The bicyclic ring-D is expected to hinder approach between the DNA base pairs, thereby if intercalation was not possible, assessment of anti-cancer activity would be undertaken, helping to decide whether intercalation was the prominant mode of action.

The whole essence behind this stereoselective synthesis is the introduction of the correct stereochemistry in ring-A. The D-glucose based diene above (made in Stoodley's group) adopts the conformation above (4) and the dienophile [in my case a quinol (5), but often an oxirane (6) derivative of quinizarin] undergoes the pericyclic Diels-Alder reaction from above, as shown in blue. Obviously, the steric hindrance from the sugar reduces underside attack. Therefore the major diastereomer in these schemes results from the favoured approach above and the minor diastereomer (seperable by fractional crystallisation) from a direction below the conjugated double bonds. The scheme below shows the standard work in Stoodley's group. The Diels-Alder reaction (in cool acetone) occurs with about 71% diastereomeric excess; the major diastereomer (8) is isolated in 74% yield. (Later on in the synthesis, the cycloadduct is reduced, treated with a Grignard reagent and hydrated, followed by removal of the sugar auxiliary on ring-A to give the aglycone idarubicinone.)

Now, modifying the above procedure to the bicyclic anologue (3), the quinol (5) is made from the naphthazarin (10) as shown below. Incidently, most of the quinols below and anthracyclines are red in colour and readily stain skin and glassware. Take precautions in handling. Indeed they act as indicators; red in acid/neutral pH and blue in alkali.

The bicyclic quinol (5) which in its stabilized form tautomerises when heated with the diene as shown below. Quinols of type (5) are known to tautomerise ("acylwanderung").

Although the cycloadduct (12) is formed by heating the quinol (5) with the diene (4) in toluene, it is unstable at this temperature in the toluene solvent and undergoes a little decomposition to a red solid (an aromatic ring-A is created). The diene (4) is unstable too. However a diastereomeric excess of the cycloadduct (12) is obtained from looking at the ¹H NMR spectrum of the mixture after several days. Crystallisation of the dry mixture from dry ether-hexanes gave the crude cycloadduct (12) in approx. 36% yield based upon the quinol (5). Further purification of compound (12) by silica-gel chromotography (ethyl acetate/hexanes eluent) and recrystallisation (chloroform/hexanes) led to isomerisation of the double dond to positions 9,10 in ring-A (see red inset). It should be noted however, that on a mmol scale, the cycloadduct (12) has been isolated pure.

The cycloadduct (12) can be shown as 3D models below:
[Legend: Carbon atom = grey; Hydrogen atom = blue; Oxygen = red and Silicon = purple.]

It should be noted that during the cycloaddition reaction, the quinol (5) in it's less stable tautomeric form forms an endo-transition state as shown below (secondary orbital interactions are shown with a dotted purple line) and the most favourable molecular orbital overlap using the HOMO of the diene and LUMO of the dienophile, which are closest in energy for the fastest reaction using the Woodward-Hoffmann rules:

This the energy level diagrams and shapes of the molecular orbitals for the diene and dienophile can be viewed here.

For a simple Diels-Alder reaction molecular orbital animation, click here.

For compound 12: 300 MHz NMR spectrum (CDCl₃ solvent^*); 0.18 and 0.22 (each 3H, s, Me₂Si), 0.95 (9H, s, Me₃C), 1.29 and 1.47 (each 2H, br d, separation 8Hz, CH₂CH₂), 1.80 (4H, br d, separation 9 Hz, CH₂CH₂), 1.73, 1.88, 1.97 and 2.09 (each 3H, s, 4 x MeCO₂), 2.16 (1H, dd, J 18.5 and 8 Hz, 10-Hb), 3.1-3.2 (1H, m, 6a-H), 3.38 (1H, t, J7 Hz, 10a-H), 3.52-3.58 (1H, m, 5'-H), 3.61-3.65 (each 1H, br s, 1- and 4-H), 4.04 (1H, dd, J 12 and 2.5 Hz, 6'-H), 4.56 (1H, t, J 5 Hz, 7-H), 4.94-4.96 (2H, m, 3'- and 4'-H), 5.15 (1H, d, J 4.5 Hz, 8-H), 11.66 and 12.28 (each 1H, s, 5- and 12-OH) (irradiation at d 4.56 caused the m at d 3.1-3.2 to simplify and the d at d 5.15 to callapse to a s). [Abbreviations: s = singlet; d = doublet; dd = doublet of doublets, t = triplet and br = broad]. The 10-Ha was not detected! Optical rotation: +231 (0.05% in dichloromethane).

* Isomerisation of the double bond in deuteriochloroform is known to slowly occur.

Hydrolysis of the cycloadduct (12) and its regioisomer occur easily using chloroform with a trace of conc. HCl to give the ketone (13).

Grignard attack of the ketone (13), at position C-9, with 30 equivalents of ethynylmagnesium chloride in THF at -5 ^oC afforded an ethynyl carbinol, with correct stereochemistry. It should be noted that some epimerisation at the ring junction occurs; this is not a problem as lead tetra-acetate oxidation removes these hydrogen atoms, to give the red aromatic compound (15). Other methods of creating the a-hydroxyketone moiety at the C-9 position, e.g. using trimethylsilyl cyanide (TMSCN) were examined with a model dihydroxytrione.

