Synthesis of TDP-3-amino-3,4,6-trideoxy-α-D-xylo-hexopyranose - The immediate precursor of TDP-α-D-desosamine

Article abstract of DOI:10.1016/S0960-894X(02)00209-3

A synthetic pathway producing the title compound starting from methyl α-D-glucose is described. This compound was shown to be a substrate for DesVI, an AdoMet-dependent methyltransferase which catalyzes N,N-dimethylation of the title compound to give a biological significant unusual sugar, desosamine.

Full text of DOI:10.1016/S0960-894X(02)00209-3

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1493–1496
Synthesis of TDP-3-Amino-3,4,6-trideoxy-ꢀ-D-xylo-
hexopyranose—The Immediate Precursor of TDP-ꢀ-D-Desosamine
Cheng-Wei T. Changa,b,y and Hung-wen Liua,b,
*
aDivision of Medicinal Chemistry, College of Pharmacy, and Department of Chemistry and Biochemistry,
University of Texas, Austin, TX 78712, USA
bDepartment of Chemistry, University of Minnesota, Minneapolis, MN 55455, USA
Received 14 January 2002; accepted 8 March 2002
Abstract—A synthetic pathway producing the title compound starting from methyl a-d-glucose is described. This compound was shown
to be a substrate for DesVI, an AdoMet-dependent methyltransferase which catalyzes N,N-dimethylation of the title compound to give
a biological significantunusual sugar, desosamine. # 2002 Elsevier Science Ltd. All rights reserved.
Aminosugars have been found in a wide variety of nat-
ural products.1 Since an amino group can be proto-
nated, alkylated, or acetylated to alter its capability to
form hydrogen-bond(s) and to bear charge, its presence
can greatly influence the hydrophobicity and surface
properties of the natural product structure as a whole.
For example, the amino group of daunosamine (1) in
doxorubicin (2) may be the recognition site for the
multidrug transporter P-glycoprotein in drug-resistant
tumor cells. Substitution of this basic amino group with
a neutral hydroxy group either partially or completely
circumvents the resistance due to the reduced transpor-
tation by P-glycoprotein.2 Indeed, aminosugars have
been found to play important roles for the biological
activities of many macrolide and aminoglycoside anti-
biotics.3 A particularly notable example is desosamine
(3), a 3-N,N-dimethylamino-3,4,6-trideoxyhexose, which
has been found in many macrolide antibiotics, including
erythromycin (4), methymycin (5), pikromycin, and
oleandomycin.
but enhanced its motor-stimulating effect on gastro-
intestinal muscle.5 In fact, the dimethylamino group of
desosamine in erythromycin can be oxidized by cyto-
chrome P-450 in liver, affecting the metabolism of other
drugs.6 The basic amino group on erythromycin and
other antibiotics of the same type may influence the
uptake of these drugs by phagocytes, thus affecting their
delivery to the infection sites.7
Studies of the biosynthesis of desosamine have estab-
lished the intermediacy of an amino-sugar, TDP-3-
amino-3,4,6-trideoxy-a-d-xylo-hexopyranose (6), in the
pathway.8 Consequently, N,N-dimethylation of 6 has
been shown to be the last step in desosamine formation.
In an effort to study the biosynthesis of methymycin (5)/
pikromycin in Streptomyces venezuellae,8c,d we have
found that N,N-dimethylation of 6 to make TDP-desos-
maine (7) is catalyzed by DesVI, which is an S-adeno-
sylmethionine (AdoMet) dependent methyltransferase
(Scheme 1).9 This enzyme has been overexpressed and
purified. To elucidate the mechanism of DesVI cata-
lyzed N,N-dimethylation steps, we have developed an
efficient sequence to prepare the substrate 6. Reported
herein is the synthesis of this compound, which has later
been converted to TDP-desosamine (7) by DesVI in a
stepwise manner via a monomethylated intermediate.
Ithas been shown htatdes- N-methylerythromycin, in
which desosamine is mono-demethylated, retains only
5% of the antibacterial activity when compared to ery-
thromycin.4 Further experiments also indicated that
modification of the N,N-dimethyl group of desosamine
resulted in the loss of this drug’s antibacterial activity,
As depicted in Scheme 2, the first stage of the synthesis
calls for the preparation of the key intermediate 10.
Preparation of this compound was initiated by using a
known protocol to remove both C-4 and C-6 hydroxyl
groups of glucose to make 4,6-dideoxyhexose 9.10
*Corresponding author. Tel.: +1-512-232-7811; fax: +1-512-471-
2746; e-mail: h.w.liu@mail.utexas.edu
yCurrent address: Department of Chemistry and Biochemistry, Utah
State University, Logan, UT 84322, USA.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00209-3

