Diastereocontrolled electrophilic fluorinations of 2-deoxyribonolactone: Syntheses of all corresponding 2-deoxy-2-fluorolactones and 2′-deoxy- 2′-fluoro-NAD+s

Article abstract of DOI:10.1021/jo900637f

(Chemical Equation Presented) Methods to construct 2′-deoxy-2′- fluoro nucleosides have undergone limited improvement in the last 20 years in spite of the substantially increased value of these compounds as pharmaceuticals and as tools for studying biological processes.We herein describe a consolidated approach to synthesize precursors to these commercially and scientifically valuable compounds via diastereocontrolled fluorination of the readily available precursor 2-deoxy-D-ribonolactone.With employment of appropriate sterically bulky silyl protecting groups at the 3 and 5 positions, controlled electrophilic fluorination of the Liribonolactone enolate by N-fluorodibenzenesulfonamide yielded the corresponding 2-deoxy-2- fluoroarabinolactone in high isolated yield (72%). The protected 2-deoxy-2,2-difluororibonolactone was obtained similarly in high yield from a second round of electrophilic fluorination (two steps, 51%from protected ribonolactone starting material). Accomplishment of the difficult ribofluorination of the lactone was achieved by the directive effects of a diastereoselectively installed α-trimethylsilyl group. Electrophilic fluorination of a protected 2-deoxy-2-trimethylsilylarabinolactone via enolate generation provided the protected 2-deoxy-2-fluororibolactone as the exclusive fluorinated product. The reaction also yielded the starting material, the desilylated protected 2-deoxyribonolactone, which was recycled to provide a 38%chemical yield of the fluorinated product (versus initial protected ribonolactone) after consecutive silylation and fluorination cycles.Using our fluorinated sugar precursors, we prepared the 2′-fluoroarabino-, 2′-fluororibo-, and 2′,2′-difluoronicotinamide adenine dinucleotides (NAD⁺) of potential biological interest. These syntheses provide the most consolidated and efficient methods for production of sugar precursors of 2′-deoxy-2′-fluoronucleosides and have the advantage of utilizing an air-stable electrophilic fluorinating agent. The fluorinated NAD⁺s are anticipated to be useful for studying a variety of cellular metabolic and signaling processes.

Full text of DOI:10.1021/jo900637f

pubs.acs.org/joc
Diastereocontrolled Electrophilic Fluorinations of 2-Deoxyribonolactone:
Syntheses of All Corresponding 2-Deoxy-2-fluorolactones and
20-Deoxy-20-fluoro-NAD+s
Yana Cen and Anthony A. Sauve*
Department of Pharmacology, Weill Medical College of Cornell University, 1300 York Avenue, New York,
New York 10065
aas2004@med.cornell.edu
Received March 25, 2009
Methods to construct 20-deoxy-20-fluoro nucleosides have undergone limited improvement in the last
20yearsinspiteofthe substantiallyincreasedvalue of thesecompounds aspharmaceuticalsand astools
for studying biological processes. We herein describe a consolidated approach to synthesize precursors
to these commercially and scientifically valuable compounds via diastereocontrolled fluorination of the
readily available precursor 2-deoxy-D-ribonolactone. With employment of appropriate sterically bulky
silyl protecting groups at the 3 and 5 positions, controlled electrophilic fluorination of the Li-
ribonolactone enolate by N-fluorodibenzenesulfonamide yielded the corresponding 2-deoxy-2-fluor-
oarabinolactone in high isolated yield (72%). The protected 2-deoxy-2,2-difluororibonolactone was
obtained similarly in high yield from a second round of electrophilic fluorination (two steps, 51% from
protected ribonolactone starting material). Accomplishment of the difficult ribofluorination of the
lactone was achieved by the directive effects of a diastereoselectively installed R-trimethylsilyl group.
Electrophilic fluorination of a protected 2-deoxy-2-trimethylsilylarabinolactone via enolate generation
provided the protected 2-deoxy-2-fluororibolactone as the exclusive fluorinated product. The reaction
also yielded the starting material, the desilylated protected 2-deoxyribonolactone, which was recycled
to provide a 38% chemical yield of the fluorinated product (versus initial protected ribonolactone) after
consecutive silylation and fluorination cycles. Using our fluorinated sugar precursors, we prepareþd the
20-fluoroarabino-, 20-fluororibo-, and 20,20-difluoronicotinamide adenine dinucleotides (NAD ) of
potential biological interest. These syntheses provide the most consolidated and efficient methods for
production of sugar precursors of 20-deoxy-20-fluoronucleosides and have the advantage of utilizing an
air-stable electrophilic fluorinating agent. The fluorinated NADþs are anticipated to be useful for
studying a variety of cellular metabolic and signaling processes.
Introduction
substituent.1-4 The selective replacement of hydrogen or oxy-
gen with fluorine can change a compound’s biological activity,
metabolic stability, chemical stability, lipophilicity, acidity,
and dipole properties with modest change in steric bulk.2-4
The efficient introduction of fluorine atoms into bioactive
organic molecules has attracted considerable attention in
recent years owing to the unique properties of the fluorine
DOI: 10.1021/jo900637f
r
Published on Web 07/07/2009
J. Org. Chem. 2009, 74, 5779–5789 5779
2009 American Chemical Society

Cen and Sauve
JOCFeatured Article
A broad class of particularly relevant compounds are the
fluorinated carbohydrates1,5,6 and nucleosides.1,3,4 Selective
fluorination in the sugar moiety of glycosides or nucleosides
has proven useful to numerous investigations of enzyme
mechanism where either sugars or nucleosides are sub-
strates.6-14 Some of these compounds are potent drugs. For
example, gemcitabine (Gemzar, Eli Lilly, 20-deoxy-20,20-di-
fluorocytidine) is used clinically to treat numerous cancers
including ovarian,15 pancreatic,16 and breast17 cancers, with
sales in excess of 1.6 billion dollars per year.18
Of the monosubstituted 20-fluoronucleosides, clofarabine
(Clorar, Genzyme, 2-chloro-20-deoxy-20-fluoroarabinoadeno-
sine), has been approved for pediatric patients with relapsed or
refractory acute lymphocytic leukemia,19 and annual sales now
exceed 100 million dollars.20
Ourlaboratoryis interested in general and flexible methods
to construct 20-fluoronucleosides, nucleotides and dinucleo-
tides, particularly derivatives of nicotinamide riboside. In
previous studies, we demonstrated that 20-deoxy-20-fluoroar-
abinonicotinamide-mononucleotide is a potent mechanism-
based inhibitor (apparent Ki=61 nM)of the signaling enzyme
cell developmental protein 38 (CD38).10,12 The 20-fluoro-
NADþs and related compounds are likely to be valuable
for the study of sirtuin enzyme mechanism11,21 and for study-
ing the chemical properties of poly-ADP-ribosylpoly-
merases.22 Fluorinated NADþs could be useful for
identifying ADP-ribosyltransfer sites on proteins as well.
