Ring-D Deleted Anthracycline

Approaches Towards a Truncated Anthracycline (1)

It was of interest to attempt the synthesis of the hitherto unknown ring-D deleted anthracycline (1) to determine it's anticancer activity and intercalation properties with DNA and interaction with topoisomerase enzymes.

The aglycone precursor (2) can be envisaged to be synthesised according to the retrosynthesis in scheme 1, following work on anthracyclinone synthesis in the Stoodley group. However, naphthazarin (5) failed to react with the D-glucose based diene (6) under reflux¹and to date, the cycloadduct (4) is unknown! Attaching daunosamine to the aglycones is a well-known process² to afford the anthracyclines. The aglycone (2) was synthesised as a racemic mixture by Krohn³ [using naphthazarin (5) in a Diels -Alder reaction; see below], although a stereoselective synthesis is required in order to prepare the C -5,7-S diastereomer so that L-Daunosamine can be attached to afford the optically pure target (1).

In this study, we decided to make use of naphthazarin glucoside (7) as the chiral dienophile in the cycloaddition. Following previous work¹ in the Stoodley group, Naphthazarin (5) was sonicated with the acetobromoglucose (8) and silver(I) oxide to afford the glucoside (7) after chromatographic work-up in 21% yield. However, the commercially cheaper copper(I) oxide failed to carry out this process.

Previously, the glucoside (7) reacted with the electron-rich diene (9) to afford the cycloadduct (10) in 87% yield¹. In this study we decided to investigate the reactions of the glucoside (7) with Danishefsky's Diene (12), in a similar fashion to the Juglone glucoside (11)⁴ in formation of the trione (13) with high yields.

Reacting the naphthazarin glucoside (7) with Danishefsky's Diene (12) in dichloromethane afforded predominantly the cycloadduct (14), in a stereoselective manner. The structure was inferred on the basis of the aforementioned results with the juglone glucoside and by considering the ¹H NMR spectrum and mass spectra data. Upon precipitation of the product from dichloromethane-diethyl ether-hexanes, a mixture containing a 7:2 ratio of the cycloadduct (14) and the presumed isomer (15) was obtained in 77% yield. Isomerisation of the double bond has been observed with other cycloadducts.

In another experiment, in which the solvent was changed to boiling benzene, the reaction was complete within 90 min; other isomers or decomposition products were not detected. Tetraacetyl diborate (16) was found to have a catalytic effect upon the reaction at room temperature, although, with time, some isomerisation of the cycloadduct (14) to the presumed isomer (15) was observed.

Again, H�ckel molecular orbital theory could shed some light on the correct side of the glucoside (7) to attack the achiral Danishefsky's Diene (12). Frontier orbital calculations of sizes of molecular orbitals are used to determine regiochemistry in reactions e.g. Diels-Alder reactions and stereo- electronic effects. The energy difference (DE = 6.66 eV) between the Highest Occupied Molecular orbital (HOMO) of the glucoside (7) and Lowest Unoccupied Molecular Orbital (LUMO) of diene (12) is much smaller than LUMO of compound (7) and HOMO of the diene (12) (21.34 eV). This would then give the quickest reaction rate with the lowest Activation Energy. Unfortunately, the shapes of the molecular orbitals for the glucoside dienophile are not easily represented and are omitted here.

The LUMO of Danishefsky's diene is shown below; notice the silyl group at the top of the image and the OMe at the bottom; each phase (+) or (-) of the orbital is shown in blue or red.

LUMO of Danishefsky's diene

The mixture of isomers (14) and (15) was hydrolysed by the action of 0.1M hydrochloric acid in THF; the hydroxytrione (17) was isolated in 44% yield after two crystallisations.

A 3-D image of the hydroxytrione is shown below; the molecule has been rotated to show the chair-like shape of the ring-A at the bottom right and the 8-OH group a the bottom left. In this image the 1-OMe group appears to be pointing up behind the molecule (key; hydrogen blue, oxygen red and carbon dark grey with double and single bonds represented).

The may 1-OMe group is seen more clearly here, with the ketone at C-3 pointing behind.

