Synthesis of (±)-calicheamicinone by two methods

10332

J. Am. Chem. Soc. 1998, 120, 10332-10349

Synthesis of (()-Calicheamicinone by Two Methods

Derrick L. J. Clive,* Yunxin Bo, Yong Tao, Sylvain Daigneault, Yong-Jin Wu, and

Ge´rard Meignan

Contribution from the Chemistry Department, UniVersity of Alberta, Edmonton, Alberta, Canada T6G 2G2

ReceiVed January 26, 1998

Abstract: Ketene acetal 25 was converted into silyl enol ether 20, which underwent a Diels-Alder reaction

with methyl (E)-3-nitropropenoate to afford ketone 27. This was converted by two routes into (()-

calicheamicinone (1). In the first, modification of the nitro, ester, and allyl substituents gave ketone 38, which

reacted stereoselectively with cerium trimethylsilylacetylide to place the acetylene unit syn to the nitrogen

function (38 f 39). Further elaboration took the route as far as aldehyde 42. A slightly different series of

reactions served to convert ketone 27 into tricyclic ketone 43. This also reacted with cerium trimethylsily-

lacetylide, but in the opposite stereochemical sense to 38, so as to place the acetylene unit anti to the nitrogen

function (43 f 50). Further elaboration took this second route as far as lactone 44. Both 42 and 44 serve as

advanced intermediates for the synthesis of (()-calicheamicinone. The monoacetylenic aldehyde 42 reacted

stereoselectively with cerium trimethylsilylacetylide to give the bis(acetylene) 53 as the major product. This

was elaborated into lactone 60, which was desaturated (60 f 68) and methoxycarbonylated on nitrogen. The

acetylenic silyl groups were then removed, so as to generate syn bis(acetylene) 72, and the terminal acetylenic

hydrogens were replaced by iodine. The resulting diiodide 81 formed the cyclic enediyne 82 on reaction with

(Z)-1,2-bis(trimethylstannyl)ethene in the presence of (Ph₃P)₄Pd, and the enediyne was converted into (()-

calicheamicinone. syn-Bis(acetylene) 72 was also synthesized from the other advanced intermediate (lactone

44), using, as a key step, free radical bromination of the derived unsaturated lactone 77. The resulting bromo

lactones 78 were converted into an aldehyde ester 79, and this reacted stereoselectively with cerium

trimethylsilylacetylide to give 80, convertible by the action of TBAF into 72.

Introduction

biochemical questions concerning the mode of action of

enediyne antitumor agents.⁶

In this paper, we describe full details¹of two related methods

for the synthesis of (()-calicheamicinone (1),²the aglycon (in

I

racemic form) of the antitumor antibiotic calicheamicin γ₁

(2).^3-5

Compounds 1 and 2 have attracted an exceptional degree of

interest, not only because they represent very difficult synthetic

targets but also because they pose fascinating and important

(1) Preliminary communication: Clive, D. L. J.; Bo, Y.; Tao, Y.;

Daigneault, S.; Wu, Y.-J.; Meignan, G. J. Am. Chem. Soc. 1996, 118, 4904.

(2) (a) Synthesis of (()-calicheamicinone: Haseltine, J. N.; Paz Cabal,

M.; Mantlo, N. B.; Iwasawa, N.; Yamashita, D. S.; Coleman, R. S.;

Danishefsky, S. J.; Schulte, G. K. J. Am. Chem. Soc. 1991, 113, 3850. (b)

Synthesis of (-)-calicheamicinone: Smith, A. L.; Pitsinos, E. N.; Hwang,

C.-K.; Mizuno, Y.; Saimoto, H.; Scarlato, G. R.; Suzuki, T.; Nicolaou, K.

C. J. Am. Chem. Soc. 1993, 115, 7612. (c) Aiyer, J.; Hitchcock, S. A.;

Denhart, D.; Liu, K. K.-C.; Danishefsky, S. J.; Crothers, D. M. Angew.

Chem., Int. Ed. Engl. 1994, 33, 855.

(3) Nonsystematic numbering is used in this manuscript, except for

chemical names in the Experimental Section.

I

(4) (a) Structure determination: Lee, M. D.; Dunne, T. S.; Chang, C.

C.; Siegel, M. M.; Morton, G. O.; Ellestad, G. A.; McGahren, W. J.; Borders,

D. B. J. Am. Chem. Soc. 1992, 114, 985. (b) For a related compound

(namenamicin) reported recently that also shows powerful antitumor and

antimicrobial activity, see: McDonald, L. A.; Capson, T. L.; Krishnamurthy,

G.; Ding, W.-D.; Ellestad, G. A.; Bernan, V. S.; Maiese, W. M.; Lassota,

P.; Discafani, C.; Kramer, R. A.; Ireland, C. M. J. Am. Chem. Soc. 1996,

118, 10898.

(5) The corresponding disulfide has also been isolated: (a) Ellestad, G.

A.; Hamann, P. R.; Zein, N.; Morton, G. O.; Siegel, M. M.; Pastel, M.;

Borders, D. B.; McGahren, W. J. Tetrahedron Lett. 1989, 30, 3033. (b)

McGahren, W. J.; Ding, W.-D.; Ellestad, G. A. In Enediyne Antibiotics as

Antitumor Agents; Borders, D. B., Doyle, T. W., Eds.; Dekker: New York,

1995; p 75.

The chemistry and biological properties of calicheamicin γ₁

have been reviewed extensively,⁶the essential details being that

calicheamicin γ₁has extremely potent anticancer activity^7,8and

I

that the mode of action involves binding in the minor groove

(6) Reviews on the chemistry and biology of enediyne anticancer

antibiotics and model systems: (a) Lee, M. D.; Ellestad, G. A.; Borders,

D. B. Acc. Chem. Res. 1991, 24, 235. (b) Nicolaou, K. C.; Dai, W.-M.

Angew. Chem., Int. Ed. Engl. 1991, 30, 1387. (c) Nicolaou, K. C.; Smith,

A. L. Acc. Chem. Res. 1992, 25, 497. (d) Nicolaou, K. C.; Smith, A. L.;

Yue, E. W. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5881. (e) Murphy, J.

A.; Griffiths, J. Nat. Prod. Rep. 1993, 10, 551. (f) Weymouth-Wilson, A.

C. Nat. Prod. Rep. 1997, 14, 99.

Published on Web 09/25/1998

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10333

Scheme 1

the resulting biradical 5 abstracts hydrogen^6a,e,10e-gfrom both

strands of the DNA sugar backbone, leading to DNA cleavage.

Each of these steps has been scrutinized carefully, and a detailed

picture has emerged. Many of the mechanistic features are

shared by other enediyne antitumor agents, and the mode of

activation (though not the sequence selectivity¹³or propensity

to cause double strand cleavage¹⁴) of the closely related

esperamicins¹⁵is likely to be similar.

I

As a synthetic target, calicheamicin γ₁represents an espe-

cially large number of complex problems that are inherent not

only in the construction of the sensitive enediyne segment and

the unusual tetrasaccharide but also in the linking of both of

these substructures, or portions of them, in a way that preserves

their integrity during the remaining steps of the synthesis. We

have dealt with the first of these taskssthe synthesis of the

aglycon.

During the course of our work, two syntheses²of calicheami-

cinone, and of calicheamicin γ₁itself,¹⁶were completed by the

I

research teams led by Danishefsky and by Nicolaou, but the

routes to the enediyne section in each of these remarkable

syntheses were so very different from the one we were exploring

that it was appropriate to continue our studies.

of DNA⁹followed by activation of the enediyne portion so as

to afford a highly reactive biradical. This, in turn, initiates

cleavage of both DNA strands. The recognition process^6a,e,10

s

certain sequences are preferentially cleavedsis effected

largely^10a-dby the aryl carbohydrate segment. Activation

(Scheme 1) involves cleavage of the trisulfide (2 f 3) followed

by intramolecular Michael addition (3 f 4).¹¹Once that

addition has taken place and C(3) (nonsystematic numbering)

becomes sp³hybridized, Bergman cyclization^11,12occurs, and

Results and Discussion

Development of the Synthetic Plan. Although no synthesis

of calicheamicinone had been reported when we began,¹⁷

a

number of papers had appeared that dealt with the most

conspicuous features of the molecule: the allylic trisulfide and

the cyclic enediyne. Published model studies¹⁸on the enediyne

emphasized the difficulties in constructing such a unit and

prompted us to take appropriate precautions in the form of

unusually extensive backup plans.

The properties of allylic trisulfides¹⁹had received little

attention, but we were greatly helped by the pioneering studies

of Magnus and his collaborators on the synthesis and thermal

(7) (a) Against murine tumors the optimum dose is 0.5-1.5 mg/kg

Thomas, J. P.; Carvagal, S. G.; Lindsay, H. L.; Citarella, R. V.; Wallace,

R. E.; Lee, M. D.; Durr, F. E. Program and Abstracts, 26th Interscience

Conference on Antimicrobial Agents and Chemotherapy, New Orleans, LA,

Sept 1986; American Society for Microbiology: Washington, DC.; 1986;

Abstr. 229 (quoted in Lee, M. D.; Dunne, T. S.; Siegel, M. M.; Chang, C.

C.; Morton, G. O.; Borders, D. B. J. Am. Chem. Soc. 1987, 109, 3464). (b)

Measurement of cytotoxic activity against numerous cancer cell lines gave

IC₅₀values in the nanomolar range: Nicolaou, K. C.; Li, T.; Nakada, M.;

Hummel, C. W.; Hiatt, A.; Wrasidlo, W. Angew. Chem., Int. Ed. Engl. 1994,

33, 183.

(12) Bergman, R. G. Acc. Chem. Res. 1973, 6, 25. See the Supporting

Information for additional references.

(8) For measurements of its antimicrobial activity, see ref 4b.

(13) (a) Sugiura, Y.; Takahashi, Y.; Uesawa, Y.; Kuwahara, J. Nucleic

Acids Symp. Ser. 1988, 20, 63. (b) Long, B. H.; Golik, J.; Forenza, S.;

Ward, B.; Rehfuss, R.; Dabrowaik, J. C.; Catino, J. J.; Musial, S. T.;

Brookshire, K. W.; Doyle, T. W. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 2.

(c) Sugiura, Y.; Uesawa, Y.; Takahashi, Y.; Kuwahara, J.; Golik, J.; Doyle,

T. W. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 7672. (d) Christner, D. F.;

Frank, B. L.; Kozarich, J. W.; Stubbe, J.; Golik, J.; Doyle, T. W.; Rosenberg,

I. E.; Krishnan, B. J. Am. Chem. Soc. 1992, 114, 8763. (e) Esperamicin-

DNA NMR studies: Ikemoto, N.; Kumar, R. A.; Dedon, P. C.; Danishefsky,

S. J.; Patel, D. J. J. Am. Chem. Soc. 1994, 116, 9387. (f) Langley, D. R.;

Golik, J.; Krishnan, B.; Doyle, T. W.; Beveridge, D. L. J. Am. Chem. Soc.

1994, 116, 15.

(14) Esperamicin A₁causes single strand cleavage (see ref 13b-d).

(15) Golik, J.; Dubay, G.; Groenewold, G.; Kawaguchi, H.; Konishi, M.;

Krishnan, B.; Ohkuma, H.; Saitoh, K.; Doyle, T. W. J. Am. Chem. Soc.

1987, 109, 3462.

(16) (a) Nicolaou, K. C.; Hummel, C. W.; Nakada, M.; Shibayama, K.;

Pitsinos, E. N.; Saimoto, H.; Mizuno, Y.; Baldenius, K.-U.; Smith, A. L. J.

Am. Chem. Soc. 1993, 115, 7625. (b) Hitchcock, S. A.; Chu-Moyer, M. Y.;

Boyer, S. H.; Olson, S. H.; Danishefsky, S. J. J. Am. Chem. Soc. 1995,

117, 5750.

(17) Taken in part from the Ph.D. Thesis (April 1991) of S.D.

(18) Reviews: (a) Lhermitte, H.; Grierson, D. S. Contemp. Org. Synth.

1996, 3, 41. (b) Review: Lhermitte, H.; Grierson, D. S. Contemp. Org.

Synth. 1996, 3, 93. See the Supporting Information for additional references.

(19) Reviews on polysulfides: (a) Harpp, D. N. PerspectiVes in the

Organic Chemistry of Sulfur; Zwanenburg, B., Klunder, A. J. H., Eds.;

Studies in Organic Chemistry 28; Elsevier: Amsterdam, 1987. (b) Kutney,

G. W.; Turnbull, K. Chem. ReV. 1982, 82, 333. Diallyl trisulfide: (c) Ariga,

T.; Oshiba, S.; Tamada, T. Lancet 1981, 1, 150. Synthetic methods for

trisulfides: (d) Harpp, D. N.; Ash, D. K. Int. J. Sulfur Chem. Part A 1971,

1, 57. (e) Cf. Sullivan, A. B.; Boustany, K. Int. J. Sulfur Chem. Part A

1971, 1, 207. (f) Cf. Harpp, D. N.; Ash, D. K. Int. J. Sulfur Chem. Part A

1971, 1, 211. (g) Mott, A. W.; Barany, G. Synthesis 1984, 657. (h) Derbesy,

G.; Harpp, D. N. Tetrahedron Lett. 1994, 35, 5381.

(9) NMR studies of the binding: (a) Walker, S.; Murnick, J.; Kahne, D.

J. Am. Chem. Soc. 1993, 115, 7954. (b) Walker, S. L.; Andreotti, A. H.;

Kahne, D. E. Tetrahedron 1994, 50, 1351. (c) Cf. Paloma, L. G.; Smith, J.

A.; Chazin, W. J.; Nicolaou, K. C. J. Am. Chem. Soc. 1994, 116, 3697. (d)

Zein, N.; McGahren, W. J.; Morton, G. O.; Ashcroft, J.; Ellestad, G. A. J.

Am. Chem. Soc. 1989, 111, 6888.

(10) (a) Drak, J.; Iwasawa, N.; Danishefsky, S.; Crothers, D. M. Proc.

Natl. Acad. Sci. U.S.A. 1991, 88, 7464. (b) Nicolaou, K. C.; Tsay, S.-C.;

Suzuki, T.; Joyce, G. F. J. Am. Chem. Soc. 1992, 114, 7555. (c) Li, T.;

Zeng, Z.; Estevez, V. A.; Baldenius, K. U.; Nicolaou, K. C.; Joyce, G. F.

J. Am. Chem. Soc. 1994, 116, 3709. (d) Ikemoto, N.; Kumar, R. A.; Ling,

T.-T.; Ellestad, G. A.; Danishefsky, S. J.; Patel, D. J. Proc. Natl. Acad.

Sci. U.S.A. 1995, 92, 10506. (e) Zein, N.; Sinha, A. M.; McGahren, W. J.;

Ellestad, G. A. Science 1988, 240, 1198. (f) Cf. De Voss, J. J.; Townsend,

C. A.; Ding, W.-D.; Morton, G. O.; Ellestad, G. A.; Zein, N.; Tabor, A. B.;

Schreiber, S. L. J. Am. Chem. Soc. 1990, 112, 9669. (g) Hangeland, J. J.;

De Voss, J. J.; Heath, J. A.; Townsend, C. A.; Ding, W.-d.; Ashcroft, J. S.;

Ellestad, G. A. J. Am. Chem. Soc. 1992, 114, 9200. See Supporting

Information for additional references.

(11) (a) Lee, M. D.; Dunne, T. S.; Chang, C. C.; Ellestad, G. A.; Siegel,

M. M.; Morton, G. O.; McGahren, W. J.; Borders, D. B. J. Am. Chem.

Soc. 1987, 109, 3466. (b) For studies of the Bergman mechanism in model

compounds, see: Magnus, P.; Carter, P. A. J. Am. Chem. Soc. 1988, 110,

1626. (c) Stability of model enediynes: Nicolaou, K. C.; Zuccarello, G.;

Ogawa, Y.; Schweiger, E. J.; Kumazawa, T. J. Am. Chem. Soc. 1988, 110,

4866. (d) Magnus, P.; Lewis, R. T.; Huffman, J. C. J. Am. Chem. Soc.

1988, 110, 6921. (e) In vitro NMR studies of the activation mechanism:

De Voss, J. J.; Hangeland, J. J.; Townsend, C. A. J. Am. Chem. Soc. 1990,

112, 4554. (f) Chatterjee, M.; Cramer, K. D.; Townsend, C. A. J. Am. Chem.

Soc. 1993, 115, 3374. (g) Cf. Myers, A. G.; Cohen, S. B.; Kwon, B. M. J.