Finally, acidic mercury (II) oxidation (in hot ethanol) of compound (15) afforded the anthracycline (16). Note that the aglycone part of this molecule has the correct framework for an anthracycline; it is necessary now to remove the D-glucose sugar and attach daunosamine via standard methods, to give the target compound (3).

The relationship between the cycloadduct and it's regioisomer, the C-10a epimer formation and the ethynylcarbinol C-10a epimers is illustrated here.

The daunosamine group is coupled to the anthracyclinone (which, by the way shows no anti-cancer activity) by refluxing with mercuric cyanide/mercuric bromide in THF (with molecular sieves) or using silver(I) triflate or using trimethylsilyl triflate as the Lewis acid catalyst to aid the coupling of the two components. The daunosamine needs to be in a protected form.

Click here to see full experimental details or download. The numbering of the compounds is different that to the schemes above for clarity!

In summary are the two schemes for you to compare; the first is Stoodley's synthesis of (+)-4-demethoxydaunomycinone (R=H) [the synthesis of (+)-daunomycinone (R=OMe) follows as similar pattern see R.C. Gupta, C.M. Raynor, A.M.Z. Slawin, D.J. Williams and R.J. Stoodley, J. Chem. Soc., Perkin Trans 1, 1988, 1773]and the second is Miller's synthesis of the ring-D (bicyclic) modified (+)-4-demethoxydaunomycin analogue.

And Miller's synthesis:

An analogous scheme to Stoodley's work with (+)-4-demethoxydaunomycinone using an epoxy bicyclic DCB-ring synthon was attempted, but the cycloaddition reaction afforded lower yields compared to the quinol dienophile (5) and discusses some molecular orbital regiochemistry in the Diels-Alder reactions.

Although not mentioned on this page, efforts are in progress to create the ring-A modified (with an additional hydroxyl group) anthracycline (2); Dr Fernando Escribano (now at in the department of organic chemistry, University of Salamanca, Spain frena@usal.es) started this work and I continued a stage further. To view the complete procedure in detail (with 3D models), click here! Interestingly, the hydroxyl group could be attached by treating the Diels-Alder cycloadduct with dimethyldioxirane, to afford a hydroxy ketone, which could be ethynylated. When I worked on this project, the stereochemistry had not been exaustively established.

More of my thesis work involved attempting to synthesize a ring-D deleted (truncated) anthracycline and some new methodology as an alternative to ethynylation to prepare the a-hydroxyketone substituent. The stereochemical nature of this Diels-Alder reaction can be viewed to discuss the exo-anomeric effect or viewed as a word document.

View my laboratory photo taken from the Department of Chemistry, University of Manchester (formerly UMIST): laboratory.gif. Allow several minutes to download (865 Kb) or view it as a smaller JPEG file (105 kb). See alternative video in Shockwave format of which the picture quality is inferior to the main video located within the home page. Also view the video with a rotating guitar clip.

My email: jonathan@jonathanpmiller.com

To read about Roche (Pharmaceuticals) Plc click here!

Prof. Richard Stoodley's international FAX: +44 161 200 4436; further departmental address details on my C.V..

Richard Stoodley's researchand dick.stoodley@manchester.ac.uk.

Anthracyclinone syntheses: R. C. Gupta, C. M. Raynor, A. M. Z. Slawin, D. J. Williams and R. J. Stoodley, J. Chem. Soc., Perkin Trans. 1, 1988, 1773; R. C. Gupta, P. A. Harland and R. J. Stoodley, Tetrahedron, 1984, 40, 4657; (related article) R. C. Gupta, D. S. Larsen and R. J. Stoodley, J. Chem. Soc., Perkin Trans. 1, 1989, 739, Final-year project or pdf file.

To see a recent example of Dr Fernando Tome Escribano's aryl silyloxy Diels-Alder chemistry, see E. Caballero, N. Longieras, E. Zausa, B. del Rey, M. Medarde and F. Tome Escribano, Tetrahedron Letters, 2001, 42, 7233. Or view web page.

For a recent publication about highly enantioselective Diels-Alder reactions of 1-amino-3-silyloxy-dienes catalyzed by Cr(III) Salen complexex, see V. H. Rawal et al, J. Am. Chem. Soc., 2000, 122, 7843. [email]

For daunosamine coupling to the anthracyclinones, see Y. Kita, H. Maeda, F. Tekahashi and S. Fukui, J. Chem. Soc., Perkin Trans 1, 1993, 2639 and references cited therein.

Art & Science of Total Synthesis at the Dawn of the Twenty-First Century: K. C. Nicolaou, D. Vourloumis, N. Winssinger and P. S. Baran, Angew. Chem. Int. Ed., 2000, 39, 44 or download PDF file.

The Nobel Prize in Chemistry 2001

What was the hardest part of the research degree?

One day, the target anthracyclines will be reached.....

My C. V..

See my video.

To leave you with a relaxing image...

Sign Guestbook View Guestbook

This website is a non-commercial personal project and you may wish to make a small financial contribution to help promote scientific knowledge in the Internet community: