A daunorubicin β-galactoside prodrug for use in conjunction with gene- directed enzyme prodrug therapy

Article abstract of DOI:10.1016/S0040-4039(00)00742-5

N-[4-(β-D-Galactopyranosyl)-3-nitrobenzyloxycarbonyl]daunomycin was synthesized as a neutral prodrug of daunomycin. Although stable in 0.05 M phosphate buffer, pH 7.4 at 37°C, it was rapidly converted to daunomycin and 3-nitro-4-hydroxybenzyl alcohol under the same conditions in the presence of E. coli β-galactosidase. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00742-5

Tetrahedron Letters 41 (2000) 4871±4874
A daunorubicin b-galactoside prodrug for use in conjunction
with gene-directed enzyme prodrug therapy
Ajit K. Ghosh, Saeed Khan, Frank Marini, J. Arly Nelson and David Farquhar*
Department of Experimental Therapeutics, University of Texas M. D. Anderson Cancer Center, Houston,
Texas 77030, USA
Received 7 March 2000; revised 27 April 2000; accepted 1 May 2000
Abstract
N-[4-(b-d-Galactopyranosyl)-3-nitrobenzyloxycarbonyl]daunomycin was synthesized as a neutral prodrug
of daunomycin. Although stable in 0.05 M phosphate bu€er, pH 7.4 at 37ꢀC, it was rapidly converted to
daunomycin and 3-nitro-4-hydroxybenzyl alcohol under the same conditions in the presence of E. coli b-
galactosidase. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: anthracycline; daunomycin; prodrug; b-galactosidase; gene therapy.
Anthracycline antibiotics, particularly doxorubicin and daunorubicin, have been used extensively
in the treatment of human malignancies.1 However, cardiotoxicity and multidrug resistance are
signi®cant problems that limit the clinical ecacy of such agents.2,3 In an e€ort to design less
toxic and more e€ective antitumor drugs, we have reported various agents4 11 that can be used in
conjunction with antibody-directed enzyme prodrug therapy12 (ADEPT) and gene-directed
enzyme prodrug therapy (GDEPT).13 In the latter approach, a transduced gene is used to encode
an enzyme that can convert a nontoxic prodrug into a cytotoxic product. The two most com-
monly used prodrug/activating enzyme combinations are: (a) ganciclovir (GCV)/herpes simplex
virus thymidine kinase (HSV tk), and (b) 5-¯uorocytosine/cytosine deaminase (CD). A limitation
of HSV tk and CD as drug-activating enzymes is that they are substrate selective and can only
activate prodrugs that closely structurally resemble the natural substrates. The vast majority of
clinically useful antitumor agents (e.g. mustards, anthracyclines, antifols, etc.) are not candidates
for `prodrug activation' by these enzymes. In an e€ort to use such agents in conjunction with
ADEPT and GDEPT, we have investigated b-galactosidase4 7 and b-glucuronidase8 11 as
potential prodrug-activating enzymes. The former is particularly appealing because it can be
readily expressed in mammalian cells transduced with the E. coli LacZ gene, and has been widely
used as a marker enzyme in gene therapy studies.14 In this communication, we report a
* Corresponding author.
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00742-5

4872
daunomycin prodrug 1 (Scheme 1) that is activated by b-galactosidase. The approach is based on
the use of certain para-substituted benzyloxycarbonyl groups as bioreversible amine protective
groups.15
Scheme 1.
A similar strategy has been reported with other drug-activating enzymes including b-galactosidase16
and b-glucuronidase.17 Conference reports of this work have appeared.4,5
Prodrug 1 was prepared as shown in Scheme 2.
Experimental: A solution of 4-hydroxy-3-nitrobenzaldehyde, 5 (1.26 g, 7.55 mmol) in CH3CN
(15 ml) was added, with stirring, to a solution of tetra-O-acetyl-a-d-galactopyranosyl bromide,
418 (3.1 g, 7.55 mmol) in CH3CN (50 ml) containing freshly-prepared, dry Ag2O (8.0 g, 32.33
mmol). After 4 h at room temperature, the mixture was ®ltered to remove Ag salts, and the
solvent was evaporated. The crude product was puri®ed by column chromatography on silica
(200±400 mesh, Merck Inc.) using CH2Cl2:MeOH (95:5) as mobile phase to give the b-anomer 6
as a white solid (2.6 g, 69%), mp 191±192ꢀC. A solution of 6 (2.49 g, 5.0 mmol) in a mixture of
dry CHCl3 (30 ml) and dry (CH3)2CHOH (7 ml) was stirred with NaBH4 (0.4 g, 10.4 mmol) for 2 h

4873
Scheme 2. Reagents: (i) Ag2O±CH3CN; (ii) NaBH4, CHCl3±iPrOH; (iii) DSC±Et3N, CH3CN; (iv) DMF±Et3N; (v)
MeOH±MeONa
at 0±5ꢀC. After work up, the alcohol 7 was obtained as a white solid (1.8 g, 73%), mp 195±196ꢀC.
A solution of 7 (0.25 g, 0.5 mmol) in anhydrous CH3CN (5 ml) was treated with a solution of
disuccinimidyl carbonate (0.25 g, 1.0 mmol) in the same solvent (5 ml) and the mixture was stirred
for 90 min. The solution was evaporated to dryness and the residue was taken up in DMF (1 ml),
and a solution of daunorubicin hydrochloride, 3 (0.15 g, 0.3 mmol) in DMF (2 ml) was added
followed by one drop of Et3N. After 25 min at room temperature, the DMF was evaporated and
the residue was chromatographed on silica (200±400 mesh, Merck Inc.) using CHCl3:MeOH
(95:5) as mobile phase to give 8 (0.2 g, 38% with respect to 7). Compound 8 (0.03 g, 0.028 mmol)
was deacylated with MeONa (2 mg) in MeOH (2.5 ml) at 0±5ꢀC for 2.5 h to give 119 as a red solid
(0.012 g, 47%).
Stability and Cytotoxicity Studies of 1. Compound 1 was completely stable when incubated at a
concentration of 10^6 M in 0.05 M phosphate bu€er, pH 7.4, at 37ꢀC for up to six days. Under
similar conditions in the presence of 2 units of E. coli b-galactosidase per mmole of prodrug, 1 was
degraded with a half-life of 8.4 min (Fig. 1). HPLC analysis20 of the reaction mixture showed
progressive disappearance of the peak attributable to 1 (retention time, 5.1 min), and the
Figure 1. Degradation of daunomycin prodrug, 1 (10^6 M) in 0.05 M phosphate bu€er, pH 7.4, in the presence of
E. coli b-galactosidase (2 units per mmole of 1)

4874
appearance of two new, progressively increasing, peaks with retention times identical to those of
authentic samples of 3-nitro-4-hydroxybenzyl alcohol and daunorubicin, (2.5 min and 3.1 min,
respectively.)
Evidence that the reaction was enzyme-mediated was obtained from the observation that the
degradation half-lives of 1 increased to 25 min and 41 min, respectively, when the concentration
of b-galactosidase in the incubation mixture was reduced by 50 and 75%. When incubated in
culture with LacZ-transduced (i.e. b-galactosidase-expressing) human prostate, breast, or leukemic
cancer cells, 1 showed a 100 to 300-fold enhancement in cytotoxicity compared to mock-transduced
control cells. These ®ndings demonstrate that 1 is a good substrate for b-galactosidase and suggests
that the compound may be therapeutically useful in an ADEPT or GDEPT setting. Antitumor
evaluation studies of 1 are in progress.
Acknowledgements
This research was supported by NIH grant CA 71527.
References
1. Myers, C. E. Cancer Chemotherapy & Biological Response Modi®ers 1992, 13, 45±52.
2. Mott, M. G. Anal. N.Y. Acad. Sci. 1997, 824, 221±228.
3. Kroger, N.; Achterrath, W.; Hegewisch-Becker, S.; Mross, K.; Zander, A. R. Cancer Treatment Reviews 1999, 25,
279±291.
4. Farquhar, D.; Ghosh, A. K.; Bakina, E.; Marini, F. Proc. Amer. Assoc. Cancer Res. 1997, 38, 613 (4121).
5. Marini, F.; Ghosh, A. K.; Bakina, E.; Nelson, J. A.; Farquhar, D. Proc. Amer. Assoc. Cancer Res. 1997, 38, 342
(2294).
6. Ghosh, A. K.; Khan, S. R.; Farquhar, D. Chem. Commun. 1999, 24, 2527±2529.
7. Bakina, E.; Farquhar, D. Anti-Cancer Drug Des. 1999, 14 (6), in press.
8. Bakina, E.; Wu, Z.; Rosenblum, M. Proc. Amer. Assoc. Cancer Res. 1997, 38, 432 (2895).
9. Farquhar, D.; Ghosh, A. K.; Nelson, J. A. Proc. Amer. Assoc. Cancer Res. 1998, 39, 219 (1494).
10. Bakina, E.; Wu, Z.; Rosenblum, M.; Farquhar, D. J. Med. Chem. 1997, 40, 4013±4018.
11. Ghosh, A. K.; Farquhar, D. Tetrahedron Lett. 1997, 38, 8795±8798.
12. Bagshawe, K. D. Br. J. Cancer 1989, 60, 275±281.
13. Gutierrez, A. A.; Lemoine, N. R.; Sikora, K. Lancet 1992, 339, 715±721.
14. Macgregor, G. R.; Nolan, G. P.; Fiering, S.; Roeder, M.; Herzenberg, L. A. Methods Mol. Biol. 1991, 7, 217±235.
15. Carl, P. L.; Chakravarty, P. K.; Katzenellenbogen, J. A. J. Med. Chem. 1981, 24, 479±480.
16. Andrianomenjanahary, S.; Dong, X.; Florent, J. C.; Gaudel, G.; Gesson, J. P.; Jacquesy, J. C.; Koch, M.; Michel,
S.; Mondon, M.; Monneret, C.; Petit, P.; Renoux, B.; Tillequin, F. J. Bioorg. Med. Chem. Lett. 1992, 2, 1093±
1096.
17. Bosslet, K.; Czech, J.; Ho€mann, D. Cancer Res. 1994, 54, 2151±2159.
18. Bollenback, G. N.; Long, J. W.; Benjamin, D. G.; Lindquist, J. A. J. Am. Chem. Soc. 1955, 77, 3310.
19. High resolution FAB-MS m/e 907.2380 (M+Na)+.
20. HPLC conditions: A C-18 reverse-phase column (Phenomenex, 150 mmÂ3.90 mm) was used. The mobile phase
was 0.05 M NH4OAc bu€er (pH 4.7)±CH3CN (3:2) at a ¯ow rate of 1.0 ml per min. Prodrug 1 and the enzyme
reaction products were monitored by a variable wavelength UV detector set at 254 nm with 0.01 AUFS sensitivity
(Waters 484 model) and quantitated electronically as a function of time using a NEC Pinwriter P6200 integrator.
The half-lives were determined by non-linear regression analysis. b-Galactosidase was used as received from Sigma
Company (catalog number G-2513).