla, CA 92093
ABSTRACT Ever since the initial re- ports of the enediyne anticancer antibiot- ics in the late 1980s, researchers from a number of disciplines have been devoting increasing attention to their chemistry, biology, and potential medical applica- tions. Synthetic chemists and molecular designers have been engaged in attempts to synthesize these molecules and to model their unique architecture. Considerable efforts have been directed at understand- ing and mimicking the various processes involved in the targeting, activation, and DNA cleavage associated with these natu- ral products. This review summarizes the main contributions to the field, with par- ticular emphasis on work from our labo- ratories. Highlights include studies of the Bergman reaction, which is central to the mechanism of action of enediynes, the design and chemical synthesis ofa number of these systems, and biological studies with selected molecules. Finally, the total synthesis of calicheamicin yI, the most 8: kedarcidin chromophore 9: neocarzinostatin chromophore prominent member of this class of natu- rally occurring compounds, is discussed. FIG. 1. Naturally occurring enediyne anticancer antibiotics. omatization when heated to give the ben- nostatin chromophore (9, Fig. 1) (12-14) Nature continually serves to provide sci- zene ring 4 with the intermediacy of the has also been grouped with the enediyne entists with new ideas, often provoking highly reactive 1,4-benzenoid diradical antibiotics due to the similarity of its us to reexamine previous studies and structure and mode of action. leading us in new directions. An example species 2. Masamune and coworkers (3) of an instance which has and Wong and Sondheimer (4) also ob- such captured served the cycloaromatization of 10- Molecular Structures, Biological the imagination of many scientists, our- and Mechanisms of Action of selves included, came in 1987 with the membered ring enediynes. However, the Properties, unveiling of a new class of natural prod- full significance of these simple observa- Naturally Occurring Enediynes ucts, the so-called "enediyne anticancer tions became apparent only when the antibiotics" (Fig. 1), which possess po- structures of calicheamicin -y (5, Fig. 1) Calicheamicin y1 (5), the most prominent tent anticancer activity and a hitherto and esperamicin A1 (6, Fig. 1), the first member of the calicheamicins, was iso- unseen biological mode of action (for a representatives of the enediyne antibiot- lated from Micromonospora echinospora previous review, see ref. 1). The story ics, were reported (5-8). For at the very ssp. calichensis and is a remarkable piece began some 15 years earlier, however, heart of these molecules is contained the of engineering by Nature, which has per- for it was in the laboratories of Robert same enediyne unit; and therein lies the fectly constructed the molecule to endow Bergman that the intriguing chemical means by which these unprecedented it with its extraordinary chemical and processes which provide the key to un- molecules exert their remarkable biolog- biological properties (15). These include derstanding the remarkable activity of ical properties. Dynemicin A (7, Fig. 1) activity in the biochemical prophage in- these antibiotics were first studied (2). (9, 10) and kedarcidin chromophore (8, duction assay at concentrations < 1 pM, The findings of Bergman are summarized Fig. 1) (11), representing structurally dis- high antibacterial activity, and extreme in Scheme I, in which the enediyne sys- tinct classes of enediyne antibiotics, potency against murine tumors such as tem 1 was observed to undergo cycloar- were subsequently reported. Neocarzi- P388 and L1210 leukemias and solid neo- plasms such as colon 26 and B16 mela- noma with optimal doses of 0.15-5 ug/ DH kg. Calicheamicin -yl (5), along with the *To whom reprint requests should be ad- 2 4 dressed. tPresent address: Merck Sharp & Dohme Re- SCHEME I. Bergman's original design and observation regarding the thermal cycloaroma- search Laboratories, Terlings Park, Harlow, tization of enediynes. Essex CM20 2QR, U.K. 5881 Downloaded by guest on September 24, 2021 5882 Review: Nicolaou et al. Proc. Natl. Acad. Sci. USA 90 (1993) trigonal bridgehead position to a tetrag- onal center, causes a significant change in structural geometry which imposes a great deal of strain on the 10-membered ring. This strain is completely relieved by the enediyne undergoing the Bergman reaction, generating a highly reactive benzenoid diradical (12). The calicheam- icin diradical abstracts hydrogen atoms from duplex DNA at the C-5' position of the cytidine in 5'-TCCT-3' and the C-4' position ofthe nucleotide three base pairs removed on the 3' side of the comple- mentary strand (21, 22), leading to cleav- age of both strands of DNA. The mech- anism of action ofthe structurally similar esperamicins is thought to be essentially the same as that of the calicheamicins. Dynemicin A (7), the first member of FIG. 2. Computer-generated model of DNA-bound calicheamicin -yI (5) along a TCCT