For example, we previously demonstrated that the 20-
deoxy-20-fluoroarabinofuranosyl modification is suitably ro-
bust for MS/MS approaches used to characterize amino acid
post-translational modifications.10 ADP-ribosylation sites
are poorly surveyed within the proteome, and yet these
modifications areof heightened interest as they are implicated
in important biological effects.22-24 The synthesis of 20-
deoxy-20-fluoroarabino-NADþ was previously described,25
but syntheses of 20-deoxy-20-fluororibo-NADþ and 20-deoxy-
20,20-difluoro-NADþ or their nucleoside precursors have not
appeared in the literature. The difluoro derivatives in parti-
cular could be quite interesting on the basis of their antici-
pated chemical stability and altered electronic properties.
Methods to synthesize the 2-deoxy-2-fluoro-D-furanose
precursors to 20-fluorinated nucleosides have not experienced
substantial recent innovation,1,3,26 despite the fact that inter-
est and demand for these compounds have only increased in
recent years. The synthetic methods currently available vary
in efficiency, and all depend on routes relying on different
initial precursors to the respective final products. For exam-
ple, the best method to make a protected 2-deoxy-2-fluor-
oarabinofuranose requires only two steps (Scheme 1, part I)
from a protected ribose (58% overall yield).25,27-29 On the
other hand, 2-deoxy-2-fluororibofuranose has no concise or
efficient synthesis30 and still requires six steps from arabinose
(Scheme 1, part II; 4% overall yield).31 Alternatively, it can be
obtained via a 10-step nondiastereoselective method in 11%
overall yield.32-34 An efficient but nondiastereoselective
route is established for the synthesis of protected 2-deoxy-
2,2-difluororibonofuranose in five steps (Scheme 1, part III;
crude yield 25%) via coupling of ethyl bromodifluoroacetate
and isopropylidene glyceraldehyde. The route is not diaster-
eoselective and depends upon crystallization of the preferred
isomer.35 This procedure was developed by Eli Lilly for
commercial synthesis of gemcitabine. 2-Deoxy-2,2-difluoror-
ibofuranose can also be obtained stereoselectively from
glucose or mannose,36 but that method involves eight steps
and very lowoverall yield (<15%). We wonderedif it might be
possible to improve and perhaps consolidate synthetic ap-
proaches to these desirable sugars by employing electrophilic
fluorine as a means to modify protected 2-deoxyribonolac-
tone, a starting material that can be obtained readily, abun-
dantly, and cheaply.
(1) Ma, J. A.; Cahard, D. Chem. Rev. 2008, 108, PR1–PR43.
(2) Hagmann, W. K. J. Med. Chem. 2008, 51, 4359–4369.
(3) Kirk, K. L. Org. Process Res. Dev. 2008, 12, 305–321.
(4) O’Hagan, D. Chem. Soc. Rev. 2008, 37 (2), 308–319.
(5) Nicolaou, K. C.; Mitchell, H. J. Angew. Chem., Int. Ed. 2001, 40,
1576–1624.
(6) Williams, S. J.; Withers, S. G. Carbohydr. Res. 2000, 327, 27–46.
(7) Vocadlo, D. J.; Withers, S. G. Carbohydr. Res. 2005, 340, 379–388.
(8) Vocadlo, D. J.; Davies, G. J.; Laine, R.; Withers, S. G. Nature 2001,
412, 835–838.
The strategy we considered is visually described in Scheme 2.
We contemplated a general solution to the introduction of
fluorine by controlling the reactivity of a 2-deoxyribonolac-
tone enolate with an electrophilic fluorinating reagent. It is
(9) Zechel, D. L.; Withers, S. G. Curr. Opin. Chem. Biol. 2001, 5, 643–649.
(10) Sauve, A. A.; Deng, H. T.; Angeletti, R. H.; Schramm, V. L. J. Am.
Chem. Soc. 2000, 122, 7855–7859.
(23) Smith, B. C.; Hallows, W. C.; Denu, J. M. Chem. Biol. 2008, 15,
1002–1013.
(11) Jackson, M. D.; Schmidt, M. T.; Oppenheimer, N. J.; Denu, J. M. J.
Biol. Chem. 2003, 278, 50985–50998.
(12) Liu, Q.; Kriksunov, I. A.; Jiang, H.; Graeff, R.; Lin, H.; Lee, H. C.;
Hao, Q. Chem. Biol. 2008, 15, 1068–1078.
(24) Haigis, M. C.; Mostoslavsky, R.; Haigis, K. M.; Fahie, K.; Chris-
todoulou, D. C.; Murphy, A. J.; Valenzuela, D. M.; Yancopoulos, G. D.;
Karow, M.; Blander, G.; Wolberger, C.; Prolla, T. A.; Weindruch, R.; Alt, F.
W.; Guarente, L. Cell 2006, 126, 941–954.
(13) Porter, D. J.; Merrill, B. M.; Short, S. A. J. Biol. Chem. 1995, 270,
15551–15556.
(25) Sleath, P. R.; Handlon, A. L.; Oppenheimer, N. J. J. Org. Chem.
1991, 56, 3608–3613.
(14) Clark, J. L.; Hollecker, L.; Mason, J. C.; Stuyver, L. J.; Tharnish, P.
M.; Lostia, S.; McBrayer, T. R.; Schinazi, R. F.; Watanabe, K. A.; Otto, M.
J.; Furman, P. A.; Stec, W. J.; Patterson, S. E.; Pankiewicz, K. W. J. Med.
Chem. 2005, 48, 5504–5508.
(26) Dax, K.; Albert, M.; Ortner, J.; Paul, B. J. Carbohydr. Res. 2000,
327, 47–86.
(27) Reichman, U.; Watanabe, K. A.; Fox, J. J. Carbohydr. Res. 1975, 42,
233–240.
(15) Ferrandina, G.; Ludovisi, M.; Lorusso, D.; Pignata, S.; Breda, E.;
Savarese, A.; Del Medico, P.; Scaltriti, L.; Katsaros, D.; Priolo, D.; Scambia,
G. J. Clin. Oncol. 2008, 26, 890–896.
(16) Vulfovich, M.; Rocha-Lima, C. Expert Rev. Anticancer Ther. 2008,
8, 993–1002.
(17) Silvestris, N.; Cinieri, S.; La Torre, I.; Pezzella, G.; Numico, G.;
Orlando, L.; Lorusso, V. Breast 2008, 17, 220–226.
(18) Eli Lilly. Eli Lilly 2007 Annual Report, 2007.
(19) Bonate, P. L.; Arthaud, L.; Cantrell, W. R., Jr.; Stephenson, K.;
Secrist, J. A. III; Weitman, S. Nat. Rev. Drug Discov. 2006, 5, 855–863.
(20) Genzyme. Genzyme 2007 Annual Report, 2007.
(21) Sauve, A. A.; Wolberger, C.; Schramm, V. L.; Boeke, J. D. Annu.
Rev. Biochem. 2006, 75, 435–465.
