Synthesis of 14-fluorodoxorubicin

Article abstract of DOI:10.1016/S0040-4039(02)00299-X

Synthesis of the novel anthracycline 14-fluorodoxorubicin is described. A key step in the synthesis is the hydrolysis of 14-bromo,14-fluoro derivative 14 with AgBF₄ and DMSO, to give the geminal fluorohydrin system.

Full text of DOI:10.1016/S0040-4039(02)00299-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 2867–2871
Synthesis of 14-fluorodoxorubicin
M. Berettoni,a,* A. Cipollone,a L. Olivieri,a D. Palomba,a F. Arcamone,b C. A. Maggic and
F. Animatia
aMenarini Ricerche S.p.A., Via Tito Speri 10, 00040, Pomezia, Roma, Italy
bCNR-ICOCEA, Via P. Gobetti, 101, 40129, Bologna, Italy
cMenarini Ricerche S.p.A., Via Sette Santi 3, 50131, Firenze, Italy
Received 20 January 2002; accepted 6 February 2002
Abstract—Synthesis of the novel anthracycline 14-fluorodoxorubicin is described. A key step in the synthesis is the hydrolysis of
14-bromo,14-fluoro derivative 14 with AgBF4 and DMSO, to give the geminal fluorohydrin system. © 2002 Elsevier Science Ltd.
All rights reserved.
Anthracycline antibiotics are well known antitumor
agents. Doxorubicin 1 (Fig. 1), widely used in clinical
therapy since the early 1970s, is still a major drug in the
treatment of solid tumors.1 Advances in the knowledge
of the mechanism of action of anthracyclines highlight
the enzyme DNA topoisomerase II as the primary
cellular target for these molecules.2 According to this
mechanism, formation of a stable DNA–enzyme–drug
complex hinders the relegation of the DNA strands and
results in irreversible DNA breaks, eventually leading to
apoptosis and cell death.3,4 The molecular interactions of
the drug in the ternary complex are still unclear and
structural requirements for optimal activity remain to be
elucidated. However, recent studies indicate the crucial
role of the sugar moiety and of the cyclohexene ring (ring
A) of the anthracyclines5–7 and a study concerning their
structural and stereochemical requirements has been
reported.8
The 14 hydroxyl group is held responsible for the higher
in vivo activity of 1 as compared with daunorubicin 2
against solid tumors.1 Even if the reason of this notable
difference in antitumor efficacy is still unknown, it was
reasonable to conclude that selected modifications in the
C-9 side chain and particularly at position 14 of the
tetracyclic quinone chromophore might favorably influ-
ence the biological activity of the new derivatives. Intro-
duction of the strongly electron-withdrawing and poorly
steric-demanding fluorine atom at C-14 to give interest-
ing derivative (3) is presently reported.
O
OH
O
O
OH
O
9
14
R'
H
A
OH
OH
R
O
CH3O
O
OH
O
CH3O
O
OH OH
O
4
NH2
HO
1) R=H, R'=OH, Doxorubicin
2) R=H, R'=H, Daunorubicin
3) R=F, R'=OH
Figure 1.
Keywords: anthracycline; doxorubicin; fluorohydrin; fluoroanthracycline.
* Corresponding author. Tel.: +39 06 91184423; fax: +39 06 9106137; e-mail: chemistry@menarini-ricerche.it
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00299-X

2868
M. Berettoni et al. / Tetrahedron Letters 43 (2002) 28670
Anthracyclines derivatives carrying a fluorinated C-9
side chain have been synthesized,9,10 but the fluorine
atom and the hydroxyl group were never simulta-
neously present on the same carbon atom. Besides, only
very few examples of biologically active compounds
carrying the geminal fluorohydrin system can be found
in the literature.11,12 In the study of Guest et al.12 the
geminal fluorohydrin was obtained in the reaction of
diethylaminosulphur trifluoride (DAST) with alde-
hydes, a reaction known to give difluoroderivatives.13
Model experiments aimed at the preparation of a 14-
fluorohydrin were thus carried out in the first instance
in the aglycon series and the synthesis of ketoaldehyde
4 from daunomycinone 7 via the known14 14-bromo
intermediate 5; this latter may undergo nucleophilic
displacement with neighboring-group participation to 6.
Hydrolysis of the oxysulfonium provides the observed
a-hydroxyketo derivative (Scheme 1).
Even compound 9, under typical oxidation conditions
(CrO3, pyridine, CH2Cl2, rt; ClCOCOCl, DMSO, Et3N,
CH2Cl2, −40°C), did not afford desired 4.
However, these results allowed the exploration of a
different synthetic strategy based on the preparation of
14-bromo,14-fluoro derivative 11. In fact mild substitu-
tion of the bromine with a hydroxyl group by reaction
with AgBF4 and DMSO as previously observed would
have afforded the desired fluorohydrin.
derivative
8
was attempted. Daunomycinone is
obtained upon acid hydrolysis of 2, a readily available
biosynthetic product. In case of success with such
model compounds we could have either coupled the
sugar moiety with the functionalized aglycone, or trans-
ferred the methodology to derivatives of a preformed
anthracycline glycoside. Unfortunately, treatment of 8
with different oxidative systems (Na2CO3, KI, DMSO,
120°C;15 DMSO, Sym Collidine, rt;16 AgBF4, DMSO,
Et3N, rt17) did not give the desired 4, but during the
reaction of compound 8 with AgBF4 and DMSO, a
mild method for the preparation of aldehydes from
alkyl halides by a mechanism similar to the Moffat/
Swern oxidation,17 we obtained 14-hydroxydaunomy-
cinone (adriamycinone, 9). In the reported oxidation of
alkyl bromides, AgBF4 facilitates the displacement of
bromine by DMSO affording an oxysulfonium salt
which generates the aldehyde in presence of base
(Scheme 1). Formation of 9 on the contrary, may be
explained by admitting a nucleophilic attack of DMSO
to the C-13 carbonyl instead of the C-14 to form
The necessary intermediate 14-fluorodaunomycinone 10
has already been synthesized upon reacting 14-bromo
derivative with tetrabutyl ammonium fluoride; but this
reaction did not work on compound 8 albeit only on its
7-deoxy derivative.9,18 It had thus been necessary to
reintroduce the 7 hydroxyl group on the 7-deoxy-14-
fluoro derivative in order to perform the coupling with
the sugar residue. Our different methodology, allowing
the substitution of bromine with a fluorine atom
directly on compound 8 provided a more practical
synthetic pathway to the final glycoside.
A
first sample of 14-fluorodaunomycinone was
obtained upon refluxing 8 and [n-Bu4N]+ [Ph3SnF2]− in
a 2:1 (by vol.) mixture of acetonitrile and tetra-
hydrofuran.19 However, the prolonged reaction time (40
h) and low reaction yield (35%) prompted us to further
investigate this step.
i
ii
R
Br
R
O-S-(CH3)2
R
O-S-(CH3)2
RCHO
(CH3)2S
Ag+
Br
Ag+
O
O
O
i
Br
Br
R
R
R
13
14
O
S
O
S
CH3
CH3
CH3
CH3
5
H
O
O
O
R
H
OH
O
S
R
DMSO
CH3
CH3
6
Scheme 1. (i) AgBF4, DMSO; (ii) base.

M. Berettoni et al. / Tetrahedron Letters 43 (2002) 2867–2871
2869
Better results were obtained using 3HF·Et3N complex
as fluorinating agent. Heating compound 8 and neat
3HF·Et3N complex to 100°C for 4 h gave, after flash
chromatography purification, 14-fluoro derivative 10 in
a 73% yield. Addition of different solvents (toluene,
dioxane, acetonitrile, dimethoxyethane, triethylamine)
did not increase the yield.