Am. Chem. Soc. 1994, 116, 1255. (h) Chatterjee, M.; Smith, P. J.; Townsend,

C. A. J. Am. Chem. Soc. 1996, 118, 1938. (i) Theoretical treatment of

trisulfide cleavage: Mulhearn, D. C.; Bachrach, S. M. J. Am. Chem. Soc.

1996, 118, 9415.

10334 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

Scheme 2^a

Scheme 4^a

^aKey: (a) AcSNa; (b) LiAlH₄.

Scheme 3

^aKey: R ) unspecified group.

Scheme 5^a

stability of such compounds. Particularly valuable guidance was

provided by the transformations summarized in Scheme 2.²⁰The

mesylate 6 was subjected to nucleophilic displacement using

AcSNa, and then hydride reduction afforded the free thiol 8.

This reacted with the phthalimido disulfide 9 to give the desired

allylic trisulfide (8 f 10). A number of other allylic trisulfides

were prepared, and their thermal stability was examined; some

of them can survive heating (85-110 °C) in toluene, which

suggested that calicheamicinone might best be synthesized by

way of allylic alcohol 11 (Scheme 3), at least in approximate

terms, since the need for protecting groups is totally ignored in

Scheme 3. The most appropriate stage at which to elaborate

the alcohol into the trisulfide would be decided once it became

clear what protecting groups still had to be removed after the

trisulfide was in place.

As far as the cyclic enediyne was concerned, the most obvious

point, in view of the mode of action of calicheamicinone, was

that the enediyne should not be assembled in a structure in which

C(3) (see 11) is sp³hybridized, as an otherwise premature

Bergman rearrangement would be likely to occur. But apart

from that, it was not very clear²¹what type of relevant structures

could accommodate the enediyne,^11cand so we decided to

assemble it at a stage that seemed convenient on the basis of

other considerations. We would then find out if the compounds

were stable enough for further elaboration.

^aKey: R ) unspecified group.

on the choice of a ring-closing step. For example, if bonds a

or e (see 12, Scheme 4) are made last to close the ring, then the

enediyne could be introduced in one segment (e.g., as in 14)

by reaction at the aldehyde or ketone carbonyl of compounds

synthetically equivalent to the hypothetical structure 13. A

prefabricated enediyne would not be essential as a smaller

acetylide, such as 15, might serve equally well, although, in

that case, the product would have to be coupled with a two-

carbon unit.

If bonds b or d are made last (Scheme 5, see structures 16

and 17), then the two-carbon and four-carbon acetylenic

components could be introduced individually, or introduced

while they are joined together by a removable or modifiable

tether. There was also the possibility of making bonds b and d

sequentially in the same reaction (cf. 18), and ultimately, this

was the procedure we used.

Last of all, we considered ways of forming bond c as the

ring-closing step, and we intended to rely for that purpose on

some type of dicarbonyl coupling (Scheme 6) such as a

McMurry reaction (path A), a pinacol coupling (path B), or a

double extrusion²²(path C). One particular advantage of

forming the enediyne at a late stage, along the lines we had in

mind, was that if one approach did not work then others would

be available without the need for redesigning the whole route.

This range of opportunities reassured us that what we suspected

might be a very difficult part of the synthesis could be tackled

in a number of ways.

Our plan was to generate the enediyne late in the synthesis

by building it onto a substructure that already had most of the

remaining features present in the final target, and it seemed clear

that this might be accomplished in a number of ways, depending

At this stage of the planning there remained the question of

how to control the stereochemistry of the C(4)dC(7) double

bond, and it seemed most sensible to form that bond within a

(20) Magnus, P.; Lewis, R. T.; Bennett, F. J. Chem. Soc., Chem. Commun.

1989, 916, and references therein.

(21) For an early demonstration with a model compound of thermal

sensitivity as a function of the state of hybridization of C(3) (present

numbering system), see: (a) Magnus, P.; Lewis, R. T. Tetrahedron Lett.

1989, 30, 1905. (b) For a counter example, see ref 11d.

(22) Cf. Back, T. G.; Barton, D. H. R.; Britten-Kelly, M. R.; Guziec, F.

S. J. Chem. Soc., Perkin Trans 1 1976, 2079, and references therein.

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10335

Scheme 8^a

Scheme 6^a

^aKey: (a) methyl (E)-2-nitropropenoate; (b) H₃O⁺, 82%.

Scheme 9^a

^aKey: R ) unspecified group.

Scheme 7

six-membered ring so that only the required geometry was

possible. On the basis of this argument, therefore, the prelimi-

nary target 11 was changed to a compound of type 19 where,

again, the need for protecting groups was ignored.

^aKey: (a) NaH, BuLi, THF; allyl bromide; 80%; (b) 2-chloroethanol,

Ti(O-i-Pr)₄; 63%; (c) K₂CO₃, DMF; 80%; (d) (Me₂PhSi)₂NLi, Me₃SiCl,

THF, -78 °C; (e) methyl (E)-2-nitropropenoate (21), THF; (f) aq

NH₄Cl; 56% from 25.

than an acyclic acetal against aromatization because expulsion

of one of the acetal oxygens from 22 (in the presence of acid

or base) would be more easily reversible than with, for example,

a dimethyl acetal.

Scheme 7 summarizes this plan, which was quickly recog-

nized as a worthwhile approach because it could build upon

the background information available in the literature. A few

cases had been reported in which 21 was used in Diels-Alder

reactions, and one paper in particular^25awas encouraging

because it showed (see Scheme 8) that the regiochemistry is

controlled by the nitro group, exactly in the sense we had hoped.

The use of ketene acetal silyl enol ether 20 also requires some

comment. A few simpler compounds of this type had been

described²⁶before, and some guidance was available, therefore,

for developing a route to our particular example.

With the above considerations as background, it was clear

that a highly functionalized six-membered ring with suitably

disposed carbonyl groups for acetylide attachment and a chain

to provide C(7) and C(8) of the target was required. We

considered several possibilities that might correspond to the

hypothetical structure 13, and after a good deal of exploratory

work,²³we settled on a compound of structure 22 (see Scheme

7), as representing the essential features implied by 13.

We felt that 22 should be accessible by a Diels-Alder

reaction between the ketene acetal silyl enol ether 20 and methyl

(E)-2-nitropropenoate^24,25(21). The use of a cyclic acetal was

based on the expectation that it would afford greater protection

Synthesis of the First Key Intermediate (27) by a Diels-

Alder reaction. To put the ideas of Scheme 7 into practice,

we first prepared the â-keto ester 23 (Scheme 9), which is a

known substance,²⁷easily made on a large scale by the allylation

(23) For exploratory studies involving radical cyclization, see the

Supporting Information.

(24) (a) Shechter, H.; Conrad, F.; Daulton, A. L.; Kaplan, R. B. J. Am.

Chem. Soc. 1952, 74, 3052. (b) Cf. McMurry, J. E.; Musser, J. H. Org.

Synth. 1977, 56, 65. (c) Cf. McMurry, J. E.; Musser, J. H.; Fleming, I.;

Fortunak, J.; Nu¨bling, C. Organic Syntheses; Wiley & Sons: New York,

1988: Collect. Vol. VI, p 799.

(26) Cf. (a) Banville, J.; Brassard, P. J. Chem. Soc., Perkin Trans. 1

1976, 1852. (b) Broadhurst, M. D. J. Org. Chem. 1985, 50, 1117. (c) Eid,

C. N., Jr.; Konopelski, J. P. Tetrahedron 1991, 47, 975. (d) Konopelski, J.

P.; Kasar, R. A. Tetrahedron Lett. 1993, 34, 4587.

(25) Cf. (a) Danishefsky, S.; Prisbylla, M. P.; Hiner, S. J. Am. Chem.

Soc. 1978, 100, 2918. (b) Danishefsky, S. Acc. Chem. Res. 1981, 14, 400.

(27) Huckin, S. N.; Weiler, L. J. Am. Chem. Soc. 1974, 96, 1082.

10336 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

of doubly deprotonated methyl acetoacetate. Next, ester

exchange with 2-chloroethanol under standard conditions of

titanium catalysis²⁸gave the chloroethyl ester 24, and this could

be cyclized to acyl ketene acetal 25 by treatment^26bwith

potassium carbonate. Surprisingly, attempts to effect cyclization

with sodium hydride were unsuccessful.

Scheme 10^a

With the acyl ketene acetal in hand, the next task was to

convert it into a silyl enol ether, preferably with Z geometry

(as shown in structure 20) because the E isomer would be less

reactive in Diels-Alder cycloadditions. Deprotonation of 25

with LDA, followed by trapping with chlorotrimethylsilane, gave

a 3:7 mixture of the Z and E isomers.²⁹The use of lithium

hexamethyldisilazide improved the ratio (3:2 in favor of the Z

isomer) but, with an even more hindered base [(Me₂PhSi)₂NLi³¹],

the desired Z enol ether 20 was formed in great preponderance,

and perhaps exclusively. Compound 20 can be made on a large

scale (ca. 40 g of starting material), but it is rather sensitive,

and attempts to isolate it by distillation (ca. 0.001 mmHg) caused

extensive decomposition. Fortunately, the compound can be

used in situ because the hindered base that is present [(Me₂-

PhSi)₂NH] does not react with our dienophile. Addition of

methyl (E)-2-nitropropenoate (21) to a solution of freshly

generated silyl enol ether 20 (Scheme 9) gave the adduct 27,³²

after mild acid hydrolysis, as a crystalline compound in 56%

overall yield from the ketene acetal 25.

Ketone 27 is a key intermediate in the synthesis of cali-

cheamicinone. Its structure and stereochemistry could be

deduced from the ¹H NMR spectrum, and the interpretation was

confirmed by X-ray methods;³³the shape of the molecule is as

shown in structure 27′. The ketone was the starting point for

a very large number of exploratory experiments, and after a

considerable effort, we learned how to take it forward to

calicheamicinone by two routes.

The first tasks were to protect the C(5) carbonyl and to modify

the ester. We decided to protect the ketone by reduction, and

it was quickly found that this could be accomplished in high

yield, but only at the expense of good stereoselectivity.

Treatment with sodium borohydride gave a 2:1 mixture of C(5)

epimeric alcohols, and these could be separated after silylation

(Scheme 10, 27 f 28 f 29). The yield of the 5â-silyl ether

from ketone 27 is 65%, and the yield of the 5R-isomer is 33%.

Scheme 10 shows only the 5â-isomer 29, but both isomers were

individually subjected to the same reactions. All yields shown

in parentheses in Scheme 10 refer to the 5R series, but only the

5â compounds are drawn. In the text, only the yields for the

5â series are specified. Elaboration of each of the C(5) epimers

eventually leads to the same compound so that both products

from the initial ketone reduction are used in the synthesis. In

principle, the C(5) epimers could be processed without separa-

tion.

^aKey: Yields in parentheses refer to the 5-R series; R ) t-BuMe₂Si;

AOM ) CH₂OC₆H₄OMe-p; (a) NaBH₄, MeOH; (b) t-BuMe₂SiOTf,

2,6-lutidine, CH₂Cl₂; 65% 5â, 33% 5R from 27; (c) DIBAL-H, CH₂Cl₂;

99% (99%); (d) t-BuCOCl, DMAP, PhMe; 99% (99%); (e) OsO₄,

NaIO₄, CCl₄, H₂O, t-BuOH; 73% (77%); (f) NaBH₄, MeOH; 96%

(92%); (g) p-MeOC₆H₄OCH₂Cl, i-PrNEt₂, DMAP, PhMe; 91% (89%);

(h) NaBH₄, NiCl₂‚6H₂O, MeOH, sonication; 95% (95%); (i) allyl

chloroformate, pyridine, THF; 82% (82%); (j) Bu₄NF, THF, 95%

(96%); (k) CrO₃, pyridine, CH₂Cl₂; 95% (90%).

starting ketone (cf. 27′), with the notable feature that the large

silyloxy group at C(5) is axial.

The methyl ester was next reduced (Scheme 10, DIBAL-H,

99%), and the resulting primary alcohol was protected as its

pivaloate (t-BuCOCl, DMAP, 99%, 29 f 30 f 31). At this

point, with the oxygen functions at C(5) and C(9) protected, it

was necessary to remove C(10) and introduce oxygen at C(8).

Cleavage of the double bond under classical conditions³⁴

(Scheme 10), using osmium tetroxide and sodium periodate,

gave an aldehyde (31 f 32, 73%), and that was reduced with

sodium borohydride (96%) to the expected primary alcohol.

Finally, the hydroxyl group was protected as its p-anisyloxym-

ethyl ether³⁵(33 f 34, 91%).

Both of the silyl ethers are crystalline, but the 5R compound

affords better crystals, and these were used for X-ray analysis.³³

The shape of the molecule is very much like the shape of the

(28) Seebach, D.; Hungerbu¨hler, E.; Naef, R.; Schnurrenberger, P.;

Weidmann, B.; Zu¨ger, M. Synthesis 1982, 138.

(29) The assignment was made by analogy with results for simple ketones

(see ref 30).

(30) Cf. Masamune, S.; Ellingboe, J. W.; Choy, W. J. Am. Chem. Soc.

1982, 104, 5526.

(31) Made from the amine (Zhinkin, D. Ya.; Mal’nova, G. N.; Gorisla-

vskaya, Zh. V. J. Gen. Chem. USSR 1968, 38, 2702) by treatment with

BuLi (THF, 0 °C; 20 min).

(32) Unlike a literature precedent (see ref 25a), the Diels-Alder reaction

proceeds with good endo selectivity. Cf. (a) Node, M.; Nishide, K.; Imazato,

H.; Kurosaki, R.; Inoue, T.; Ikariya, T. J. Chem. Soc., Chem. Commun.

1996, 2559. (b) Stoodley, R. J.; Yuen, W.-H. Chem. Commun. 1997, 1371.

(33) See the Supporting Information for X-ray data.

Having reached the stage of 34, the nitro group now had to

be reduced. While reduction of aromatic nitro compounds is

(34) Pappo, R.; Allen, D. S., Jr.; Lemieux, R. U.; Johnson, W. S. J. Org.

Chem. 1956, 21, 478.

(35) Masaki, Y.; Iwata, I.; Mukai, I.; Oda, H.; Nagashima, H. Chem.

Lett. 1989, 659.

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10337

Scheme 12^a

Scheme 11^a

^aKey: AOM ) CH₂OC₆H₄OMe-p; (a) lithium trimethylsilylacetyl-

ide, CeCl₃, THF; 91%; (b) t-BuMe₂SiOTf, 2,6-lutidine; 97%; (c)

DIBAL-H, CH₂Cl₂; 95%; (d) CrO₃, pyridine, CH₂Cl₂; 90%.

^aKey: AOM ) CH₂OC₆H₄OMe-p.

THF and sonicated overnight^39d,eto produce a white suspension.

This is then diluted at a low temperature (-78 °C) with a THF

solution of the lithium salt of trimethylsilylacetylene, and the

mixture is kept for 40 min at -78 °C before use. Quite recently,

the nature of cerium chloride has been subjected to various

dehydration procedures as described in the literature,⁴⁰and

though it is clear that the material we use is not pure cerium

trichloride, we have not gone back to repeat our experiments

with what is now judged to be the pure anhydrous salt.

The tertiary alcohol resulting from the acetylide addition was

protected by silylation (39 f 40, 97%), and the pivaloyl group

was removed by the action of DIBAL-H (95%). The resulting

alcohol 41 was then oxidized (90%) to aldehyde 42 by the

Collins procedure. This aldehyde is a key advanced intermedi-

ate because it brought us directly to several critical points in

the synthesis: the attachment of an acetylene at C(9) in a

stereocontrolled manner and the introduction of a double bond

at C(2)-C(3) and also at C(4)-C(7). In the latter case, the

stereochemistry must be controlled.

usually straightforward, the same process in the aliphatic series

is often inefficient;³⁶however, we found that the reagent

generated by sonication of sodium borohydride and nickel(II)

chloride hexahydrate³⁷works very well and affords the corre-

sponding amine in high (95%) yield. The latter was then

protected as its allyl carbamate (35 f 36, 82%). To prepare

for construction of the enediyne, the C(5) oxygen function was

deprotected in the usual way with tetrabutylammonium fluoride

(36 f 37, 95%), and finally, Collins oxidation of the resulting

alcohol gave ketone 38 (95%). At this stage, both the 5R and

5â series converge to a single compound, and with the exception

of the double bond cleavage 31 f 32 (73% for the 5R

compound and 77% for the 5â), the yields are excellent

throughout each of the two sequences.