(28) Tewson, T. J.; Welch, M. J. J. Org. Chem. 1978, 43, 1090–1092.
(29) Watts, J. K.; Damha, M. J. Can. J. Chem. 2008, 86, 641–656.
(30) Pankiewicz, K. W. Carbohydr. Res. 2000, 327, 87–105.
(31) Larsen, C. H.; Ridgway, B. H.; Shaw, J. T.; Smith, D. M.; Woerpel,
K. A. J. Am. Chem. Soc. 2005, 127, 10879–10884.
(32) Mikhailopulo, I. A.; Poopeiko, N. E.; Sivets, G. G.; Khripach, N. B.
Nucleosides Nucleotides 1995, 14, 383–384.
(33) Wright, J. A.; Taylor, N. F.; Fox, J. J. J. Org. Chem. 1969, 34, 2632–
2636.
(34) Anderson, C. D.; Goodman, L.; Baker, B. R. J. Am. Chem. Soc.
1958, 80, 5247–5252.
(35) Chou, T. S.; Heath, P. C.; Patterson, L. E.; Poteet, L. M.; Lakin, R.
E.; Hunt, A. H. Synthesis 1992, 6, 565–570.
(22) Hakme, A.; Wong, H. K.; Dantzer, F.; Schreiber, V. EMBO Rep.
2008, 9, 1094–1100.
(36) Fernandez, R.; Matheu, M. I.; Echarri, R.; Castillon, S. Tetrahedron
1998, 54, 3523–3532.
5780 J. Org. Chem. Vol. 74, No. 16, 2009

Cen and Sauve
JOCFeatured Article
SCHEME 1. Existing Methods for the Syntheses of 2-Deoxy-2-fluoroarabino/ribofuranoses
SCHEME 2. General Strategy for Diastereoselective Electrophi-
SCHEME 4. Attempted Fluorination of Lactone 5
lic Fluorination of Protected 2-Deoxyribonolactone
SCHEME 5. Synthesis of TBDMS-Protected 2-Deoxy-2-fluor-
oarabinolactone
SCHEME 3. Electrophilic Fluorination of 2,3-Dideoxylactone 1
and 2-Deoxylactone 3
successful, we needed (1) to solve the competing β-elimina-
tion of the enolate species and (2) to rigorously control
stereochemistry for the electrophilic substitution. We herein
describe a successful solution to these problems with signifi-
cant attenuation of the β-elimination as the undesired side
reaction by utilization of bulky silyl protecting groups.
Furthermore, we establish a novel approach to the ribo-
isomer using an R-silyl group. These methods allow us
to synthesize the corresponding 2-deoxy-2-fluoroarabino-,
2-deoxy-2-fluororibo-, and 2-deoxy-2,2-difluorolactones and
furanoses in a diastereoselective manner with markedly im-
proved synthetic efficiency. We also report the successful
well-known that a lactone enolate can be functionalized in this
manner, although stereocontrol is an issue and the studied
systems predominantly are not subject to elimination.1 To be
J. Org. Chem. Vol. 74, No. 16, 2009 5781

Cen and Sauve
JOCFeatured Article
SCHEME 6. Synthesis of Silyl-Protected 2-Deoxy-2-fluoroarabinofuranosea
aThe inset shows the Li-enolate of compound 10 as minimized by MM2 calculations. The bond angle is shown after minimization. Triisopropyl groups, lithium
atom, and other hydrogens have been omitted for clarity.
SCHEME 7. Synthesis of Silyl-Protected 2-Deoxy-2,2-difluororibofuranose
(Scheme 3).38 Although the authors did not specifically address
SCHEME 8. Debromination Approach to 2-Deoxy-2-fluoroarabi-
the reason for the poor outcome, we suspected competing
elimination was likely the cause, since the alkoxide anion is
significantly less basic than the enolate (ΔpKa ≈10) providing a
strong thermodynamic driving force for elimination. We began
our study by using the p-chlorobenzoyl-protected deoxyribo-
nolactone 5, which was reacted under Liotta’s procedure
(NFSi, LiHMDS in THF at -78 °C) and furnished only the
R,β-unsaturated γ-lactone 6 in 65% yield (Scheme 4). Since the
driving force for elimination in our case was so much greater
than in the Dehoux case, we were not surprised by this result.
We speculated that the elimination process could be
diminished by moving to silyl protecting groups, which form
leaving groups less basic than alkoxides (alcohol pKa =
16 versus silanol pKa =1139,40) but can be potentially bulkier
and more readily installed. We hypothesized that a suffi-
ciently bulky group might sterically demand a 3-endo-ring
pucker of the deoxyribonolactone ring, which would mini-
mize steric interactions of the bulky group. Such a ring
conformation would organize the 3-oxygen into a pseudoe-
quatorial position, thus diminishing the tendency of the
silanoate to undergo elimination upon enolate formation.
Protection of deoxyribonolactone with tert-butyldi-
methylsilyl (TBDMS) groups provided lactone 7. When 7
and NFSi were dissolved in THF and cooled to -78 °C, slow
addition of LiHMDS resulted in the formation of 8 and 9
(Scheme 5). 2-Deoxy-2-fluoroarabinolactone 8 was obtained
in 58% yield after silica gel chromatography, supplying a
more than 3-fold improvement in yield for electrophilic
fluorination versus the Dehoux result, encouraging our
strategy. Favorably, the reaction yielded only the arabino
isomer, which can be attributed to the sterically bulky
TBDMS group preventing a syn approach of the bulky
fluorinating agent to the enolate. Although promising, the
bulky TBDMS did not completely inhibit β-elimination as
nolactone
synthesis of the corresponding 20-fluoro-NADþs. Impor-
tantly, the successful development of a consolidated and
flexible synthetic methodology to obtain all three types of
2-deoxy-2-fluoro(ribo or arabino)furanoses is anticipated
to improve access to a large family of medicinally important
20-fluoro-substituted nucleosides.
Results and Discussion
Diastereoselective electrophilic fluorination of protected
γ-lactone was encouraged by existing precedents wherein
β-elimination is not a problem. For example, Liotta and
co-workers reported fluorination of 2,3-dideoxylactone 1 with
N-fluorodibenezenesulfonimide (NFSi) to furnish the 2-fluor-
olactone 2 in 50-70% yield (Scheme 3).37 However, very
limited success has been achieved when the β-position is
substituted with oxygen. In one of the very few examples,
Dehoux et al. applied these conditions to compound 3 and
obtained the fluorinated compound 4 in only 16% yield
(37) McAtee, J. J.; Schinazi, R. F.; Liotta, D. C. J. Org. Chem. 1998, 63,
2161–2167.
(39) Tripp, C. P.; Hair, M. L. Langmuir 1995, 11, 149–155.
(40) Damrauer, R.; Simon, R.; Krempp, M. J. Am. Chem. Soc. 1991, 113,
4431–4435.
(38) Dehoux, C.; Gorrichon, L.; Baltas, M. Eur. J. Org. Chem. 2001, 6,
1105–1113.