made. If 14-fluoro,14-hydroxy daunomycinone 12 is
used in the condensation reaction, the 14-hydroxyl
group should be selectively protected. Moreover, obser-
vations made during the preparation of 12 showed that
the geminal fluorohydrin rapidly decomposes in even
mild basic conditions. On the other hand, a strong
acidic medium could not be tolerated by the newly
formed glycosidic linkage. Taking account of the
acceptable yield of the glycosidation reaction (over
50%) and of that of the hydroxylation (practically
quantitative) the condensation reaction was performed
on 11, using a suitably protected sugar derivative. The
selected glycosylating residue was 1-O-p-nitrobenzoyl-
The standard bromination procedure already used for
the preparation of 814 appeared not to be applicable to
10. However, 14-bromo,14-fluorodaunomycinone 11
was obtained in a 61% yield using a large excess of
pyridinium bromide perbromide (6 equiv. every 3 h five
times) and refluxing for 20 h in tetrahydrofuran. Com-
pound 11 was then dissolved in DMSO and added to a
solution of AgBF4 in DMSO. After a few minutes
stirring at room temperature water was added and the
mixture extracted with dichloromethane to give 14-
fluoro,14-hydroxy daunomycinone 12, as a mixture of
diastereoisomers in a 95% yield (Scheme 2).
3-N,4-O-di-allyloxycarbonyl-
L
-daunosamine
13,
obtained in a three-step sequence from commercially
available 2.† Both amino and hydroxyl functionalities
were protected with an allyloxycarbonyl group, because
it is easily removable under neutral conditions.20
Compounds 11 and 13 were allowed to react in the
presence of TMSOTf as activating agent in a mixture of
diethyl ether and dichloromethane according to a well
known procedure21 and the crude reaction mixture
purified by flash chromatography to give 14 with a 54%
Having achieved the synthesis of the geminal fluorohy-
drin system, our attention focused on compound 3.
With regard to the coupling reaction of the sugar
moiety with the aglycone, some considerations must be
O
OH
O
O
OH
O
Br
i
OH
OH
7
CH3O
O
OH OH
CH3O
O
OH OH
8
7
iii
ii
O
O
OH
O
O
O
OH
O
F
OH
OH
OH
CH3O
OH OH
CH3O
OH OH
10
9
iv
O
O
OH
O
O
O
OH
O
F
OH
F
ii
OH
OH
Br
CH3O
OH OH
CH3O
OH OH
12
11
Scheme 2. (i) Pyridinium bromide perbromide (PBP), THF, rt, 70%; (ii) AgBF4, DMSO, rt, 95%; (iii) Et3N·3HF, CH3CN, reflux,
73%; (iv) PBP, THF, reflux, 61%.
† Daunorubicin 2 was treated with allylchloroformiate in dichloromethane and pyridine and then submitted to hydrolysis in dioxane/6N HCl at
60°C to give daunomycinone 7 and 3-N,4-O-diallyloxy carbonyl-
p-nitro benzoylchloride in pyridine to give 13.
L-daunosamine which was activated at the anomeric position by treatment with

2870
M. Berettoni et al. / Tetrahedron Letters 43 (2002) 2867–2871
(1H, m, H-7), 4.8 (1H, d, 7-OH), 4.15 (3H, s, OCH3),
3.6–3.3 (2H, m, H-10), 3.4 (1H, s, 9-OH), 3.15–2.8 (1H,
m, H-10), 2.5–2.2 (2H, m, H-8). ESMS m/z calcd for
C21H16BrFO8 (M+H)+ 496.26, found 496.41.
yield. The obtained glycoside was then dissolved in
DMSO and added to a solution of AgBF4 in DMSO.
After 5 min stirring at room temperature the mixture
was added with water and extracted with
dichloromethane to give 15 with a 95% yield. The final
deprotection step was performed in dichloromethane in
the presence of tetrakis (triphenylphosphine) palla-
dium(0), trimethylsilyl acetate and N,N-dimethyl,
Compound 12 (mixture of two diastereoisomers)
1H NMR (CDCl3): l 13.2 (2H, d, Ph-OH), 8.05 (1H, d,
H-1), 7.8 (1H, t, H-2), 7.4 (1H, d, H-3), 7.2 (0.5H, d,
H-14, J=52 Hz), 7.05 (0.5H, d, H-14, J=52 Hz), 5.4
(1H, m, H-7), 4.75 (1H, d, 7-OH), 4.1 (3H, s, OCH3),
3.4 (1H, s, 9-OH), 3.3 (1H, dd, H-10), 3.05 (1H, dd,
H-10), 2.75–2.3 (2H, m, H-8). ESMS m/z calcd for
C21H17FO9 (M+H)+ 433.37, found 433.55.
trimethylsilylamine20 affording
(Scheme 3).