site. the dynemicin subclass of enediyne anti- [Reprinted with permission from ref. 1 (copyright VCH, Weinheim, FRG).] biotics to be reported (9, 10), was discov- other enediyne antibiotics, is believed to iodo substituent of the hexasubstituted ered in the fermentation broth of Mi- exert its biological activity by aromatic ring and the cromonospora chersina. It exhibits very damaging exocyclic amino potent activity against a variety ofcancer DNA. Indeed, it is a highly potent DNA- substituents of the two guanines in the cell lines and significantly prolongs the cleaving agent giving rise primarily to 3'-AGGA-5' tract. Experimental evidence lifespan of mice inoculated with P388 sequence-selective double-strand cuts from our own laboratories supports the leukemia and B16 melanoma cells. Fur- (16). idea of an important role for the iodine in thermore, it exhibits promising in vivo The calicheamicin yA molecule contains the binding affinity ofthis oligosaccharide antibacterial activity with low toxicity. two distinct structural regions. The larger to DNA (20). Like the calicheamicins and esperami- of the two consists of an extended sugar The second region of calicheamicin yi cins, the dynemicins include in their mo- residue comprising four monosaccharide is its aglycon, a rigid, highly functional- lecular structures a 10-membered ring units and., one hexasubstituted benzene ized bicyclic core termed calicheamici- with a 1,5-diyne-3-ene bridge. They are, ring which are joined together through a none, which acts as the "warhead" ofthe however, unique in combining the highly unusual series of glycosidic, molecule. The enediyne functionality of enediyne unit with the anthraquinone thioester, and hydroxylamine linkages. It the molecule is locked within a rigid chromophore of the anthracycline anti- is this aryloligosaccharide which serves to 10-membered bridged ring awaiting acti- cancer antibiotics. deliver the molecule to its target, binding vation to undergo the Bergman reaction. A mechanism for the antitumor activ- tightly in the minor groove of double- Also forming part of the aglycon is a ity of dynemicin A (7) has been proposed helical DNA and displaying high specific- trisulfide, which serves as the trigger. which combines elements of the mecha- ity for sequences such as 5'-TCCT-3' and Once the molecule is in the vicinity of nisms of action of the calicheamicin/ 5'-TTTT-3' (16, 17) through significant DNA (whether prior or subsequent to esperamicin, neocarzinostatin, and an- hydrophobic interactions and other forces binding is not known), a series of chem- thracycline classes of antibiotics and (Fig. 2). This binding is thought to be ical events unfold which ultimately leads which is supported by the observation facilitated by substantial preorganization to DNA cleavage (Scheme II) (5-8). A that DNA strand cleavage by dynemicin of the oligosaccharide into a rigid, ex- nucleophile (e.g., glutathione) attacks the A is enhanced by the presence of thiols. tended conformation (18). Molecular central sulfur atom of the trisulfide In this mechanism (Scheme Im) (10), the modeling calculations by Schreiber and group, causing the formation of a thiolate anthraquinone nucleus intercalates into coworkers (19) suggested that a signiflcant or a thiol (10) which adds intramolecu- the DNA and undergoes bioreduction portion of the sequence selectivity for larly to the adjacent a,,4unsaturated ke- (not necessarily in that order), facilitating 5'-TCCT-3' arises from a favorable inter- tone embedded within the framework of opening of the epoxide. This causes a action between the large and polarizable the aglycon. This reaction, converting a significant conformational change in the molecule and introduces considerable strain into the enediyne system which is relieved by the new system undergoing the Bergman reaction, thus generating the DNA-damaging diradical species 17 (10). Molecular Design, Chemical Synthesis, and Biological Actions of Enediynes The reports of the structures of the cali- cheamicins and esperamicins, and their fascinating mode of action, prompted several groups to undertake investigation of the Bergman reaction in simple cyclic systems. The synthesis of the monocy- clic 10-membered ring enediyne and its SCHEME H. Mechanism by which calicheamicin yI (5) cleaves DNA. Nu, nucleophile. homologues was the initial focus of a Downloaded by guest on September 24, 2021 Review: Nicolaou et al. Proc. Natl. Acad. Sci. USA 90 (1993) 5883 strated that the crucial factor in deter- mining the ease with which a particular Anthraquinone nucleus system undergoes the Bergman reaction intercalates OH ~OH HN is the relative strain energies of the into DNA and Me ground and transition states for the reac- undergoes bioreduction tion. These findings should always be OH OH OH borne in mind when the empirical cd distance rule is applied. Given that the simple 10-membered ring enediyne 20b underwent the Berg- man reaction at physiological tempera- tures at a reasonable rate, we proceeded to attempt to mimic the DNA-cleaving action of the calicheamicins and espe- ramicins by using such simple systems. The diol 22 (Scheme V) was designed in order to endow the molecule with some degree of water solubility and also to provide for the option of attachment to delivery systems (24, 27). It was cor- rectly predicted from the calculated cd distance of 3.20 A that this molecule would be sufficiently stable for isolation and handling at ambient temperatures but DNA double would undergo the Bergman reaction at strand cleavage physiological temperature at a sufficient rate to cause DNA cleavage. Thus SCHEME Iml. Mechanism by which dynemicin A (7) cleaves DNA. enediyne 22 caused significant cleavage