5782 J. Org. Chem. Vol. 74, No. 16, 2009

Cen and Sauve
JOCFeatured Article
SCHEME 9. Synthesis of 2-Deoxy-2-fluororibolactone
SCHEME 10. Synthesis of 20-Deoxy-20-fluoroarabino-NADþ
we repeat the fluorination procedure to achieve this valuable
compound. Favorably, the second fluorination under the
same reaction conditions provided the desired 2-deoxy-2,2-
difluororibonolactone 13 in 71% yield (Scheme 7). Reduc-
tion with DIBAL-H provided lactol 14 in three steps with
47% yield from 10. This straightforward method surpassed
the nondiastereoselective method of Chou (five steps, 25%
yield, Scheme 1) for selectivity, brevity, and yield.
The absence from the literature of a convenient synthesis
of 2-deoxy-2-fluororibofuranose30 prompted us to investi-
gate the possibility of converting the 2-deoxyribonolactone
to the 2-deoxy-2-fluororibo isomer. We supposed that a
bulky protecting group at the 3 position would enforce steric
arrangements in enolate reactions with electrophiles. To
force stereochemistry of the fluoro-electrophile into the ribo
configuration we thought of using a removable directing
group introduced at the R position. One attempt along these
lines is shown in Scheme 8 and is based on a previous strategy
of Kirk.41 2-Deoxy-2-bromolactone 15 was prepared by
direct bromination of 7 (Scheme 8). Fluorination of 15
occurred stereoselectively, providing 16, of which the stere-
ochemistry was assigned based upon coupling constants.
Somewhat surprisingly, debromination of 16 with tributyltin
hydride gave only the undesired arabinofluoro compound 8
as the major product with no ribo isomer formed (Scheme 8).
This failure led us to consider an R-silyl lactone as a means
for diastereocontrolled formation of the 2-deoxy-2-fluo-
roribolactone. Purified R-silyl ketones have been used as
substrates in electrophilic fluorinations,42,43 mostly for re-
giocontrolled fluorination away from the silyl group,
although fluorination at the silylated carbon suggested the
potential workability of our strategy.43 We imagined that
fluorination of an R-silyl lactone would generate the desired
ribofluoro configuration as a consequence of the tendency
of the bulky R-silyl and 3 position protecting group to
move into a trans disposition in the enolate, thus permitting
electrophilic fluorination syn to the 3-substitutent at the
R-carbon. Subsequent proteo-desilylation with retention
of stereochemistry was envisioned to provide the desired
compound.
evident by the formation of 9. To increase silyl bulk, we
turned to the triisopropylsilyl (TIPS) group and synthesized
the TIPS-protected lactone 10. Lactone 10 was reacted as
before and β-elimination was diminished, enabling diaster-
eoselective fluorination to give only the arabino isomer 11 in
72% isolated yield (Scheme 6). Stereochemistry was con-
firmed by NOE measurements (see the Experimental Sec-
tion). Subsequently, 11 was reduced with DIBAL-H to
obtain the TIPS-protected 2-deoxy-2-fluoroarabinofura-
nose 12 in 91% yield.
These results demonstrate that β-elimination in electro-
philic fluorination of 2-deoxyribonolactone can be suc-
cessfully mitigated by silyl steric bulk at the susceptible
β-position. An MM2 minimization of the lithium enolate
provided evidence that the steric bulk of the TIPS group
organizes the 3-siloxy group into a geometry that is
40 degrees from the ideal angle for elimination (inset
Scheme 6). The effect of the TIPS on preventing the
elimination is impressive, as approximately 14 orders of
magnitude of anion stability (ΔpKa) favor elimination of
the silanoate from the enolate. Finally, in two steps from a
protected 2-deoxyribonolactone, diastereoselective pro-
duction of a 2-deoxy-2-fluoroarabinofuranoside was
achieved in 66% yield, surpassing the yield of the previous
methodology (58% yield in two steps)25,28,29 and avoid-
ing the use of the caustic substance diethylaminosulfur
trifluoride (DAST).
R-Silyl lactone 17 was generated diastereoselectively (the
arabino configuration was determined by NOE correlation
between H2 and H4) from 7 in 71% yield (Scheme 9).
Interestingly, treatment of 17 with NFSi afforded the desired
ribofluoro product 18 in 33% yield with 2-deoxyribonolac-
tone 7 in 61% yield. The formation of 2-fluoroarabinolac-
(41) Ge, P.; Kirk, K. L. J. Org. Chem. 1997, 62, 3340–3343.
(42) Enders, D.; Potthoff, M.; Raabe, G.; Runsink, J. Angew. Chem., Int.
Ed. 1997, 36, 2362–2364.
(43) Enders, D.; Faure, S.; Potthoff, M.; Runsink, J. Synthesis 2001, 15,
2307–2319.
We considered the monofluoro-substituted lactone 11
to be a possible precursor to the corresponding 2-deoxy-
2,2-difluororibonolactone. We felt it was only required that
J. Org. Chem. Vol. 74, No. 16, 2009 5783

Cen and Sauve
JOCFeatured Article
SCHEME 11. Synthesis of 20-Deoxy-20-fluororibo-NADþ
tone 8 was not observed. The recovered compound 7 was
resubmitted to silylation-fluorination conditions to give
18 in 38% combined yield after one recycle (Scheme 9).
Subsequent reduction of 18 to lactol 19 was effected in 95%
yield. The three-step silylation, fluorination, and reduction
sequence afforded the 2-deoxy-2-fluororibofuranose in 22%
yield. This method is by far the most efficient chemical
synthesis of 2-deoxy-2-fluororibofuranose ever reported.
The result confirmed our expectation that the R-trimethyl-
silyl (TMS) substituent can control the stereochemistry of the
fluorine substituent. The coincident desilylation observed in
the process, although desirable and stereoselective, raises the
question of the origin of the selectivity. Is it determined at the
enolate reaction to the fluorine, or in the desilylation process,
or both? At this point, we do not have a clear answer.
Additionally, we have limited insight into the cause of the
desilylation (without fluorination) that limits yield. We spec-
ulate that the reaction conditions stimulate a significant
percentage of the C-silyl starting material to isomerize to
the O-silyl ketene acetal, before it can react with NFSi, and
when quenched the O-silyl derivative decomposes to 7. Low-
temperature interconversions of C-silylated esters and O-
silylated ketene acetals have been reported.44 Further inves-
tigations into this reaction are ongoing in our laboratory.
At this point, with all possible 2-deoxy-2-fluorofuranoses
in hands, we were prepared to couple nicotinamide to sugars
to obtain the corresponding 20-deoxy-20-fluoronicotinamide
nucleosides and 20-deoxy-20,20-difluoronucleosides, prefer-
ably with stereocontrol to yield the desired β-isomers. Our
main concern was that the silyl groups restrict the levels of
acidity we could employ in the coupling reaction.
describe the synthesis of the 1-(20-deoxy-20-fluoroarabino-
furanosyl)nicotinamide in five steps, 41% isolated yield from
protected 2-deoxyribonolactone. Compound 12 was acti-
vated with methanesulfonylchloride and triethylamine,
which formed only the R-chloro sugar 20 (Scheme 10).
