Stereoselective synthesis of nicotinamide β-riboside and nucleoside analogs

Article abstract of DOI:10.1016/j.bmcl.2004.06.093

A stereoselective synthesis of β-anomers of nicotinamide riboside (NAR), its deamidated analog (NaR), and of a nicotinamide C-methylated riboside derivative (NA3′-C-MeR) was developed. The β-anomers of N-ribofuranosylnicotine-3-carboxamide (β-NAR) and its nicotinic acid analog (β-NaR) were obtained by stereoselective synthesis via glycosylation of the presilylated bases under Vorbruggen's protocol. A NAR analog, methylated in position 3 of the ribosylic moiety, is also reported.

Full text of DOI:10.1016/j.bmcl.2004.06.093

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4655–4658
Stereoselective synthesis of nicotinamide b-riboside and
nucleoside analogs
Palmarisa Franchetti,* Michela Pasqualini, Riccardo Petrelli, Massimo Ricciutelli,
Patrizia Vita and Loredana Cappellacci
`
Dipartimento di Scienze Chimiche, Universita di Camerino, 62032 Camerino, Italy
Received 8 April 2004; revised 29 June 2004; accepted 30 June 2004
Available online 27 July 2004
Abstract—The b-anomers of N-ribofuranosylnicotine-3-carboxamide (b-NAR) and its nicotinic acid analog (b-NaR) were obtained
by stereoselective synthesis via glycosylation of the presilylated bases under Vorbruggenꢀs protocol. A NAR analog, methylated in
position 3 of the ribosylic moiety, is also reported.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
Kinetic and structural studies of both human and bacte-
rial NMNAT require the b-anomer of nicotinamide
monoribotide as a substrate and of nicotinamide ribo-
side as a product of the NAD degradation pathway.
The stereospecificity of NAD-mediated reactions is
determinant in all living organisms because the pyridine
ribotide moiety of NAD reacts only in the b-
configuration.
Nicotinamide riboside (b-NAR) is an intermediate in
one biosynthetic pathway by which nicotinamide is con-
verted into NAD. In fact, nicotinamide or nicotinamide
riboside derived from degradation of NAD can be re-
used by nicotinamide phosphoribosyl transferase or ribo-
sylnicotinamide kinase, respectively, to form the ribotide
NMN that is adenylated to NAD by nicotinamide
mononucleotide adenylyltransferase (NMNAT).1 Hu-
man NMNAT is an indispensable enzyme in both de
novo and salvage/recycling pathways for NAD biosyn-
thesis, catalyzing the conversion of NMN or its deami-
dated form NaMN into NAD or NaAD, respectively.
On the other hand, it is known that many bacterial
NMNATs strongly prefer the deamidated NaMN as a
substrate.2 NAD is a co-factor in numerous enzyme-cat-
alyzed redox reactions in all living organisms and plays
a fundamental role in cellular metabolic processes. It is
crucial thus that proper levels of NAD are regulated
and maintained for cellular survival.2 There is evidence
that hNMNAT is weakly expressed in tumor cells, deter-
mining thus interest in the enzyme as a target for anti-
cancer drugs. In this respect, some nicotinamide
riboside analogs have been described as antitumor
agents involving their metabolic conversion to NAD
analogs by NMNAT activity.3
In order to develop the study of ribonucleoside and
nucleotide analogs as chemotherapeutic agents, we
needed to synthesize pyridinium ribosidic derivatives.
From a survey of the literature, we found that nicotin-
amide and nicotinic acid ribosides could be obtained
from their respective mononucleotides by enzymatic
reaction.4 In addition, a number of synthetic approaches
to form NAR and its derivatives have been described.5
However, most of the reported procedures gave the
nicotinamide riboside as an a/b anomeric mixture.
Recently, a simple but nonstereoselective chemical syn-
thesis of b-NAR starting from 1,2,3,5-tetra-O-acetyl-b-
D-ribofuranose was reported by Tanimori et al.6 in this
journal. At first, we repeated this synthetic strategy,
but all attempts to isolate the NAR were unsuccessful.
Thus we decided to reinvestigate the above methodology
making changes in some parameters to set up a stereo-
selective synthesis of both nicotinamide and nicotinic
acid b-ribosides (b-NAR 1, and b-NaR 2, respectively).
