A methodological comparison: The advantage of phosphorimidates in expanding the sugar nucleotide repertoire

Full text of DOI:10.1021/jo981265n

7568
J . Org. Chem. 1998, 63, 7568-7572
molecules.6 For example, TDP-Rha (4) is the progenitor
A Meth od ologica l Com p a r ison : Th e
Ad va n ta ge of P h osp h or im id a tes in
Exp a n d in g th e Su ga r Nu cleotid e
Rep er toir e
of L-rhamnose, which is found in numerous Gram-
negative bacterial cell surface antigens and is also an
integral ligand of a variety of antibiotic/antitumor agents
produced by Gram-positive species.6,7 Unfortunately, 4
is only accessible via a tedious and expensive multien-
zymatic method.8 Consequently, in an effort to expand
the repertoire of available sugar nucleotides, we report
a chemical synthesis of 4 via the one pot generation and
reaction of TMP-imidazolide with rhamnose-1-phos-
phate (2). A comparison of the TMP-imidazolide method
to two distinct “state of the art” methods5n,osthe coupling
of TMP-morpholidate with 2 or the reaction of TDP with
2,3,4-tri-O-benzyl-L-rhamnopyranosyl bromide (3)9sre-
veals that 1,1′-carbonyldiimidazole (CDI)-catalyzed cou-
pling gives rise to reaction times and yields comparable
to the morpholidate method with significantly less ex-
pensive precursors and more flexibility in the overall
reaction parameters.
Yongxin Zhao and J on S. Thorson*
Laboratory for Biosynthetic Chemistry, Molecular
Pharmacology & Therapeutics Program, Memorial
Sloan-Kettering Cancer Center and the Sloan-Kettering
Division, Graduate School of Medical Sciences,
Cornell University, 1275 York Avenue, Box 309,
New York, New York 10021
Received J une 30, 1998
In tr od u ction
Recent research has revealed that complex glycosyla-
tion patterns mediate or modulate a variety of biological
processes.1 Thus, a major focus in glycobiology is on
methods to access biologically relevant oligosaccharides
via chemical or enzymatic synthesis.2 Of the available
enzymatic methods, glycosyltransferase-catalyzed oli-
gosaccharide synthesis is generally preferred since the
reactions are regio- and stereoselective and eliminate the
typical multiple protection/deprotection schemes required
for chemical synthesis.3 As a result, access to the
nucleotide sugar substrates of glycosyltransferases has
become a major limiting factor in glycosyltransferase-
catalyzed synthesis, and the chemical or enzymatic
syntheses of the eight primary sugar nucleotides utilized
by mammalian transferases (UDP-Glc, UDP-GlcNAc,
UDP-Gal, UDP-GalNAc, GDP-Man, GDP-Fuc, UDP-
GlcUA, and CMP-NeuAc)4 have been reported.5
Resu lts a n d Discu ssion
Formation of the activated TMP-imidazolide was
accomplished by reacting the triethylammonium salt of
TMP (1) with 3 equiv of 1,1′-carbonyldiimidazole5d,g,m,11
in DMF-d7 or CD3CN, and the reaction was monitored
by 31P NMR. Immediately after the addition of CDI, a
complete disappearance of the resonance signal of the
starting material (δ 3.31 in DMF-d7) and the formation
of the 5′-phosphocarbonylimidazolide (δ -6.26 in DMF-
d7 and -5.82 in CD3CN) were observed followed by the
eventual disappearance of the TMP 5′-phosphocarbon-
ylimidazolide with concomitant formation of the desired
In contrast, bacterial glycosyltransferases utilize an
almost endless repertoire of nucleotide sugars in con-
structing the cell wall and cell surface antigens as well
as for the glycosylation of numerous small bioactive
(5) (a) Moffatt, J . G.; Khorana, H. G. J . Am. Chem. Soc. 1959, 81,
1265. (b) Moffatt, J . G.; Khorana, H. G. J . Am. Chem. Soc. 1961, 83,
649-658. (c) Roseman, S.; Distler, J . J .; Moffatt, J . G.; Khorana, H.
G. J . Am. Chem. Soc. 1961, 83, 659-663. (d) Cramer, F.; Neunhoeffer,
H. Chem. Ber. 1962, 95, 1664-1669. (e) Clark, V. M.; Hutchinson, D.
W.; Kirby, A. J .; Warren, S. G. Angew. Chem., Int. Ed. Engl. 1964, 76,
704-712. (f) Ginsburg, V. Adv. Enzymol. 1964, 26, 35. (g) Hoard, D.