Compound 20 was coupled with nicotinamide in acetonitrile
and dichloromethane mixed solvent with stoichiometric
amount of AgSbF6. Crude product 21 was a mixture of both
isomers with β being the major isomer. After deprotection
with fluoride, β- and R-isomers were separated by prepara-
tive HPLC (β/R=3.5). The 1H NMR spectrum obtained for
22 (β) agreed with literature data.25 The yield of 22 (β) from
12 was 62% and required only one purification step.
A description of the synthesis of 1-(20-deoxy-20-fluoror-
ibofuranosyl)nicotinamide has not appeared in the chemical
literature, although the compound’s stability to hydrolysis
has been described.45 We herein describe the synthesis of
this compound in six steps and 15% isolated yield from
2-deoxyribonolactone. The furanose 19 was converted to the
chloro sugar 23 by treatment with methanesulfonylchloride
and triethylamine, and the reaction occurred quantitatively
but produced both R- and β-isomers (Scheme 11). Coupling
of the chloro sugar to nicotinamide with AgSbF6 provided a
mixture of nicotinamide adducts 24, followed by depro-
tection and HPLC purification to provide the desired nico-
tinamide-substituted 20-deoxy-20-fluororibonucleoside 25 in
45% yield from the lactol 19. The corresponding R isomer
was isolated in 46% yield. Stereochemistries were assigned
by NOEs (see the Experimental Section).
The synthesis of the 1-(20-deoxy-20,20-difluororibofura-
nosyl)nicotinamide has never been reported. The comple-
tion of the synthesis of the β-isomer in six steps with 18%
isolated yield and 23% for the R-isomer are reported here.
We prepared the 1-mesylate 26 similarly to a reported
The only reported synthesis of 1-(20-deoxy-20-fluoroara-
binofuranosyl)nicotinamide was from 1,2:5,6-di-O-isopro-
pylidene-D-allofuranose (eight steps in 24% yield).25 We here
(44) Welch, J. T.; Plummer, J. S.; Chou, T. S. J. Org. Chem. 1991, 56, 353–
359.
(45) Handlon, A. L.; Xu, C.; Mullersteffner, H. M.; Schuber, F.; Oppen-
heimer, N. J. J. Am. Chem. Soc. 1994, 116, 12087–12088.
5784 J. Org. Chem. Vol. 74, No. 16, 2009

Cen and Sauve
JOCFeatured Article
SCHEME 13. Syntheses of Protected 2-Deoxy-2-fluorofuranose
SCHEME 12. Synthesis of 20-Deoxy-20,20-difluoro-NADþ
from 2-Deoxyribonolactone
synthesized as well as 20-fluororibo-NMN and NADþ 33, 34
(eight steps 13% yield from 7).
method.35 It appears that the additional fluorine at the
2-position deactivates the mesylate from nucleophilic
chlorination which occurs for the arabino- and ribofluoro
analogues (Schemes 10 and 11). Reaction with nicotina-
mide yielded nucleoside 27 in both R and β-configurations.
Subsequent deprotection and HPLC purification provided
β-28 in 38% yield and the R-anomer in 50% yield. Poor
stereochemical control from the mesylate is known, with
Conclusions
We have developed general methods of achieving diaster-
eoselective electrophilic fluorination of 2-deoxyribonolac-
tone to produce each of the corresponding R-fluoro-
substituted isomers (Scheme 13). Fluorinated furanoses,
appropriate for nucleoside synthesis, are made in increased
yield and efficiency versus previously reported methods with
the advantage of avoiding use of DAST. In addition, con-
solidation and synthetic brevity are achieved with complete
control of diastereoselectivity of fluorination. In the case of
2-deoxy-2,2-difluororibofuranose, the yield almost doubles
the previously reported yield (47% versus 25%) while the
number of required synthetic steps decreases from five to
three (Scheme 13). Moreover, the synthesis here is diaster-
eochemically pure. With respect to the ribofluoro deriva-
tives, we found that we could control stereochemistry into
the difficult ribo configuration by utilization of an R-silyl
group. The corresponding ribofuranose was made in three
steps (with one recycle), slashing in half the number of steps
from the previous shortest synthesis while increasing yield by
32% (from 4% to 36%, Scheme 13).
Central to the success of our strategy was the ability of
bulky silyl protecting groups at the 3 and 5 positions of the
lactone to attenuate the tendency of the lactone enolate to
undergo elimination prior to reacting with the electrophilic
fluorinating agent. The silyl groups could have forced late-
stage manipulations of the protecting groups for synthesis of
nucleosides, but we found that the silyl-protected lactols
were easily activated to chloro sugars (in arabino and ribo),
and in the arabino case, only the R-isomer was generated,
allowing preferential synthesis of the β-nicotinamide sub-
stituted nucleoside. The gem-difluoro sugar was activated
to mesylate.35 We anticipate that the methods reported here
will improve and simplify synthetic access to a variety of
20-fluorosubstituted nucleosides, particularly the 20-deoxy-
20-fluororibofuranoside derivatives. In addition, we are explor-
ing other types of diastereocontrolled electrophilic modifica-
tion of 2-deoxyribonolactones and are investigating
introduction of chlorine, nitrogen, and other heteroatoms.
few preferable alternatives.35 NOEs between the H1 and
0
0
0
0
H4 in the β-nucleoside and NOEs between the H1 and H3
0
and nicotinamide H2 and H4 for the R isomer confirmed
the stereochemistries.
The complete syntheses of the nicotinamide-substituted
mononucleotides and dinucleotides were straightforward,
and only the syntheses of the difluoronucleotide and difluor-
odinucleotide are described. The monofluoro derivatives
were prepared by similar methods, and the preparation of
these compounds is described in the Experimental Section
and in Schemes 10 and 11. 1-(2-Deoxy-2,2-difluororibosyl)
nicotinamide 28 (β-isomer) was phosphorylated with POCl3
in trimethylphosphate in the presence of 6-methylnico-
tinamide, a hindered weak base which controls acidity, to
yield 20-deoxy-20,20-difluoronicotinamide mononucleotide
(20-deoxy-20,20-difluoro-NMN) 29 in 78% isolated yield
(Scheme 12). Although previously unstudied, we found that
the 2-fluoro-NMN compounds could be adenylated enzy-
matically with yeast nicotinamide mononucleotide adenylyl-
transferase46 (NMNAT-1, see the Supporting Information).
