Simultaneous removal of benzyl and benzyloxycarbonyl protective groups in 5′-O-(2-deoxy-α-D-glucopyranosyl)uridine by catalytic transfer hydrogenolysis

Article abstract of DOI:10.1080/15257770802458303

Synthesis of N3,2′,3′-O-tris-(benzyloxycarbonyl)uridine and its use in the synthesis of 5′-O-(2-deoxy-α-d-glucopyranosyl)uridine is described. Simultaneous removal of benzyl and benzyloxycarbonyl groups was accomplished by catalytic transfer hydrogenolysis in the presence of Pearlman′s catalyst without competing side reactions. Copyright Taylor & Francis Group, LLC.

Full text of DOI:10.1080/15257770802458303

Nucleosides, Nucleotides and Nucleic Acids, 27:1250–1256, 2008
Copyright ꢀC Taylor & Francis Group, LLC
ISSN: 1525-7770 print / 1532-2335 online
DOI: 10.1080/15257770802458303
SIMULTANEOUS REMOVAL OF BENZYL
AND BENZYLOXYCARBONYL PROTECTIVE GROUPS
IN 5ꢀ-O-(2-DEOXY-α-D-GLUCOPYRANOSYL)URIDINE BY CATALYTIC
TRANSFER HYDROGENOLYSIS
Ilona Wandzik, Tadeusz Bieg, and Monika Kadela
Silesian University of Technology, Faculty of Chemistry, Department of Organic Chemistry,
Bioorganic Chemistry and Biotechnology, Gliwice, Poland
Synthesis of N3,2ꢁ,3ꢁ-O-tris-(benzyloxycarbonyl)uridine and its use in the synthesis of 5ꢁ-O-(2-
2
deoxy-α-d-glucopyranosyl)uridine is described. Simultaneous removal of benzyl and benzyloxycar-
bonyl groups was accomplished by catalytic transfer hydrogenolysis in the presence of Pearlman’s
catalyst without competing side reactions.
Keywords Benzyl; benzyloxycarbonyl protective groups; hydrogenolysis; uridine; 2-
deoxy-D-glucopyranose
INTRODUCTION
Chemical synthesis of biologically important complex carbohydrates
usually requires using protective groups for the protection and differenti-
ation of various O-H and N-H groups. Among them, benzyl-type protective
groups are of great interest due to their stability to a variety of reagents and
easy removal by catalytic hydrogenolysis.
Recently, we reported on the synthesis of 2-deoxyhexopyranosyl deriva-
tives of uridine as donor substrate analogues of glycosyltransferases.[1] As
an example compound 1 which possesses three type of protective groups:
benzyl, benzoyl and acetonide is shown in Figure 1.
We originally attempted to prepare totally deprotected title compound
3 in order to ensure better interaction with the active site of glycosyl-
transferase, similar to that of natural substrate; therefore, three deblocking
Received 18 July 2008; accepted 4 September 2008.
Address correspondence to Ilona Wandzik, Silesian University of Technology, Faculty of Chemistry,
Department of Organic Chemistry, Bioorganic Chemistry and Biotechnology, Krzywoustego 4, 44-100
Gliwice, Poland. E-mail: ilona.wandzik@polsl.pl
1250

Benzyl and Benzyloxycarbonyl Protective Groups
1251
FIGURE 1 2-Deoxy-α-D-glucopyranosyl derivatires of uridine.
procedures should be undertaken. Saponification of acyl ester followed
by catalytic transfer hydrogenolysis led to compound 2. There was a
concern that acetonide derivative will not fit into the active site of the
enzyme owing to the steric hindrances created by the isopropylidene group.
Unfortunately removal of acetonide group performed in the presence of
trifluoroacetic or formic acid occurred in low yield due to the poor stability
of 2-deoxyglycosidic linkage in acid medium. We considered, therefore,
another way of protecting the uridine moiety. Our goal was to apply a
protective group that could be removed in neutral conditions, preferably
by catalytic hydrogenolysis, together with O-benzyl protection in 2-deoxy-D-
glucopyranose part.