Preparation of the First Monoacetylenic Advanced In-

termediate (42). Treatment of ketone 38 with the lithium salt

of trimethylsilylacetylide gave a complex mixture. However,

when we changed to the cerium salt^38,39(Scheme 11), not only

was the yield high but the reaction was stereoselective, and in

fact, the only product isolated (91%) was 39, in which the

acetylene group is syn to the nitrogen. The stereochemistry of

39 was inferred from the X-ray structure³³of a more advanced

intermediate (see 62) derived by further elaboration of 39.

Acetylide addition is very straightforward, but initially it

presented some difficulties; indeed, the optimized procedure for

generating and for using the cerium reagent (see the Experi-

mental Section) must be followed exactly. The cerium chloride

is first dehydrated by heating at 130 °C under an oil pump

vacuum for 3-12 h. The cooled material is then covered with

The reactions leading to aldehyde 42 were selected after

considerable exploratory work, and in fact, we encountered so

many unexpected difficulties during our examination of potential

routes from the Diels-Alder adduct to calicheamicinone that

we took the precaution of synthesizing an alternative advanced

intermediate (44, Scheme 12) that again brought us to the stage

where attachment of an acetylene and introduction of two double

bonds was required. The relationship between the two routes

is summarized in Scheme 12. Both start with ketone 27, derived

by the Diels-Alder reaction, and the main structural difference

is at the stage of the intermediate ketones 38 and 43. In the

former, C(7), C(8), and C(9) are part of acyclic units; in the

latter, they are in a ring and the stereochemical outcome of the

first acetylide addition is different as the acetylide now enters

anti to the nitrogen (43 f 44).

Preparation of a Second Monoacetylenic Advanced In-

termediate (44). The first three steps from the original Diels-

Alder adduct 27 are the same as before (see Scheme 10, 27 f

28 f 29 f 30). Once again, the two C(5) epimers were

individually subjected to the same reactions (Scheme 13) until

the sequences converge to a single compound. Only the 5â

series is shown in Scheme 13; the yields in parentheses refer to

the 5R series.

(36) (a) For a listing of many available methods, see: Larock, R. C.

ComprehensiVe Organic Transformations; VCH publishers: New York,

1989; p 411. (b) Ganem, B.; Osby, J. O. Chem. ReV. 1986, 86, 763. (c)

Kende, A. S.; Mendoza, J. S. Tetrahedron Lett. 1991, 32, 1699.

(37) (a) Osby, J. O.; Ganem, B. Tetrahedron Lett. 1985, 26, 6413. (b)

Cf. Yoon, N. M.; Choi, J. Synlett 1993, 135. Petrini, M.; Ballini, R.; Rosini,

G. Synthesis 1987, 713.

(38) Stoichiometry: Me₃SiCtCLi/CeCl₃) 1:1.

(39) Cf. (a) Imamoto, T.; Sugiura, Y.; Takiyama, N. Tetrahedron Lett.

1984, 25, 4233. (b) Suzuki, M.; Kimura, Y.; Terashima, S. Chem. Lett.

1984, 1543. (c) Jung, P. M. J.; Burger, A.; Biellmann, J.-F. Tetrahedron

Lett. 1995, 36, 1031. (d) Greeves, N.; Lyford, L. Tetrahedron Lett. 1992,

33, 4759. (e) Bailey, P. D.; Morgan, K. M.; Smith, D. I.; Vernon, J. M.

Tetrahedron Lett. 1994, 35, 7115. (f) Review on organocerium reagents:

Imamoto, T. In ComprehensiVe Organic Synthesis, Trost, B. M., Fleming,

I., Eds.; Pergamon Press: New York, 1991; Vol. 1, p 231.

(40) (a) Dimitrov, V.; Kostova, K.; Genov, M. Tetrahedron Lett. 1996,

37, 6787. (b) Evans, W. J.; Feldman, J. D.; Ziller, J. W. J. Am. Chem. Soc.

1996, 118, 4581.

10338 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

Scheme 13^a

was oxidized to a ketone (49 f 43, 91%). Once again the two

series of 5R and 5â epimers converged to a single ketone, and

for both series, identical reactions have been used, with very

high yields (above 90%) throughout.

From ketone 43, introduction of the acetylene was again

accomplished with the cerium salt of trimethylsilylacetylene,

and we obtained in high yield (91%) alcohol 50, in which the

acetylide had entered anti to the nitrogensa different stereo-

chemical relationship from the first route. Alcohol 50 is

crystalline, and X-ray analysis³³confirmed our structural

assignment and established the conformation as shown in 50′.

The tertiary hydroxyl was protected by silylation (50 f 51,

93%), and the product was treated with DIBAL-H to afford a

mixture of lactols (51 f 52, 96%). Finally, Collins oxidation

gave lactone 44 (97%), bringing the work once more to the

stage of an advanced intermediate where the next tasks were

desaturation at C(2)-C(3) and C(4)-C(7) and the attachment

of an acetylene at C(9).

Introduction of a double bond at C(2)-C(3) involves no

geometrical considerations because the bond will be generated

within a six-membered ring for both 44 and the first advanced

intermediate (42, see Scheme 12) and can have only the desired

geometry. Formation of a double bond at C(4)-C(7), in the

case of compound 44, is likewise free of any geometrical

problems, but introducing the C(4)-C(7) double bond in 42 is

more difficult because there are no obvious constraints to control

the stereochemistry. We decided, therefore, to delay operating

on that part of the molecule until we had incorporated the C(7)-

C(8) chain of 42 into a ring.

The remaining key step for both routes was the attachment

of another acetylene unit at C(9) in the proper stereochemical

sense. In the case of 42, a suitable functional group (-CHO)

was already present to react with an acetylide anion, but with

44 attachment of an acetylene required that C(9) be oxidized

to the aldehyde level. In principle, for both 42 and 44, there is

also the choice of using direct bimolecular reaction with an

acetylide or of first attaching the second acetylenic unit to the

one that is already present. For bimolecular processes, the

stereochemical outcome at C(9) is difficult to predict.

Elaboration of Advanced Intermediate 42: Attachment

of the Second Acetylene Unit and Introduction of the C(4)-

C(7) Double Bond. Treatment (Scheme 14) of aldehyde 42

with the cerium salt of trimethylsilylacetylene, generated exactly

as in our earlier experiments, gave the desired alcohol 53 in

79% yield, together with the C(9) epimer 54 (16%). We tried

to increase the stereoselectivity by running the reaction at an

even lower temperature, but achieved little, if any, improvement.

We also evaluated the corresponding ytterbium acetylide because

there is reason⁴¹to expect that use of ytterbium organometallics

may generally result in greater stereoselection, but in the present

case the reaction was too slow. With the lithium salt of the

acetylene, the yield was very poor. Fortunately, alcohol 54

^aKey: Yields in brackets refer to the 5R series; R ) t-BuMe₂Si;

(a) OsO₄, NaIO₄, CCl₄, H₂O, t-BuOH; 98% (98%); (b) t-BuCOCl,

pyridine, CH₂Cl₂; 96% (96%); (c) NaBH₄, NiCl₂‚6H₂O, MeOH,

sonication; 95% (91%); (d) allyl chloroformate, pyridine, THF; 94%

(93%); (e) Bu₄NF, THF; 97% (95%); (f) PCC, 4 Å sieves CH₂Cl₂;

91% (92%); (g) lithium trimethylsilylacetylide, CeCl₃, THF; 91%; (h)

t-BuMe₂SiOTf, 2,6-lutidine, CH₂Cl₂; 93%; (i) DIBAL-H, CH₂Cl₂; 96%;

(j) CrO₃, pyridine, CH₂Cl₂; 97%.

In the earlier route, the C(9) hydroxyl of 30 was protected

before cleavage of the double bond; this time the double bond

was cleaved (OsO₄, NaIO₄, 98%) without hydroxyl protection,

and the result, as expected, was a mixture of epimeric lactols

(30 f 45). These were then protected using pivaloyl chloride.

In this acylation, a single product is formed in high yield (96%)

from a mixture of epimers, and so it is clear that equilibration

occurs during the reaction and that one epimer reacts faster than

the other, so as to afford a product (46) in which the bulky

pivaloyl group is equatorial.

At this point, we were ready to reduce the nitro group to an

amine, protect the amine, and regenerate the carbonyl at C(5).

All of these tasks were achieved by the same reactions that had

served us well in the first route. Nitro group reduction (46 f

47) with the sodium borohydride-nickel chloride combination

again proceeded in excellent yield (95%), and the amine was

protected under standard conditions with allyl chloroformate

(94%). Then the silicon group was removed by the action of

fluoride ion (48 f 49, 97%), and the liberated hydroxyl at C(5)

(41) Utimoto, K.; Nakamura, A.; Matsubara, S. J. Am. Chem. Soc. 1990,

112, 8189.

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10339

Scheme 15^a

Scheme 14^a

^aKey: AOM ) CH₂OC₆H₄OMe-p; (a) lithium trimethylsilylacetyl-

ide, CeCl₃, THF; 79% for 53, 16% for 54; (b) PCC; (c) NaBH₄, MeOH;

82% from 54.

could be oxidized easily (PCC, ca. 93%) to the corresponding

ketone (55), and then reduction with sodium borohydride gave

(ca 96%) an 11.6:1 mixture of C(9) epimeric alcohols in favor

of the desired isomer 53. The components of the mixture are

easily separated, so that the conversion of 42 into 53 is efficient,

but requires a recycling step.

The next tasks were to incorporate C(7), C(8), and C(9) into

a ring and then to form the C(4)-C(7) double bond. For this

purpose, temporary protection of the C(9) hydroxyl as a

chloroacetate proved ideal. Acylation with chloroacetic anhy-

dride (Scheme 15) gave the expected chloroacetate (53 f 56,

99%), and then exposure to ceric ammonium nitrate served to

remove³⁵the p-anisyloxymethyl group and release the parent

alcohol (56 f 57, 89%). Collins oxidation converted the

alcohol into the corresponding aldehyde (57 f 58, 91%) and

set the stage for generating the desired ring structure.

^aKey: AOM ) CH₂OC₆H₄OMe-p, CLA ) ClCH₂C(O); (a)

(ClCH₂CO)₂O, pyridine; 99%; (b) (NH₄)₂Ce(NO₃)₆, pyridine, MeOH,

H₂O; 89%; (c) CrO₃, pyridine, CH₂Cl₂; 91%; (d) NH₄OH; (e) CrO₃,

pyridine, CH₂Cl₂; 92% from 58; (f) LDA, THF, PhSeBr; dimethyl-

dioxirane; 67%; (g) (Ph₃P)₄Pd, dimedone, THF; 80%.

When chloroacetate 58 was treated with aqueous ammonia,⁴²

the C(9) hydroxyl was released, and it immediately closed onto

the C(8) aldehyde to afford a mixture of lactols which, in turn,

were oxidized efficiently by the Collins reagent (58 f 59 f

60, 92% overall). The double bond at C(4)-C(7) was then

introduced by phenylselenenylation⁴³followed by oxidation with

dimethyldioxirane⁴⁴(60 f 61, 67%). The intermediate phenyl

selenide is rather unstable, and we have no spectroscopic

evidence for its stereochemistry, but assume that the phenylse-

leno group is syn to the C(4) hydrogen, so as to accommodate

the standard syn mechanism of selenoxide fragmentation.

The nitrogen-protecting group was next removed in order to

prepare for desaturating the system at C(2)-C(3). While the

use of tetrakis(triphenylphosphine)palladium(0) is well estab-

lished for the deprotection of allyl esters, the choice of an added

nucleophile in the present case was important, and dimedone⁴⁵

proved best (61 f 62, 80%). Amine 62 is a crystalline

substance; X-ray analysis³³confirmed the desired structure and

showed that the two acetylene pendants are axial to their

respective rings.

Further Elaboration of Advanced Intermediate 42: In-

troduction of the C(2)-C(3) Double Bond. With amine 62

in hand, we came to what was initially a very troublesome job:

introducing the C(2)-C(3) double bond. We had planned to

do that by forming a double bond between C(2) and the nitrogen:

i.e., we would make an imine and then allow the double bond

to move into conjugation. An appropriate method for converting

a primary amine into an imine⁴⁶seemed to be the classical

procedure of N-chlorination followed by base treatment, and

we applied this method to a number of compounds, such as

63-66,⁴⁷as well as to some others that were all made during

our exploratory studies. In these early experiments, we obtained

either complex mixtures or substances that we suspected (but

did not prove) to be aziridines. All of the test compounds

contained the basic six-membered ring (carbons 1-6 of our

target), but differed in the nature of the appendages at C(3),

C(4), and C(5). Eventually, we gained the impression that the

presence of a C(4)-C(7) double bond and incorporation of C(8)

(42) Cf. Reese, C. B.; Stewart, J. C. M.; van Boom, J. H.; de Leeuw, H.

P. M.; Nagel, J.; de Rooy, J. F. M. J. Chem. Soc., Perkin Trans. 1 1975,

934.

(43) Some selenenylation probably also occurs on nitrogen.

(44) Preparation and estimation: Adam, W.; Chan, Y.-Y.; Cremer, D.;

Gauss, J.; Scheutzow, D.; Schindler, M. J. Org. Chem. 1987, 52, 2800.

(45) Kunz, H.; Unverzagt, C. Angew. Chem., Int. Ed. Engl. 1984, 23,

436.

(46) Cf. Dayagi, S.; Degani, Y. In The Chemistry of the Carbon-

Nitrogen Double Bond; Patai, S., Ed.; Interscience: New York, 1970; p

61. See Supporting Information for additional references.

(47) These compounds were used without full characterization.

10340 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

Scheme 16^a

and C(9) into a ring are features that would permit desaturation

at C(2)-C(3) by the N-chlorination method. Careful examina-

tion of our exploratory work then allowed us to select the most

felicitous combination of steps that yielded such a structure,

and this analysis led us to the route shown in Scheme 15 and

described above.

^aKey: (a) t-BuOCl, Et₂O-THF; (b) DABCO, PhMe; (c) Cl₃COCOO-

CCl₃, pyridine, CH₂Cl₂; MeOH; 78% from 62.

Scheme 17

Exposure of amine 62 to freshly prepared tert-butyl hy-

pochlorite⁴⁸gave the N-chloro compound 67 (Scheme 16),

which was immediately treated at room temperature with a

5-fold excess of DABCO or DBU. Those operations gave the

fully conjugated enamine as an 8:1 mixture of C(9) epimers

(67 f 68, 78% from 62). We suspect (on the basis of a single

experiment) that the use of a larger amount of DABCO gives

an improved ratio, but it turned out that the epimerization is

not so important, as the stereochemistry can be corrected easily

(see below). The enamine is unstable and must be used without

delay. It was, therefore, converted directly into the correspond-

ing isocyanate by treatment with triphosgene, and the isocyanate

was quenched with methanol.^2a,b,49That experiment gave the

methyl carbamate 69 in almost 80% yield from the parent amine

62, again as an 8:1 mixture of C(9) epimers. Neither of the

last two epimer mixtures were chromatographically separable,

but separation and stereochemical correction were easily

achieved at the next stage. We assume that the epimerization,

which can take place with both 68 and the derived carbamate

(69) in the presence of an acid or a base, occurs by the

mechanism summarized in Scheme 17.

Scheme 18^a

Having reached carbamate 69, we had next to remove the

acetylenic silyl groups in order to attempt formation of the

enediyne ring. The desilylation was done with tetrabutylam-

monium fluoride (Scheme 18) under standard conditions, but

during the course of the reaction there was further epimerization

at C(9) (69 f 71 and 72), attributable to the basic nature of the

reagent. Starting from 69 (as an 8:1 mixture in favor of the

syn isomer shown), it was possible to isolate the desired syn

bis-acetylene 72 in 46% and the anti bis-acetylene 71 in 39%

yield. However, when the anti compound is treated with

tetrabutylammonium acetate, it is converted quantitatively into

a 6:4 mixture of the two epimers, with the desired one

predominating. The syn and anti bis-acetylenes are easily

separated, and so, after one recycling, the yield for conversion

of 69 into 72 is 69%.

^aKey: (a) Bu₄NF, THF; 46% for 72, 39% for 71; 69% for 72 after

one recycling; (b) Bu₄NOAc; 100%, 71:72 ) 4:6.

An X-ray structure determination³³showed that the distance

between the terminal acetylenic carbons of 72 is 4.154 ( 0.005

Å.

With a small supply of 72 in hand, the stage was set for

construction of both the enediyne ring and the trisulfide,

followed by removal of residual protecting groups. At about

this time, however, we completed a different route to the syn

bis-acetylene, and it was material from this new source that

was actually taken on to calicheamicinone.

(48) (a) Mintz, M. J.; Walling, C. Organic Syntheses; Wiley & Sons:

New York, 1973, Collect. Vol. V, p 184. (b) Hazard warning: Organic

Syntheses; Wiley & Sons: New York, 1973, Collect. Vol. V, p 183.