Keywords: Nicotinamide b-riboside; Nicotinic acid b-riboside; NAD.
Corresponding author. Tel.: +39-0737-402228; fax: +39-0737-
637345; e-mail: palmarisa.franchetti@unicam.it
*
The key elements of our synthesis involved the forma-
tion of silylated nicotinamide as starting material, and
0960-894X/$ - see front matter ꢀ 2004Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.093

4656
P. Franchetti et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4655–4658
then its coupling with peracylated sugars in the presence
of trimethylsilyl trifluoromethanesulfonate. However,
we found that when a large excess of trimethylsilyl chlo-
ride in hexamethyldisilazane was used to silylate the
nicotinamide, a poor yield of the desired nucleoside
was obtained and the purification of 3-(carbamoyl)-1-
(2,3,5-tri-O-acetyl-b-D-ribofuranosyl) pyridinium triflate
(9) was very difficult. In addition, the amount of
TMSOTf proved to be crucial in this reaction.
In order to validate the above stereoselective procedure,
we investigated the ribosylation of 3-nicotinic acid using
both the sugars 7 and 8. In this case, for the O-silylation
of 5, 1equiv of TMSCl was used. Both the glycosyl-
ations provided the tri-O-acylated ribosides 1011 and
1212 uniquely as b-anomers. Nucleoside 213 was ob-
tained as a white solid after deprotection of 10 or 12
as reported for 1.
The stereoselectivity of the synthesis may be explained in
the following way (Fig. 1). In the first step, the TMSOTf
catalyst converts peracylated sugars 7 and 8 into the cor-
responding 1,2-acyloxonium salts as a unique electro-
philic sugar moiety. Under these reversible and thus
thermodynamically controlled conditions, the nucleo-
philic silylated bases 4 and 6 can attack only the fura-
nose cations from the top (the b-side) to afford the
b-nucleosides 9–12. Only the silylated bases will react
with the electrophilic sugar cations to form the final
products; since an excess of TMSCl blocks the N1 of
nicotinic bases, a slow and poor reactivity was observed
when more than 2equiv of TMSCl were used.
We obtained a stereoselectivity in the synthesis of NAR
by condensation of nicotinamide with both 1,2,3,5-tetra-
O-acetyl-b-D-ribofuranose (7) and 1-O-acetyl-2,3,5-tri-
O-benzoyl-b-D-ribofuranose (8) using carefully control-
led conditions. In this procedure we used 2equiv of
TMSCl to silylate the dry nicotinamide under reflux.
The intermediate 4 was directly coupled with the pro-
tected sugars (1 eqiv) under Vorbruggenꢀs7 conditions
in the presence of a catalytic amount of TMSOTf in
1,2-dichloroethane, to give 98 and 119 in high yield
(Scheme 1). Only b-anomers of the protected N-nucleo-
sides were obtained. b-NAR10 (1) was then obtained by
basic hydrolytic deblocking of compounds 9 or 11 at
ꢀ5ꢁC. The low temperature was required to minimize
cleavage of the glycosidic linkage. In the case of com-
pound 11, for removal of benzoyl groups, two days were
required under methanolic ammonia. The nucleoside 1
was then purified by chromatography on activated char-
coal and isolated as a white solid.
Anomeric purity of b-NAR and b-NaR was revealed to
be 100% by reverse phase HPLC and the structure of
these nucleosides was confirmed by mass spectrometric
analysis. Unambiguous assignment of the configuration
1
of ribonucleosides was best accomplished by H NMR
data and proton nuclear Overhauser enhancement
1
(NOE) effects.9–13 It is worth noting that the H NMR
spectrum of compound 1 was different from that re-
ported by Tanimori et al.,6 but similar to that described
by Jarman and Ross14 for the corresponding chloride.
O
R
0
0
Moreover, the coupling constant J1 –2 and the proton
chemical shift for NAR (1) corresponded to values re-
ported by Oppenheimer and Japlan.15 The proton spec-
trum of both the protected compounds 9–128,9,11,12 and
nucleosides 110 and 213 contained a unique doublet,
which was attributed to the anomeric proton (H10) of
the ribose moiety. The absence of a second doublet in
this region provided strong evidence for the essential
anomeric purity of the nucleosides. The b-configuration
N
a
4 R = N[Si(CH3)3]2
6 R = OSi(CH3)3
3 R = NH2
5 R = OH
AcO
BzO
OAc
O
OAc
O
AcO
b
BzO
OBz
OAc
7
8
O
O
R
0
0
)
was confirmed through the coupling constants (J1 –2
value and by proton NOE data. Because of the overlap-
ping of the proton signals of H20, and H40 in the de-
protected ribosides 1 and 2, the NOE spectrum was
performed in the corresponding protected nucleosides.
In the one-dimensional NOE spectrum, irradiation of
the anomeric protons of 9–12 showed an enhancement
of H20 and H40 in the ribose moiety, while the intensity
enhancement of the H30 signal was zero, supporting a
spatial arrangement where H10 and H40 protons are
on the same side as ribofuranose, as would be the case
in b-configuration. Moreover an enhancement of H2
and H6 signals in the pyridine ring was also observed.
The correct structure of b-NAR was also confirmed
by the 13C NMR spectrum on the basis of reported
values for NMN and NAD,16,17 mass spectrometric
analysis, and through a CH COSY diagram (D2O,
300MHz).
R
+
+
N
N
BzO
AcO
O
O
TfO-
TfO-
BzO
OBz
AcO
OAc
c
9 R = NH2 (96%)
10 R = OH (64%)
d
11 R = NH2 (93%)
12 R = OH (74%)
O
R
+
N
HO
O
TfO-
HO
OH
1 R = NH2 (45%)
2 R = OH (48%)
Scheme 1. Reagents and conditions: (a) TMSCl, HMDS, 120ꢁC, 5h;
(b) TMSOTf, ClCH2CH2Cl anhydrous, 45ꢁC, 2h; (c) NH3/CH3OH,
ꢀ5ꢁC, 6h; (d) NH3/CH3OH, ꢀ5ꢁC, 48h.
A sugar C-methylated in position 3 of the ribofuranosyl
moiety was also used to obtain a new class of nicotin-

P. Franchetti et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4655–4658
4657
O
R
N
RO
RO
OAc
4 R = N(TMS)2
6 R = OTMS
O
O
TMSOTf
O
9-12
+
RO
OR
O
RO
R
TfO-
7 R = Ac
8 R = Bz
Figure 1. Glycosylation mechanism of pyridinic nucleosides.
BzO
Grifantini for his helpful scientific discussion. We would
also like to thank G. Rafaiani, F. Lupidi, and M. Brandi
for technical assistance.
OAc
O
H3C
4
+
AcO
OAc
13
a
O
References and notes
NH2
+
1. Zhang, X.; Kurnasov, O. V.; Karthikeyan, S.; Grishin, N.
V.; Osterman, A. L.; Zhang, H. J. Biol. Chem. 2003, 278,
13503.
2. Magni, G.; Amici, A.; Emanuelli, M.; Orsomando, G.;
Raffaelli, N.; Ruggieri, S. Curr. Med. Chem. 2004, 11, 873,
and references cited therein.
3. Franchetti, P.; Cappellacci, L.; Grifantini, M.; Jayaram,
H. N.; Goldstein, B. M. In Inosine Monophosphate
Dehydrogenase; Pankiewicz, K. W., Goldstein, B. M.,
Eds.; ACS: Washington, DC, 2003; pp 212–230, references
cited therein.
N
BzO
H3C
AcO
TfO-
O
14
OAc
b
O
NH2
+
N
HO
O
TfO-
H3C
4. Friedlos, F.; Jarman, M.; Davies, L. C.; Boland, M. P.;
Hnox, R. J. Biochem. Pharmacol. 1992, 44, 25.
15
HO
OH
5. (a) Haynes, L. J.; Hughes, N. A.; Kenner, G. W.; Todd, A.
J. Chem. Soc. 1957, 3727; (b) Jarman, M. J. Chem. Soc.
1969, 41, 918; (c) Mikhailopulo, I. A.; Pricota, T. I.;
Timoshchuk, V. A.; Akhrem, A. A. Synthesis 1981,
388.
6. Tanimori, S.; Ohta, T.; Kirihata, M. Bioorg. Med. Chem.
Lett. 2002, 12, 1135.
7. Vorbruggen, H.; Krolikiewicz, K.; Bennus, B. Chem. Ber.
1981, 114, 1234.
Scheme 2. Reagents and conditions: (a) TMSOTf, ClCH2CH2Cl
anhydrous, 45ꢁC, 2h; (b) NH3/CH3OH, ꢀ5ꢁC, 48h, (yield 46%).
amide riboside derivatives using the above-described
methodology. Thus, the synthesis of N-pyridinium 3-
(carbamoyl)-1-(30-C-methyl)-b-D-ribofuranose (NA30-
C-MeR) is reported (Scheme 2). Coupling of 1,2,3-tri-
O-acetyl-5-O-benzoyl-b-D-(3-C-methyl)ribofuranose18
(13) with silylated nicotinamide (4) followed by direct
basic deblocking of nucleoside 14, gave nucleoside 15
as a white solid, after purification as reported above.
8. Compound 9: TLC (BuOH/H2O/AcOH, 7:2:1) Rf =0.40.
1H NMR (300MHz DMSO-d6): d (ppm) 2.08 (s, 9H,
CH3), 4.45 (d, J=3.3Hz, 2H, H50), 4.70 (m, 1H, H40), 5.40
(t, J=6.0Hz, 1H, H30), 5.65 (dd, J=3.5, 5.7Hz, 1H, H20),
6.60 (d, J=3.3Hz, 1H, H10), 8.20 (br s, 1H, NH), 8.40 (dd,
J=6.2, 8.1Hz, 1H, H5), 8.60 (br s, 1H, NH), 9.0 (d,
J=8.4Hz, 1H, H4), 9.35 (d, J=6.2Hz, 1H, H6), 9.40 (s,
1H, H2). MS (API-ESI) m/z 381.2 [MꢀTfOꢀ]. HPLC
[DAD and MS detector HP 1100 MSD, G19/46 A];
conditions: column: C18 reverse phased 3.9·300mm,
4lm, Nova Pak (Waters), isocratic flow rate 0.7mL/min,
injection volume: 1ll/min, 30% MeOH in 100mM ammo-
nium acetate buffer; 254nm DAD detection: retention
time: 3.35min. The other compounds were followed the
same above conditions for HPLC. Anal. calcd for
C17H21N2O8 ÆCF3SO3ꢀ: C, 40.76; H, 3.99; N, 5.28. Found:
C, 40.95; H, 4.11; N, 5.14.
9. Compound 11: TLC (BuOH,/H2O/AcOH, 8:1:1) Rf =0.56.
1H NMR (300MHz DMSO-d6): d (ppm) 4.90 (m, 2H,
H50), 5.20 (m, 1H, H40), 6.0 (t, J=6.1Hz, 1H, H30), 6.15
(m, 1H, H20), 7.0 (d, J=3.4Hz, 1H, H1 0), 7.20–8.0 (m,
15H, arom.), 8.10 (br s, 1H, NH), 8.30 (t, J=7.3Hz, 1H,
H5), 8.60 (br s, 1H, NH), 9.05 (d, J=8.3Hz, 1H, H4), 9.45
(d, J=5.9Hz, 1H, H6), 9.65 (s, 1H, H2); MS (API-ESI) m/
z 567.2 [MꢀTfOꢀ]. HPLC: retention time: 3.62min. Anal.
1
The structure of 1519 was determined by H NMR and
mass spectrometric analysis and b-configuration of this
nucleoside was confirmed by proton NOE.
In summary, a stereoselective reaction for b-NAR and
its analogs via TMSOTf-mediated glycosylation of the
silylated nicotinic bases using controlled conditions
was shown.
In order to obtain novel NAD analogs as substrate/
inhibitors of key enzymes involved in nucleotide biosyn-
thesis, other NAR derivatives are currently under
investigation.