RESULTS AND DISCUSSION
Successful synthesis of compound 1 was realized in two-step regio-
and stereoselective addition of uridine to D-glucal using Falck-Mioskowski
protocol.[2] Our previous experiments revealed beneficial influence of
using a large group as N-imide protection on α-selectivity in the addition
of uridine derivatives to D-glucal.[1] Therefore, there was a need to protect
not only two secondary hydroxyl groups of uridine, but also the N -imide
function of the aglycone. First, we tried out benzyl protection of N -imide
nitrogen hoping that in the final step it would be cleaved simultaneously to
benzyl protection of the hydroxyl groups in 2-deoxy-D-glucopyranose ring.
Unfortunately, we found that treatment of compound 6 with cyclohexene
and Pearlman’s catalyst (20% Pd(OH)2/C)[5] in MeOH under reflux for 3
hours resulted in removing the O-benzyl groups only (Scheme 1).
Several reports discussed a similar problem in imide N -debenzylation
by catalytic hydrogenolysis. Johnson and Widlanski[4] observed that the
deprotection of N -imide in per-benzyluridine did not occur in the presence
of hydrogen on Pearlman’s catalyst.[5] Only O-benzyl ethers were deblocked.

1252
I. Wandzik et al.
SCHEME 1 Reagents and conditions: (a) TPHB (0.1 equiv.), CH2Cl2, room temperature, 3 hours, (b)
20% Pd(OH)2,/C, cyclohexene, MeOH, reflux, 3 hours, (c) 20% Pd(OH)2/C, cyclohexene, MeOH,
reflux, 15 minutes, (d) 20% Pd(OH)2/C, 1,4-cyklohexadiene, MeOH, room temperature, 20 hours.
Mandal and coworkers revealed similar problem in thymidine derivatives
when hydrogenolysis was carried out on 10% Pd/C in the presence of
hydrogen.[6] On the other hand, Reitz and Pfleiderer[7] noted that reduc-
tion of the pyrimidine 4,5 double bond during hydrogenolytic cleavage of
O-benzyl-substituted diuridylphosphates occurred on palladium black in the
presence of hydrogen.
We revealed that hydrogenolysis of N 3-benzyl-2,3-O-isopropylideneuri-
dine (5) by hydrogen transfer from cyclohexene or 1,4-cyclohexadiene in
the presence of Pearlman’s catalyst or 10% Pd/C did not lead to the desired
2,3-O-isopropylideneuridine (9). Instead of this, a preferential formation of
N 3-benzyluridine (10) was observed (Scheme 2).
Moreover, 2,3-O-isopropylideneuridine (9) underwent deprotection to
free uridine under the same conditions. This result shows that presence
of N 3-benzyl group is not necessary for the cleavage of acetonide. The
most striking observation was that in the presence of HCO2NH4 as a
hydrogen donor,[8] the catalytic transfer hydrogenolysis of N3-benzyl-2,3-O-
isopropylideneuridine (5) proceeds with reduction of the 4,5 double bond
of nucleobase with unaffected isopropylidene block. Compounds 10[4] and
11[9] were identified by comparison with literature data.
The aforementioned experiments showed that N3-benzyl-2,3-O-
isopropylideneuridine (5) is not suitable substrate for the synthesis of

Benzyl and Benzyloxycarbonyl Protective Groups
1253
SCHEME 2 Reagents and conditions: (a) 20% Pd(OH)2/C, cyclohexene, MeOH, reflux, 3 hours; (b)
10% Pd/C, HCO2NH4, MeOH, reflux, 2 hours then next portion of HCO2NH4, 2 hours.