In this case, reaction with ATP furnished 20-deoxy-20,20-
difluoro-NADþ 30 in 90% yield versus ATP, which was
limiting, with recovery of unreacted 29. These steps com-
plete the first reportþed syntheses of 20,20-difluoro-NMN and
20,20-difluoro-NAD . The dinucleotide was completed in 8
steps with 12.4% overall yield from 2-deoxyribonolactone
10. Similarly, with these methods 20-fluoroarabino-NMN
and NADþ 31, 32 (seven steps, 26% yield from 10) were
(46) Natalini, P.; Ruggieri, S.; Raffaelli, N.; Magni, G. Biochemistry
1986, 25, 3725–3729.
(47) Fox, J. J.; Hoffer, M.; Wempen, I.; Yung, N. C. J. Am. Chem. Soc.
1961, 83, 4066–4070.
J. Org. Chem. Vol. 74, No. 16, 2009 5785

Cen and Sauve
JOCFeatured Article
We expect to report the chemical and biochemical properties of
the 20-fluorinated NADþ derivatives in future publications.
-4.9, -4.8, 17.9, 18.2, 25.7, 25.8, 39.0, 62.5, 69.6, 88.1, 175.8;
HRMS (ESI) calcd for C17H36O4Si2 360.2152, found 360.2155.
2-Deoxy-2-fluoro-3,5-di-O-(tert-butyldimethylsilyl)-D-ribono-
lactone (8). To a flame-dried 100 mL round-bottom flask were
added 7 (1.8 g, 5 mmol) and NFSi (2.36 g, 7.5 mmol) in 20 mL of
anhydrous THF. The solution was cooled to -78 °C, and 6.5 mL
(6.5 mmol) of a 1 M solution of LiHMDS in THF was added
dropwise over a period of 10 min. This was allowed to stir at
-78 °C for an additional 1 h and was quenched by saturated
NH4Cl. The mixture was allowed to warm to room temperature,
the water layer was extracted by ethyl acetate (3ꢀ10 mL), and
the organic layers were combined, washed with saturated NaH-
CO3, water, and brine, dried over anhydrous Na2SO4, and
concentrated in vacuo. The crude product was purified by flash
chromatography (hexanes/ethyl acetate 20:1) to afford both 8
(1.1 g, 29 mmol, 58%) and 9 (0.5 g, 1.9 mmol, 38%) also as a
Experimental Section
2-deoxy-3,5-di-O-(p-chlorobenzoyl)-D-ribonolactone (5). A sol-
ution of methyl 2-deoxy-3,5-di-O-(p-chlorobenzoyl)-D-ribofur-
anoside47 (500 mg, 1.18 mmol) in 20 mL of 80% acetic acid
aqueous solution and 2 mL of 10% aqueous HCl was heated to
reflux for 1 h and then cooled to room temperature. Water was
added, and the organic phase was washed with water, saturated
NaHCO3 solution, and brine and dried over anhydrous Na2SO4.
Solvent was removed under vacuo, and crude product was
dissolved in 10 mL of CH2Cl2; to this solution was added PCC
(294 mg, 1.36 mmol). The reaction was stirred at room tempera-
ture and monitored by TLC; once it was completed, PCC was
filtered off, and filtrate was concentrated and purified by column
chromatography (hexanes/ethyl acetate 4:1) to afford 5 (190 mg,
0.46 mmol, 40% yield for two steps) as a white solid: mp=89-
90 °C; 1H NMR (CDCl3, 400 MHz) δ (ppm) 0.2.81 (dd, J=2.1,
18.8 Hz, 1H), 3.12 (dd, J=7.5, 18.9 Hz, 1H), 4.60 (dd, J=3.7,
12.3 Hz, 1H), 4.68 (dd, J=3.8, 12.3 Hz, 1H), 4.92 (m, 1H), 5.58
(dt, J=1.8, 7.5 Hz, 1H), 7.42 (m, 4H), 7.93 (m, 4H); 13C NMR
(125 MHz, CDCl3) δ (ppm) 34.9, 63.9, 71.9, 82.2, 127.0, 127.4,
129.09, 129.11, 131.0, 131.2, 140.3, 140.6, 164.98, 165.02, 173.5;
HRMS (ESI) calcd for C19H14Cl2O6 408.0167, found 408.017.
(S)-4-Hydroxymethyl-2-buten-4-olide (6). To a flame-dried
round-bottom flask were added 5 (155 mg, 0.38 mmol) and
NFSi (120 mg, 0.38 mmol) in 5 mL of anhydrous THF. The
solution was cooled to -78 °C, and LiHMDS in THF solution
(0.456 mL, 0.456 mmol) was added dropwise. The reaction
mixture was allowed to stir at -78 °C for an additional 1 h
and then quenched with saturated NH4Cl solution. The organic
layer was washed with saturated NaHCO3 solution, water, and
brine and dried over anhydrous Na2SO4. Column chromatog-
raphy (hexanes/ethyl acetate 4:1-2:1) gave 6 (62 mg, 0.24 mmol,
1
white solid. 8: mp=49-50 °C; H NMR (CDCl3, 400 MHz)
δ (ppm) 0.060 (s, 3H), 0.065 (s, 3H), 0.107 (s, 3H), 0.128 (s, 3H),
0.87 (s, 9H), 0.89 (s, 9H), 3.76 (dd, J=2.5, 12.4 Hz, 1H), 3.95 (dt,
J=2.1, 12.4 Hz, 1H), 4.10 (dt, J=2.0, 7.7 Hz, 1H), 4.70 (dt, J=
7.8, 18.9 Hz, 1H), 5.09 (dd, J=8.0, 51.7 Hz, 1H); 13C NMR
(100 MHz, CDCl3) δ (ppm) -5.5, -5.4, -5.2, -4.8, 17.9, 18.2,
25.5, 25.7, 59.4, 71.3, 71.5, 80.6, 80.7, 91.3, 93.3, 168.5, 168.7;
HRMS (ESI) calcd for C17H35FO4Si2 378.2058, found 378.206.
4-[(tert-Butyldimethylsiloxy)methyl]-4-fluoro-2-buten-4-olide
(9): mp=85-87 °C; 1H NMR (CDCl3, 500 MHz) δ (ppm) 0.045
(s, 3H), 0.057 (s, 3H), 0.85 (s, 9H), 3.86 (dd, J=11.3, 15.9 Hz,
1H), 4.04 (dd, J=7.8, 11.3 Hz, 1H), 6.25 (d, J=5.7 Hz, 1H), 7.33
(d, J=5.7 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ (ppm) -5.62,
-5.58, 18.1, 25.6, 63.5, 63.8, 114.4, 116.3, 125.01, 125.05, 149.7,
149.8, 168.39, 168.41; HRMS (ESI) calcd for C11H19FO3Si
246.1087, found 246.1085.
2-Deoxy-3,5-di-O-(triisopropylsilyl)-D-ribonolactone (10).