E.; Ott, D. G. J . Am. Chem. Soc. 1965, 87, 1785-1788. (h) Moffatt, J .
G. Methods Enzymol. 1966, 8, 136-142. (i) Tochikura, T.; Kawaguchi,
K.; Kawai, H.; Mugibayashi, Y.; Ogata, K. J . Ferment. Technol. 1968,
46, 970. (j) Tochikura, T.; Kawai, H.; Tobe, S.; Kawaguchi, K.; Osugi,
M.; Ogata, K. J . Ferment. Technol. 1968, 46, 957. (k) Kawaguchi, K.;
Kawai, H.; Tochikura, T. Methods Carbohydr. Chem. 1980, 8, 261-
269. (l) Wong, C.-H.; Haynie, S. L.; Whitesides, G. M. J . Org. Chem.
1982, 47, 5416-5418. (m) Simon, E. S.; Grabowski, S.; Whitesides, G.
M. J . Org. Chem. 1990, 55, 1834-1841. (n) Arlt, M.; Hindsgaul, O. J .
Org. Chem. 1995, 60, 14-15. (o) Wittmann, V.; Wong, C.-H. J . Org.
Chem. 1997, 62, 2144-2147.
* To whom correspondence should be addressed. E-mail: jthorson@
sbnmr1.ski.mskcc.org. Tel: (212) 639-6404. Fax: (212) 717-3066.
(1) (a) Varki, A. Glycobiology 1993, 3, 97-130. (b) Dennis, R. P.
Glycoimmun 1995, 1-11. (c) Dwek, R. A. Biochem. Soc. Trans. 1995,
23, 1-25. (d) In Molecular Glycobiology: Frontiers in Molecular Biology
Series; Fukuda, M., Hindsgaul, O., Eds.; Oxford University Press:
Oxford, UK, 1994.
(2) (a) Nilsson, K. G. I. Trends Biotech. 1988, 6, 256-264. (b) Meldal,
M. Curr. Opin. Struct. Biol. 1994, 4, 710-718. (c) Boons, G.-J . Drug
Discovery Today 1996, 1, 331-342. (d) Gijsen, H. J . M.; Qiao, L.; Fitz,
W.; Wong, C.-H. Chem. Rev. 1996, 96, 443-473. (e) Sofia, M. J . Drug
Discovery Today 1996, 1, 27-34.
(3) (a) Khan, S. H.; Hindsgaul, O. In Molecular Glycobiology:
Frontiers in Molecular Biology Series; Fukuda, M., Hindsgaul, O., Eds.;
Oxford University Press: Oxford, UK, 1994; pp 206-223. (b) Crawley,
S. C.; Palcic, M. M. In Modern Methods in Carbohydrate Synthesis;
Khan, S. H., O’Neill, R. A., Eds.; Harwood Academic Publishers:
Amsterdam, 1996; pp 492-517. (c) Wong, C.-H. In Modern Methods
in Carbohydrate Synthesis; Khan, S. H., O’Neill, R. A., Eds.; Harwood
Academic Publishers: Amsterdam, 1996; pp 467-491 (d) Auge, C.;
Gautheron-Le Narvor, C. In Preparative Carbohydrate Chemistry;
Hanessian, S., Ed.; Marcel Dekker: New York, 1997; pp 469-484. (e)
Fitz, W.; Wong, C.-H. In Preparative Carbohydrate Chemistry; Han-
essian, S., Ed.; Marcel Dekker: New York, 1997; pp 485-504.
(4) Abbreviations: Ara, L-arabinose; CMP, cytidine monophosphate;
Fuc, L-fucose; Gal, D-galactose; GDP, guanosine diphosphate; Glc,
D-glucose; GlcNAc, 2-N-acetyl-D-glucose; GalNAc, 2-N-acetyl-D-galac-
tose; GlcUA, D-glucuronic acid; Man, D-mannose; NeuAc, N-acetyl-
neuramic acid (sialic acid); NMP, nucleotide monophosphate; Rha,
L-rhamnose (6-deoxy-L-mannose); TDP, thymidine diphosphate; TMP,
thymidine monophosphate; UDP, uridine diphosphate; UMP, uridine
monophosphate.