the desired title compound 3. We considered, therefore, other protecting
groups for the imide function and the secondary hydroxyls of uridine. In
this report, we turn our attention on benzyloxycarbonyl (Cbz) group. An
N -Cbz group is one of the most common amino protective group and can
be useful in protection of hydroxyl group as well. There is no report on
protection of N -imide nitrogen in uridine by N -Cbz; instead, protection
of two secondary hydroxyl groups for 5-O-tert-butyldimethylsilyluridine is
reported.[4] Reaction is carried out using benzylchloroformate (CbzCl) and
4-(N,N-dimethylamino)pyridine (DMAP) in CH2Cl2 at room temperature
for 3 days. We attempted to carry out the synthesis of totally tris-Cbz
protected uridine since we believed that such derivative would be
more relevant for our subsequent studies. We found that application of
N,N-diisopropylethylamine (DIPEA), instead of DMAP in refluxing CH2Cl2,
led to the desired tris-Cbz derivative 14 as a major product associated with
minor di-Cbz derivative 15 (Scheme 3).[10]
To test catalytic transfer hydrogenolysis of Cbz-derivative 14, Pearlman
catalyst and cyclohexene were applied. Deprotection of 14 to free uridine
12 occurred in 20 hours at room temperature. Felix and coworkers
observed beneficial influence of 1,4-cyclohexadiene over cyclohexene in
deprotection of peptides.[11] We applied 1,4-cyclohexadiene as a source of
hydrogen in our reaction and hydrogenolysis of 14 required 2 hours for
complete deprotection (Table 1).
Uridine derivative 16 enabled addition reaction to 3,4,6-tri-O-benzyl-
D-glucal (4) to be performed at 5-OH according to the described

1254
I. Wandzik et al.
SCHEME 3 Reagents and conditions: (a) TrCl, pyridine, room temperature, 4 days, (b) CbzCl,
DIPEA, CH2Cl2, room temperature, 20 hours, (c) 40% HBF4, MeCN, room temperature, 1 hour, (d)
20% Pd(OH)2/C, cyclohexene, MeOH, reflux, 15 minutes or room temperature, 20 hours, (e) 20%
Pd(OH)2/C, 1,4-cyklohexadiene, MeOH, room temperature, 2 hours.
1
procedure (Scheme 1).[1] With the aid of H NMR and 13C NMR we
have estimated that desirable α-product 8 was contaminated with small
amount of β-anomer (<5%). Subsequently the total deprotection of 8
without the reduction of the 5,6 double bond of uridine was carried out.
As anticipated, N -Cbz and O-Cbz protected nucleoside and O-Bn protected
2-deoxy-D-glucopyranose could be totally deblocked by transfer catalytic
hydrogenolysis on Pearlman’s catalyst affording title compound 3. It was
found that 1,4-cyclohexadiene as a hydrogen donor was more effective
than cyclohexene. 1,4-cyclohexadiene was applied at room temperature[12]
while cyclohexene under reflux in methanol (Table 1). It is noteworthy that
the yield of deprotection was significantly better in the case of the reaction
carried out at room temperature, since thermal decomposition of 8 may
TABLE 1 Reaction conditions for the hydrogenolysis on Pearlman’s catalyst
Entry
1
Substrate
Product
Hydrogen donor
Reaction time
Temperature
5
10
cyclohexene
3 hours
reflux
1,4-cyclohexadiene
HCO2NH4
cyclohexene
3 hours
reflux
reflux
reflux
a
2
3
4
5
6
8
11
7
3
4 hoursb
3 hours
cyclohexene
1,4-cyclohexadiene
15 minutes
20 hours
reflux
room
temerature
reflux
room
5
6
9
14
12
12
cyclohexene
cyclohexene
3 hours
20 hours
temerature
reflux
room
cyclohexene
1,4-cyclohexadiene
15 minutes
2 hours
temperature
aWhen 10% Pd/C instead of Pd(OH)2/C was applied better conversion was observed.
bReaction not completed.