(49) Cf. Eckert, H.; Forster, B. Angew. Chem., Int. Ed. Engl. 1987, 26,

894.

Elaboration of Advanced Intermediate 44: Introduction

of the C(4)-C(7) Double Bond, Attachment of the Second

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10341

Scheme 20^a

Scheme 19^a

^aKey: R ) t-BuMe₂Si. (a) Lithium trimethylsilylacetylide, CeCl₃,

THF; 91%; (b) Bu₄NF, THF; 46% for 72, 42% for the C(9) epimer

(71), 71% for 72, after one recycling of syn bis(acetylene); (c) NIS,

AgNO₃, acetone; 89%; (d) (Z)-1,2-bis(trimethylstannyl)ethene, (Ph₃P)₄Pd,

DMF; 72%.

we treated 77 with N-bromosuccinimide under standard free

radical conditions we obtained a mixture of bromides in good

yield (77 f 78). These could be hydrolyzed with aqueous silver

nitrate to an aldehyde acid,⁵⁰which is easily trapped in the form

of its methyl ester 79 by reaction with diazomethane. During

the hydrolysis, the acetylenic trimethylsilyl group is lost, and

that was very convenient, as we would have had to remove it

anyway. The overall yield for the steps from 77 to 79 was

77%.

^aKey: R ) t-BuMe₂Si. (a) LDA, THF, PhSeBr; dimethyldioxirane,

CH₂Cl₂, acetone; 85% + 5% 73; (b) (Ph₃P)₄Pd, dimedone, THF; 93%;

(c) t-BuOCl, Et₂O-THF; DBU, PhMe; 81%; (d) Cl₃COCOOCCl₃,

pyridine, CH₂Cl₂; MeOH; 91%; (e) NBS, (PhCO)₂O₂, light, CCl₄; (f)

AgNO₃, THF, H₂O, pyridine; CH₂N₂; 77% from 77.

The next task was to introduce the second acetylenic unit

(Scheme 20). Fortunately, the potential problem of stereo-

chemical control did not arise, because treatment of aldehyde

79 with the cerium salt of trimethylsilylacetylene gave in high

yield (91%) the desired lactone 80, which forms by stereose-

lective acetylide addition to the aldehyde and lactonization of

the resulting alkoxide. Removal of the acetylenic trimethylsilyl

group with tetrabutylammonium fluoride afforded the syn bis-

acetylene 72 in 46% yield and almost as much (42%) of the

anti isomer 71 (see Scheme 20). Treatment of that material

with tetrabutylammonium acetate gave more of the desired syn

compound; with one such recycling, the yield of 72 is 71%.

The acetylenic hydrogens were next replaced by iodine

(Scheme 20, 72 f 81, 89%), using N-iodosuccinimide with a

catalytic amount of silver nitrate,⁵¹and we were ready to close

the enediyne ring by means of a double Stille coupling.⁵²Such

couplings had been used⁵³on a few occasions to generate large

rings, and one example was actually reported⁵⁴in the case of

an enediyne (dynemicin) shortly before our own experiment.⁵⁵

A DMF solution of freshly distilled (Z)-1,2-bis(trimethylstan-

Acetylene Unit, and Completion of the Synthesis. As

indicated earlier, we developed an alternative route that proceeds

by way of the advanced intermediate 44. In both routes, we

deal with racemic compounds, but represent each of them by a

single enantiomer. It is permissible, therefore, to depict

compound 44 by the enantiomeric form 73 (see Scheme 19),

so that elaboration to the syn bis-acetylene 72 (see Schemes 19

and 20) is more conveniently described; at the stage of the syn

bis-acetylene, both approaches via 42 and 73 converge to a

single compound from which (()-calicheamicinone is easily

reached.

Compound 73 was desaturated at C(4)-C(7) by the method

that had served us well in the first route. Phenylselenenylation

under standard conditions gave an unstable selenide, and as

before, we assumed that the phenylseleno group was on the

same face as the C(4) hydrogen. Oxidation with dimethyldiox-

irane then afforded the selenoxide, and that, in turn, collapsed

to generate the required double bond [73 f 74, 85%, uncor-

rected for recovered 73 (5%)]. Removal of the allyloxycarbonyl

group (74 f 75) by the action of tetrakis(triphenylphosphine)-

palladium(0) in the presence of dimedone again liberated the

amino group in high yield (93%), and that group was chlorinated

with tert-butyl hypochlorite. Prompt exposure to a hindered

basesin this case DBUsgave an imine which immediately

isomerized to the unstable conjugated enamine (75 f 76, 81%).

For making the derived methyl carbamate, we again used

triphosgene and quenched the resulting isocyanate with methanol

(76 f 77, 91%).

(50) The compound exists as two hydroxy lactones, epimeric at C(9).

(51) (a) Hofmeister, H.; Annen, K.; Laurent, H.; Wiechert, R. Angew.

Chem., Int. Ed. Engl. 1984, 23, 727. (b) For a listing of other methods,

see: Brunel, Y.; Rousseau, G. Tetrahedron Lett. 1995, 36, 2619. (c) Bis-

iodide 81 can be made directly from 69 (see Scheme 16), but desilylation

in a separate step (69 f 72 f 81) is more efficient overall: Cf. Nishikawa,

T.; Shibuya, S.; Hosokawa, S.; Isobe, M. Synlett 1994, 485.

(52) Farina, V.; Krishnamurthy, V. Org. React. (N.Y.) 1997, 50, 1.

(53) E.g. (a) Pattenden, G.; Thom, S. M. Synlett 1993, 215. (b) Nicolaou,

K. C.; Chakraborty, T. K.; Piscopio, A. D.; Minowa, N.; Bertinato, P. J.

Am. Chem. Soc. 1993, 115, 4419. (c) Barrett, A. G. M.; Boys, M. L.; Boehm,

T. L. J. Chem. Soc., Chem. Commun. 1994, 1881.

To introduce an acetylene at C(9), it was necessary to oxidize

that carbon, and the oxidation was accomplished by taking

advantage of the fact that the C(9) hydrogens are allylic. When

(54) Shair, M. D.; Yoon, T.; Danishefsky, S. J. J. Org. Chem. 1994, 59,

3755.

10342 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

Scheme 21^a

Further reduction (84 f 85, 76%) with sodium borohydride

gave a triol, and the secondary hydroxyl was protected by a

two-step procedure. Treatment with triethylsilyl triflate served

to place a silicon group on both the C(8) and C(9) oxygens (85

f 86, 95%), and when the doubly silylated material was stored

for a short time in a mixture of acetic acid, THF, and water at

0 °C, the primary allylic alcohol was released (86 f 87, 94%),

but the other silicon unit was not disturbed.

The primary hydroxyl was next replaced by a thioacetyl group

(87 f 88, 94%)^2aby using typical Mitsunobu conditions, and

the free thiol was then liberated by the action of DIBAL-H (88

f 89).

The next step, elaboration to the trisulfide, required that

freshly chromatographed thiol be used, and as soon as that was

established, we found that treatment with a large excesssabout

8-foldsof freshly crystallized N-(methyldithio)phthalimide^2a,19d-f

then gave the desired trisulfide⁵⁸90 in nearly 90% yield from

the thioacetate (88).

Finally, mild acid hydrolysis served to disengage the remain-

ing two protecting groups and release racemic calicheamicinone

as a white foam (90 f 1, 84%).

Conclusion

The two syntheses illustrate the remarkable level of stereo-

selectivity that can be obtained with alkynyl cerium reagents

and also show the usefulness of a set of oxygen protecting

groups, especially the p-anisyloxymethyl group.

In 42 and 44, the stereochemical relationship between the

acetylene unit at C(5) and the nitrogen at C(2) is different, but

the only stereogenic center that is preserved after elaboration

to (()-calicheamicinone is C(5); consequently, both compounds

give the same racemic end product. The stereochemical

relationship between the two routes confers on the present work

the very unusual characteristic that, with corresponding optically

pure compounds, either enantiomer of calicheamicinone can be

reached irrespectiVe of the absolute configuration of the initial

Diels-Alder adduct.⁵⁹In a potential synthesis of (-)-cali-

cheamicinone (of natural stereochemistry, as actually depicted

in diagram 1), intermediates corresponding to the Diels-Alder

adduct 27 with a (2R) absolute configuration would have to be

processed by the methods based on 42, while the route via 44

would be used for the (2S) isomer. We also note the curious

fact that a calicheamicinone of unnatural stereochemistry is

probably⁶⁰more efficient at producing double strand cuts in

DNA than material of natural stereochemistry.

^aKey: (a) DIBAL-H, CH₂Cl₂; 98%; (b) Bu₄NF, THF; 94%; (c)

NaBH₄, MeOH; 76%; (d) Et₃SiOTf, 2,6-lutidine; 95%; (e) AcOH, THF,

H₂O; 94%; (f) i-PrO₂CNdNCO₂Pr-i, Ph₃P, AcSH; 94%; (g) DIBAL-

H, CH₂Cl₂; (h) N-(Methyldithio)phthalimide (9), CH₂Cl₂; 88% from

88; (i) TsOH‚H₂O, THF, H₂O; 84%.

nyl)ethene⁵⁶was added slowly to a warm solution of the bis-

iodide 81 and a catalytic amount of tetrakis(triphenylphosphine)-

palladium(0) in the same solvent. A few hours after the

addition, it was possible to isolate the cyclic enediyne 82 as a

white solid in 72% yield^52,57(Scheme 20).

By the time we had reached this point, publications from the

laboratories of Danishefsky^2aand Nicolaou^2breported how

compounds very similar to 82sonly the status of the tertiary

hydroxyl was differentscould be converted into calicheamici-

none, and we used a method similar to theirs.^2a,b,20

We first reduced the lactone to a mixture of lactols, using

DIBAL-H (Scheme 21, 82 f 83, 98%), and then removed the

silicon group (83 f 84) in the usual way (Bu₄NF, 94%). The

timing of these simple operations is critical. The silicon group

has to be removed before making the trisulfide, because the

trisulfide is not stable to tetrabutylammonium fluoride, at least

under conditions needed for deprotection of a tert-butyldim-

ethylsilyloxy ether. And the DIBAL-H reduction must be done

before desilylation, because if the silicon is removed first, the

product is not soluble enough in dichloromethane, which is the

solvent we use for the reduction.

Experimental Section

General Procedures. The same general procedures as used

previously⁶¹were followed. DMF was stirred overnight with crushed

CaH₂and then distilled under a water pump vacuum with protection

from moisture.

The symbols s′, d′, t′, and q′ used for ¹³C NMR signals indicate

zero, one, two, or three attached hydrogens, respectively. In cases

where the number of signals is less than expected, we assumed that

this was due to coincident chemical shifts.

(55) Recent examples of the use of Stille coupling for construction of

large rings: (a) Boden, C.; Pattenden, G. Synlett 1994, 181. (b) Smith, A.

B., III.; Condon, S. M.; McCauley, J. A.; Leazer, J. L., Jr.; Leahy, J. W.;

Maleczka, R. E., Jr. J. Am. Chem. Soc. 1995, 117, 5407. (c) Boyce, R. J.;

Pattenden, G. Tetrahedron Lett. 1996, 37, 3501. (d) Critcher, D. J.;

Pattenden, G. Tetrahedron Lett. 1996, 37, 9107. (e) Paterson, I.; Man, J.

Tetrahedron Lett. 1997, 38, 695.

(56) Mitchell, T. N.; Amamria, A.; Killing, H.; Rutschow, D. J.

Organomet. Chem. 1986, 304, 257.

(57) We have not examined the effect of additives or ligands (see the

Supporting Information for references) other than triphenylphosphine.

(58) ¹H NMR chemical shifts for MeS_xMe (x ) 2-4) are characteristic:

MeSSMe (δ 2.41), MeSSSMe (δ 2.56), MeSSSSMe (δ 2.64) (Douglas, I.

B.; Douglas, M. L. Sulfur Rep. 1995, 17, 129.) See also ref 5a. Our product

had δ 2.54 (CDCl₃).

(59) Cf. Clive, D. L. J.; Selvakumar, N. J. Chem. Soc., Chem. Commun.

1996, 2543.

(60) Such a difference has been found using an analogue of 1 (in which

the group SSSMe is replaced by SAc): See ref 2c.

(61) Clive, D. L. J.; Wickens, P. L.; Sgarbi, P. W. M. J. Org. Chem.

1996, 61, 7426.

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10343

Compound numbers followed by “R” refer to the stereochemistry

at C(5), the substituent being below the plane of the paper. The R-series

is the minor isomer series arising from hydride reduction of the C(5)

carbonyl in 27, and the experimental procedures, which are identical

to those for the 5â series, are given in the Supporting Information.

Procedures for the Advanced Intermediate (30) Common to Both

Routes. 2-Chloroethyl 3-Oxo-6-heptenoate (24). 2-Chloroethanol

(750 mL, 11.19 mol) was added to methyl 3-oxo-6-heptenoate (23)

(53 g., 0.34 mol), and Ti(OPr-i)₄(5 mL, 16.9 mmol) was added to the

resulting solution. The flask was closed with a CaSO₄guard tube,

and the mixture was stirred at 55 °C for 16 h and then at 75 °C for 24

h. The solvent was then distilled off under reduced pressure (ca. 50

°C, water pump vacuum), and the residue was filtered through a pad

(5.5 × 20 cm) of silica gel using 3:7 EtOAc-hexane. Evaporation of

the solvents and distillation of the yellowish residue under vacuum

(0.02 mmHg) gave a colorless liquid, which contained (¹H NMR, 200

MHz) ca. 15% of the starting ester. Fractional distillation gave first

the starting material (bp 40-50 °C, 0.02 mmHg) and then 24 as a pure

(¹H NMR, 200 MHz) colorless liquid (43.8 g, 63%).

temperature (ca. 20 min) and then was stirred for a further 10 min.

The solution was then cooled in an ice bath, and methyl (E)-3-

nitropropenoate (11.94 g, 91.1 mmol) in dry THF (12 mL plus 3 mL

as a rinse) was added by cannula over 5 min. The ice bath was

removed, and the mixture was stirred for 1.5 h. At this point, it was

quenched by addition of saturated aqueous NH₄Cl (500 mL). The

resulting mixture was stirred for 2 h and then extracted with EtOAc (3

× 300 mL). The combined organic extracts were washed with brine

(1 × 200 mL) and dried (NaSO₄). Evaporation of the solvent and flash

chromatography of the residue over silica gel (10 × 40 cm), using 1:4

EtOAc-hexane, followed by crystallization from EtOAc-hexane gave

pure (¹H NMR, 300 MHz) 27 (12.73, 56% overall from ketene acetal

25) as large colorless crystals. The crystallization was done by covering

the solid with EtOAc and warming the mixture (ca. 50 °C); more solvent

was added until the solid just dissolved. The solution was then allowed

to cool to room temperature; finally, the solution was kept overnight

in a refrigerator. The compound can also be crystallized from 1:4

EtOAc-hexane by dissolving the solid in EtOAc and then adding

hexane until turbidity is observed. The solution is then left at room

temperature.

The 2-chloroethanol was recycled. We suspect that the yield could

have been raised if, at the end of ca. 12 h, some of the MeOH was

distilled off, fresh batches of 2-chloroethanol and Ti(OPr-i)₄were added,

and the reaction was allowed to continue.

An X-ray structure was determined on 27 crystallized from 1:4

EtOAc-hexane.³³

Methyl (6r,7â,8â,9r)- and (6r,7â,8â,9â)-9-Hydroxy-6-nitro-8-

(2-propenyl)-1,4-dioxaspiro[4.5]decane-7-carboxylate (28). NaBH₄

(1.39 g, 36.75 mmol) was added in one portion to a stirred and cooled

(0 °C) solution of 27 (10.00 g, 33.41 mmol) in MeOH (200 mL), and

the mixture was stirred at 0 °C for 10 min. Water (400 mL) was added,

and the aqueous solution was extracted with EtOAc (4 × 200 mL).

The combined organic extracts were dried (MgSO₄) and evaporated.

Flash chromatography of the residue over silica gel (5.5 × 20 cm),

using 1:1 EtOAc-hexane, gave a 2:1 mixture of 28 and the C(5) epimer

28R as a white solid (10.02 g, 99%).