Acknowledgements
This study was supported by the Italian MIUR funds
(PRIN 2002). We are very grateful to Professor Mario

4658
P. Franchetti et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4655–4658
calcd for C32H27N2O8 ÆCF3SO3ꢀ: C, 55.31; H, 3.80; N,
3.91. Found: C, 55.50; H, 3.68; N, 4.07.
13. Compound 2: TLC (BuOH/H2O/AcOH, 5:3:2) Rf =0.36.
1H NMR (300MHz DMSO-d6): d (ppm) 3.60–3.80 (dou-
ble q, J=12.1Hz, 2H, H50), 4.10 (m, 2H, H30, H40), 4.22
(d, J=3.6Hz, 1H, H20), 5.20 (br s, 1H, OH), 5.25 (br s,
1H, OH), 6.15 (d, J=5.1Hz, 1H, H10), 6.25 (br s, 1H,
OH), 8.10 (t, J=6.2Hz, 1H, H5), 8.80 (d, J=7.6Hz, 1H,
H4), 9.20 (d, J=6.2Hz, 1H, H6), 9.30 (s, 1H, H2); 1H
NMR (300MHz D2O): d (ppm) 3.86 (dd, J=3.8, 13.0Hz,
1H, H50), 4.02 (dd, J=2.9Hz, 12.8Hz, 1H, H50), 4.30–4.36
(t, J=4.6Hz, 1H, H30), 4.41–4.50 (m, 2H, H20, H40), 6.20
(d, J=4.8Hz, 1H, H10), 8.15 (t, J=6.8Hz, 1H, H5), 8.90
(d, J=8.1Hz, 1H, H4), 9.15 (d, J=5.9Hz, 1H, H6), 9.40
(s, 1H, H2); MS (API-ESI) m/z 256.1 [MꢀTfOꢀ]. HPLC:
retention time: 2.81min. Anal. calcd for C11H14NO6 Æ
CF3SO3ꢀ: C, 33.59; H, 3.59; N, 3.56. Found: C, 33.47; H,
3.68; N, 3.41.
10. Compound 1: TLC (BuOH/H2O/AcOH, 5:3:2) Rf =0.38.
1H NMR (300MHz DMSO-d6): d (ppm) 3.60–3.80 (dou-
ble q, J=12.2Hz, 2H, H50), 4.10 (m, 1H, H30), 4.20–4.35
(m, 1H, H40, H20), 5.40 (t, J=4.4Hz, 1H, OH), 5.50 (d,
J=5.1Hz, 1H, OH), 5.90 (d, J=5.9Hz, 1H, OH), 6.18 (d,
J=4.4Hz, 1H, H10), 8.20 (br s, 1H, NH), 8.35 (t, J=
7.1Hz, 1H, H5), 8.60 (br s, 1H, NH), 9.0 (d, J=8.4Hz,
1H, H4), 9.35 (d, J=6.2Hz, 1H, H6), 9.55 (s, 1H, H2); 1H
NMR (300MHz D2O): d (ppm) 3.75 (dd, J=3.6, 12.8Hz,
1H, H50), 3.90 (dd, J=2.9, 13.2Hz, 1H, H50), 4.20 (t,
J=4.6Hz, 1H, H30), 4.25–4.40 (m, 2H, H20, H40), 6.08 (d,
J=4.2Hz, 1H, H10), 8.10 (dd, J=6.2, 8.0Hz, 1H, H5),
8.80 (d, J=7.7Hz, 1H, H4), 9.10 (d, J=5.9Hz, 1H, H6),
9.45 (s, 1H, H2); 13C NMR (300MHz D2O): d (ppm) 60.35
(C50), 69.91 (C30), 77.61 (C20), 87.80 (C40), 100.10 (C10),
128.59 (C5), 134.10 (C3), 140.53 (C2), 142.78 (C6), 145.81
(C4). MS (API-ESI) m/z 255.2 [MꢀTfOꢀ]. HPLC: reten-
tion time: 2.86min. Anal. calcd for C11H15N2O5 Æ
CF3SO3ꢀ: C, 35.64; H, 3.74; N, 6.93. Found: C, 35.88;
H, 3.61; N, 6.78.
14. Jarman, M.; Ross, W. C. J. Chem. Soc. 1969, 199.
15. Oppenheimer, N. J.; Japlan, N. O. Biochemistry 1976, 15,
3981.
