Synthesis of the Aglycones of Altromycins and Kidamycin from a Common Intermediate

Abstract

The aglycone structures 1 and 2, respectively corresponding to the antitumor antibiotic natural products altromycin and kidamycin, have been efficiently synthesized from a common advanced intermediate 3. A series of Claisen condensations and aromatizations affords the anthracene section of 3, followed by annulation of the pyrone ring. The functional groups of 3 can be manipulated for enantioselective introduction of the epoxide side-chain of altromycin aglycone 1, as well as synthesis of the kidamycin aglycone 2.

The altromycins (4 and 5, Figure 1) and kidamycin (6) are secondary metabolites of soil Streptomyces. The altromycins were first reported in the early 1990s as antibiotics with selective activity against Gram-positive bacteria (Staphylococci and Streptococci)¹ and were later shown to have anticancer activity, including in vivo activity in P388 leukemia, as well as colon, lung, and ovarian tumors. Kidamycin (6) also exhibits cytotoxicity against leukemia L-1210 and Ehrlich ascites tumors.²

Figure 1.

Structures of altromycin and kidamycin antitumor antibiotics.

Hurley has shown that altromycin B (4, R₁ = R₂ = Me; X = Y = OH) preferentially inhibits RNA and DNA synthesis relative to protein synthesis by covalent modification of guanine by alkylation of the epoxide at C16.³ Although some stereochemical elements of the altromycins 4 and 5 remain undetermined, the kidamycin structure (6) has been fully elucidated by a combination of NMR and crystallographic studies.⁴ Previous synthetic work to date in this general area has been limited to the synthesis of the branched C-glycoside substructure of altromycin B^5,6 and preparation of the kidamycin aglycone as the O-methyl ether.⁷ In this communication, we describe an efficient synthesis of the altromycin and kidamycin aglycones 1 and 2 as the initial stage for projected total syntheses of both families of natural products 4–6.

Our synthesis began with the preparation of the known naphthalene diester 8 (Scheme 1) by applying biomimetic Claisen condensation methodology, as pioneered by the laboratories of Harris and Yamaguchi.⁸ We found that freshly drying the calcium acetate monohydrate^8c was crucial for this procedure to reliably provide the naphthyl diester product. After methylation of 8 to afford the protected naphthalene diester 9,^8c the enolizable ester was temporarily masked as an enolate and the dianion of N-(trimethylsilyl)-acetamide selectively added to the aryl ester,^8a generating naphthalene amide 10. Although Dieckmann condensations of 10 to form the third aromatic ring were unsuccessful, we unexpectedly obtained the methyl ketone 11 upon refluxing 10 in acetic acid, and ketone 11 then underwent Dieckmann condensation to provide anthracene 12. As it is unlikely that the amide could be hydrolyzed under these nonaqueous conditions, a decarboxylation process is presumably not involved, but rather an analogous acid-promoted deamidation process (loss of HN[unk]C[unk]O) might be occurring. To the best of our knowledge, this is the first reported example of a decarbamidation. Selective monoprotection⁹ of the C5 hydroxyl of crude anthracene 12 with MOMCl afforded anthracene 13.

Scheme 1.

Synthesis of 12.

The plan for annulation of the pyrone ring was to first form the carbon–oxygen bond by addition–elimination¹⁰ followed by an intramolecular Friedel–Crafts acylation, as established by Bycroft in a simpler system.⁹ Thus, deproto-nation of anthracene 13 and addition of 3-chloro-2,4-dienoate 14¹¹ provided the desired product 15 but mixed with (E)- and (Z)-isomers of 16 (ratio 15:16E:16Z = 2:1:1, Scheme 2). However, saponification provided acid 17, which was easily separated from isomers 18-E and 18-Z by column chromatography. Initial attempts to close the C4—C4a bond by generating the acid chloride from 17 with oxalyl chloride resulted in Friedel–Crafts oxalylation at C7, but activation of acid 17 with 1-chloro-N,N,2-trimethyl-1-propenylamine¹² resulted in spontaneous cyclization to 3 in 86% yield. Interestingly, treatment of the mixture of isomers 18-E and 18-Z under the same conditions also provided pyrone 3, although in low yield (23%).¹³ In both cases, the methoxy-methyl (MOM) O-protective group was lost under the pyrone closure conditions.

Scheme 2.

Annulation of Pyrone Ring

To prepare the altromycin aglycone from 3, the 5-acetate derivative 19 (Scheme 3) was treated with AD-mix-β,¹⁴ providing diol 20 with an enantiomeric ratio of 13:1, as determined by Mosher ester analysis.^15,16 Global O-deprotection was achieved by AlCl₃ and tert-butanethiol,¹⁷ providing anthrone tetraol 21, which in turn was transformed into the partially protected quinone 22 by a one-pot procedure, by initial reaction with excess TIPSOTf and 2,6-lutidine in THF to give a tris-TIPS ether,¹⁸ followed by addition of water and PhI(OAc)₂.¹⁹ The second stage of this one-pot reaction selectively removed both aryl TIPS ethers while retaining the alkyl C16 TIPS ether. This partially protected quinone 22 was not purified until after the two aryl hydroxyl groups were protected as the bispivaloate ester 23, featuring two types of differentially protected hydroxyl groups. Although the protective group pattern of 23 was projected to avoid undesired mesylation of the phenols in the following epoxide formation steps, we observed that removal of the TIPS protective group was always accompanied by loss of the C5 pivaloate, to afford 24.²⁰ Fortunately, the derived bismesylate compound 25 in the presence of potassium carbonate in methanol underwent epoxide formation by intramolecular S_N2 displacement as well as methanolysis of both mesylate and pivaloate ester phenolic protective groups to provide the crystalline altromycin aglycone 1. The general lability of C5 O-protective groups may result from neighboring group participation of the C4 pyrone carbonyl oxygen. This work is the first synthesis of the altromycin aglycone 1. Although 1 is not a known degradation product of altromycins, spectroscopic comparison of ¹H and ¹³C NMR spectra as well as UV absorptions with altromycins B, H, and I are consistent with the structure of 1.

Scheme 3.

Synthesis of Altromycin Aglycone (1) from 3

The kidamycin aglycone could also be prepared from pyrone 3 via modified Kumada coupling²¹ of the 5-triflate derivative 26 with dimethylzinc to provide 27 (Scheme 4).²² Removal of O11 and O12 methyl ether protective groups afforded anthrone 28, followed by oxidation²³ to the anthraquinone of kidamycin aglycone 2. This constitutes the second synthesis of the kidamycin aglycone, and spectral data for 2 compare favorably with reported data for the synthetic O-11 methyl ether⁷ as well as kidamycin natural product structure.

Scheme 4.

Synthesis of Kidamycin Aglycone (2) from 3

In conclusion, the efficient synthesis of tetracyclic pyrone 3 and formation of the aglycones of both altromycins and kidamycin from this common intermediate 3 provide an effective entry to projected total syntheses of various altromycin natural products 4 and 5, as well as kidamycin 6.