Methyl (6r,7â,8â,9â)-9-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-

6-nitro-8-(2-propenyl)-1,4-dioxaspiro[4.5]decane-7-carboxylate (29)

and Methyl (6r,7â,8â,9r)-9-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-

6-nitro-8-(2-propenyl)-1,4-dioxaspiro[4.5]decane-7-carboxylate (29r).

t-BuMe₂SiOSO₂CF₃(11.4 mL, 49.7 mmol) was added over ca. 10 min

(syringe pump) to a stirred solution of 28 and 28R (10.00 g, 33.19

mmol) and 2,6-lutidine (7.7 mL, 66.11 mmol) in CH₂Cl₂(300 mL).

Stirring was continued for 2 h at room temperature, and water (400

mL) was then added. The mixture was extracted with EtOAc (3 ×

150 mL), and the combined organic extracts were washed with brine

(300 mL), dried (Na₂SO₄), and evaporated. Flash chromatography of

the residue over silica gel (5.5 × 38 cm), using 1:20 EtOAc-hexane,

gave 29 (8.97 g, 65%) and the C(5) epimer 29R (4.50 g, 32%). Both

compounds are solids.

1-(1,3-Dioxolan-2-ylidene)-5-hexen-2-one (25). A solution of 24

(20.4 g, 0.1 mol) in dry DMF (40 mL) was added to a magnetically

stirred suspension of K₂CO₃(Aldrich item no. 34,782-5, -325 mesh,

98%, dried overnight at 80 °C under oil pump vacuum, 16.58 g, 0.12

mol) in dry DMF (80 mL, Argon atmosphere). Stirring was continued

for 9 h, and the suspension was then allowed to settle. The supernatant

liquid was transferred by cannula to another flask (protection from

moisture, Ar atmosphere). Dry Et₂O (20 mL) was added to the solid

residue, and the mixture was stirred for 2 min and again allowed to

settle. The supernatant Et₂O layer was transferred to the other flask

as before. (We did not repeat the Et₂O rinse, but in future experiments

would do so.) The Et₂O-DMF solution was evaporated by connection,

via a bent adaptor with stopcock, to a large trap cooled in liquid nitrogen

and connected to an oil pump. The evaporation was done first (briefly)

at room temperature (for removal of the Et₂O) and then at 50 °C (oil

bath). From this point on, do not expose the compound to air, only

to argon. [Some of us who have done this experiment used a rotary

evaporator (with provision for protection from moisture), but felt that

the compound obtained was then less pure.] Kugelrohr distillation of

the residue (diffusion pump, 0.001 mmHg, oven temperature 116-

125 °C) gave 25 (13.44 g, 80%) as a pure (¹H NMR, 200 MHz)

colorless liquid, which solidified in a freezer but melted again at room

temperature. The compound can be kept in a refrigerator for several

weeks. Brief exposure to air is not harmful after the distillation.

Note: During the distillation, the pressure must be at least 0.001

mmHg in order to obtain a white distillate.

[[1-(1,3-Dioxolan-2-ylidenemethyl)-1,4-pentadienyl]oxy]trimeth-

ylsilane (20) and Methyl (6r,7â,8aâ)-6-Nitro-9-oxo-8-(2-propenyl)-

1,4-dioxaspiro[4.5]decane-7-carboxylate (27). (a) 1,1,3,3-Tetramethyl-

1,3-disilazane [(Me₂PhSi)₂NH]. NH₃was passed for 6 h directly from

an ammonia tank (via a Dreschel safety bottle) into a mechanically

stirred solution of commercial Me₂PhSiCl (200 g, 1.17 mol) in dry

PhH (1200 mL). During this time, NH₄Cl precipitated. The mixture

was filtered, and the solid was washed with bench benzene (300 mL).

Evaporation of the solvent and distillation of the residue (oil pump,

0.5 mmHg) gave 1,1,3,3-tetramethyl-1,3-disilazane (150.5 g, 90%).

(b) [[1-(1,3-Dioxolan-2-ylidenemethyl)-1,4-pentadienyl]oxy]tri-

methylsilane (20) and Methyl (6r,7â,8aâ)-6-Nitro-9-oxo-8-(2-pro-

penyl)-1,4-dioxaspiro[4.5]decane-7-carboxylate (27). n-BuLi (2.5 M,

32.8 mL, 82 mmol) was added over ca. 10 min to a stirred and cooled

(ice bath) solution of (Me₂PhSi)₂NH (24.4 g, 86 mmol) in dry THF

(300 mL) (Ar atmosphere). Stirring was continued for a further 15

min, and then the ice bath was replaced by a dry ice-acetone bath. A

solution of 20 (12.76 g, 76 mmol) in THF (50 mL) was added by

cannula over 1 h (stirring). The mixture was stirred for a further 10

min and then quenched by rapid addition (over ca. 1 min) of Me₃SiCl

(fresh Aldrich material, used as received, 12 mL, 95 mmol). The cold

bath was removed, and the solution was allowed to attain room

An X-ray structure was determined on 29R crystallized from

EtOAc-hexane.³³

(6r,7â,8â,9â)-9-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-6-nitro-

8-(2-propenyl)-1,4-dioxaspiro[4.5]decane-7-methanol (30). DIBAL-H

(1.0 M in CH₂Cl₂, 39.0 mL, 39.0 mmol) was added over ca. 10 min

(syringe pump) to a stirred and cooled (-78 °C) solution of 29 (6.50

g, 15.64 mmol) in dry CH₂Cl₂(200 mL). Stirring at -78 °C was

continued for 40 min, and then Na₂SO₄‚10H₂O (100 g) and Celite (20

g) were added sequentially at -78 °C. The cold bath was removed,

and the mixture was stirred for 1 h. The mixture was then filtered

through a pad of Celite (5 × 1 cm) with EtOAc. Evaporation of the

filtrate and flash chromatography of the residue over silica gel (4.5 ×

17 cm), using 1:3 EtOAc-hexane, gave 30 (6.00 g, 99%) as a colorless

oil.

Procedures for route via 44 (≡73). (3r,4ar,5â,8r,8ar)- and (3r,-

4aâ,5r,8â,8aâ)-5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]hexahydro-

8-nitrospiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-3-ol (45). OsO₄

(2.5% w/w in t-BuOH, 4.0 mL, 0.319 mmol) was added to a stirred

solution of 30 (2.13 g, 5.5 mmol) in 2:2:1 CCl₄-water-t-BuOH (75

mL) (the starting material was dissolved in CCl₄-t-BuOH, and the water

was added last). The mixture was stirred for 15 min, and NaIO₄(2.93

g, 13.75 mmol) was added in one portion. After 3 h, the resulting

suspension was diluted with water (20 mL) and extracted with Et₂O

(400 mL). The organic extract was washed with water (20 mL), 10%

aqueous NaHSO₃(20 mL), and water (20 mL), dried (MgSO₄), and

10344 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

evaporated. Flash chromatography of the residue over silica gel (6 ×

18 cm), using 1:3 to 3:7 EtOAc-hexane, gave 45 (2.10 g, 98%) as a

white solid consisting of a 3:1 mixture [¹H NMR (400 MHz)] of

anomeric lactols.

7H₂O used in this experiment was washed with THF and air-dried to

give white material.

Hydrated cerium chloride (CeCl₃‚7H₂O, 11.70 g, 31.35 mmol)

contained in a round-bottomed flask closed by a bent adaptor with a

tap was dried for 3 h (oil bath at 130 °C, 0.05 mmHg) with magnetic

stirring and then was cooled to room temperature (protection from

moisture).

THF (50 mL) was added with vigorous stirring to the resulting

anhydrous⁴⁰CeCl₃, and the suspension was sonicated (Branson, model

B-12, 80 W) overnight to obtain a fine suspension free of any lumps,

which was then cooled to -78 °C. Freshly made lithium (trimethyl-

silyl)acetylide in THF [prepared by addition of n-BuLi (1.6 M in

hexane, 16.3 mL, 26.0 mmol) to (trimethylsilyl)acetylene (Aldrich, 4.08

mL, 28.60 mmol) in THF (50 mL) at -78 °C, followed by stirring for

15 min at this temperature] was added to the above suspension by

cannula over ca. 10 min. After the addition, stirring was continued at

-78 °C for 30 min. During this procedure, the suspension remained

white.

The above organocerium reagent in THF was added by cannula over

ca. 15 min to a stirred and cooled (-78 °C) solution of 43 (1.07 g,

2.60 mmol) in THF (30 mL). Stirring was continued for 30 min, and

the reaction was quenched by addition of saturated aqueous NH₄Cl

(50 mL) at -78 °C. The cold bath was removed, and the mixture was

stirred for 30 min. The mixture was extracted with EtOAc (3 × 100

mL), and the combined organic extracts were washed with brine (50

mL) and dried (Na₂SO₄). (If an emulsion forms, more water is added;

should the emulsion not crack, the mixture is then filtered through a

pad of Celite.) Evaporation of the solvent and flash chromatography

of the residue over silica gel (4 × 10 cm), using 3:2 EtOAc-hexane,

gave 50 (1.21 g, 91%) as a white solid. An X-ray structure was

determined on 50 crystallized from EtOH.³³

(3r,4aâ,5â,8â,8aâ)-5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-

hexahydro-8-[[(2-propenyloxy)carbonyl]amino]-5-[(trimethylsilyl)-

ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-3-yl 2,2-

Dimethylpropanoate (51). t-BuMe₂SiOTf (1.35 mL, 5.60 mmol) was

added at a fast, dropwise rate from a syringe to a stirred and cooled

(ice bath) solution of 50 (2.58 g, 5.06 mmol) and 2,6-lutidine (1.33

mL, 11.20 mmol) in dry CH₂Cl₂(15 mL). Stirring was continued for

30 min (ice bath) (TLC control; silica, 1:4 EtOAc-hexane), and then

MeOH (2 mL) was added, followed by Et₂O (300 mL). The mixture

was washed with 10% aqueous CuSO₄(2 × 30 mL) and then with

brine (50 mL). The organic extract was dried (Na₂SO₄) and evaporated.

Flash chromatography of the residue over silica gel (4 × 16 cm), using

3:17 EtOAc-hexane, gave 51 (2.94 g, 93%) as a pure (¹H NMR, 400

MHz), white solid.

2-Propenyl (3r,4aâ,5â,8â,8aâ)- and (3r,4ar,5r,8r,8ar)-[5-[[(1,1-

Dimethylethyl)dimethylsilyl]oxy]hexahydro-3-hydroxy-5-

[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-

8-yl]carbamate (52). DIBAL-H (1.0 M solution in CH₂Cl₂, 13.4 mL,

13.4 mmol) was added dropwise by syringe over ca. 15 min to a stirred

and cooled (-78 °C) solution of 51 (2.78 g, 4.46 mmol) in CH₂Cl₂

(100 mL). Stirring at -78 °C was continued for 30 min, and then

MeOH (2 mL) was added, followed successively by Na₂SO₄(2 g),

Celite (5 g), and water (2 mL). The cold bath was removed, and stirring

was continued for 30 min. The mixture was then filtered through a

sintered disk, and the solid was washed with EtOAc. Evaporation of

the filtrate and flash chromatography of the residue over silica gel (4

× 15 cm), using 2:3 EtOAc-hexane, gave 52 (2.31 g, 96%) as a pure

(TLC, silica, 1:1 EtOAc-hexane), white solid.

2-Propenyl (4ar,5r,8r,8ar)-[5-[[(1,1-Dimethylethyl)dimethylsi-

lyl]oxy]hexahydro-3-oxo-5-[(trimethylsilyl)ethynyl]spiro[1H-2-ben-

zopyran-7(3H),2′-[1,3]dioxolan]-8-yl]carbamate (44). Note that 44

≡ 73. CrO₃(6.53 g, 65.3 mmol) was added in portions over ca. 5 min

to a stirred solution of pyridine (10.5 mL, 130.6 mmol) in dry CH₂Cl₂

(50 mL) at room temperature. Stirring was continued for 10 min, and

the resulting Collins reagent was poured (over ca. 1 min) into a stirred

solution of 52 (2.20 g, 4.08 mmol) in dry CH₂Cl₂(50 mL). Stirring

was continued for 20 min, and the mixture was then filtered through a

column (4 × 18 cm) of flash silica gel, using a 2:3 mixture of EtOAc-

hexane. Evaporation of the solvent gave 44 (2.12 g, 97%) as a pure

(¹H NMR, 400 MHz), white solid.

(3r,4aâ,5r,8â,8aâ)-5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-

hexahydro-8-nitrospiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-

3-yl 2,2-Dimethylpropanoate (46). t-BuCOCl (6.68 mL, 55.66 mmol)

was injected over ca. 20 min into a stirred solution of lactols 45 (4.33

g, 11.12 mmol) and dry pyridine (9.10 mL, 111.3 mmol) in dry CH₂-

Cl₂(100 mL). Stirring was continued at room temperature for 30 h,

and then saturated aqueous NaHCO₃(100 mL) was added. The organic

phase was separated, and the aqueous phase was extracted with CH₂-

Cl₂(2 × 100 mL). All the organic phases were combined and dried

(Na₂SO₄). Evaporation of the solvent and flash chromatography of

the residue over silica gel (4 × 30 cm), using 3:17 EtOAc-hexane,

gave 46 (5.06 g, 96%) as a pure (¹H NMR, 400 MHz) solid.

(3r,4aâ,5r,8â,8aâ)-8-Amino-5-[[(1,1-dimethylethyl)dimethylsilyl]-

oxy]hexahydrospiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-3-

yl 2,2-Dimethylpropanoate (47). Bench MeOH (100 mL) was added

to a flask containing NiCl₂‚6H₂O (1.24 g, 5.23 mmol), and the mixture

was sonicated open to the air for 30 min (Branson, model B-12, 80

W). Then NaBH₄(594 mg, 15.71 mmol) was added in one portion,

and sonication was continued for 30 min. A solution of 46 (4.95 g,

10.45 mmol) in bench MeOH (100 mL plus 20 mL as a rinse) was

added over 30 min from a dropping funnel (continued sonication, no

protection from air). After the addition, more NaBH₄(2.77 g, 73.27

mmol) was added in portions over 30 min (continued sonication).

Sonication was continued for 15 min after the end of the addition, and

the solvent was then evaporated at room temperature. The residue was

immediately filtered through a column (4 × 25 cm) of flash chroma-

tography silica gel, using 9:1 CH₂Cl₂-MeOH. Evaporation of the fil-

trate and flash chromatography of the residue over silica gel (4 × 30

cm), using 9:1 CH₂Cl₂-MeOH, gave 47 (4.41 g, 95%) as a pure (¹H

NMR, 400 MHz) solid, which must be used the same day it is made.

(3r,4aâ,5r,8â,8aâ)-5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-

hexahydro-8-[[(2-propenyloxy)carbonyl]amino]spiro[1H-2-benzopy-

ran-7(3H),2′-[1,3]dioxolan]-3-yl 2,2-Dimethylpropanoate (48). Allyl

chloroformate (Aldrich material, used directly, 1.48 mL, 13.92 mmol)

was injected dropwise over ca. 15 min into a stirred and cooled (ice

bath) solution of 47 (4.11 g, 9.26 mmol) and dry pyridine (1.51 mL,

18.56 mmol) in dry THF (100 mL). After the addition, the cold bath

was removed, and stirring was continued for 1 h. EtOAc (200 mL)

was added, and the mixture was washed successively with 1:1 brine-

water (50 mL), and brine (50 mL). The organic phase was dried (Na₂-

SO₄) and evaporated. Flash chromatography of the residue over silica

gel (4 × 26 cm), using 3:7 EtOAc-hexane, gave pure (¹H NMR, 400

MHz) 48 (4.59 g, 94%) as a foam.

(3r,4aâ,5r,8â,8aâ)-Hexahydro-5-hydroxy-8-[[(2-propenyloxy)-

carbonyl]amino]spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-3-

yl 2,2-Dimethylpropanoate (49). Bu₄NF (1.0 M in THF, 10.2 mL,

10.2 mmol) was added dropwise from a syringe over 10 min to a stirred

solution of 48 (4.49 g, 8.51 mmol) in THF (100 mL). The mixture

was stirred for 5 h, and the solvent was then evaporated. EtOAc (300

mL) was added, and the solution was washed with 1:1 brine-water

(50 mL) and brine (50 mL). The organic extract was dried (Na₂SO₄)

and evaporated. Flash chromatography of the residue over silica gel

(4 × 20 cm), using 4:1 EtOAc-hexane, gave 49 (3.42 g, 97%) as a

pure (¹H NMR, 400 MHz) white solid.