ci MeU-C. of phage 4X174 double-stranded super- coiled DNA in the absence of any addi- H2n -W- CH)n 0 0~~ \- J tives at as at - concentrations as low 10 ,uM
1ga-h n = 1-8 20a-h 21a-h 37°C, with the extent of cleavage being dependent upon concentration, incuba- SCHEME IV. Synthesis and cycloaromatization of monocyclic enediynes. tion time, and temperature (24, 27). As a control, it was demonstrated that the program in these laboratories (23, 24). underwent the Bergman reaction (Table corresponding Bergman cyclized product The parent series of 10- through 16- 1) showed a clear trend in which a de- 24 (Scheme V) caused no DNA cleavage membered ring enediynes (20b-h) were creased cd distance reflected, in addition (24, 27). These molecules thus consti- conveniently prepared via the Ramberg- to closer intimacy between the acetylenic tuted the first designed DNA-cleaving Backlund reaction of the corresponding groups, an increased ring torsion and agents based upon the mechanism of ac- a-chlorosulfones 19b-h (Scheme IV). hence an increased tendency to undergo tion of the calicheamicins/esperamicins. The 10-membered ring enediyne 20b the Bergman reaction in order to relieve Later on, the thermally reactive diols 25 readily underwent the Bergman reaction the strain. For these simple systems a and 26 were similarly demonstrated to at room temperature with a half-life of 18 critical upper limit for the cd distance of effect DNA cleavage whereas the con- hr (Table 1), while the larger ring around 3.2-3.3 A appeared to be required formationally locked and thermally sta- enediynes (20c-h) were found to be sta- for the Bergman reaction to occur at a ble derivative 27 (Fig. 3) failed to cleave ble. By contrast, the 9-membered ring measurable rate at ambient tempera- DNA (28). Under basic conditions, how- 20a could not be prepared, although tures. This finding provided a useful ever, compound 27 became active via the products formally arising from a Berg- means of predicting the thermal stability release of system 26 thus exhibiting both man reaction were identified. Compari- of compounds in subsequent work in our DNA cleavage and cytotoxic properties. son of the distances cd between the ter- laboratories. More sophisticated calcula- Since the naturally occurring enediyne mini of the enediyne moiety of these tions and kinetics experiments by Snyder antibiotics are triggered to exert their systems and the ease with which they (25) and Magnus et al. (26) later demon- biological actions by bioreductive pro- cesses, a system was designed to control Table 1. Calculated cd distances and stabilities of cyclic enediynes the Bergman cyclization by a hydroqui- Compound n Ring size cd distance, A Stability none = quinone redox process (29). It was postulated that hydroquinone sys- 20a 1 9 2.84 Unknown tems such as 28 should be more stable 20b 2 10 3.25 ti/2= 18 hr at 250C toward cycloaromatization than the cor- 20c 3 11 3.61 Stable at 25°C responding quinones 29 and 30 (Fig. 3). 20d 4 12 3.90 Stable at 250C This was indeed found to be the case. 20e 5 13 4.14 Stable at 250C Thus the hydroquinone 28 had a half-life 20f 6 14 4.15 Stable at 25°C of 74 hr at 110°C, while 29 and 30 had 20g 7 15 4.33 Stable at 25°C half-lives of 2.6 hr and 32 min, respec- 20h 8 16 4.20 Stable at 25°C tively at 55°C. These results were further 22 10 3.20 ti/2 = 11.8 hr at 370C emphasized by the finding that, while 28 25 10 3.29 ti/2= 4 hr at 500C exhibited no DNA-cleaving properties, 26 10 3.34 ti/2 = 2 hr at 50°C 29 and 30 were able to cause DNA dam- 27 10 3.42 Stable at 250C age and death of tumor cells. Downloaded by guest on September 24, 2021 5884 Review: Nicolaou et al. Proc. Natl. Acad. Sci. USA 90 (1993)
DNAO 0 0 XOH 37°C [OH ___L N OH SAc OH OH I OH HO OH H OH DNA 22 23 cleavage 24
SCHEME V. Enediyne 22 as a designed DNA-cleaving agent. 35 (Danishefsky) 36 (Kende)
0 Me Me Me Me Me Me SSSBn K,O%OH <,kYOH HO" H Q0° TBSO H...>'N. OTBS -