16. Birdsall, B.; Feeney, J. J. C. S. Perkin II 1972, 1643.
17. Johnson, R. W.; Marschner, T. M.; Oppenheimer, N. J.
J. Am. Chem. Soc. 1988, 110, 2257.
11. Compound 10: TLC (BuOH/H2O/AcOH, 7:2:1) Rf =0.38.
1H NMR (300MHz DMSO-d6): d (ppm) 2.10 (s, 9H,
CH3), 4.42 (d, J=2.9Hz, 2H, H50), 4.70 (m, 1H, H40), 5.40
(t, J=5.9Hz, 1H, H30), 5.60 (dd, J=3.3, 5.9Hz, 1H, H20),
6.70 (d, J=2.9Hz, 1H, H10), 8.40 (m, 1H, H5), 9.10 (d,
J=8.5Hz, 1H, H4), 9.30 (d, J=6.2Hz, 1H, H6), 9.50 (s,
1H, H2); MS (API-ESI) m/z 382.1 [MꢀTfOꢀ]. HPLC:
retention time: 3.32min. Anal. calcd for C17H20NO9 Æ
CF3SO3ꢀ: C, 40.68; H, 3.79; N, 2.63. Found: C, 40.75; H,
3.91; N, 2.55.
18. (a) Ong, S. P.; Nelson, L. S.; Hogenkamp, H. P. C.
Biochemistry 1992, 11210; (b) A novel synthesis of 13 will
be reported by us elsewhere.
19. Compound 15: TLC (BuOH/H2O/AcOH, 5:3:2) Rf =0.44.
1H NMR (300MHz DMSO-d6): d (ppm) 1.25 (s, 3H,
CH3), 3.60–3.80 (m, 2H, H50), 4.0 (d, J=6.9Hz, 1H, H40),
4.10 (s, 1H, H20), 4.40–4.60 (br s, 1H, OH), 5.40 (br s, 1H,
OH), 5.50 (br s, 1H, OH), 6.18 (d, J=6.9Hz, 1H, H10),
8.20 (br s, 1H, NH), 8.35 (t, J=7.1Hz, 1H, H5), 8.70 (br s,
1H, NH), 9.0 (d, J=7.7Hz, 1H, H4), 9.40 (d, J=6.6Hz,
1H, H6), 9.65 (s, 1H, H2); 1H NMR (300MHz D2O): d
(ppm) 1.23 (s, 3H, CH3), 3.70 (dd, J=4.1, 12.7Hz, 1H,
H50), 3.85 (dd, J=3.2, 12.7Hz, 1H, H50), 4.18 (d,
J=6.7Hz, 1H, H20), 4.30 (t, J=3.7Hz, 1H, H40), 6.0 (d,
J=7.0Hz, 1H, H10), 8.15 (q, J=6.2Hz, 1H, H5), 8.85 (d,
J=8.2Hz, 1H, H4), 9.15 (d, J=6.1Hz, 1H, H6), 9.44 (s,
1H, H2); MS (API-ESI) m/z 269.2 [MꢀTfOꢀ]. HPLC:
retention time: 2.95min. Anal. calcd for C12H17N2O5 Æ
CF3SO3ꢀ: C, 37.32; H, 4.09; N, 6.70. Found: C, 37.48; H,
3.91; N, 6.82.
12. Compound 12: TLC (BuOH/H2O/AcOH, 8:1:1) Rf =0.50.
1H NMR (300MHz DMSO-d6):
d (ppm) 4.85 (d,
J=3.3Hz, 2H, H50), 5.20 (m, 1H, H40), 6.0 (t, J=5.6Hz,
1H, H30), 6.15 (dd, J=3.3, 5.8Hz, 1H, H20), 7.05 (d,
J=3.6Hz, 1H, H10), 7.20–8.15 (m, 15H, arom.), 8.25 (t,
J=7.6Hz, 1H, H5), 9.0 (d, J=8.4Hz, 1H, H4), 9.40 (d,
J=6.6Hz, 1H, H6), 9.70 (s, 1H, H2); MS (API-ESI) m/z
568.1 [MꢀTfOꢀ]. HPLC: retention time: 3.62min. Anal.
calcd for C32H26NO9 ÆCF3SO3ꢀ: C, 54.47; H, 3.71; N,
1.98. Found: C, 54.59; H, 3.60; N, 2.16.