2-Deoxy-D-ribonolactone (1.1 g, 8.3 mmol) was dissolved in
10 mL of anhydrous DMF, and to this solution were added
imidazole (3.4 g, 50 mmol) and triisopropylsilyl chloride (6.4 g,
33 mmol). The resulting solution was stirred at room tempera-
ture for 24 h and quenched by addition of water. The water layer
was extracted with ethyl acetate (3 ꢀ 10 mL), and the organic
layers were combined, washed with brine, and dried over
anhydrous Na2SO4. The crude product was concentrated in
vacuo. Column chromatography (hexanes/ethyl acetate 25:1-
20:1) provided 10 (3.4 g, 7.7 mmol, 92%) as a colorless oil:
1H NMR (CDCl3, 400 MHz) δ (ppm) 1.05 (stack, 42H), 2.41
(dd, J=1.9, 17.6 Hz, 1H), 2.86 (dd, J=6.6, 17.6 Hz, 1H), 3.83
(dd, J=2.6, 9.4 Hz, 1H), 3.91 (dd, J=3.0, 11.4 Hz, 1H), 4.39
(s, 1H), 4.65 (d, J = 6.5, 1H); 13C NMR (125 MHz, CDCl3)
δ (ppm) 11.9, 12.1, 18.0, 39.7, 63.4, 70.1, 88.9, 176.1; HRMS
(ESI) calcd for C23H48O4Si2 444.3091, found 444.3094. MM2
calculations on the lithium enolate of 10 were performed using
CHEM3D version 10.0 with a minimized energy to minimum
rms gradient of 0.100.
2-Deoxy-2-fluoro-3,5-di-O-(triisopropylsilyl)-D-ribonolactone
(11). Compound 11 was obtained according to the fluorination
procedure to synthesize 8, using 10 (2 g, 4.5 mmol) as the starting
material. Column chromatography (hexanes/ethyl acetate 30:1)
provided 11 (1.5 g, 3.25 mmol, 72%) as a white solid: mp=38-
39 °C; 1H NMR (CDCl3, 400 MHz) δ (ppm) 1.07 (stack, 42H),
3.91 (dd, J=2.4, 12.1 Hz, 1H), 4.08 (dt, J=2.1, 12.1 Hz, 1H),
4.16 (dt, J=2.1, 7.0 Hz, 1H), 4.92 (dt, J=7.2, 18.8 Hz, 1H), 5.10
(dd, J=7.4, 51.3 Hz, 1H); C NMR (100 MHz, CDCl3) δ (ppm)
11.9, 12.1, 17.7, 17.80, 17.84, 17.86, 60.3, 71.6, 71.8, 81.8, 81.9,
91.7, 93.7, 168.6, 168.8; NOE identified between H2 and H4 in
NOESY; HRMS (ESI) calcd for C23H47FO4Si2 462.2997, found
462.2993.
1
65%) as a white solid: mp=115-116 °C; H NMR (CDCl3,
400 MHz) δ (ppm) 4.55 (dd, J=4.8, 12.0 Hz, 1H), 4.61 (dd, J=
3.7, 12.1 Hz, 1H), 5.34 (m, 1H), 6.22 (dd, J=2.1, 5.8 Hz, 1H),
7.40 (d, J=8.6 Hz, 2H), 7.48 (dd, J=1.5, 5.7 Hz, 1H), 7.91 (d, J=
8.6 Hz, 2H); 13C NMR (125 MHz, CDCl3) δ (ppm) 63.1, 76.9,
80.8, 123.6, 127.5, 128.91, 128.96, 131.1, 131.6, 140.2, 152.1,
165.2, 172.1; HRMS (ESI) calcd for C12H9ClO4 252.0189, found
252.0188.
2-Deoxy-3,5-di-O-(tert-butyldimethylsilyl)-D-ribonolactone (7).
To a solution of 2-deoxy-D-ribose (1.0 g, 7.45 mmol) in 6 mL of
water was added Br2 (2 mL). The flask was sealed, and the
contents were stirred at room temperature for 5 days. The
resulting mixture was neutralized by adding silver carbonate
until the pH was 7. The mixture was filtered, and the filtrate was
concentrated under reduced pressure to yield 2-deoxyribono-
lactone as a yellow oil. Without further purification, the crude
product was dissolved in 20 mL of anhydrous DMF, and
imidazole (2.53 g, 37.3 mmol) and tert-butyldimethylsilyl chlor-
ide (4.5 g, 29.8 mmol) were added. The resulting solution was
stirred at room temperature for 24 h and quenched by addition
of water. The water layer was extracted with ethyl acetate (3 ꢀ
10 mL), and the organic layers were combined, washed with
brine, and dried over anhydrous Na2SO4. The crude product
was concentrated in vacuo. Flash chromatography (hexanes/
ethyl acetate 20:1) afforded 7 (3.2 g, 8.9 mmol, 89% yield after
two steps) as a white solid: mp=72-73 °C; 1H NMR (CDCl3,
400 MHz) δ (ppm) 0.038 (s, 3H), 0.051 (s, 3H), 0.062 (s, 6H),
0.085 (s, 18H), 2.36 (dd, J=2.6, 17.7 Hz, 1H), 2.79 (dd, J=6.7,
17.7 Hz, 1H), 3.73 (dd, J=2.5, 11.5 Hz, 1H), 3.78 (dd, J=3.4,
11.5 Hz, 1H), 4.30 (dd, J=2.5, 5.2 Hz, 1H), 4.48 (dt, J=2.3,
6.6 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ (ppm) -5.7, -5.5,
2-Deoxy-2-fluoro-3,5-di-O-(triisopropylsilyl)-D-arabinofuranose
(12). Compound 11 (200 mg, 0.43 mmol) was dissolved in 1.5 mL
of anhydrous toluene and cooled to -78 °C. To this solution was
5786 J. Org. Chem. Vol. 74, No. 16, 2009

Cen and Sauve
JOCFeatured Article
added 3.02 mL of DIBAL-H in THF solution (3.02 mmol). The
reaction mixture was held at -78 °C at all times. Two hours
later, the mixture was quenched with methanol at -20 °C, and
additional cold methanol was added. The mixture was then
allowed to warm slowly to room temperature and was washed
with 0.1 M HCl. The aqueous layer was extracted with ether,
and the combined organic layer was washed with saturated
NaHCO3, water, and brine, dried over anhydrous Na2SO4, and
concentrated in vacuo. Column chromatography (hexanes/ethyl
acetate 15:1) afforded 12 (181 mg, 0.39 mmol, 91%) as a color-
less oil: 1H NMR (CDCl3, 400 MHz) δ (ppm) 1.07 (stack, 54H),
3.46 (d, J=10.9 Hz, 1H), 3.60 (m, 1.27H), 3.78 (m, 1.57H), 3.95
(q, J=3.7 Hz, 0.28H), 4.32 (m, 1H), 4.49 (dd, J=0.9, 12.6 Hz,
1H), 4.63 (t, J=4.0 Hz, 0.14H), 4.67 (t, J=4.0 Hz, 0.14 H), 4.81
(t, J=4.1 Hz, 0.28H), 4.83 (dd, J=0.8, 50.2 Hz, 1H), 5.31 (m,
0.3H), 5.35 (t, J=10.1 Hz, 1H); 13C NMR (125 MHz, CDCl3)
δ (ppm) 12.0, 12.2, 18.0, 63.2, 75.4, 75.5, 75.6, 87.6, 87.7, 95.9,
97.5, 97.7, 98.8, 98.9, 100.8, 101.0; HRMS (ESI) calcd for
C23H49FO4Si2 464.3153, found 464.3162.