-- ber of normal cell lines. coworkers (34, 35) which utilized acety- Scheme VII (41/42 -+ 43/44 -* 45/46 Furthermore, lene cobalt complexes as intermediates 47/48 -+ 49/50). The intermediacy of the preliminary in vivo studies using 40 and a to arrive at the targeted enediynes. highly reactive 1,4-benzenoid diradicals tritiated analogue of40 with mice infected When the structure and mode ofaction 47 and 48 is presumably responsible for with leukemia and solid tumors showed of dynemicin A (7) were reported in 1989 the high cytotoxicity of these com- encouraging results (unpublished results). (9), several groups, including ours, un- pounds, although the precise target of Interestingly, treatment of MOLT-4 dertook programs directed toward the this species within the cell has not been cells under appropriate conditions with design and synthesis of models which defined unambiguously. As expected, enediyne 40 followed by observation of cell morphology and cell death revealed the phenomenon of programmed cell death (apoptosis) (48) as the prevailing cause of cell destruction (49). Further- more, competition experiments using enediynes with relatively low toxicities resulted in the identification of certain 31 32 inhibitors of apoptosis. Specifically, the methoxy enediyne 55 (Fig. 7), which dis- played diminished tendency to undergo 0 the Bergman reaction, exhibited inhibi- tion of the cytotoxic action of compound 40. Thus, when MOLT-4 cells were pre- OH OMe 34 33 incubated with enediyne 55 at 0.1 mM for 1 hr prior to treatment with the cytotoxic SCHEME VI. Myers' approach to redox activation of enediyne systems. compound 40, a dramatic reduction (fac- Downloaded by guest on September 24, 2021 Review: Nicolaou et al. Proc. Natl. Acad. Sci. USA 90 (1993) 5885 r hS"
Base: J cd = 3.64A R 51 39: R = H 40: R = OCH2CH20H
0. ~0' Ph'.1 J0' Ph" -' .O N HN DNA
Nu DNA 53 54 cleavage R 050 0OH 00 H H - cd= 3.10 3.20A h"SIPh0 0 Ph Y,)< OAN0 49: R = H 47: R = H 45: R = H ~~~~~~~R,R, 50: R = OCH2CH20H 48: R = OCH2CH20H 46: R = OCH2CH20H
SCHEME VII. Design of dynemicin models with triggering devices and their postulated 55 56: RI =Me; R2=H mechanism of action. 57: RI = H; R2 MO 58: RI = R2 = Me tor of _105) in the cytotoxicity of 40 was (46). These experiments demonstrated FIG. 7. Designed enediynes with triggering observed. Similar reductions (factors of dramatic differences in potencies de- devices and modulated reactivity. 1O2-104) were observed in the cytotoxic- pending on the enantiomeric form of the ities of the naturally occurring enediynes enediyne 39 and the degree and stereo- Fig. 8) (52), Magnus (61, Fig. 8) (53), and dynemicin A (9) and calicheamicin 'y (5) chemistry of methyl substitution in com- Isobe (62, Fig. 8) (54). in the presence of the methoxy enediyne pounds 56-58 and raised intriguing ques- Inspired by the molecular architecture 55. Particularly intriguing was the obser- tions: Is there an intracellular receptor and the mechanism ofaction of neocarzi- vation that 55 inhibited apoptotic mor- for these enediynes other than DNA? nostatin chromophore (9), a number of phology of cell death by powerful induc- Could this putative receptor serve as a have ers of apoptosis such as actinomycin D capturing and delivery system for these groups designed, synthesized, and and cycloheximide, although cell viabil- enediynes to specific sequences ofDNA? studied a variety of model systems. A ity was not affected in these cases (49). Is there a prevailing biological mecha- selection ofthese model compounds cov- Further insight into the remarkable nism in certain cell types which facilitates ering the rather wide spectrum of struc- cell-type selectivity of these designed the /3elimination? These questions raise tural types synthesized in this field is enediynes was obtained by comparing even more interesting issues concerning shown in Fig. 9 [63 (55), 64 (56), 65 (57), the cytotoxicities of enantiomerically the regulation of cell death. Exploitation 66 (58), 67 (59), and 68 (60)]. pure compounds (+)-39 and (-)-39 of such observations and elucidation of (Scheme VII) (50) and the methyl- the mechanism of action of these agents Designed Enediynes Tethered to Delivery substituted compounds 56-58 (Fig. 7) may lead to new approaches to drug Systems design. pH 6.0 pH 7.0 pH 7.4 pH 8.5 a r-*i n-i r--i r-i_ In addition to our own efforts in the Compared with the naturally occurring compound: - 40 - 40 - 40 - 40 dynemicin area, a number of other compounds (e.g., calicheamicin Ay), the groups have contributed to the design designed enediynes reported so far ex- I1: and synthesis of model systems. Notable hibit significantly lower potencies as among these contributions are those of DNA-cleaving agents. In the absence of lane: 1 2 3 4 5 6 7 8 Schreiber (59, Fig. 8) (51), Wender (60, suitable delivery systems these observa- b compound: - 41 39 40 55 54 51 49 Table 2. Cytotoxicities of designed enediyne 40 against a panel of tumor cell lines and normal cell lines III Cell type Cell line IC50, M lane: 1 2 3 4 5 6 7 8 Tumor cell lines Melanoma M-14 1.6 x 10-6 FIG. 5. (a) DNA cleavage by enediyne 40. Colon carcinoma (b) DNA cleavage by enediynes 39-41, 51, 54, HT-29 1.6 x 10-6 and 55 and Bergman product 49. For condi- Ovarian carcinoma Ovcar-3 7.8 x 10-7 tions, see ref. 42. [Reproduced with permis- Breast carcinoma MCF-7 7.8 x 10-7 sion from ref. 42 (copyright 1992, AAAS, Lung carcinoma UCLA P-3 9.8 x 10-8 Washington, DC).] Pancreatic carcinoma Capan-1 3.1 