Benzyl and Benzyloxycarbonyl Protective Groups
1255
occur at elevated temperatures (90% compared with 75% in reflux). All
new compounds 3, 7, and 14 were purified by column chromatography
1
and their structures were elucidated with the aid of H and 13C NMR
spectroscopy data and mass spectrometry analysis.[13]
CONCLUSIONS
Synthesized tris-Cbz uridine 14 can be valuable building block in the
synthesis of complex carbohydrates or oligoribonucleotides when imide
moiety of uridine must be protected during the course of synthesis. Its
application in the synthesis of 5ꢁ-O-(2-deoxy-α-D-glucopyranosyl)uridine (3)
was performed. Simultaneous removal of N -Cbz, O-Cbz, and O-Bn groups in
compound 8 was accomplished by catalytic transfer hydrogenolysis on Pearl-
man’s catalyst in the presence of 1,4-cyclohexadiene (room temperature) or
cyclohexene (reflux) affording title compound 3 without side reactions.
REFERENCES
1. Wandzik, I.; Bieg, T. Adducts of uridine and glycals as potential substrates for glycosyltransferases.
Bioorg. Chem. 2007, 35, 401–416.
2. Bolitt, V.; Mioskowski, Ch.; Lee, S.-G.; Falck, J.R. Direct preparation of 2-deoxy-D-glucopyranosides
from glycals without Ferrier rearrangement. J. Org. Chem. 1990, 55, 5812–5813.
3. Maguire, A.R.; Hladezuk, I.; Ford, A. New methods for the synthesis of N-benzoylated uridine
and thymidine derivatives; a convenient method for N-debenzoylation. Carbohydr. Res. 2002, 337,
369–372.
4. Johnson II, D.C.; Widlanski, T.S. Facile Deprotection of O-Cbz-Protected Nucleosides by Hy-
drogenolysis: An Alternative to O-Benzyl Ether-Protected Nucleosides. Org. Lett. 2004, 4643–4646.
5. Pearlman, W.M. Noble metal hydroxides on carbon nonpyrophoric dry catalysts. Tetrahedron Lett.
1967, 17, 1663–1664.
6. Bar, N.C.; Patra, R.; Achari, B.; Mandal, S.B. An Entry to Unusual Classes of Nucleoside Analogues.
Tetrahedron 1997, 53, 4727–4738.
7. Reitz, G.; Pfleiderer, W. Synthese und Eigenschvaften von Oꢁ-benzyl-substituierten Diuridylphos-
phaten. Chem. Ber. 1975, 108, 2878–2894.
8. Bieg, T.; Szeja, W. Regioselective hydrogenolysis of benzyl glycosides. Carbohydr. Res. 1990, 205,
c10–c11.
ˇ
9. Skaric´, V.; Gaˇspert, B.; Hohnjec, M. Thio-analogues of 5,6-Dihydrouridine. J. Chem. Soc. (C). 1970,
2444–2447.
10. Procedure for the carbamoylation of uridine (Scheme 3): Trityl chloride (1.37 g, 4.91 mmol) was
added to a solution of uridine (1.00 g, 4.09 mmol) in pyridine (5 mL) and reaction mixture was
kept at room temperature. After 4 days, the reaction mixture was diluted with CH2Cl2 (300 mL),
washed with water (2 × 100 mL), the organic layer was separated, dried (MgSO4), concentrated
in vacuo and co-evaporated with toluene (3 × 100 mL). Crude 5ꢁ-O-trityluridine (13) was dissolved
in CH2Cl2 (15 mL), DIPEA (4.05 ml, 24.54 mmol) was added and solution was cooled to 10◦C.
A CbzCl (3.50 mL, 24.54 mmol) was added dropwise and the mixture was refluxed for 0.5 hour.