(6) Liu, H.-w.; Thorson, J . S. Annu. Rev. Microbiol. 1994, 48, 223-
256.
(7) In Enediyne Antibiotics as Antitumor Agents; Borders, D. B.,
Doyle, T. W., Eds.; Marcel Dekker: New York, 1995.
(8) Marumo, K.; Lindqvist, L.; Verma, N.; Weintraub, A.; Reeves,
P. R.; Lindberg, A. A. Eur. J . Biochem. 1992, 204, 539-545.
(9) (a) Frechet, J . M.; Baer, H. H. Carbohydr. Res. 1975, 42, 369-
372. (b) Paulsen, H.; Kutscher, W.; Lockhoff, O. Chem. Ber. 1981, 114,
3233-3241. (c) Ottosson, H. Carbohydr. Res. 1990, 197, 101-107.
(10) Zhao, Y.; Biggins, J . B.; Thorson, J . S. J . Am. Chem. Soc.,
manuscript submitted.
(11) (a) Cramer, F.; Schaller, H.; Staab, H. A. Chem. Ber. 1961, 94,
1612-1621. (b) Schaller, H.; Staab, H. A.; Cramer, F. Chem. Ber. 1961,
94, 1621-1634. (c) Cramer, F.; Schaller, H. Chem. Ber. 1961, 94, 1634-
1640. (d) Cramer, F.; Neunhoeffer, H.; Scheit, K. H.; Schneider, G.;
Tennigkeit, J . Angew. Chem., Int. Ed. Engl. 1962, 1, 331. (e) Zatorski,
A.; Goldstein, B. M.; Colby, T. D.; J ones, J . P.; Pankiewicz, K. W. J .
Med. Chem. 1995, 38, 1098-1105.
(12) Similar intermediates were observed by both 1H and 31P NMR
in the CDI-catalyzed synthesis of NAD analogues.11e.
S0022-3263(98)01265-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/25/1998

Notes
Ta ble 1. Com p a r a tive Yield s of 4 via th e in Situ
TMP -Im id a zolid e Meth od (Meth od I), th e Cou p lin g of
TMP -Mor p h olid a te w ith 2 (Meth od II), a n d th e Rea ction
of TDP w ith 3 (Meth od III)
J . Org. Chem., Vol. 63, No. 21, 1998 7569
reaction in either DMF or ethanol resulted in a <2%
change in the overall yield, while a slightly larger effect
was observed (∼10% increase in yield) upon the addition
of equivalent amounts of 1H-tetrazole in methanol.
To confirm the biological relevance of our synthetic 4,
we examined its ability to serve as a substrate for the in
vitro biosynthesis of a precursor for the Salmonella group
E1 O antigen repeat unit (11, Scheme 2).14 The biosyn-
thesis of 11 is initiated by enzyme(Gal-1-PT)-catalyzed
galactose-1-phosphate transfer from UDP-Gal to unde-
caprenyl phosphate (5),15 with subsequent loss of UMP
to provide 7, followed by enzyme (RhaTR3)-catalyzed
rhamnosyl transfer from TDP-Rha and loss of TDP to
give 9, and subsequent enzyme (ManTâ4)-catalyzed man-
nosyl transfer from GDP-Man with loss of GDP.10 To
accomplish the assay, a mixture containing a truncated
analogue of 5 (heptaprenyl phosphate, 6), Gal-1-PT,
UDP-Gal, RhaTR3, 4, ManTâ4, and 0.33 µM GDP-[U-
14C]Man (15 000 cpm) was allowed to react for 2 h, and
the reaction was subsequently quenched by combining
with 4 mL of CHCl3/CH3OH (2:1). Excess GDP-[U-14C]-
Man was removed from the glycolipid products by selec-
tive extraction,16 the glycolipids were subsequently dried
by evaporation and dissolved in scintillant, and the extent
of radioactivity recovered (directly proportional to the
amount of 12 produced) was determined by liquid scintil-
lation counting. Under these conditions, significant
ManTâ4-catalyzed mannosyltransfer to provide 12 was
observed (819.5 ( 40.5 cpm), and parallel control reac-
tions lacking 4 (164.5 ( 18.1 cpm) clearly demonstrate
that chemically synthesized 4 is required to complete the
biosynthetic cascade. Control reactions in the absence
of Gal-1-PT (89.5 ( 13.4 cpm), RhaTR3 (215.0 ( 20.7 cpm),
or ManTâ4 (153.0 ( 17.5 cpm) are also consistent with
the requirement of these reagents for the in vitro
enzymatic construction of 12.