x 10-9 T-cell leukemia TCAF 1.1 x 10-9 1 2 3 4 5 6 TCAF-DAX* 1.7 x 10-9 Promyelocytic leukemia HL-60 3.6 x 10-11 Form II 10-1 Form IlIl T-cell leukemia MOLT-4 2.0 x Form I Normal cell lines Bone marrow HNBM 5.0 x Human mammary epithelial cells HMEC 6.3 x 10-6 FIG. 6. DNA cleavage by enediyne 42. For Normal human dermal fibroblast NHDF 5.0 x 10-6 conditions see ref. 47. with [Reproduced per- Chinese hamster ovary CHO 3.1 x mission from ref. 47 (copyright Am. Chem. 10-6 Soc., Washington, DC).] *Multiple-drug-resistant cell line. Downloaded by guest on September 24, 2021 5886 Review: Nicolaou et al. Proc. Natl. Acad. Sci. USA 90 (1993) tions are not surprising, but they did cule would be coupled together in the oftheir thoughts. This led them to under- prompt several attempts to improve the final stages of the synthesis (Fig. 11). take several model studies for the con- DNA-cleaving properties of these mole- However, degradation studies on the nat- struction of the B ring which culminated cules by tethering them to suitable mol- ural product by other groups had failed to in providing a strategy (Scheme VIII) for ecules known to bind DNA. The hybrid furnish either the intact aglycon or oligo- the pivotal step of the synthesis of the molecules shown in Fig. 10 are represen- saccharide due to their inherent instabil- oligosaccharide fragment (65). Thus, tative examples ofthe work carried out in ities, and so there would be no opportu- heating the thionoimidazolide 75 in re- this area [69 (1), 70 (1), 71 (61), 72 (62), 73 nity for establishing the correct condi- fluxing toluene effected a clean allylic (63), and 74 (64)]. Despite these efforts, tions necessary for achieving this final transposition of functionality on the B however, the increase in potency ofthese union of the two domains without first ring via a Ferrier-like rearrangement as agents as DNA cleavers is not as dra- carrying out the mammoth task of their illustrated by the arrows in Scheme VIII. matic as expected. Reasons for the lack si- of success in this area so far may be due separate syntheses. This [3,3] sigmatropic rearrangement to the requirement for precise docking of In tackling the synthesis of the oligo- multaneously introduced the 4-thio sub- the enediyne into the DNA site in order saccharide portion of calicheamicin yly, stituent at C-4 with the correct stereo- for the generated radicals to abstract hy- Nicolaou and coworkers were fortunate chemistry and deoxygenated the 2-posi- drogen from the DNA backbone. As in in being able to draw upon their experi- tion. Nicolaou et al. (65) were thus able to the case of the natural enediynes, much ence in the field of sugar chemistry, en- report the first total synthesis of the oli- fine tuning and evolution needs to be abling them to recognize the most chal- gosaccharide portion of calicheamicin VyI done before we can reach sophisticated lenging features of the fragment and de- toward the end of 1990. Significantly, the molecular designs with highly potent and sign their synthetic strategy around the efficient and convergent nature of this efficient DNA-cleaving profiles. Such de- solution of these problems. They per- strategy provided a means of readily ob- signs may be accelerated by molecular ceived that the B ring (Fig. 11), with its taining the multi-gram quantities of suit- modeling and human ingenuity. synthetically demanding 2-deoxy-,8- ably advanced precursors to the oligo- hydroxylamino glycosidic linkage and saccharide which would be needed in Total Synthesis of Naturally Occurring 4-thio substituent, should be at the center attempts to reach calicheamicin yI itself. Enediynes: Calicheamicin yi OCOtBu c When the structures of calicheamicin iy, (5) and esperamhicin A1 (6) were revealed DLOH LY=S=. OH to the scientific community, our group, HO 63 (Myers) 64 (hNicolaou) 65 (Nicolaou) like many other synthetic groups around 0 the world, recognized the formidable Me PPh2 synthetic challenge presented by these HO'. M unprecedented molecular systems and 66(Hirama) 67T(S7D + OH undertook the task of achieving the total OH synthesis of calicheamicin Vyi Such a 66(Hirama) 67 (T7oshima) 68 (Nicolaou) daring endeavor would require careful planning in order to overcome the con- FIG. 9. Selected neocarzinostatin models and related systems. siderable difficulties inherent in con- structing a molecule of such complexity. The total synthesis of calicheamicin VyI would not only require control of the absolute stereochemistry at 19 chiral cen- ters but also require insight into the po- tential instability and reactivity of the multitude of functionalities which com- prise the molecule, in order to judge the correct timing for their introduction. Not unreasonably, the strategy adopted by our group was one in which suitably advanced precursors to the aglycon and oligosaccharide fragments of the mole-
MeO N0.N
OH
59 (Schreiber) 60 (Wender)
0 EtO N MMe
OH
61 (Magnus) 62 (Isobe) FIG. 8. Dynemicin model systems. FIG. 10. Enediynes tethered to delivery systems. Downloaded by guest on September 24, 2021 Review: Nicolaou et al. Proc. Natl. Acad. Sci. USA 90 (1993) 5887