(3r,4aâ,8â,8aâ)-Hexahydro-5-oxo-8-[[(2-propenyloxy)carbonyl]-

amino]spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-3-yl 2,2-Dim-

ethylpropanoate (43) from 49. A mixture of PCC (10.4 g, 48.23

mmol) and powdered 4 Å molecular sieves (2.0 g) was tipped into a

stirred solution of 49 (3.32 g, 8.03 mmol) in dry CH₂Cl₂(100 mL),

and stirring was continued for 2 h. The mixture was then filtered

through a column of flash chromatography silica gel (4 × 22 cm), using

1:1 EtOAc-hexane, to afford 43 (3.01 g, 91%) as a pure (¹H NMR,

400 MHz) solid. The filtration must be done quickly to avoid

epimerization R to the ketone carbonyl.

(3r,4aâ,5â,8â,8aâ)-Hexahydro-5-hydroxy-8-[[(2-propenyloxy)-

carbonyl]amino]-5-[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-

7(3H),2′-[1,3]dioxolan]-3-yl 2,2-Dimethylpropanoate (50). The CeCl₃‚

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10345

2-Propenyl (5r,8r,8ar)-[5-[[(1,1-Dimethylethyl)dimethylsilyl]-

oxy]-5,6,8,8a-tetrahydro-3-oxo-5-[(trimethylsilyl)ethynyl]-

spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-8-yl]carbamate (74).

(a) 2-Propenyl (4r,4ar,5r,8r,8ar)-[5-[[(1,1-Dimethylethyl)dimeth-

ylsilyl]oxy]hexahydro-3-oxo-4-(phenylseleno)-5-[(trimethylsilyl)-

ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-8-yl]car-

bamate. n-BuLi (1.6 M in hexane, 2.44 mL, 3.91 mmol) was added

dropwise over ca. 10 min to a stirred and cooled (ice bath) solution of

i-Pr₂NH (0.69 mL, 4.69 mmol) in THF (25 mL). Stirring was continued

for 15 min, and the solution was then cooled to -78 °C. The resulting

solution of LDA was added over ca. 1 min by cannula to a stirred and

cooled (-78 °C) solution of 73 (≡ 44) (700 mg, 1.30 mmol) in THF

(25 mL), and stirring was continued for 40 min.

(all the material from the previous experiment) in PhMe (15 mL) at

room temperature. Stirring was continued for 45 min. Next, Et₂O (50

mL) and then 10% aqueous CuSO₄(30 mL) were added. The mixture

was shaken vigorously, and the aqueous phase was extracted with Et₂O

(2 × 50 mL). The combined organic extracts were washed with brine

(30 mL), dried (Na₂SO₄), and evaporated. Flash chromatography of

the residue over silica gel (2 × 22 cm), using 1:1 EtOAc-hexane,

gave 76 (476 mg, 81% over the two steps) as a pure (¹H NMR, 400

MHz), yellowish foam. Some starting amine (75) (36 mg, 6%) was

also recovered. The material (76) should be stored in a refrigerator

and used within 2 days.

Methyl [5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-5,6-dihydro-3-

oxo-5-[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]-

dioxolan]-8-yl]carbamate (77). Dry pyridine (1.29 mL, 15.57 mmol)

was injected into a stirred and cooled (0 °C) solution of 76 (466 mg,

1.04 mmol) in dry CH₂Cl₂(50 mL). Then, triphosgene (927 mg, 3.11

mmol) was tipped into the mixture. The cold bath was removed, and

stirring was continued for 50 min. The mixture was recooled to 0 °C.

Pyridine (1.29 mL, 15.57 mmol) and dry MeOH (12.8 mL) were

injected, and stirring at 0 °C was continued for 1 h. At this stage,

brine (30 mL) was added, and the organic phase was separated. The

aqueous phase was extracted with CH₂Cl₂(2 × 50 mL). The combined

organic extracts were dried (Na₂SO₄) and evaporated. Flash chroma-

tography of the residue over silica gel (2 × 20 cm), using 3:7 EtOAc-

hexane, gave 77 (479 mg, 91%) as a pure (¹H NMR, 400 MHz) white

solid.

During this time, PhSeBr was prepared by adding bromine (234 µL,

4.55 mmol) dropwise to a stirred solution of PhSeSePh (1.63 g, 5.20

mmol) in THF (10 mL) at room temperature. The mixture was stirred

for 10 min, cooled to -78 °C, and then added over ca. 20 s by cannula

to the main reaction mixture. Stirring was continued for 15 min,

saturated aqueous NH₄Cl (50 mL) was added, and the mixture was

extracted with EtOAc (2 × 100 mL). The combined organic extracts

were washed with brine (50 mL), dried (Na₂SO₄), and evaporated at

room temperature. Flash chromatography of the residue over silica

gel (2 × 30 cm), using first 1:10 and then 3:7 EtOAc-hexane, gave

the crude phenyl selenide, which was used directly in the next step

without purification or characterization. The compound is not stable,

and should be used the same day.

Methyl trans- and cis-[1-Bromo-5-[[(1,1-dimethylethyl)dimeth-

ylsilyl]oxy]-5,6-dihydro-3-oxo-5-[(trimethylsilyl)ethynyl]spiro[1H-

2-benzopyran-7(3H),2′-[1,3]dioxolan]-8-yl]carbamate (78). Lactone

77 (200 mg, 0.39 mmol) and (PhCO)₂O₂(two small grains; less than

1 mg) were covered with dry CCl₄(20 mL), and the mixture was

refluxed under Ar and illuminated by a 100 W tungsten filament bulb

held close to the flask. The lamp, flask, and oil bath were surrounded

by aluminum foil. The condenser was removed momentarily, and a

small portion of NBS was added. Addition of NBS was repeated at

intervals of 10-15 min after each batch had reacted (disappearance of

the reagent from the bottom of the flaskswhich could be detected

clearly) until all the NBS (176 mg, 0.99 mmol) had been added. The

addition took ca. 1 h. Refluxing and irradiation was continued for a

further 15 min after the end of the last addition. At this point, the last

batch of NBS had reacted, and none was left at the bottom of the flask.

The mixture was cooled and filtered through a pad (2 × 3 cm) of Celite,

using CCl₄(ca. 10-15 mL). Evaporation of the filtrate at room

temperature gave crude 78, which was used directly without charac-

terization: The H NMR spectrum (CDCl₃, 400 MHz) shows disap-

pearance of the C(9)-H₂signal and new signals at δ 8.05, 7.25, and

6.5.

We suspect, from the stoichiometry of reagents used (2.5 equiv of

NBS), that the nitrogen of the carbamate has been brominated, but we

have no spectral data to confirm this point. When we used 1.2 equiv

of NBS, the overall yield, after conversion to 79, was the same, and

presumably, the nitrogen has not been brominated.

Methyl (8E)-[9-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-9-ethynyl-

7-formyl-6-[(methoxycarbonyl)amino]-1,4-dioxaspiro[4.5]dec-6-en-

8-ylidene]acetate (79). (a) Methyl trans- and cis-[5-[[(1,1-Dimeth-

ylethyl)dimethylsilyl]oxy]-5-ethynyl-5,6-dihydro-1-hydroxy-3-

oxospiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-8-

yl]carbamate. The crude product from the above experiment was

dissolved in dry THF (15 mL), and dry pyridine (0.16 mL, 1.97 mmol)

was injected. A solution of AgNO₃(201 mg, 1.18 mmol) in water (2

mL) was then added dropwise with stirring over ca. 1.5 min. The flask

was then immediately wrapped with aluminum foil, and stirring was

continued for 3 h. Brine (30 mL) was added, and stirring was continued

for 15 min. The mixture was then filtered through a pad (2 × 3 cm)

of Celite, which was washed with CH₂Cl₂. The organic phase was

separated, and the aqueous phase was extracted with CH₂Cl₂(2 × 15

mL). The combined organic phases were dried (Na₂SO₄) and evapo-

rated. Flash chromatography of the residue over silica gel (1 × 20

cm), using 1:1 EtOAc-hexane, gave the desired hydroxy lactone (147

mg, ca. 83%) as a slightly impure (¹H NMR, 400 MHz) oil, which

was suitable for the next step.

Our impression from the integration of ¹H NMR spectra run on the

crude product is that C,N-diselenated material (almost 50% of the total)

is present, but the crude material is still suitable for the next step.

(b) Selenoxide Fragmentation. Ice-cold dimethyldioxirane⁴⁴(0.07

M in acetone, 28.0 mL, 1.96 mmol) was added over ca. 10 min from

a syringe to a stirred and cooled (-45 °C) solution of the above crude

selenide in CH₂Cl₂(30 mL). Stirring at -45 °C was continued for 30

min (TLC control, silica, 3:7 EtOAc-hexane), by which time the

initially yellow solution had become almost colorless, and no starting

material was detected by TLC. The solution was evaporated at room

temperature, and flash chromatography of the residue over silica gel

(2 × 30 cm), using 1:3 EtOAc-hexane, gave 74 (592 mg, 85% over

two steps) as a pure (¹H NMR, 400 MHz) white solid. Some starting

material (36 mg, 5%) was also recovered. The yield of product,

corrected for recovered starting material, was 89%.

(5r,8r,8ar)-8-Amino-5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-

5,6,8,8a-tetrahydro-5-[(trimethylsilyl)ethynyl]spiro[1H-2-benzopy-

ran-7(3H),2′-[1,3]dioxolan]-3-one (75). (Ph₃P)₄Pd (173 mg, 0.150

mmol) was tipped into a stirred solution of 74 (800 mg, 1.49 mmol)

and dimedone (1.26 g, 8.97 mmol) in dry THF (25 mL) (Ar atmosphere,

protection from light by aluminum foil). Stirring in the dark at room

temperature was continued for 3 h, and the mixture was then evaporated.

The residue was dissolved in CH₂Cl₂(50 mL), and the solution was

washed with saturated aqueous Na₂CO₃(30 mL). The aqueous phase

was extracted with CH₂Cl₂(2 × 30 mL), and the combined organic

extracts were dried (Na₂SO₄) and evaporated. Flash chromatography

of the residue over silica gel (2 × 25 cm), using 3:2 EtOAc-hexane,

gave 75 (628 mg, 93%) as a pure (¹H NMR, 400 MHz), faintly yellow

solid.

1

8-Amino-5-[[(1,1-dimethylethyl)dimethylsilyl]oxy]-5,6-dihy-

dro-5-[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]-

dioxolan]-3-one (76). (a) (5r,8r,8ar)-8-Chloramino-5-[[(1,1-

Dimethylethyl)dimethylsilyl]oxy]-5,6,8,8a-tetrahydro-5-

[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]diox-

olan]-3-one. Freshly prepared t-BuOCl⁴⁸(156 mg, 1.44 mmol) in dry

Et₂O (18 mL) was added by syringe pump over 1 h to a stirred and

cooled (-45 °C) solution of 75 (590 mg, 1.31 mmol) in a mixture of

THF (10 mL) and Et₂O (20 mL). Stirring at -45 °C was continued

for 15 min after the end of the addition, and the mixture was then

evaporated at room temperature to afford the crude N-chloro compound.

The material appeared to be homogeneous by TLC (silica, 1:1 EtOAc-

hexane), but was used directly in the next step without characterization.

(b) Elimination of HCl. DBU (1.55 mL, 10.40 mmol) was added

over 3 min to a stirred solution of the above crude N-chloro compound

10346 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

(b) Methyl (8E)-[9-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-9-ethy-

nyl-7-formyl-6-[(methoxycarbonyl)amino]-1,4-dioxaspiro[4.5]dec-

6-en-8-ylidene]acetate (79). Ethereal CH₂N₂(made by the procedure

supplied with the Aldrich Diazald kit, 1.1 mL) was added to a stirred

and cooled (0 °C) solution of the above hydroxy lactone (127 mg, ca.

0.282 mmol) in shelf Et₂O (5 mL). After the addition, a faint yellow

color persisted. Stirring was continued for 1 h. Evaporation of the

yellowish solution, and flash chromatography of the residue over silica

gel (1 × 15 cm), using 2:3 EtOAc-hexane, gave the crude product.

Rechromatography, under the same conditions, gave 79 (121 mg, ca.

93%, or 77% over three steps) as a pure (¹H NMR, 400 MHz) white

solid.

Methyl trans-[5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-5-ethy-

nyl-5,6-dihydro-3-oxo-1-[(trimethylsilyl)ethynyl]spiro[1H-2-ben-

zopyran-7(3H),2′-[1,3]dioxolan]-8-yl]carbamate (80). The CeCl₃‚

7H₂O used in this experiment was washed with THF and air-dried to

give a white material.

Hydrated cerium chloride (CeCl₃‚7H₂O, 1.12 g, 3.01 mmol) con-

tained in a round-bottomed flask closed with a bent adaptor with a tap

was dried for 3 h (oil bath at 130 °C, 0.05 mmHg) with magnetic stirring

and then cooled to room temperature (protection from moisture).

THF (6 mL) was added with vigorous stirring to the resulting

anhydrous CeCl₃, and the suspension was sonicated (Branson, model

B-12, 80 W) overnight to obtain a fine suspension free of any lumps,

which was then cooled to -78 °C. Freshly made lithium (trimethyl-

silyl)acetylide in THF [prepared by addition of n-BuLi (1.6 M in

hexane, 1.45 mL, 2.32 mmol) to (trimethylsilyl)acetylene (Aldrich, 0.35

mL, 2.55 mmol) in THF (6 mL) at -78 °C, followed by stirring for 15

min at this temperature] was added to the above suspension by cannula

over ca. 3 min. After the addition, stirring was continued at -78 °C

for 45 min. During this procedure, the suspension remained white.

The above organocerium reagent in THF was added by cannula over

ca. 3 min to a stirred and cooled (-78 °C) solution of 79 (108 mg,

0.232 mmol) in THF (5 mL). Stirring was continued for 30 min, and

the reaction was quenched by the addition of saturated aqueous NH₄-

Cl (6 mL) at -78 °C. The cold bath was removed, and the mixture

was stirred for 10 min. The mixture was extracted with EtOAc (3 ×

10 mL), and the combined organic extracts were washed with brine (5

mL) and dried (Na₂SO₄). Evaporation of the solvent and flash

chromatography of the residue over silica gel (1 × 15 cm), using 3:7

EtOAc-hexane, gave 80 (112 mg, 91%) as a pure (¹H NMR, 400

MHz), white solid.

Methyl trans-[5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-1,5-di-

ethynyl-5,6-dihydro-3-oxospiro[1H-2-benzopyran-7(3H),2′-[1,3]di-

oxolan]-8-yl]carbamate (72) and Methyl cis-[5-[[(1,1-Dimethylethyl)-

dimethylsilyl]oxy]-1,5-diethynyl-5,6-dihydro-3-oxospiro[1H-2-

benzopyran-7(3H),2′-[1,3]dioxolan]-8-yl]carbamate (71). Bu₄NF (1.0

M in THF, 62 µL, 0.062 mmol⁶²) was added to a stirred and cooled (0

°C) solution of 80 (66 mg, 0.124 mmol⁶²) in THF (3 mL), and stirring

was continued at 0 °C for 30 min. A 1:1 water-brine mixture (2 mL)

was added, and the mixture was extracted with EtOAc (3 × 5 mL).

The combined organic extracts were washed with brine (3 mL), dried

(Na₂SO₄), and evaporated. Flash chromatography of the residue over

silica gel (1 × 10 cm), using 2:3 EtOAc-hexane, gave the less polar

isomer 71 (24 mg, 42%) and the more polar isomer 72 (26 mg, 46%),

identical to the corresponding compounds obtained from a mixture of

69 and its anti isomer.

AgNO₃(1.5 mg, 0.0088 mmol) were added to a stirred solution of 72

(82 mg, 0.179) in acetone (3 mL) contained in a flask protected from

light by being wrapped in aluminum foil. Stirring at room temperature

was continued for 5 h. The mixture was poured into ice water (5 mL)

and extracted with EtOAc (3 × 10 mL). The combined organic extracts

were washed with brine (5 mL), dried (Na₂SO₄), and evaporated. Flash

chromatography of the residue over silica gel (1 × 12 cm), using 7:13

EtOAc-hexane, gave 81 (113 mg, 89%) as a pure (¹H NMR, 400

MHz), white solid.

Methyl (1′R*,5′S*,11′Z)-[5′[[(1,1-Dimethylethyl)dimethylsilyl]-

oxy]-5′,6′-dihydro-3′-oxospiro[1,3-dioxolane-2,7′(3′H)-[1,5][3]hexene-

[1,5]diyno[1H-2]benzopyran]-8′-yl]carbamate (82). Freshly distilled

(Z)-1,2-bis(trimethylstannyl)ethene⁵⁶(54.7 mg, 0.155 mmol) in dry

DMF (10 mL) was added by syringe pump over 30 min to a stirred

and warmed (60 °C) solution of 81 (110 mg, 0.155 mmol) and (Ph₃P)₄-

Pd (from a freshly opened ampule, 17.9 mg, 0.0155 mmol) in DMF

(20 mL). Stirring at 60 °C was continued, and the reaction was

monitored by TLC (silica, 1:1 EtOAc-hexane). (First a polar spot

appears and then a less polar spot of similar R_fto that of the starting

material. This is the desired product.) After 3 h, the reaction was

complete, and the solvent was then evaporated under oil pump vacuum

at 40 °C. Flash chromatography of the residue over silica gel (1 × 15

cm), using 2:3 EtOAc-hexane, gave 82 (54 mg, 72%) as a pure (¹H

NMR, 400 MHz), white solid.