method I
method II
yield (%)a,c
method III
yield (%)a,d
time (h)
yield (%)a,b
5
12
24
48
NDe
57.5
71.9
75.5g
NDe
53.3
67.7
74.7g
18.3f
NDe
NDe
NDe
a
Reactants and products were separated by analytical HPLC,
and the reported yields were calculated directly from peak
b
integration. Coupling solvent DMF, stoichiometry of TMP-
imidazolide and 2 was 1.23:1. c Coupling solvent pyridine, stoi-
chiometry of TMP-imidazolide, 2 and 1H-tetrazole was 1.33:1:3.
d
Coupling solvent CH2Cl2, stoichiometry of TDP and 3 was 1:3.5.
e Not determined. f TLC indicated complete consumption of 3 in
less than 5 h. The reported yield was a complex mixture of
g
products based upon 31P NMR. A slight decrease in yield was
observed after 48 h attributed to slow product degradation.
TMP-imidazolide (δ -7.51 in DMF-d7 and -6.75 in CD3-
CN).12 Excess CDI was quenched by the addition of
MeOH, and the solvent evacuated followed by the addi-
tion of 2. Based upon analytical HPLC, the coupling of
2 with TMP-imidazolide was complete in less than 24
h. Chromatography provided spectroscopically pure 4 as
either the disodium or dipotassium salt in >70% yield.
As an alternative, we also examined the coupling of
commercially available TMP-morpholidate with 2, in the
presence of 1H-tetrazole, using conditions previously
described for the synthesis of GDP-Fuc, GDP-Man, and
UDP-Gal5o and the reaction of TDP with 3 using
conditions previously described for the construction of
GDP-Fuc, UDP-Ara, and UDP-Gal.5n Surprisingly,
the CDI method provided reaction times and yields
comparable to or slightly better than the morpholidate
methodology (Table 1) while both the TMP-imidazolide
and -morpholidate coupling reactions were superior in
providing 4 to the coupling of TDP with 3.
Con clu sion s
Previous studies revealed a large effect upon the
addition of an acidic catalyst to morpholidate-catalyzed
coupling reactions.5o In particular, the extent of stimula-
tion of morpholidate-catalyzed coupling upon the addition
of various additives followed the order 1H-tetrazole (pKa
) 4.9) g 4-(dimethylamino)pyridine hydrochloride (DMAP‚
HCl, pKa ) 6.1) > perchloric acid > N-hydroxysuccinim-
ide (pKa ) 6.1) = acetic acid (pKa ) 4.75) > 1,2,4-triazole
(pKa ) 10.0), suggesting that 1H-tetrazole activates
NMP-morpholidate by a combination of protonation of
the leaving group nitrogen (pKa = 8.33)13 and nucleophilic
catalysis via the formation of a highly reactive phospho-
tetrazolide. In contrast, we found that permutations of
the TDP-imidazolide (pKa = 6.95)13 reaction conditions
illustrate this reaction to be much less dependent upon
additives or solvent. For example, yields similar to those
reported in Table 1 were observed with ethanol, pyridine,
DMF, or DMSO as the reaction solvent, while a decrease
of approximately 60% was observed in methanol. Fur-
thermore, the addition of 3 equiv of 1H-tetrazole to the
In conclusion, this work provides the first direct
comparison of the “state of the art” nucleotide sugar
synthetic methodologies5n,o and reveals the single-flask
activation of TMP and subsequent coupling of TMP-
imidazolide with rhamnosyl-1-phosphate as the best
method for the production of 4. These results also
suggest CDI-catalyzed nucleotide sugar synthesis may
prove advantageous in the production of other nucleotide
sugars since the choice of solvent is flexible and the
method is not limited by the commercial availability of
expensive NMP-morpholidates. In addition, by provid-
ing significant quantities of 4, this work sets the stage
for our future studies of various glycosyltransferases
required for bacterial O antigen biosynthesis as well as
the unique glycosyltransferases responsible for tailoring
steps in calicheamicin biosynthesis.17 Work is continuing
in these exciting areas.