3 in a 94% yield
In conclusion, we have described the synthesis of the
novel anthracycline 14-fluorodoxorubicin. Its cytotoxic
activity against A2780 and GLC-4 cell lines turned out
to be about ten-fold lower than that of doxorubicin 1
(internal report), while antitumor efficacy against
human tumor cell lines grown in athymic nude mice, is
currently under investigation. We have also described a
useful method to introduce both fluorine atom and
hydroxyl group in the C9-side chain of anthracyclines
and we believe that this approach should also prove
convenient for the synthesis of geminal fluorohydrins a
to a carbonyl group.
Compound 14 (mixture of two diastereoisomers)
1H NMR (CDCl3): l 13.2 (2H, d, Ph-OH), 8.0 (1H, d,
H-1), 7.8 (1H, t, H-2), 7.44 (1H, d, H-14, J=52 Hz), 7.4
(1H, d, H-3), 6.0–5.8 (2H, m, CHꢀCH2), 5.55 (1H, s,
H-1%), 5.4–5.15 (5H, m, CHꢀCH2, H-7), 5.0 (1H, s,
H-4%), 4.75–4.5 (5H, m, OCH2-CH, H-3%), 4.2 (3H, s,
OCH3), 4.2–4.1 (1H, m, H-5%), 3.55 (1H, d, H-10), 3.2
(1H, d, H-10), 2.4–2.25 (2H, m, H-8), 1.85 (2H, m,
H-2%), 1.2 (3H, d, H-6). ESMS m/z calcd for
C35H35BrFNO14 (M+H)+ 793.57, found 793.70.
Selected experimental data
Compound 15 (mixture of two diastereoisomers)
1H NMR (CDCl3): l 13.2 (2H, d, Ph-OH), 8.0 (1H, d,
H-1), 7.8 (1H, t, H-2), 7.4 (1H, d, H-3), 7.15 (0.5H, d,
H-14, J=52 Hz), 7.1 (0.5H, d, H-14, J=52 Hz), 6.0–5.8
(2H, m, CHꢀCH2), 5.55 (1H, s, H-1%), 5.4–5.1 (5H, m,
Compound 11 (mixture of two diastereoisomers)
1H NMR (CDCl3): l 13.2 (2H, d, Ph-OH), 8.05 (1H, d,
H-1), 7.8 (1H, t, H-2), 7.46 (0.5H, d, H-14, J=52 Hz),
7.42 (1H, d, H-3), 7.40 (0.5H, d, H-14, J=52 Hz), 5.4
O
OH
O
O
OH
O
F
F
OH
OH
Br
Br
CH3O
O
OH OH
CH3O
O
OH
O
O
i
11
NHCO2All
AllCO2O
O-p-NO2-Bz
14
O
NHCO2All
AllCO2O
ii
13
O
O
OH
O
O
O
OH
O
F
OH
F
OH
iii
OH
OH
CH3O
OH
O
O
CH3O
OH
O
O
NH2
NHCO2All
HO
AllCO2O
3
15
Scheme 3. (i) TMSOTf, CH2Cl2/Et2O, −20°C, m.s., 54%; (ii) AgBF4, DMSO, rt, 95%; (iii) [(Ph3P)4]Pd, (CH3)2NTMS, AcOTMS,
CH2Cl2, rt, dark, 94%.

M. Berettoni et al. / Tetrahedron Letters 43 (2002) 2867–2871
2871
CHꢀCH2, H-7), 5.0 (1H, s, H-4%), 4.8 (1H, m, H-3%),
4.6–4.5 (4H, m, OCH2-CH), 4.3–4.15 (1H, m, H-5%),
4.05 (3H, s, OCH3), 3.4–2.65 (2H, m, H-10), 2.4–2.25
(2H, m, H-8), 2.0 (1H, d, 14-OH), 1.9 (2H, m, H-2%),
1.25 (3H, d, H-6). ESMS m/z calcd for C35H36FNO15
(M+H)+ 730.67, found 730.78.