Methyl (1′R*,3′R*,5′S*,11′Z)- and (1′R*,3′S*,5′S*,11′Z)-[5′-[[(1,1-

Dimethylethyl)dimethylsilyl]oxy]-5′,6′-dihydro-3′-hydroxyspiro[1,3-

dioxolane-2,7′(3′H)-[1,5][3]hexene[1,5]diyno[1H-2]benzopyran]-8′-

yl]carbamate (83). DIBAL-H (1.0 M in CH₂Cl₂, 0.25 mL, 0.25 mmol)

was added dropwise to a stirred and cooled (-78 °C) solution of 82

(40 mg, 0.0828 mmol) in CH₂Cl₂(3 mL). Stirring was continued at

-78 °C for 30 min. MeOH (0.2 mL) was added, followed by Na₂SO₄

(0.5 g), Celite (2 g), and water (4 drops). The cold bath was removed,

and stirring was continued for 30 min. The mixture was filtered through

a sintered disk, and the insoluble material was washed with EtOAc.

The combined filtrates were evaporated, and flash chromatography over

silica gel (1 × 8 cm), using 2:3 EtOAc-hexane, gave 83 (39.2 mg,

98%) as a pure (¹H NMR, 400 MHz), white solid that was a mixture

of anomers.

Methyl (1′R*,3′R*,5′S*,11′Z)- and (1′R*,3′S*,5′S*,11′Z)-(5′,6′-

Dihydro-3′,5′-dihydroxyspiro[1,3-dioxolane-2,7′(3′H)-[1,5][3]hexene-

[1,5]diyno[1H-2]benzopyran]-8′-yl)carbamate (84). Bu₄NF (1.0 M

in THF, 0.089 mL, 0.089 mmol) was added to a stirred and cooled (0

°C) solution of 83 (39 mg, 0.080 mmol) in THF (2 mL). Stirring was

continued at 0 °C for 30 min, and saturated aqueous NH₄Cl (5 mL)

was added. The mixture was extracted with EtOAc (3 × 10 mL), and

the combined organic extracts were washed with brine (5 mL), dried

(Na₂SO₄), and evaporated. Flash chromatography of the residue over

silica gel (1 × 8 cm), using 4:1 EtOAc-hexane, gave 84 (28.2 mg,

94%) as a pure (¹H NMR, 400 MHz), colorless glass that was a mixture

of epimers.

Methyl (1R*,4Z,8S*,13E)-[1,8-Dihydroxy-13-(2-hydroxyethylidene)-

spiro[bicyclo[7.3.1]trideca-4,9-diene-2,6-diyne-11,2′-[1,3]dioxolan]-

10-yl]carbamate (85). NaBH₄(57.5 mg, 1.52 mmol) was added to a

stirred and cooled (0 °C) solution of 84 (28.2 mg, 0.076 mmol) in

bench MeOH (1.5 mL). Stirring was continued for 2 h, and then AcOH

(0.6 mL) was added dropwise at 0 °C. Stirring was continued for 5

min, and the solution was evaporated. THF (1 mL), MeOH (2 mL),

and water (4 drops) were added sequentially, and the mixture was stirred

for 20 min and then evaporated. THF (3 mL), MeOH (3 drops) and

water (3 drops), were again added sequentially. The mixture was stirred

for 20 min and then filtered through a pad (0.6 × 2 cm) of Celite,

which was washed well with THF. Evaporation of the filtrate and flash

chromatography of the residue over silica gel (1 × 8 cm), using first

4:1 EtOAc-hexane and then 1:19 MeOH-EtOAc, gave 85 (21.5 mg,

76%) as a pure (¹H NMR, 400 MHz), white glass.

Methyl trans-[5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-1,5-di-

ethynyl-5,6-dihydro-3-oxospiro[1H-2-benzopyran-7(3H),2′-[1,3]di-

oxolan]-8-yl]carbamate (72) from 71. Bu₄NOAc (7.5 mg, 0.022

mmol) was added to a solution of 71 (20 mg, 0.044 mmol) in THF (2

mL), and the mixture was stirred at room temperature for 5 h.

Evaporation of the solvent and flash chromatography of the residue

over silica gel (0.6 × 8 cm), using 2:3 EtOAc-hexane, gave 72 (12

mg, 60%) and recovered 71 (8 mg, 40%).

Methyl (1R*,4Z,8S*,13E)-[1-Hydroxy-8-[(triethylsilyl)oxy]-13-[2-

[(triethylsilyl)oxy]ethylidene]spiro[bicyclo[7.3.1]trideca-4,9-diene-

2,6-diyne-11,2′-[1,3]dioxolan]-10-yl]carbamate (86). Dry 2,6-lutidine

(30.1 µL, 0.254 mmol) and then Et₃SiOTf (30.3 µL, 0.127 mmol) were

added dropwise to a stirred and cooled (0 °C) solution of 85 (21.5 mg,

0.0576 mmol) in dry CH₂Cl₂(2 mL). Stirring at 0 °C was continued

Methyl trans-[5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-5,6-di-

hydro-1,5-bis(iodoethynyl)-3-oxospiro[1H-2-benzopyran-7(3H),2′-

[1,3]dioxolan]-8-yl]carbamate (81). NIS (88 mg, 0.393 mmol) and

(62) A deficiency of Bu₄NF is used: Cf. Nishikawa, T.; Ino, A.; Isobe,

M. Tetrahedron 1994, 50, 1449.

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10347

for 20 min, and then 10% aqueous CuSO₄(3 mL) was added. The

mixture was extracted with Et₂O (2 × 20 mL). The combined organic

extracts were washed with brine (5 mL), dried (Na₂SO₄), and

evaporated. Flash chromatography of the residue over silica gel (1 ×

8 cm), using 1:3 EtOAc-hexane, gave 86 (32.9 mg, 95%) as a pure

(¹H NMR, 400 MHz) oil.

Methyl (1R*,4Z,8S*,13E)-[1-Hydroxy-13-(2-hydroxyethylidene)-

8-[(triethylsilyl)oxy]spiro[bicyclo[7.3.1]trideca-4,9-diene-2,6-diyne-

11,2′-[1,3]dioxolan]-10-yl]carbamate (87). A 3:6:1 mixture (0.6 mL)

of AcOH, THF, and water was added to 86 (32.9 mg, 0.0547 mmol)

contained in a flask immersed in an ice bath at 0 °C, and the resulting

solution was stirred for 30 min. Saturated aqueous NaHCO₃(5 mL)

was added, the mixture was extracted with CH₂Cl₂(3 × 5 mL), and

the combined organic extracts were dried (Na₂SO₄) and evaporated.

Flash chromatography of the residue over silica gel (1 × 8 cm), using

4:1 EtOAc-hexane, gave 87 (25 mg, 94%) as a pure (¹H NMR, 400

MHz) oil.

(1R*,4Z,8S*,13E)-S-[2-[1-Hydroxy-10-[(methoxycarbonyl)amino]-

8-[(triethylsilyl)oxy]spiro[bicyclo[7.3.1]trideca-4,9-diene-2,6-diyne-

11,2′-[1,3]dioxolan]-13-ylidene]ethyl] Ethanethioate (88). Diisopro-

pyl azodicarboxylate (40.6 µL, 0.205 mmol) was added to a stirred

and cooled (0 °C) solution of Ph₃P (70.0 mg, 0.267 mmol) in THF (1

mL). Stirring was continued for 30 min, and then AcSH (15 µL, 0.205

mmol) followed immediately by a solution of 87 (12.5 mg, 0.0256

mmol) in THF (0.4 mL plus 0.2 mL as a rinse) were added dropwise.

Stirring at 0 °C was continued for 30 min. Saturated aqueous NaHCO₃

(4 mL) was added, and the mixture was extracted with EtOAc (3 × 5

mL). The combined organic extracts were washed with brine (5 mL),

dried (Na₂SO₄), and evaporated. Flash chromatography of the residue

over silica gel (0.6 × 8 cm), using 7:13 EtOAc-hexane, gave 88 (13.1

mg, 94%) as a pure (¹H NMR, 400 MHz) white solid.

Methyl (1R*,4Z,8S*,13E)-[1-hydroxy-13-(2-mercaptoethylidene)-

8-[(triethylsilyl)oxy]spiro[bicyclo[7.3.1]trideca-4,9-diene-2,6-diyne-

11,2′-[1,3]dioxolan]-10-yl]carbamate (89). DIBAL-H (1.0 M in

CH₂Cl₂, 0.097 mL, 0.097 mmol) was added dropwise to a stirred and

cooled (-78 °C) solution of 88 (6.60 mg, 0.0121 mmol) in CH₂Cl₂(2

mL). Stirring was continued for 30 min, and then MeOH (0.03 mL),

Na₂SO₄(0.5 g), Celite (1 g), and water (3 drops) were added. The

cold bath was removed, and the mixture was stirred for 30 min and

then filtered through a sintered disk. The insoluble material was washed

with EtOAc. Evaporation of the combined filtrates and flash chroma-

tography of the residue over silica gel (0.6 × 6 cm), using 3:7 EtOAc-

hexane, gave crude 89, suitable for the next stage. No spectroscopic

data were obtained. Chromatography is necessary to ensure a pure

product in the next step, which was carried out immediately.

Methyl (1R*,4Z,8S*,13E)-[1-Hydroxy-13-[2-(methyltrithio)eth-

ylidene]-8-[(triethylsilyl)oxy]spiro[bicyclo[7.3.1]trideca-4,9-diene-

2,6-diyne-11,2′-[1,3]dioxolan]-10-yl]carbamate (90). In this experi-

ment, it is important to use freshly chromatographed thiol and a large

(ca. 8-fold) excess of N-(methyldithio)phthalimide.

obtain pure (¹H NMR, 400 MHz) (()-calicheamicinone (1) (3.8 mg,

84%) as a white foam.

Selected Procedures for Route via 42. (6r,7â,8â,9â)-[6-Amino-

9-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-8-[2-[(4-methoxyphenoxy)-

methoxy]ethyl]-1,4-dioxaspiro[4.5]dec-7-yl]methyl 2,2-Dimethyl-

propanoate (35). NiCl₂‚6H₂O (1.12 g, 4.71 mmol) in commercial

MeOH (200 mL) was sonicated (Branson, model B-12, 80 W) to

complete dissolution (ca. 2 min). NaBH₄(535.0 mg, 14.14 mmol) was

then added in one portion, and the mixture was sonicated for ca. 10

min to give a fine black suspension. A solution of 34 (5.25 g, 8.58

mmol) in MeOH (50 mL plus 2 × 10 mL as a rinse) was added over

ca. 30 min, and more NaBH₄(4.87 g, 128.7 mmol) was then added

portionwise over 30 min (continuous sonication). Sonication was

continued for a further 30 min after the addition of NaBH₄. Most of

the solvent was then evaporated. The mixture was diluted with CHCl₃

(60 mL) and filtered through a column of silica gel (4.5 × 20 cm) by

gravity, using 5:1 CHCl₃-MeOH, to give a crude material (all black

Ni₂B was left on the column). This was purified by flash chromatog-

raphy over silica gel (4.5 × 25 cm), using 10:1 CHCl₃-MeOH, to

give 35 (4.75 g, 95%) as a colorless oil.

(6r,7â,8â)-[8-[2-[(4-Methoxyphenoxy)methoxy]ethyl]-9-oxo-6-

[[(2-propenyloxy)carbonyl]amino]-1,4-dioxaspiro[4.5]dec-7-yl]me-

thyl 2,2-Dimethylpropanoate (38) from 37. CrO₃(10.00 g, 100.0

mmol) was added at room temperature in three portions to a vigorously

stirred solution of dry pyridine (16.2 mL, 200.0 mmol) in CH₂Cl₂(130

mL), and the mixture was stirred for 10 min. The resulting freshly

made Collins reagent was then poured slowly into to a stirred solution

of 37 (3.57 g, 6.47 mmol) in CH₂Cl₂(250 mL). Stirring was continued

for 10 min, and the mixture was then filtered through a pad (4.5 × 4.0

cm) of silica gel, using EtOAc. The filtration and subsequent

chromatography must be done quickly to avoid epimerization R to the

ketone carbonyl. Evaporation of the solvent and flash chromatography

of the residue over silica gel (3.5 × 18 cm), using 3:2 EtOAc-hexane,

gave 38 (3.39 g, 95%).

(6r,7â,8â,9â)-[9-Hydroxy-8-[2-[(4-methoxyphenoxy)methoxy]-

ethyl]-6-[[(2-propenyloxy)carbonyl]amino]-9-[(trimethylsilyl)ethy-

nyl]-1,4-dioxaspiro[4.5]dec-7-yl]methyl 2,2-Dimethylpropanoate (39).

The following is the maximum scale for this reaction (at least with our

Branson sonicator (model B-12, 80 W). If more powerful sonication

is used, it is possible that the reaction could be scaled up, but we have

not tried this.

Hydrated cerium chloride (CeCl₃‚7H₂O, 10.7309 g) contained in a

round-bottomed flask closed with a bent adaptor with a tap was dried

overnight (oil bath at 130 °C, 0.05 mmHg) with vigorous magnetic

stirring and then cooled to room temperature (protection from moisture)

to obtain anhydrous⁴⁰CeCl₃(7.1850 g).

THF (25 mL) was added with vigorous stirring to the resulting

anhydrous CeCl₃(7.1850 g, 29.15 mmol), and the suspension was

sonicated (Branson, model B-12, 80 W) for 4 h to obtain a fine

suspension free of any lumps, which was then cooled to -42 °C.

Freshly made lithium (trimethylsilyl)acetylide in THF [prepared by

addition of n-BuLi (1.6 M in hexane, 13.0 mL, 20.8 mmol) to

(trimethylsilyl)acetylene (Aldrich, 3.3 mL, 23.00 mmol) in THF (14

mL) at -78 °C, followed by stirring for 5 min at this temperature]

was added to the above suspension by cannula (over ca. 5 min). After

the addition, the suspension partially dissolved. Stirring was continued

at -42 °C for 40 min. (If the suspension is yellowish and the color

does not fade after stirring for 40 min, do not use the reagent, because

the yield in the reaction will be low.)

The chromatographed crude thiol from the last experiment was

dissolved in dry CH₂Cl₂(0.5 mL), and the solution was cooled to 0

°C. N-(Methyldithio)phthalimide^2a(recrystallized⁶³from CH₂Cl₂-

EtOH, 21.8 mg, 0.0968 mmol) was tipped in, and stirring was continued

for 2 h. Hexane (1 mL) was added, and the mixture was applied directly

to a flash chromatography column (0.6 × 8 cm) packed with silica

gel, a little 3:7 EtOAc-hexane being used as a rinse. The column

was developed with 3:7 EtOAc-hexane to obtain pure (¹H NMR, 400

MHz) 90 (6.2 mg, 88% over two steps) as an oil.

The above organocerium reagent in THF was added by cannula (over

ca. 5 min) to a stirred and cooled (-78 °C) solution of 38 (1.19 g,

2.17 mmol) in THF (35 mL). Stirring was continued for 10 min, and

the reaction was quenched by addition of saturated aqueous NH₄Cl

(20 mL) at -78 °C. The cold bath was removed, and the mixture was

allowed to attain room temperature (over ca. 45 min). The mixture

was diluted with saturated aqueous NH₄Cl (40 mL) and extracted with

EtOAc (3 × 60 mL). The combined organic extracts were washed

with brine (100 mL) and dried (Na₂SO₄). Evaporation of the solvent

and flash chromatography of the residue over silica gel (2.0 × 22 cm),

using 3:2 and then 3:1 EtOAc-hexane, gave 39 (1.28 g, 91%) as a

colorless oil.