Exp er im en ta l Section
Gen er a l Meth od s. Infrared spectra were recorded on a
Perkin-Elmer 1600 series FTIR spectrophotometer. 1H NMR
spectra were obtained on a Bruker AMX 400 (400 MHz) and
(13) Weast, R. C., Ed. In CRC Handbook of Chemistry and Physics;
CRC Press: Cleveland, OH, 1984; pp D163-D164.
(14) (a) Liu, D.; Haase, A. M.; Lindqvist, L.; Lindberg, A. A.; Reeves,
P. R. J . Bacteriol. 1993, 175, 3408-3413. (b) Wang, L.; Romana, L.
R.; Reeves, P. R. Genetics 1992, 130, 429-443. (c) Raetz, C. R. H. Annu.
Rev. Biochem. 1990, 59, 129-170. (d) J iang, X.-M.; Neal, B.; Santiago,
F.; Lee, S. J .; Romana, L. K.; Reeves, P. R. Mol. Microbiol. 1991, 5,
695-713. (e) Schnaitman, C. A.; Kiens, J . P. Microbiol. Rev. 1993, 57,
655-682.
(15) Wang, L.; Reeves, P. R. J . Bacteriol. 1994, 176, 4348-4356.
(16) Osborn, M. J .; Cynkin, M. A.; Gilbert, J . M.; Muller, L.; Singh,
M. Methods Enzymol. 1972, 28, 583-601.
(17) Thorson, J . S.; Shen, B.; Whitwam, R. E.; Liu, W.; Li, Y. Bioorg.
Chem., manuscript submitted.

7570 J . Org. Chem., Vol. 63, No. 21, 1998
Notes
Sch em e 1. Syn th etic Rou tes to 4: (a ) CDI, CH3CN; 2, DMF ; (b) TMP -Mor p h olid a te, 1H-Tetr a zole,
P yr id in e; (c) (i) TDP (NBu 4), CH2Cl2; (ii) H2, P d C
Sch em e 2. Sch em a tic Rep r esen ta tion of th e En zym e-ca ta lyzed Biosyn th esis of th e Sa lm on ella Gr ou p E1
O An tigen P r ecu r sor 11: (a ) Ga l-1-P T/UDP -Ga l; (b) Rh a Tr3/4; (c) Ma n Tâ4/GDP -Ma n a
a
Enzymatic step that requires 4 is highlighted.
are reported in parts per million (δ) relative to either tetra-
methylsilane (0.00 ppm) or CDCl3 (7.25 ppm) for spectra run in
CDCl3, D2O (4.82 ppm), or CD3OD (3.35 ppm). Coupling
constants (J ) are reported in hertz. 13C NMR are reported in δ
relative to CDCl3 (77.00 ppm) or CD3OD (49.05 ppm) as an
internal reference, and 31P NMR spectra are reported in δ
relative to H3PO4 (0.00 ppm in D2O). Mass spectra were
recorded on a PE SCIEX API 100 LC/MS mass spectrometer.
Optical rotations were recorded on a J asco DIP-370 polarimeter
using a 1.0 or 0.5 dm cell at room temperature (25 °C) and the
reported concentrations. Melting points were measured with
an Electrothermal 1A-9100 digital melting point instrument.
Chemicals used were reagent grade and used as supplied
except where noted. Analytical TLC was performed on either
E. Merck silica gel 60 F254 plates (0.25 mm) or Whatman AL Sil
G/UV silica gel 60 plates. Compounds were visualized by
spraying I2/KI/H2SO4 or by dipping the plates in a cerium
sulfate-ammonium molybdate solution followed by heating.
Liquid column chromatography was performed using forced flow
of the indicated solvent on E. Merck silica gel 60 (40-63 µm),
and high-pressure liquid chromatography was performed on a
RAININ Dynamax SD-200 controlled with Dynamax HPLC
software.
Diben zylp h osp h or yl-2,3,4-tr i-O-a cetyl-â-L-r h a m n op yr a -
n osid e. Activated 4 Å molecular sieves (powder, 5 g) were
added to a solution of dibenzyl phosphate (4.0 g, 14.38 mmol),
silver triflate (3.28 g, 12.7 mmol), and 2,4,6-collidine (2.0 g, 16.5
mmol) in dry CH2Cl2 (15 mL), and the mixture was stirred at

Notes
J . Org. Chem., Vol. 63, No. 21, 1998 7571
room temperature for 1.0 h under argon atmosphere in the
absence of light. The temperature was decreased to -40 °C, and
2,3,4-tri-O-acetyl-R-L-rhamnopyranosyl bromide (4.5 g, 12.7
mmol)10 in 10 mL of dry CH2Cl2 was added in dropwise fashion.