Aglycon fragment m m 0 0 0 0 ,Ns 0 HO~... 0 BzOw v Me O MeSSS _ HCO2Me BzOKj' OMEM OMEM 77 78 O~~y~OMeoHct ~HON o several steps MeOMeM H 0 HO glycoside n e Me_Na1:j ~~~~~~~~~Strategicbond for couling 0 0 OH MeO j NHCO2Me the aglycon and HO.... = several 4 NPhth oligosaccharide steps ll fragments of the TESO7 CHO Oligosaccharide fragment target molecule. OH /C02Me FIG. 11. Retrosynthetic analysis of calicheamicin Ay (5). MeSSS 80 79 Furthermore, this strategy provided the the full functionality of the aglycon but SCHEME IX. Strategy for the enantioselec- means by which several modified oligo- also allowed ultimate control of its abso- tive synthesis of calicheamicinone (80) and saccharides became available for binding lute stereochemistry to provide cali- other calicheamicin precursors. MEM, studies with DNA fragments (20). cheamicinone (80) as well as an advanced 2-methoxyethoxymethyl; Phth, phthaloyl; Bz, By this time thoughts were turning precursor, 82 (Scheme X), suitable for benzoyl. toward the synthesis of the other portion coupling with the oligosaccharide frag- of calicheamicin AyI its aglycon. This ment and thus opening the way for fur- aglycon precursor 82 coupled smoothly synthesis would require a strategy which ther progress. and stereoselectively with 81 (Scheme would provide large amounts of a suit- Now that both domains of calicheam- X), and the careful choice we had made of ably advanced precursor in homochiral icin y' were available, it remained for us protecting groups for the various sensi- form (i.e., as a single enantiomer). Sev- to couple them together and complete the tive functionalities of the entire molecule eral groups had spent the previous 4 synthesis. By this time Nicolaou et al. meant that several steps later we were years examining the chemistry of the (69) had already demonstrated the syn- able to unveil the final target in its full aglycon and synthetic approaches toward thetic utility of the carbohydrate precur- glory to complete the synthesis of cali- its various structural features. Danishef- sor 81 (Scheme X) as a coupling partner cheamicin 'y (5) (71), the first total syn- sky (31, 32), Kende (33), Magnus (34, 35), in model studies, but the question re- thesis of any member of the enediyne and Schreiber (36) and their colleagues mained as to whether it were among the first of many to demon- would be possible class of natural products. strate that the construction ofthe bicyclic to find a precursor to the aglycon which framework ofthe aglycon is not as daunt- was suitable for coupling to the oligosac- Concluding Remarks and Future ing as at first appeared. Meanwhile, Mag- charide and yet sufficiently advanced to Prospects nus et al. (66) had studied the chemistry allow the final unveiling of its features. of the allylic trisulfide group, and Dan- Examination of molecular models indi- Much original work has been carried out ishefsky and coworkers (67), in a mag- cated that it would be sterically demand- in the area ofenediynes since 1987, when nificent achievement, had succeeded in ing to bring the two pieces together, the first structures ofthe naturally occur- assembling the full functionality to although only the required f3glycosidic ring calicheamicins and esperamicins achieve the first synthesis of calicheam- linkage would be formed ifit was possible were disclosed. Even though some ofthe icinone (80) as its racemate. Thus at the at all. Indeed, at this very time Danishef- fundamental chemistry had been known beginning of 1991, we embarked upon a sky and coworkers (70) were reporting prior to the 1987 reports on these novel new approach to the aglycon which difficulty in their attempts to couple a anticancer antibiotics, it took a bold les- would hopefully provide the molecule in sufficiently advanced precursor of the son from Nature again to show us the way enantiomerically pure form and in suffi- aglycon to their oligosaccharide, which to bring together chemistry and biology cient quantities for coupling to the oligo- they had also synthesized by this time, in arriving at such elegant molecular sys- saccharide and completing the total syn- although they were able to couple an tems and with such specific biological thesis of calicheamicin 'y. Indeed, prog- immature version of the aglycon. How- actions. Research in this field is continu- ress was rapid, since we were able to ever, our fears were unfounded, for the ing unabated in chemistry, biology, and benefit during the latter stages of our synthesis from the lessons learned by our Me 0 peers, allowing us to complete the first N Me 0TESO... . enantioselective synthesis of the aglycon Me EHBzOT NHCO2Me towards the end ofthe same year (68). At Me ~ OMe FMOC OyNH the heart of our strategy was an intramo- Meo OTES MeO --O lecular dipolar cycloaddition reaction of 81 82 the nitrile oxide 77 (Scheme IX), in which [carbohydrate precursor] [aglycon precursor] chirality had already been established, to I several steps give the adduct 78. This strategy not only 0 led directly to the efficient installation of HO..._ 2M Me 0 MeSSS NHCO2Me Me Me 0 Me 0 0oAN , 110°C N O S~MeO HOlm