Methyl (1R*,4Z,8S*,13E)-(()-[1,8-Dihydroxy-13-[2-(methyltrithio)-

ethylidene]-11-oxobicyclo[7.3.1]trideca-4,9-diene-2,6-diyne-10-yl]-

carbamate (1) [(()-calicheamicinone)]. TsOH‚H₂O (4.1 mg, 0.0213

mmol) was added to a stirred solution of 90 (6.2 mg, 0.0107 mmol) in

a mixture of THF (0.6 mL) and water (1 drop), and stirring was

continued at room temperature for 15 h. Hexane (1 mL) was added,

and the mixture was applied directly to a flash chromatography column

(0.6 × 8 cm) packed with silica gel, a little 1:1 Et₂O-hexane being

used as a rinse. The column was developed with 1:1 Et₂O-hexane to

(63) Pure reagent can be used even after storage at room temperature

for 1 year.

10348 J. Am. Chem. Soc., Vol. 120, No. 40, 1998

CliVe et al.

2-Propenyl (6r,7â,8â,9â)-[9-[[(1,1-Dimethylethyl)dimethylsilyl]-

oxy]-7-formyl-8-[2-[(4-methoxyphenoxy)methoxy]ethyl]-9-[(trimeth-

ylsilyl)ethynyl]-1,4-dioxaspiro[4.5]dec-6-yl]carbamate (42). CrO₃

(10.00 g, 100.0 mmol) was added at room temperature in three portions

to a vigorously stirred solution of dry pyridine (16.2 mL, 200.0 mmol)

in dry CH₂Cl₂(150 mL), and the mixture was stirred for 10 min. The

resulting freshly made Collins reagent was then poured slowly into a

stirred solution of 41 (4.47 g, 6.59 mmol) in CH₂Cl₂(150 mL). Stirring

was continued for 10 min, and the mixture was then filtered through a

pad (4.5 × 3.5 cm) of silica gel, using EtOAc. Evaporation of the

solvent and flash chromatography of the residue over silica gel (4.5 ×

20 cm), using 3:1 EtOAc-hexane, gave 42 (4.04 g, 90%) as a colorless

oil.

2-Propenyl [6r,7â(R*),8â,9â]- and [6r,7â(S*),8â,9â]-[9-[[(1,1-

Dimethylethyl)dimethylsilyl]oxy]-7-[1-hydroxy-3-(trimethylsilyl)-2-

propynyl]-8-[2-[(4-methoxyphenoxy)methoxy]ethyl]-9-[(trimethylsilyl)-

ethynyl]-1,4-dioxaspiro[4.5]dec-6-yl]carbamate (53) and (54).

Hydrated cerium chloride (CeCl₃‚7H₂O, 10.7002 g) contained in a

round-bottomed flask closed with a bent adaptor with a tap was dried

overnight (oil bath at 130 °C, 0.05 mmHg) with vigorous magnetic

stirring and then cooled to room temperature (protection from moisture)

to obtain CeCl₃(7.0849 g).

THF (25 mL) was added by syringe to the resulting anhydrous CeCl₃

(7.0849 g, 28.74 mmol) with vigorous stirring, and stirring was

continued for 30 min. The suspension was then sonicated (Branson,

model B-12, 80 W) for 4 h to obtain a fine suspension, which was

then cooled to -42 °C. (We suspect that a more powerful sonicator

would have been better.) Freshly made, cold (-78 °C) lithium

(trimethylsilyl)acetylide in THF [prepared by addition of n-BuLi (1.6

M in hexane, 13.0 mL, 20.8 mmol) to (trimethylsilyl)acetylene (3.3

mL, 23.00 mmol) in THF (15 mL) at -78 °C, followed by stirring for

5 min at this temperature] was added to the above suspension by cannula

(over ca. 2 min). After the addition, the yellowish suspension partially

dissolved (sometimes the color was nearly white; this color indicates

that the reagent is of good quality). Stirring was continued at -42 °C

(dry ice-MeCN) for 40 min, by which time the color had almost faded.

(If the suspension is yellowish and the color does not fade after stirring

for 40 min, the reagent should not be used, because the yield in the

reaction will be low.) Then the mixture was cooled in an acetone/dry

ice bath.

The above organocerium reagent in THF was added by cannula (over

ca. 3 min) to a stirred and cooled (MeOH/liquid N₂, bath temperature

-90 °C) solution of 42 (1.29 g, 1.91 mmol) in THF (30 mL). Stirring

was continued for 10 min, and the reaction was quenched by addition

of saturated aqueous NH₄Cl (20 mL) at -90 °C. The cold bath was

removed, and the mixture was stirred for ca. 30 min, during which

time it attained room temperature. The mixture was diluted with

saturated aqueous NH₄Cl (50 mL) and extracted with EtOAc (3 × 60

mL). The combined organic extracts were washed with brine (100

mL) and dried (Na₂SO₄). After evaporation of the solvent, the ¹H NMR

spectrum of the residue showed that the crude material was a mixture

of two isomers in a ratio of 5:1 (desired compound to undesired

compound). Flash chromatography of the material over silica gel (3.5

× 30 cm), using 1:15 and then 1:5 EtOAc-hexane, gave 53 (1.168 g,

79%) as a colorless oil and its C(9) epimer 54 (233 mg, 16%) as a

colorless oil.

aqueous NH₄Cl (5 mL) was added, and the mixture was extracted with

EtOAc (3 × 20 mL). The combined organic extracts were dried (Na₂-

SO₄) and evaporated. Flash chromatography of the residue over silica

gel (1 × 15 cm), using 1:4 EtOAc-hexane, gave 53 (72 mg, 88%)

and the C(9) epimer 54 (6.2 mg, 7.5%).

2-Propenyl (1r,3r,4ar,5â,8r,8ar)- and (1r,3â,4ar,5â,8r,8ar)-

[5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]hexahydro-3-hydroxy-5-

[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-

8-yl]carbamate (59). Aqueous ammonia (28-30% w/w, 6.0 mL) was

added to a stirred and cooled (0 °C) solution of 58 (1.92 g, 2.69 mmol)

in commercial MeOH (150 mL). Stirring was continued at 0 °C for 1

h. Water (300 mL) was then added and the mixture was extracted

with EtOAc (3 × 150 mL). The combined organic extracts were

washed with brine (300 mL), dried (MgSO₄), and evaporated, and the

crude material (59) was used directly for the next step.

2-Propenyl (1r,4r,4ar,5â,8r,8ar)-[5-[[(1,1-Dimethylethyl)dim-

ethylsilyl]oxy]hexahydro-3-oxo-4-(phenylseleno)-1,5-bis-

[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-7(3H),2′-[1,3]dioxolan]-

8-yl]carbamate

and

2-Propenyl

(1r,5â,8r,8ar)-[5-[[(1,1-

Dimethylethyl)dimethylsilyl]oxy]-5,6,8,8a-tetrahydro-3-oxo-1,5-

bis[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-7(3H),2′-

[1,3]dioxolan]-8-yl]carbamate (61). (a) Selenation. n-BuLi (1.6 M

in hexane, 5.3 mL, 8.48 mmol) was added to a stirred and cooled (0

°C) solution of i-Pr₂NH (1.25 mL, 9.54 mmol) in THF (10.0 mL), and

the solution was stirred at 0 °C for 10 min.

A portion of the above freshly made LDA solution (14.0 mL, ca.

7.0 mmol) was added to a stirred and cooled (-78 °C) solution of 60

(1.47 g, 2.32 mmol) by syringe at a fast dropwise rate, and the reaction

mixture was stirred for 40 min at -78 °C.

During this period, a freshly made solution of PhSeBr in THF was

prepared by the following procedure: Br₂(300 µL, 5.82 mmol) was

added to a stirred solution of PhSeSePh (2.05 g, 6.57 mmol) in THF

(15 mL) at room temperature, and the mixture was stirred for 5 min.

A portion of the above solution of PhSeBr (ca. 0.77 M, 13.6 mL,

10.44 mmol) was added by syringe at a fast dropwise rate (ca. 2-3

min) to the enolate solution at -78 °C, and stirring was continued for

a further 10 min. The reaction was quenched by addition of saturated

aqueous NH₄Cl (40 mL) at -78 °C. The cold bath was removed, and

the mixture was stirred for 40 min. Another portion of saturated

aqueous NH₄Cl (100 mL) was added, and the mixture was extracted

with EtOAc (3 × 60 mL). The combined organic extracts were washed

with brine (1 × 100 mL) and dried (Na₂SO₄). Evaporation of the

solvent and flash chromatography of residue over silica gel (3.5 × 24

cm), using first 1:10 EtOAc-hexane to remove selenium species and

then 1:2 EtOAc-hexane, gave the crude C-selenide, which was used

without characterization.

(b) Selenoxide Fragmentation. Dimethyldioxirane (prepared and

estimated as described in the literature⁴⁴) (ca. 0.05 M in acetone, 56

mL, ca. 2.8 mmol) was added over ca. 5 min by syringe to a stirred

and cooled (-42 °C) solution of the above crude selenide (ca. 2.32

mmol) in CH₂Cl₂(150 mL), and stirring was continued for a further

30 min at -42 °C. Evaporation of the solvent and flash chromatog-

raphy of residue over silica gel (2.5 × 24 cm), using first 1:5 EtOAc-

hexane and then 1:4 EtOAc-hexane, gave 61 (0.98 g, 67% over the

two steps) as a white solid.

trans- and cis-8-Amino-5-[[(1,1-dimethylethyl)dimethylsilyl]oxy]-

5,6-dihydro-1,5-bis[(trimethylsilyl)ethynyl]spiro[1H-2-benzopyran-

7(3H),2′-[1,3]dioxolan]-3-one (68) and Its C(9)³Epimer. A solution

of t-BuOCl⁴⁸(48.9 mg, 0.450 mmol) in dry Et₂O (4.5 mL) was added

over 10 min (syringe pump) to a stirred and cooled (-42 °C, MeCN-

dry ice) solution of 62 (149.4 mg, 0.273 mmol) in 1:2 THF-Et₂O (9

mL), with both syringe and cold bath wrapped in aluminum foil to

protect the contents from light. The mixture was stirred at -42 °C for

a further 10 min after the end of the addition, and the solvent was then

evaporated (rotary evaporator, room temperature) to give the crude

monochloride 67 [which was not weighed, but which contains one major

component (TLC, silica, 1:1 hexane-EtOAc)]. The material was used

directly, without characterization.

2-Propenyl [6r,7â,8â,9â]-[9-[[(1,1-Dimethylethyl)dimethylsilyl]-

oxy]-7-[1-oxo-3-(trimethylsilyl)-2-propynyl]-8-[2-[(4-methoxyphenoxy)-

methoxy]ethyl]-9-[(trimethylsilyl)ethynyl]-1,4-dioxaspiro[4.5]dec-6-

yl]carbamate (55). PCC (147 mg, 0.682 mmol) and powdered 4 Å

molecular sieves (50 mg) were added to a stirred solution of 54 (88

mg, 0.114 mmol) in dry CH₂Cl₂(10 mL). Stirring was continued

overnight, and the mixture was then applied directly to a column (1 ×

15 cm) of silica gel. The column was developed with 1:4 EtOAc-

hexane to give 55 (82 mg, 93%) as a pure (¹H NMR, 400 MHz) oil.

2-Propenyl [6r,7â(R*),8â,9â]-[9-[[(1,1-Dimethylethyl)dimethyl-

silyl]oxy]-7-[1-hydroxy-3-(trimethylsilyl)-2-propynyl]-8-[2-[(4-meth-

oxyphenoxy)methoxy]ethyl]-9-[(trimethylsilyl)ethynyl]-1,4-dioxaspiro-

[4.5]dec-6-yl]carbamate (53) by reduction of 55. NaBH₄(20 mg,

0.531 mmol) was added to a stirred and cooled (0 °C) solution of 55

(82 mg, 0.106 mmol) in MeOH (2 mL). After 30 min, saturated

The crude monochloride was dissolved in dry PhMe (10 mL), and

DABCO (purified by sublimation under oil pump vacuum, 153 mg,

1.365 mmol) was added at room temperature with stirring. Stirring

Synthesis of (()-Calicheamicinone by Two Methods

J. Am. Chem. Soc., Vol. 120, No. 40, 1998 10349

was continued for 2 h at room temperature. Then, saturated aqueous

NH₄Cl (20 mL) was added, and the mixture was extracted with EtOAc

(3 × 20 mL). The combined organic extracts were washed with brine

(30 mL), dried (Na₂SO₄), and evaporated to afford 68 as a glassy solid

(140 mg), which was an ca. 8:1 mixture of syn and anti isomers. This

crude material [68 and its C(9) epimer] was not characterized.

Compound 68 is easily epimerized at C(9), and so the material must

be used immediately in the next step without further purification by

The mixture immediately turned yellow. Stirring was continued for 1

h (ice bath), and then water (30 mL) was added. (We suspect, from

TLC monitoring, that the reaction is over in less than 5 min, and we

later found that the mixture should be worked up as soon as the reaction

is over.) The mixture was extracted with EtOAc (3 × 20 mL). The

combined organic extracts were washed with brine (30 mL), dried (Na₂-

SO₄), and evaporated. Flash chromatography of the residue over silica

gel (1 × 15 cm), using first 1:2 EtOAc-hexane to separate the minor

isomer 71 (47.5 mg, 39%) and then 2:3 EtOAc-hexane to separate

the major (syn) isomer 72 (56.5 mg, 46%).

1

flash chromatography. If the H and ¹³C NMR spectra are measured,

dissolve only the spectral sample, and not the whole supply, in the

NMR solvent.

An X-ray structure was determined on 72 crystallized from EtOAc-

In this experiment, use of DBU or DBN gave a mixture of isomers.

Use of 1-3 equiv DABCO also gave mixtures because the reaction

took a long time (more than 4 h). A bigger excess (5-fold) of DABCO

gives an improved isomer ratio (better than 8:1). Acid also appears to

cause isomerization since, when a sample of the 8:1 syn/anti isomer

mixture was kept in CDCl₃for 36 h, we obtained a 1:1 mixture of the

isomers. When we dissolved the 8:1 syn/anti isomer mixture in CD₂-

Cl₂, we observed rapid (<15 min) isomerization. We did not test any

other batches of CD₂Cl₂, but we think that the isolated enamine should

not be exposed for more than a few seconds to dichloromethane. We

suggest that, in the future, the NMR spectrum should be run in CDCl₃

(that has been passed through a column of basic alumina) or in some

none acidic solvent, such as benzene-d₆. We do not know if pyridine-

d₅would also be suitable.

Methyl trans-[5-[[(1,1-Dimethylethyl)dimethylsilyl]oxy]-1,5-di-

ethynyl-5,6-dihydro-3-oxospiro[1H-2-benzopyran-7(3H),2′-[1,3]di-

oxolan]-8-yl]carbamate (72) and Methyl cis-[5-[[(1,1-Dimethylethyl)-

dimethylsilyl]oxy]-1,5-diethynyl-5,6-dihydro-3-oxospiro[1H-2-

benzopyran-7(3H),2′-[1,3]dioxolan]-8-yl]carbamate (71) by desilylation

of 69 and Its C(9)³Epimer. The starting material for this experiment

was made by the above procedure, the only difference being that DBU

was used as the base. We did not establish which stereoisomer (syn

or anti) was the major component.

hexane.³³

Acknowledgment. We thank the Natural Sciences and

Engineering Research Council of Canada, the Merck Frosst

Therapeutic Research Centre, and the Alberta Cancer Board for

financial support. Y.T. held an AHFMR Postdoctoral Fellow-

ship, Y.-J.W. an NSERC Postdoctoral Fellowship, S.D. a 1967

NSERC Graduate Fellowship, and G.M. a Postdoctoral Scholar-

ship (Ministe`re de la Recherche et de la Technologie, France).

The assistance of Professor G. V. J. da Silva, Dr. N. Selvakumar,

and Professor Y.-Z. Hu is also gratefully acknowledged. We

thank Professor W. A. G. Graham for a supply of hexameth-

ylditin, Professor D. N. Harpp for advice on the properties of

N-(methyldithio)phthalimide, and Dr. R. McDonald for X-ray

measurements.

Supporting Information Available: X-ray data for 27, the

5R epimer of 29, and 50, 62, and 72; a description of exploratory

studies, involving radical cyclization; experimental procedures

and characterization data not given in the main text; and

preparation of dimethyldioxirane (160 pages, print/PDF). See

any current masthead page for ordering information and Web

access instructions.

Bu₄NF (1.0 M in THF, 0.56 mL, 0.56 mmol) was added dropwise

over ca. 1 min to a stirred and cooled (ice bath) solution of 69 [2:1

mixture of C(9) stereoisomers, 160 mg, 0.265 mmol] in THF (10 mL).

JA980292S

Article Doi

Synthesis of (±)-calicheamicinone by two methods

DOI: 10.1021/ja980292s

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja980292s

Full text of DOI:10.1021/ja980292s

Products guided by the article

R&D Labs maybe for 126999-95-9

Relevant to this article

Hot Product