The mixture was diluted with CH2Cl2 (100 mł) and washed with 1 M HCl and then water. The
organic phase was dried (MgSO4) and concentrated. The residue was dissolved in acetonitrile
(10 mL) and 48% HBF4 (0.65 mL, 4.09 mmol) was added. The reaction mixture was kept at room
temperature for 1 hour, neutralised with NaHCO3, dried (MgSO4), concentrated and purified on a
column of silica gel 60 (70–230 mesh) developed with the hexane/AcOEt 7:1 → 1:1 (v/v) solvent
system to yield 14 (1.41 g, 58%) as a first eluting fraction (RF toluene/AcOEt 1:1: 0.57) and 15 (0.46

1256
I. Wandzik et al.
g, 22%) as the second fraction (RF: 0.17). Analytical data for 15 are in accordance with Johnson and
Widlanski.[4]
11. Felix, A.M.; Heimer, E.P.; Lambros, T.J.; Tzougraki, C.; Meienhofer, J. Rapid Removal of Protecting
Groups from Peptides by Catalytic Transfer Hydrogenation with 1,4-Cyclohexadiene. J. Org. Chem.
1978, 43, 4194–4196.
12. Procedure for final O-Cbz, N-Cbz and O-Bn deprotection (Scheme 1, procedure d): To a solution of
8 (109 mg, 0.10 mmol) in MeOH (10 mL) 1,4-cyclohexadiene (0.6 mL) and Pearlman’s catalyst
(75 mg) were added. Reaction mixture was stirred at room temperature overnight. After re-
moval of the catalyst by filtration crude product were purified by column chromatography with
CHCl3/MeOH 10:1 → 5:1 (v/v) solvent system to yield 3 (35 mg, 90%) as a white solid. Compound
3 was prepared using protected uridine 14 and glucal 4 according to previously described
procedure.[1]
13. Analytical data for new compounds: 5ꢁ-O-(2-deoxy-a-D-glucopyranosyl)uridine, 3: solid; [α]D20+36.0
(MeOH, c 1.0); 1H NMR (300 MHz, CD3OD): δ 1.70 (ddd, 1H, J 3.5, 11.8, 13.2 Hz, H-2ꢁꢁax), 2.09
(ddd, 1H, J 1.1, 5.0, 13.2 Hz, H-2ꢁꢁeq), 3.26 (t, 1H, J 9.0 Hz, H-4ꢁꢁ), 3.53 (ddd, 1H, J 2.3, 5.6, 9.0 Hz,
H-5ꢁ), 3.67–3.74 (m, 2H, H-6ꢁꢁa,b), 3.79 (ddd, J 5.0, 9.0, 11.8 Hz, H-3ꢁꢁ), 3.83 (dd, 1H, J 2.3, 11.5
Hz, H-5ꢁa), 3.94 (dd, 1H, J 2.6, 11.5 Hz, H-5ꢁb), 4.12–4.19 (m, 3H, H-2ꢁ, 3ꢁ, 4ꢁ), 4.99 (br d, 1H, J
3.2 Hz, H-1ꢁꢁ), 5.80 (d, J 8.2 Hz, H-5), 5.88 (d, 1H, J 4.1 Hz, H-1ꢁ), 7.98 (d, 1H, J 8.2 Hz, H-6); 13