1494
C.-W. T. Chang, H. Liu / Bioorg. Med. Chem. Lett. 12 (2002) 1493–1496
Attempts to carry out this transformation directly on
a-d-glucose proved to be problematic. However, much
improved yield was achieved when methyl a-d-glucoside
(8) was used as the starting material. Subsequent ben-
zylation of 9 mediated by bis(tributyltin) oxide11 led to a
2:1 mixture of monobenzylated 10 and 11. The regios-
electivity for C-2 alkylation is lower than expected and
may be a resultof C-4 deoxygenation, which renders the
C-3 hydroxyl group more nucleophilic and less sterically
hindered for alkylation. Since 10 and 11 had nearly
identical chromatographic properties, it was difficult to
isolate the desired 2-O-benzylated 10 in pure form.
Acetylation of this mixture, however, provided a pair of
regioisomers, 12 and 13, which could be resolved by
repeated chromatography. The desired isomer 12 was
collected and subjected to base hydrolysis to give 10.
The configuration of the 3-OH group in 10 was inverted
from equatorial to axial in 14 by Swern oxidation and
NaBH4 reduction in sequence. The masked amino func-
tionality was introduced into 14 atC-3 under Mitsunobu
conditions using diphenylphosphoryl azide as the azide
source to afford 15.12 Removal of the anomeric methoxyl
Scheme 1.
Scheme 2.

C.-W. T. Chang, H. Liu / Bioorg. Med. Chem. Lett. 12 (2002) 1493–1496
1495
Scheme 3.
group was accomplished by treatment of 15 with BBr3
at low temperature13 and the product 16 was benzyl-
phosphorylated to generate compound 17. The unpro-
tected aminosugar-1-phosphate 18 was obtained in one
pot via hydrogenation catalyzed by Dugassa type
Pd(OH)2.14 The final step involved coupling of 18 with
TMP-morpholidate in the presence of 1H-tetrazole.15
The crude productwas purified using P2 column chro-
matography (eluted with 25 mM ammonium bicarbo-
nate) and FPLC equipped with a MonoQ column
(eluted with a linear gradient of 0–0.5 M ammonium
bicarbonate buffer over 15 min) to give 6, whose struc-
ture was confirmed by spectral analyses.16
tography on silica gel (10–30% ether in hexanes). This
procedure had been successfully applied to obtain pure
10 from a mixture of 24 g of 12 and 13 using 5 g of
lipase.18 Since the recovered isomer 13 could be con-
verted to the desired 10 via a recycling route as illu-
strated in Scheme 3, the combined yield of 10 was
greater than 56% based on the total mass of the original
mixture of isomers 12 and 13.
In summary, we have developed a practical method for
the preparation of TDP-3-amino-3,4,6-trideoxyhexose
6, which is the immediate precursor of desosamine (3).
The synthetic sequence combines chemical and enzy-
matic reactions to achieve dideoxygenation at C-4 and
C-6, and differentiation of the C-2 and C-3 hydroxyl
groups. Separation of the structural isomers, 12 and 13,
was facilitated by regioselective hydrolysis catalyzed by
lipase. The use of a dialysis bag to confine the lipase dur-
ing the incubation18 is worthy of mentioning. Since it not
only alleviates the common emulsion problem associated
with the chemoenzymatic synthesis, but also enables us to
recycle the lipase. Using the synthetic 6 as the substrate,
we have unequivocally established the catalytic role of
DesVI to be an AdoMet-dependent methyltransferase.
The N,N-dimethylation proceeds via a stepwise manner,
and the Km and kcat values for 6 were determined to be
307.4 mM and 92.0 minÀ1, respectively, when the con-
centration of AdoMet was maintained at 3.0 mM.
Although synthesis according to the above reaction
sequence had provided sufficient quantity of compound
6 for initial enzymatic study, attempts to carry out
large-scale preparation were hampered by the tedious-
ness of isolation of multi-grams of 10 via repetitive
chromatography. Thus, efforts were directed to develop
a more efficient method to separate 10/11 or 12/13.
Since lipases are known to be able to catalyze chemo-,
regio- and/or stereoselective hydrolysis of a variety of
esters and/or amides,17 the feasibility of using a lipase
catalyzed regioselective deacylation to achieve differ-
entiation between 12 and 13 was explored. Among three
lipases tested, Type I from wheat germ, Type II from
porcine pancreas, and Type VII from Candida rugosa,
only Type I lipase from wheatgerm showed selecitve
activity toward 12. Using the wheat germ lipase, com-
pound 12 could be converted to 10, while compound 13
remained intact. The optimal selectivity and turnover
ratio were achieved using 10% acetonitrile as the co-
solvent (in 50 mM potassium phosphate buffer, pH 8.0).
Since compounds 10 and 13 are no longer a pair of
regioisomers, they can be readily separated by chroma-
Acknowledgements
This work was supported in part by a grant from the
National Institutes of Health (GM54346). H.-w.L. also
thanks the National Institute of General Medical Sciences
for a MERIT Award.