vN4N~~~~_- OR meHO7o OMe H.N~~HO~l.-I HO_ Me,_N 147 \=J R = SituMe2 Meo OH MeO 75 76 5: calicheamicin y1I SCHEME VIII. Ferrier-like rearrange- ment strategy utilized in the synthesis of the SCHEME X. Coupling of oligosaccharide and aglycon fragments and completion of the total calicheamicin 'y oligosaccharide. synthesis of calicheamicin y' (5). Downloaded by guest on September 24, 2021 5888 Review: Nicolaou et al. Proc. Natl. Acad. Sci. USA 90 (1993) medicine, with some compounds already 10. Konichi, M., Ohkuma, H., Tsuno, T., Oki, T., S. V., Smith, A. L., Torisawa, Y., Maligres, P. VanDuyne, G. D. & Clardy, J. (1990) J. Am. & Hwang, C.-K. (1991) Angew. Chem. Int. Ed. in clinical trials. Chem. Soc. 112, 3715-3716. Engl. 30, 1032-1036. In chemistry, research has taken sev- 11. Leet, J. E., Schroeder, D. R., Hofstead, S. J., 42. Nicolaou, K. C., Dai, W.-M., Tsay, S.-C., Es- eral directions, including molecular de- Golik, J., Colson, K. L., Huang, S., Klohr, tevez, V. A. & Wrasidlo, W. (1992) Science S. E., Doyle, T. W. & Matson, J. A. (1992) J. 256, 1172-1178. sign and chemical synthesis. Several 43. Nicolaou, K. C., Maligres, P., Suzuki, T., model systems related to the natural Am. Chem. Soc. 114, 7946-7948. Wendebom, S. V., Dai, W.-M. & Chadha, products have been synthesized, demon- 12. Edo, K., Mizugaki, Y., Koide, Y., Seto, H., R. K. (1992) J. Am. Chem. Soc. 114, 8890- strating new synthetic principles and Furihata, K., Otake, N. & Ishida, N. (1985) 8907. Tetrahedron Lett. 26, 331-334. 44. Nicolaou, K. C. & Dai, W.-M. (1992) J. Am. strategies. Designed enediynes demon- 13. Myers, A. G. & Proteau, P. J. (1989) J. Am. Chem. Soc. 114, 8908-8921. strated abilities to cleave DNA and ex- Chem. Soc. 111, 1146-1147. 45. Nicolaou, K. C., Hong, Y.-P., Tsay, S.-C. & hibited selective cytotoxicity against tu- 14. Goldberg, I. H. (1991) Acc. Chem. Res. 24, Dai, W.-M. (1991) J. Am. Chem. Soc. 113, 191-198. 9878-9880. mor cells versus normal cells (39, 42). 15. Lee, M. D., Ellestad, G. A. & Borders, D. B. 46. Nicolaou, K. C., Dai, W.-M., Tsay, S.-C. & Enediynes have been implicated in the (1991) Acc. Chem. Res. 24, 235-240. Wrasidlo, W. (1992) Bioorg. Med. Chem. Lett. puzzling but important phenomenon of 16. Zein, N., Poncin, M., Nilakantan, R. & Elie- 2, 1155-1160. programmed cell death (apoptosis) (49). stad, G. A. (1989) Science 244, 697-699. 47. Nicolaou, K. C. & Smith, A. L. (1992) Acc. 17. Walker, S., Landovitz, R., Ding, W.-D., Elle- Chem. Res. 25, 497-503. And the total synthesis ofthe most prom- stad, G. A. & Kahne, D. (1992) Proc. Natl. 48. Vaux, D. L. (1993) Proc. Natl. Acad. Sci. USA inent and complex member of the Acad. Sci. USA 89, 4608-4612. 90, 786-789. enediyne class, calicheamicin yI (5), has 18. Walker, S., Valentine, K. G. & Kahne, D. 49. Nicolaou, K. C., Stabila, P., Esmaeli-Azad, B., been achieved (71). (1990) J. Am. Chem. Soc. 112, 6428-6429. Wrasidlo, W. & Hiatt, A. (1993) Proc. Natl. 19. Hawley, R. C., Kiessling, L. L. & Schreiber, Acad. Sci. USA 90, 3142-3146. The next phase of research in the S. L. (1989) Proc. Natl. Acad. Sci. USA 86, 50. Nicolaou, K. C., Hong, Y.-P., Dai, W.-M., enediyne field will undoubtedly include 1105-1109. Zeng, Z.-J. & Wrasidlo, W. (1992) J. Chem. further synthetic attempts at the natu- 20. Nicolaou, K. C., Tsay, S.-C., Suzuki, T. & Soc. Chem. Commun., 1542-1544. rally occurring targets, new designed Joyce, G. F. (1992) J. Am. Chem. Soc. 114, 51. Wood, J. L., Porco, J. A., Jr., Taunton, J., 7555-7557. Lee, A. Y., Clardy, J. & Schreiber, S. L. (1992) enediynes with sophisticated mecha- J. Am. Chem. Soc. 114, 5898-5900. nisms of in vitro and in vivo activation, 21. De Voss, J. J., Townsend, C. A., Ding, W.-D., 52. Wender, P. A. & Zercher, C. K. (1991) J. Am. Morton, G. O., Ellestad, G. A., Zein, N., Ta- Chem. Soc. 113, 2311-2313. and attachment of these systems to suit- bor, A. B. & Schreiber, S. L. (1990) J. Am. 53. Magnus, P. & Fortt, S. M. (1991)J. Chem. Soc. able delivery systems. Targeting devices Chem. Soc. 112, 9669-9670. Chem. Commun., 544-546. may include antibodies, oligonucleo- 22. Hangeland, J. J., De Voss, J. J., Heath, J. A., 54. Nishikawa, T., Ino, A., Isobe, M. & Goto, T. and Townsend, C. A., Ding, W.-D., Ashcroft, J. S. (1991) Chem. Lett. 1271-1274. tides, oligosaccharides, peptides pro- & Ellestad, G. A. (1992) J. Am. Chem. Soc. 55. Myers, A. G., Harrington, P. M. & Kuo, E. Y. teins, DNA intercalators, DNA groove 114, 9200-9202. (1991) J. Am. Chem. Soc. 113, 694-695. binders, hormones, and other ligands. 23. Nicolaou, K. C., Zuccareilo, G., Ogawa, Y., 56. Nicolaou, K. C., Skokotas, G., Furuya, S., Hybrid molecules between enediyne Schweiger, E. J. & Kumazawa, T. (1988) J. Suemune, H. & Nicolaou, D. C. (1990) Angew. Am. Chem. Soc. 110, 4866-4868. Chem. Int. Ed. Engl. 29, 1064-1067. "molecular warheads" and such delivery 24. Nicolaou, K. C., Zuccarello, G., Riemer, C., 57. Nicolaou, K. C., Wendeborn, S., Maligres, P., systems should provide new insights into Estevez, V. A. & Dai, W.