7. Binaschi, M.; Capranico, G.; Dal Bo, L.; Zunino, F. Mol.
Pharmacol. 1997, 51, 1053–1059.
8. Arcamone, F.; Animati, F.; Capranico, G.; Lombardi, P.;
Pratesi, G.; Manzini, S.; Supino, R.; Zunino, F. Pharma-
col. Ther. 1997, 76, 117–124.
9. Matsumoto, T.; Ohsaki, M.; Yamada, K.; Matsuda, F.;
Terashima, S. Chem. Pharm. Bull. 1988, 36, 3793–3804.
10. Matsuda, F.; Matsumoto, T.; Ohsaki, M.; Terashima, S.
Tetrahedron Lett. 1989, 32, 4259–4262.
11. Fokin, A. V.; Astashkin, E. I.; Prikhodko, A. Z.;
Ryabichenko, V. V.; Nikolaeva, I. S.; Slavich, Y. V.;
Kolomiets, A. F. Dokl. Akad. Nauk. SSSR 1985, 284,
1492–1494.
12. Guest, A.; Milner, P. H.; Southgate, R. Tetrahedron Lett.
1989, 30, 5791–5794.
13. Middleton, W. J. J. Org. Chem. 1975, 40, 574.
14. (a) Tamoto, K.; Sugimori, M.; Terashima, S. Tetrahedron
1984, 40, 4617; (b) Suzuki, M.; Kimura, Y.; Terashima,
S. Chem. Pharm. Bull. 1986, 34, 1531.
15. Bauer, D. P.; Macomber, R. S. J. Org. Chem. 1975, 40,
1990–1992.
16. Wilcox, C. F., Jr.; Weber, K. A. J. Org. Chem. 1986, 51,
1088–1094.
Compound 3 (mixture of two diastereoisomers)
1H NMR (CDCl3): l 13.2 (2H, d, Ph-OH), 8.0 (1H, d,
H-1), 7.8 (1H, t, H-2), 7.4 (1H, d, H-3), 7.1 (0.5H, d,
H-14, J=52 Hz), 7.05 (0.5H, d, H-14, J=52 Hz), 5.55
(1H, s, H-1%), 5.35 (1H, s, H-7), 4.1 (3H, s, OCH3), 4.0
(1H, m, H-5%), 3.5–3.4 (2H, m, H-10, H-4%), 3.2–3.05
(2H, m, H-10, H-3%), 2.75–2.2 (2H, m, H-8), 1.85–1.6
(2H, m, H-2%), 1.25 (3H, d, H-6).). ESMS m/z calcd for
C27H29ClFNO11 (M+H)+ 598.98, found 599.13.
References
1. Arcamone, F. Doxorubicin, Anticancer Antibiotics;
Medicinal Chemistry Series, 1981; Vol. 17, pp. 1–369.
2. Wang, J. C. Annu. Rev. Biochem. 1996, 65, 635–692.
3. Capranico, G.; Binaschi, M.; Borgnetto, M. E.; Zunino,
F.; Palumbo, M. Trends Pharmacol. Sci. 1997, 18, 323–
329.
17. Ganem, B.; Boeckman, R. K., Jr. Tetrahedron Lett. 1974,
11, 917–920.
18. Matsumoto, T.; Ohsaki, M.; Matsude, F.; Terashima, S.
Tetrahedron Lett. 1987, 28, 4419–4422.
19. Gingras, M. Tetrahedron Lett. 1991, 32, 7381–7384.
20. Merzouk, A.; Guibe´, F. Tetrahedron Lett. 1992, 33, 477–
480.
4. Gatto, B.; Capranico, G.; Palumbo, M. Curr. Pharm.
Des. 1999, 5, 195–215.
5. Capranico, G.; Zunino, F. Anticancer Drug Des. 1990, 5,
307–317.
21. Kimura, Y.; Suzuki, M.; Matsumoto, T.; Abe, R.;
Terashima, S. Bull. Chem. Soc. Jpn. 1986, 59, 415–423.
6. Capranico, G.; Butelli, E.; Zunino, F. Cancer Res. 1995,
55, 312–317.