1496
C.-W. T. Chang, H. Liu / Bioorg. Med. Chem. Lett. 12 (2002) 1493–1496
9. (a) Zhao, L.; Sherman, D. H.; Liu, H.-w. J. Am. Chem.
References and Notes
Soc. 1998, 120, 10256. (b) Chang, C.-W. T.; Zhao, L.; Yamase,
H.; Liu, H.-w. Angew. Chem. Int. Ed. Engl. 2000, 39, 2160.
10. (a) Szarek, W. A.; Kong, X. In Preparative Carbohydrate
Chemistry; Hanessian, S., Ed.; Marcel Dekker: New York,
1997; p 105. (b) Jennings, H. J.; Jones, J. K. N. Can. J. Chem.
1962, 40, 1408.
11. Dasgupta, F.; Garegg, P. J. Synthesis 1994, 1121.
12. Hughes, D. L. Org. React. 1992, 42, 335.
13. Demuynck, M.; De Clercq, P.; Vandewalle, M. J. Org.
Chem. 1979, 44, 4863.
1. (a) Grisebach, H. Adv. Carbohydr. Chem. Biochem. 1978,
35, 81. (b) Williams, N.; Wander, J. In The Carbohydrates:
Chemistry and Biochemistry; Pigman, W., Horton, D., Eds.;
Academic: New York, 1980; Vol. 1B, p 761. (c) Robyt, J. F.
Essentials of Carbohydrate Chemistry; Springer: New York,
1997. (d) Kennedy, J. F. Carbohydrate Chemistry; Clarendon:
Oxford, 1988. (e) He, X.; Agnihotri, G.; Liu, H.-w. Chem. Rev.
2000, 100, 4615.
2. (a) Priebe, W.; Van, N. T.; Burke, T. G.; Perez-Soler, R.
Anticancer Drugs 1993, 4, 37. (b) Lothstein, L.; Sweatman,
T. W.; Priebe, W. Bioorg. Med. Chem. Lett. 1995, 5, 1807.
3. (a) von Ahsen, U.; Noller, H. F. Science 1993, 260, 1500.
(b) Hendrix, M.; Alper, P. B.; Priestley, E. S.; Wong, C. H.
Angew. Chem., Int. Ed. Engl. 1997, 36, 95. (c) Van Dyke, M.;
Dervan, P. B. Biochemistry 1983, 22, 2373. (d) Tronchet,
J. M. J.; Tronchet, J. F. Helv. Chim. Acta 1981, 64, 425. (e)
Varki, A. Glycobiology 1993, 3, 97. (f) Weymouth-Wilson,
A. C. Nat. Prod. Rep. 1997, 14, 99.
14. Wischnat, R.; Martin, R.; Wong, C.-H. J. Org. Chem.
1998, 63, 8361.
15. Wittmann, V.; Wong, C.-H. J. Org. Chem. 1997, 62, 2144.
16. 1H NMR (D2O) d 1.24 (3H, d, J=6.3 Hz, 5-Me), 1.50
(1H, ddd, J=11.7, 12.3, 12.7 Hz, 4-Hax), 1.93 (3H, s, 500-Me),
2.27 (1H, ddd, J=3.0, 4.0, 12.7 Hz, 4-Heq), 2.40 (2H, m, 20-
Hs), 3.59 (1H, ddd, J=4.0, 10.7, 12.2 Hz, 3-H), 3.74 (1H, ddd,
J=3.4, 2.9, 10.7 Hz, 2-H), 4.20 (3H, m, 40- and 50-Hs), 4.29
(1H, ddq, J=3.0, 11.7, 6.3 Hz, 5-H), 4.62 (1H, dt, J=3.0,