-M. (1992) J. Am. Isshiki, K., Zein, N. & Eliestad, G. (1991) biological phenomena and may facilitate Chem. Soc. 114, 7360-7371. Angew. Chem. Int. Ed. Engl. 30, 418-420. and 25. Snyder, J. P. (1989) J. Am. Chem. Soc. 111, 58. Hirama, M., Gomibuchi, T., Fujiwara, K., Sug- drug design development. 7630-7632. iura, Y. & Uesugi, M. (1991)J. Am. Chem. Soc. 26. Magnus, P., Carter, P., Elliott, J., Lewis, R., 113, 9851-9853. We thank our many talented associates, Harling, J., Pitterna, T., Bauta, W. E. & Fortt, 59. Toshima, K., Ohta, K., Ohtake, T. & Tatsuda, whose names appear in the original articles S. (1992) J. Am. Chem. Soc. 114, 2544-2559. K. (1991) Tetrahedron Lett. 32, 391-394. cited below, for their invaluable contributions 27. Nicolaou, K. C., Ogawa, Y., Zuccareilo, G. & 60. Nicolaou, K. C., Maligres, P., Shin, J., de to this research effort. Our work was finan- Kataoka, H. (1988) J. Am. Chem. Soc. 110, Leon, E. & Rideout, D. (1990) J. Am. Chem. the National Institutes of 7247-7248. Soc. 112, 7825-7826. cially supported by 28. Nicolaou, K. C., Sorensen, E. J., Discordia, 61. Nicolaou, K. C., Schreiner, E. P., Iwabuchi, Health, the National Science Foundation, The R., Hwang, C.-K., Minto, R. E., Bharucha, Y. & Suzuki, T. (1992) Angew. Chem. Int. Ed. Scripps Research Institute, and the University K. N. & Bergman, R. G. (1992) Angew. Chem. Engl. 31, 340-341. of California, San Diego. Int. Ed. Engl. 31, 1044-1046. 62. Toshima, K., Ohta, K., Ohashi, A., Ohtsuka, 29. Nicolaou, K. C., Liu, A., Zeng, Z. & McComb, A., Nakata, M. & Tatsuta, K. (1992) J. Chem. 1. Nicolaou, K. C. & Dai, W.-M. (1991) Angew. S. (1992) J. Am. Chem. Soc. 114, 9279-9282. Soc. Chem. Commun., 1306-1308. Chem. Int. Ed. Engl. 30, 1387-1416. 30. Myers, A. G. & Dragovich, P. S. (1992) J. Am. 63. Tokuda, M., Fujiwara, K., Gomibuchi, T., 2. Bergman, R. G. (1973) Acc. Chem. Res. 6, Chem. Soc. 114, 5859-5860. Hirama, M., Uesugi, M. & Sugiura, Y. (1993) 25-31. 31. Danishefsky, S. J., Mantlo, N. B., Yamashita, Tetrahedron Lett. 34, 669-672. 3. Darby, N., Kim, C. U., Salailn, J. A., Shelton, D. S. & Schulte, G. (1988) J. Am. Chem. Soc. 64. Boger, D. L. & Zhou, J. (1993) J. Org. Chem. K. W., Takada, S. & Masamune, S. (1971) J. 110, 6890-6891. 58, in press. Chem. Soc. Chem. Commun., 1516-1517. 32. Hazeltine, J. N., Danishefsky, S. J. & Schulte, 65. Nicolaou, K. C., Groneberg, R. D., Miyazaki, 4. Wong, H. N. C. & Sondheimer, F. (1980) Tet- G. (1989) J. Am. Chem. Soc. 111, 7638-7640. T., Stylianides, N. A., Schulze, T. J. & Stahl, rahedron Lett. 21, 217-220. 33. Kende, A. S. & Smith, C. A. (1988) Tetrahe- W. (1990) J. Am. Chem. Soc. 112, 8193-8195. 5. Lee, M. D., Dunne, T. S., Siegel, M. M., dron Lett. 29, 4217-4220. 66. Magnus, P., Lewis, R. T. & Bennett, F. (1989) Chang, C. C., Morton, G. 0. & Borders, D. B. 34. Magnus, P., Lewis, R. T. & Huffman, J. C. J. Chem. Soc. Chem. Commun., 916-919. (1987) J. Am. Chem. Soc. 109, 3464-3466. (1988) J. Am. Chem. Soc. 110, 6921-6923. 67. Haseltine, J. N., Cabal, M. P., Mantlo, N. B., 6. Lee, M. D., Dunne, T. S., Chang, C. C., Elle- 35. Magnus, P., Lewis, R. & Bennett, F. (1992) J. Iwasawa, N., Yamashita, D. S., Coleman, stad, G. A., Siegel, M. M., Morton, G. O., Am. Chem. Soc. 114, 2560-2567. R. S., Danishefsky, S. J. & Schulte, G. K. McGahren, W. J. & Borders, D. B. (1987) J. 36. Schoenen, F. J., Porco, J. A., Jr., & Schreiber, (1991) J. Am. Chem. Soc. 113, 3850-3866. Am. Chem. Soc. 109, 3466-3468. S. L. (1989) Tetrahedron Lett. 30, 3765-3768. 68. Smith, A. L., Hwang, C.-K., Pitsinos, E., Scar- 7. Golik, J., Clardy, J., Dubay, G., Groenewold, 37. Mantlo, N. B. & Danishefsky, S. J. (1989) J. lato, G. R. & Nicolaou, K. C. (1992) J. Am. G., Kawaguchi, H., Konishi, M., Krishnan, B., Org. Chem. 54, 2781-2783. Chem. Soc. 114, 3134-3136. Ohkuma, H., Sithoh, K. & Doyle, T. W. (1987) 38. Drak, J., Iwasawa, N., Danishefsky, S. J. & 69. Nicolaou, K. C., Schreiner, E. P., Iwabuchi, J. Am. Chem. Soc. 109, 3461-3462. Crothers, D. M. (1991) Proc. Natl. Acad. Sci. Y. & Suzuki, T. (1992) Angew. Chem. Int. Ed. 8. Golik, J., Clardy, J., Dubay, G., Groenewold, USA 88, 7464-7468. Engl. 31, 340-342. G., Kawaguchi, H., Konishi, M., Krishnan, B., 39. Nicolaou, K. C., Smith, A. L., Wendeborn, 70. Halcomb, R. L., Boyer, S. H. & Danishefsky, Qhkuma, H., Sithoh, K. & Doyle, T. W. (1987) S. V. & Hwang, C.-K. (1991) J. Am. Chem. S. J. (1992) Angew. Chem. Int. Ed. Engl. 31, J. Am. Chem. Soc. 109, 3462-3464. Soc. 113, 3106-3114. 338-340. 9. Konishi, M., Ohkuma, H., Matsumoto, K., 40. Nicolaou, K. C., Hwang, C.-K., Smith, A. L. 71. Nicolaou, K. C., Hummel, C. W., Pitsinos, Tsuno, T., Kamei, H., Miyaki, T., Oki, T., & Wendeborn, S. V. (1990) J. Am. Chem. Soc. E. N., Nakada, M., Smith, A. L., Shibayama, Kawaguchi, H., Van Duyne, G. D. & Clardy, J. 112, 7416-7418. K. & Saimoto, H. (1992) I. Am. Chem. Soc. (1989) J. Antibiot. 42, 1449-1452. 41. Nicolaou, K. C., Dai, W.-M., Wendeborn, 114, 10082-10084. Downloaded by guest on September 24, 2021