The reaction mixture was stirred at -40 °C for 4 h under argon
atmosphere and the absence of light and then allowed to warm
to room temperature. After approximately 12 h, a few drops of
methanol were added to quench unreacted bromide, the mixture
was stirred for 1 h and partitioned between CHCl3 (40 mL) and
H2O (40 mL), and the organics were dried over MgSO4, filtered,
concentrated, and purified by silica gel chromatography (33%f
40% EtOAc/hexanes) to give 5.33 g (76%) of the desired
between 5 mL of 100 mM NH4HCO3/CHCl3 (1:4), and 20 µL of
the corresponding aqueous layer was utilized directly for chro-
matographic analysis. Reactants and products (TMP, 6.5 min;
TDP-Rha, 15.0 min; TDP, 22.5 min) were separated by Sphere-
clone SAX analytical chromatography (0.46 × 250 cm, 1.5 mL
min-1, λ ) 268 nm) with a mobile phase of H2O (A) and 200
mM KH2PO4, pH 5.5 (B); 20% A (3 min), 20%f40% B (20 min),
40%f100% B (3 min), 100% B (5 min).
Upon completion, the reaction was diluted with 100 mM NH4-
HCO3 and extracted with CHCl3, the aqueous layer concen-
trated, and the desired product isolated by Sphereclone SAX
semipreparative chromatography (1.0 × 250 cm, 7.0 mL min-1
,
product: Rf ) 0.33 (40% EtOAc/hexanes); [R]25 ) -46° (c 1.0,
λ ) 268 nm) with a mobile phase of H2O (A) and 200 mM KH2-
PO4, pH 5.5 (B); 20% A (5 min), 20%f40% B (15 min),
40%f100% B (10 min), 100% B (3 min). The product-containing
fractions were collected (TMP, 14.2 min; TDP-Rha, 23.6 min;
TDP, 30.4 min), concentrated, and desalted upon Sephadex G-10
resin (1.5 × 120 cm) to give the dipotassium salt of the title
product as a white solid after lyophilization (mp 147-150 °C).
Alternatively, the concentrated reaction mixture was loaded
directly upon a Sephadex G-10 column (1.5 × 170 cm), and the
product was eluted with water, concentrated, and subsequently
crystallized from 1.0 M NaClO4 acetone solution to provide the
disodium salt of TDP-Rha (mp 158-161 °C): Rf ) 0.80 (25%
1.0 M NH4OAc/2-propanol); [[R]25D ) -32° (c 0.5, H2O); IR (thin
film) 3423 (s), 2940, 1701 (s), 1477, 1373, 1249, 1048, 1052, 925;
1H NMR (D2O) δ 7.68 (s, 1H), 6.31 (t, J ) 6.9 + 7.0 Hz, 1H),
5.37 (d, J ) 7.2 Hz, 1H), 4.54 (m, 1H), 4.12 (m, 3H), 3.97 (s,
1H), 3.88-3.78 (m, 2H), 3.39 (t, J ) 9.7 + 9.7 Hz, 1H), 2.32 (t,
J ) 5.0 + 6.1 Hz, 1H), 1.87 (m, 3H), 1.22 (d, J ) 6.6 Hz, 3H);
13C NMR (D2O) δ 166.56, 151.71, 137.28, 111.71, 95.52, 85.24,
84.94, 72.80, 72.72, 71.59, 70.95, 69.11, 65.44, 60.23, 38.52, 11.61;
31P NMR (D2O) δ -10.57, -12.11; MS 663.0 (M + K), 647.1 (M
+ Na).