C
NMR: δ 38.86 (C-2ꢁꢁ), 62.87, 67.11, 69.99, 71.32, 73.21, 74.53, 76.12, 84.46 (C-2ꢁ, C-3ꢁ, C-4ꢁ, C-5ꢁ, C-3ꢁꢁ,
C-4ꢁꢁ, C-5ꢁꢁ, C-6ꢁꢁ), 90.91 (C-1ꢁ), 98.84 (C-1ꢁꢁ), 102.60 (C-5), 142.03 (C-6), 152.37 (C-2), 166.19 (C-4);
ESI-HRMS: Calcd for C15H22N2O10Na ([M+Na]+): m/z 413.1167, found: m/z 413.1150. N 3-benzyl-
2ꢁ,3ꢁ-O-isopropylidene-5ꢁ-(2-deoxy-α-D-glucopyranosyl)uridine, 7: solid; [α]2D0+27.3 (MeOH, c 0.5);
1H NMR (CDCl3): δ 1.35; 1.54 (2s; 6H; C(CH3)2); 1.61 (ddd, 1H, J 3.7, 12.0, 13.4 Hz; H-2ꢁꢁax),
1.98 (ddd, 1H, J ≈ 0, 5.2, 13.4 Hz, H-2ꢁꢁeq), 3.22 (t, 1H, J 9.0 Hz, H-4ꢁꢁ), 3.44–3.91 (m, 6H, H-5ꢁa,b,
H-3ꢁꢁ, H-5ꢁꢁ, H-6ꢁꢁa,b), 4.43 (m, 1H, H-4ꢁ), 4.82–4.86 (m, 2H, H-2ꢁ, H-3ꢁ), 4.91 (d, 1H, J 2.9 Hz, H-1ꢁꢁ),
5.02–5.14 (qAB, 2H, N-CH2Ph), 5.83 (d, 1H, J 8.1 Hz, H-5), 5.86 (d, 1H, J 1.7 Hz, H-1ꢁ), 7.19–7.38
(m, 5H, Ph), 7.79 (d, 1H, J 8.1 Hz, H-6); 13C NMR: d 25.55, 27.57 (C(CH3)2), 38.63 (C-2ꢁꢁ), 45.14
(N-CH2Ph), 62.79 (C−5ꢁ), 68.22, 69.84, 73.13, 74.58 (C-3ꢁꢁ, C-4ꢁꢁ, C-5ꢁꢁ, C-6ꢁꢁ), 82.59, 86.78, 87.29
(C-2ꢁ, C-3ꢁ, C-4ꢁ), 95.47 (C-1ꢁ), 98.87 (C-1ꢁꢁ), 101.77 (C-5), 114.94 (C(CH3)2), 128.55, 129.27, 129.49
(Ph), 138.14 (Phq), 141,50 (C-6), 152.29 (C-2), 164.91 (C-4), ESI-HRMS: Calcd for C25H32N2O10Na
([M+Na]+): m/z 543.1949, found: m/z 543.1968. N 3,2ꢁ,3ꢁ-O-tris-(benzyloxycarbonyl)uridine, 14: oil;
[α]2D0-22.3 (CHCl3, c 3.0); 1H NMR: δ 2.78 (dd, 1H, J 3.6, 5.9 Hz, 5-OH), 3.78 (ddd, 1H, J 1.8,
5.9, 12.1, Hz, H-5ꢁa), 3,89 (m, 1H, H-5ꢁb), 4.25 (m, 1H, H-4ꢁ), 5.04–5.42 (m, 6H, 3CH2Ph), 5.39
(dd, 1H, J 3.5, 5.7 Hz, H-2ꢁ), 5.48 (dd, 1H, J 5.7, 5.7 Hz, H-3ꢁ), 5.76 (d, 1H, J 8.2 Hz, H-5), 5.95
(d, 1H, J 5.7 Hz, H-1ꢁ), 7.25–7.44 (m, 15H, Ph), 7.64 (d, J 8.2 Hz 1H, H-6), 13C NMR: d 61.85
(C-5ꢁ), 70.69, 70.87 (2PhCH2OC(O)O), 71.78 (PhCH2OC(O)N), 74.24, 75.92 (C-2ꢁ, C-3ꢁ), 83.34
(C-4ꢁ), 89.02 (C-1ꢁ), 102.88 (C-5), 128.65–129.20 (Ph), 133.84, 134.70, 134.78 (Phq), 140.68 (C-6),
148.56 (N-C(O)O), 149.89 (C-2), 154.03, 154.36 (2O-C(O)O), 160.30 (C-4); ESI-HRMS: Calcd for
C
33H30N2O12Na ([M+Na]+): m/z 669.1691, found: m/z 669.1687.