In order to pursue the synthesis of the aglycone (2) (after removing the sugar the acidic hydrolysis), ethynylation followed by the typical oxidation and hydration would be necessary to set up the functionality in the chromophore. Thus the action of ethynylation reagents upon the C-3 ketone moiety of the hydroxytrione (17) was studied. Treating the hydroxytrione (17) with an excess (ca. 30 mol equiv.) of ethynylmagnesium chloride in THF at -10 ^oC and at room temperature afforded, after lead(IV) acetate oxidation, a complex mixture of materials. Chromatography and crystallisation failed to provide any pure products. Addition of ca. 5 mol of nickel(II) acetylacetonate had no effect upon the ethynylation reaction.
2-Trimethylsulylethynylcerium dichloride⁵ has been used as a reagent for introducing the acetylene moiety into anthracycline systems. Unfortunately, after subjecting the hydroxytrione (17) to ca. 30 mol equiv. of the reagent in THF at -78 and -15 ^oC, a complex mixture was obtained. Although the 300 MHz ¹H NMR spectrum was too complex to resolve any individual components, the FAB mass spectrum indicated the presence of a quasimolecular ion (C₃₉H₅₂O₁₅Si₂Na+) with a relative abundance of 100% and a m/z ratio of 839 attributed to the bis-ethynylated material (17).

Another approach to the novel anthracycline (1) was carried out using the more typical Diels-Alder reaction to afford a ketone (18) that was difficult to ethynylate yielding a mixture of products. It is inherent that more effective practical ethynylation methods are required in anthracycline synthesis, or better still new ways of introducing the a-hydroxy ketone group.

This may be compared to a synthesis of (rac)-4- demethoxydaunomycinone devised by Krohn⁶ and co-workers. Naphthazarin (19) was heated with the diene (20) to afford, after hydrolysis, the (rac)-ketone (21) in 91% yield. Fuctionalisation of ring-A, followed by reaction of the less-stable tautomer (22a) with 1-methoxybuta-1,3-diene afforded a cycloadduct which was converted into (rac)-4-demethoxydaunomycinone (23) (an aglycone precursor to the chemotherapeutic agent Idarubicin) in an overall yield of 29% from naphthazarin (19). Alternatively³, the more stable tautomer (22) was hydrated and the derived (rac)-product (2) [also the aglycone precursor to the truncated anthracycline (1)] was reacted with 1- methoxycyclohexa-1,3-diene to afford the racemic daunomycinone (23) in an overall yield of 33% from naphthazarin (19).

References
1. A. D. Curtis and R. J. Stoodley, Ph.D. thesis, University of Manchester, 1990. Naphthazarin glucoside has also been used in an enantioselective synthesis of (+)-Hatomarubigin B using Diels-Alder chemistry, G. B. Caygill, D. S. Larsen and S. Brooker, J. Org. Chem., 2001, 66, 7427.
2. M. J. Broadhurst, C. H. Hassall and G. J. Thomas, J. Chem. Soc., Chem. Commun., 1982, 158; Y. Kita, H. Maeda, F. Tekahashi and S. Fukui, J. Chem. Soc., Perkin Trans 1, 1993, 2639.
3. K. Krohn and K. Tolkiehn, Chem. Berichte, 1979, 112, 3453.
4. B. Beagley, A. D. M. Curtis, R. G. Pritcard and R. J Stoodley, J. Chem. Soc., Perkin Trans. 1, 1992, 1981; see also a citation in a review, editor: G. R. Stephenson, Advanced Asymmetric Synthesis, Chapman & Hall, London, 1996, pages 126-145 [i.e. chapter 7, 'Asymmetric Diels-Alder reactions', A. Whiting (University of Manchester), author].
5. E. Ghera and Y. Ben-David, J. Org. Chem., 188, 52, 2972; T. Imamoto, N. Takiyama, K. Nakamura, T. Hatajima and Y. Kamiya, J. Am. Chem. Soc., 1989, 111, 4392.
6. K. Krohn and K. Tolkiehn, Tetrahedron Letters, 1978, 4023.

jonathan@jonathanpmiller.com

Relevant experiments from Experimental :
[Labelling of the compounds here differs to the experimental!]
Preparation of Naphthazarin Glucoside
Reaction of Naphthazarin Glucoside with Danishefsky's Diene
Preparation of the Hydroxytrione
Attempted ethynylation of the Hydroxytrione
Attempted ethynylation of the Hydroxytrione using a cerium(III) reagent