D
CHCl3); IR (thin film) 1749 (s), 1372, 1219, 1161, 1015, 957; 1H
NMR (CDCl3) δ 7.29 (m, 10H), 5.48 (d, J ) 6.1 Hz, 1H), 5.21 (m,
1H), 5.16 (s, 1H), 5.07-4.98 (m, 5H), 3.88 (ddd, J ) 6.2, 12.4,
15.1 Hz, 1H), 2.06 (s, 3H), 1.97 (s, 3H), 1.91 (s, 3H), 1.04 (d, J )
6.2 Hz, 3H); 13C NMR δ 169.74, 169.51, 128.58, 128.06, 127.89,
95.18, 70.15, 69.74, 68.92, 68.22, 20.59, 17.04; 31P NMR δ -4.78;
MS 573.0 (M + Na).
Tr ieth yla m m on iu m â-L-r h a m n op yr a n osylp h osp h a te (2).
Dibenzylphosphoryl-2,3,4-tri-O-acetyl-â-L-rhamnopyranoside (2.0
g, 3.63 mmol) was hydrogenated over 10% Pd/C (600 mg) in CH3-
OH (25 mL) and 1.0 M NaHCO3 (10 mL) under 50 psi hydrogen
atmosphere at room temperature until the reaction was deter-
mined complete by TLC (7 h). The catalyst was removed by
filtering the reaction through Celite, and the filtrate was
concentrated and partitioned between CH2Cl2 (25 mL) and H2O
(25 mL). The product-containing aqueous layer was isolated and
concentrated to 10 mL. This solution was subsequently cooled
to 0 °C, and 1.0 M NaOH (15 mL) was added in dropwise fashion
while keeping the temperature < 5 °C. After the NaOH had
been added, the reaction was stirred and allowed to warm to
room temperature (3 h). The pH of the mixture was carefully
adjusted to pH 7.5 by the addition of cold 1.0 M HOAc, and the
mixture was diluted to 250 mL with H2O and submitted to anion
exchange chromatography (Dowex 1 × 8, HCO3-, 3 × 15 cm).
The product was eluted with H2O (250 mL), 0.1 M NH4HCO3
(250 mL), 0.2 M NH4HCO3 (250 mL), and 0.3 M NH4HCO3 (250
mL), during which the desired compound eluted between 0.2 and
0.3 M NH4HCO3. The product-containing fractions were pooled,
concentrated, and coevaporated several times with water and
ethanol to remove remaining NH4HCO3 to give 0.81 g (85%) of
Meth od s II a n d III. The coupling of commercially available
TMP-morpholidate with 2 was accomplished using conditions
previously described5o for the synthesis of GDP-Fuc, GDP-Man,
and UDP-Gal, and the reaction of TDP with 310 was ac-
complished using conditions previously described5n for the
construction of GDP-Fuc, UDP-Ara, and UDP-Gal. In each
case, the analysis of the reaction progress and final purification
was accomplished as described for method I.
P r ep a r a tion of th e Ga l-1-P T a n d Rh a Tr3 Cr u d e Ex-
tr a cts. PCR amplification of wbaP (Gal-1-PT) and wbaN
(RhaTR3) was accomplished from genomic template DNA isolated
from Salmonella (group LT2) using flanking primers based upon
the previously identified sequence.14a-d The amplified genes
were cloned into T7-driven pET11a-BL21 systems and grown
to an OD600 ) 0.55 at 37 °C with shaking (250 rpm). The growth
temperature was subsequently reduced to 20 °C, protein expres-
ammonium â-L-rhamnopyranosylphosphate: Rf ) 0.43 (33%
1
H2O/CH3CN); [R]25 ) -19° (c 2.0, H2O); H NMR (D2O) δ 5.22
D
(d, J ) 8.2 Hz, 1H), 3.92 (s, 1H), 3.87 (m, 2H), 3.37 (t, J ) 9.7 +
9.7 Hz, 1H), 1.23 (d, J ) 6.1 Hz, 3H);13C NMR (D2O) δ 95.22,
73.22, 71.02, 70.70, 69.89, 69.02, 17.07; 31P NMR (D2O) δ -1.26.