5.8 Hz, 30-H), 5.63 (1H, dd, J=3.4, 7.3 Hz, 1-H), 6.35 (1H, t,
J=6.8 Hz, 10-H), 7.71 (1H, s, 600-H); 13C NMR (D2O) d 11.6,
19.5, 35.3, 38.5, 48.9, 65.4, 69.5 (d, J=6.6 Hz), 70.9, 84.9, 85.0,
85.3 (d, J=10.1 Hz), 95.0 (d, J=8.0 Hz), 111.7, 137.3, 151.7,
166.5. 31P NMR (D2O) d À12.4 (d, J=20.9 Hz), À10.5 (d,
4. Flynn, E. H.; Murphy, H. W.; McMahon, R. E. J. Am.
Chem. Soc. 1955, 77, 3104.
5. Omura, S.; Tsuzuki, K.; Sunazuka, T.; Marui, S.; Toyota,
H.; Inatomi, N.; Itoh, Z. J. Med. Chem. 1987, 30, 1943.
6. (a) Delaforge, M.; Strtori, E.; Mansuy, D. Chem.-Biol.
Interact. 1988, 68, 179. (b) Delaforge, M.; Ladam, P.; Bouille,
G.; Gharbi-Benarous, J.; Jaouen, M.; Girault, J.-P. Chem.-
Biol. Interact. 1992, 85, 215. (c) Ladam, P.; Gharbi-Benarous,
J.; Delforge, M.; Van Calsteren, M. R.; Jankowski, C. K.;
Girault, J.-P. Bioorg. Med. Chem. 1995, 3, 587.
J=20.9 Hz).
High-resolution
FABMS
calcd
for
C16H28N3O13P2 (M+H+) 532.1097, found 532.1098.
17. Wong, C.-H.; Whitesides, G. Enzymes in Synthetic
Organic Chemistry; Elsevier Science: New York, 1994; p 70.
18. The wheatgerm lipase (5 g) dissolved in 50 mM potassium
phosphate buffer (pH 7.0, 20 mL) was put in a dialysis bag and
placed in a solution of 12 and 13 (24.0 g, 81.6 mmol) in the
same phosphate buffer (800 mL) containing 10% CH3CN
(80 mL). After gentle stirring for 2–3 days at 4 ꢀC, the reaction
solution was extracted with chloroform. The combined
organic extracts were dried over Na2SO4 and concentrated
under reduced pressure. The desired product 10 and the
unreacted 13 were then separated by gradient column chro-
matography on silica gel (10–30% ether/hexanes).
7. Gladue, R. P.; Bright, G. M.; Isaacson, R. E.; Newborg,
M. F. Antimicrob. Agents Chemother. 1989, 33, 277.
8. (a) Gaisser, S.; Bohm, G. A.; Cortes, J.; Leadlay, P. F. Mol.
Gen. Genet. 1997, 256, 239. (b) Summers, R. G.; Donadio, S.;
Staver, M. J.; Wendt-Pienkowski, E.; Hutchinson, C. R.;
Katz, L. Microbiology 1997, 143, 3251. (c) Zhao, L.; Sherman,
D. H.; Liu, H.-w. J. Am. Chem. Soc. 1998, 120, 9374. (d) Xue,
Y.; Zhao, L.; Liu, H.-w.; Sherman, D. H. Proc. Natl. Acad.
Sci. U.S.A. 1998, 95, 12111.