A 0.80 g (3.06 mmol) sample of this product was subsequently
dissolved in H2O (15 mL), applied to an AG 50W-X8 cation-
exchange column (Et3NH+, 2.5 × 15 cm), and eluted with H2O
(150 mL). The eluate was evaporated and coevaporated with
methanol (3 × 25 mL) to give the triethylammonium salt of â-L-
rhamnopyranosylphosphate (1.01 g, 90%): [R]25D ) -30° (c. 2.0,
H2O); 1H NMR (D2O) δ 5.29 (dd, J ) 1.7, 7.6 Hz, 1H), 3.63 (t, J
) 3.0 + 2.1 Hz, 1H), 3.87-3.80 (m, 2H), 3.39 (t, J ) 9.7 + 9.7
Hz, 1H), 3.16 (m, 10.0H, 1.6 equiv of triethylamine), 1.23 (m,
17H);13C NMR (D2O) δ 96.06, 72.34, 70.90, 69.99, 69.70, 46.93,
17.08, 8.54; 31P NMR (D2O) δ -1.36; MS 392.1 (M + Na2).
sion was induced with 1.0 mM IPTG, rifampicin (75 µg mL-1
)
was added 2 h after induction, and the cultures were allowed to
grow for an additional 12-14 h. Cells were harvested (2000g,
15 min), resuspended in 50 mM Tris-HCl and 1 mM EDTA, pH
8.5, and lysed by sonication. The cellular debris was removed
by centrifugation (2000g, 15 min), and the membrane fractions
were isolated from the supernatant by centrifugation (30000g,
1 h), resuspended in 50 mM Tris-HCl and 1 mM EDTA, pH 8.5,
and stored at -80 °C until used. SDS-PAGE revealed the
desired glycosyltransferases as approximately 70-80% of the
total membrane associated protein.
Th ym id in e
5′-d ip h osp h o-â-L-r h a m n op yr a n ose
(4).
Meth od I. The TMP sodium salt was exchanged to triethylam-
monium salt by passage through a column of Dowex-1 × 8 (Et3-
NH+) in water, and eluent containing 135 mg (419 µmol) of 1
was dried and dissolved in either DMF-d7 or CD3CN. Three
equivalents of 1,1′-carbonyldiimidazole was added, and the
progress of the reaction monitored by 31P NMR. Upon comple-
tion (80-90 min), several drops of dry methanol were added to
quench the excess CDI. The reaction mixture was concentrated,
coevaporated with dry ethanol and toluene, and dried in vacuo,
and the crude solid residue was used directly for the coupling
reaction without further purification. To accomplish the cou-
pling reaction, a mixture of 2 (118 mg, 342 µmol) and the
thymidine 5′-phosphoimidazolide was coevaporated three times
with the dry reaction solvent before the solution (concentration
of the reactants kept at 0.2-0.3 M) was stirred at room
temperature under argon atmosphere. To monitor reactions at
various time points, 10 µL aliquots was removed and partitioned
Assa y for th e P r od u ction of 12. A mixture containing 6
(6 nmol, Sigma), 25 µL of Gal-1-PT crude membrane extracts,
25 µL of RhaTR3 crude membrane extracts, 10 mM UDP-Gal,
10 mM 4, 50 mM Tris-acetate (pH 8.5), 10 mM MgCl2, 1 mM
EDTA, 25 µL of ManTâ4 crude membrane extracts,10 and 0.33
µM GDP-[U-14C]Man (15 000 cpm) in a final volume of 250 µL
was allowed to react at 37 °C for 2 h and subsequently stopped
by combining with 4 mL of CHCl3/CH3OH (2:1). The organic
phase was washed three times with 800 µL of pure solvent upper
phase (CHCl3/130 mM NaCl solution/CH3OH, 3%:48%:49%),
dried by evaporation, and dissolved in scintillant, and the extent
of radioactivity recovered was determined by liquid scintillation
counting.
Ack n ow led gm en t. We gratefully acknowledge Pro-
fessor Hung-wen Liu (Department of Chemistry, Uni-

7572 J . Org. Chem., Vol. 63, No. 21, 1998
Notes
Su p p or tin g In for m a tion Ava ila ble: 1H, 13C, and 31P
NMR spectra for 2 and 4 (6 pages). This material is contained
in libraries on microfiche, immediately follows this article in
the microfilm version of the journal, and can be ordered from
the ACS; see any current masthead page for ordering
information.
versity of Minnesota) for the generous gift of Salmonella
genomic DNA and Dr. George Sukenick (NMR Core
Facility, Sloan-Kettering Institute) for NMR and mass
spectral analyses. J .S.T. is a Rita Allen Foundation
Scholar. This work was supported in part by the Cancer
Center Support Grant (CA-08748) and a grant from the
Special Projects Committee of The Society of Memorial
Sloan-Kettering Cancer Center.
J O981265N