for C17H35BrO4Si2 438.1257, found 438.1261. Ribono-15: 1H
NMR (CDCl3, 500 MHz) δ (ppm) 0.058 (s, 3H), 0.062 (s, 3H),
0.116 (s, 3H), 0.172 (s, 3H), 0.87 (s, 9H), 0.88 (s, 9H), 3.79 (dd,
J=3.0, 12.0 Hz, 1H), 3.92 (dd, J=3.3, 12.0 Hz, 1H), 4.21 (m,
1H), 4.37 (d, J=6.9 Hz, 1H), 4.67 (dd, J=6.1, 6.6 Hz, 1H);
13C NMR (125 MHz, CDCl3) δ (ppm) -5.5, -5.4, -5.0, -4.1,
17.8, 18.2, 25.6, 25.7, 46.1, 60.2, 75.6, 85.3, 170.1; HRMS (ESI)
calcd for C17H35BrO4Si2 438.1257, found 438.1258.
(2R)-2-Deoxy-2-bromo-2-fluoro-3,5-di-O-(tert-butyldimethyl-
silyl)-D-ribonolactone (16). Compound 16 was obtained accord-
ing to the fluorination procedure to synthesize 8, using 15
(400 mg, 0.91 mmol) as the starting material. Column chroma-
tography (hexanes/ethyl acetate 30:1) provided 16 (230 mg,
0.5 mmol, 55%) as a pale yellow oil: 1H NMR (CDCl3,
500 MHz) δ (ppm) 0.058 (s, 3H), 0.065 (s, 3H), 0.13 (s, 3H),
0.17 (s, 3H), 0.86 (s, 9H), 0.93 (s, 9H), 3.77 (dd, J=1.9, 12.7 Hz,
1H), 4.00 (m, 2H), 4.53 (dd, J=8.0, 15.4 Hz, 1H); 13C NMR
(125 MHz, CDCl3) δ (ppm) -5.5, -5.4, -5.2, -4.6, 18.0, 18.2,
25.5, 25.7, 58.3, 72.0, 72.2, 80.6, 80.7, 98.2, 100.5, 165.6, 165.8;
HRMS (ESI) calcd for C17H34BrFO4Si2 456.1163, found
456.1178.
2-Deoxy-2,2-difluoro-3,5-di-O-(triisopropylsilyl)-D-ribonolac-
tone (13). Compound 13 was obtained according to the fluor-
ination procedure to synthesize 8, using 11 (92 mg, 0.2 mmol) as
the starting material. Column chromatography (hexanes/ethyl
acetate 40:1) provided 13 (68 mg, 0.14 mmol, 71%) as a pale
Debromination of 16 Yielding 8. Compound 16 (60 mg, 0.13
mmol), tributyltin hydride (83 mg, 0.28 mmol), and AIBN
(3 mg, 0.018 mmol) were dissolved in 1 mL of toluene and
stirred at 90 °C for 24 h. Solvent was evaporated, and the residue
was dissolved in acetonitrile and washed with hexanes to remove
organotin compounds. The solvent was again concentrated in
vacuo, and an 1H NMR spectrum identified it as compound 8
described previously.
1
yellow oil: H NMR (CDCl3, 400 MHz) δ (ppm) 1.08 (stack,
42H), 3.96 (dd, J=2.4, 12.0 Hz, 1H), 4.08 (dt, J=2.6, 12.1 Hz,
1H), 4.31 (m, 1H), 4.76 (dt, J=6.0, 11.2 Hz, 1H); 13C NMR
(100 MHz, CDCl3) δ (ppm) 11.4, 11.45, 11.8, 12.1, 17.2, 17.3,
17.36, 17.43, 59.6, 68.3, 68.5, 68.6, 68.7, 82.5, 82.6, 109.9, 112.4,
115.0, 163.3, 163.6, 164.0; HRMS (ESI) calcd for C23H46F2O4-
Si2 480.2903, found 480.2901.
2-Deoxy-2-trimethylsilyl-3,5-di-O-(tert-butyldimethylsilyl)-D-
arabinolactone (17). To a solution of 7 (1 g, 2.78 mmol) and
triethylamine (1.68 g, 16.68 mmol) in 28 mL of CH2Cl2 at 0 °C
was added TMSOTf (1.85 g, 8.34 mmol) dropwise. The solution
was stirred at this temperature for another 2 h and then
quenched with saturated NH4Cl. The mixture was allowed to
warm to room temperature, the water layer was extracted by
CH2Cl2 (3ꢀ10 mL), and the organic layers were combined,
washed with saturated NaHCO3, water, and brine, dried over
anhydrous Na2SO4, and concentrated in vacuo. The crude
product was purified by flash chromatography (hexanes/ethyl
acetate 25:1) to afford unreacted 7 (240 mg, 0.67 mmol) and 17
(850 mg, 1.97 mmol, 71%) as a white solid: mp=70-71 °C; 1H
NMR (CDCl3, 400 MHz) δ (ppm) 0.067 (s, 6H), 0.074 (s, 6H),
0.19 (s, 9H), 0.85 (s, 9H), 0.89 (s, 9H), 2.17 (d, J=2.5 Hz), 3.50
(dd, J=7.7, 10.7 Hz, 1H), 3.76 (dd, J=7.1, 12.2 Hz, 1H), 4.27
(m, 1H), 4.45 (t, J=2.2 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ
(ppm) -5.4, -5.3, -4.7, -4.2, -1.7, -0.8, 17.7, 18.4, 25.6, 25.8,
25.90, 25.94, 42.2, 62.2, 71.8, 87.4, 177.7; HRMS (ESI) calcd for
C20H44O4Si3 432.2547, found 432.2553.
2-Deoxy-2-fluoro-3,5-di-O-(tert-butyldimethylsilyl)-D-ribono-
lactone (18). Compound 18 was obtained according to the
fluorination procedure to synthesize 8, using 17 (780 mg, 1.81
mmol) as the starting material. Column chromatography (hex-
anes/ethyl acetate 30:1-10:1) provided both 18 (226 mg, 0.60
mmol, 33%) and 7 (400 mg,1.11 mmol, 61%) as white solid:
mp =75-77 °C; 1H NMR (CDCl3, 400 MHz) δ (ppm) 0.046 (s,
3H), 0.061 (s, 3H), 0.087 (s, 6H), 0.86 (s, 9H), 0.87 (s, 9H), 3.77
(dd, J=1.7, 11.9 Hz, 1H), 3.84 (dd, J=2.6, 12.0 Hz, 2H), 4.37 (d,
J=2.1 Hz, 1H), 4.43 (d, J=5.2 Hz, 1H), 5.21 (dd, J=5.3, 50 Hz,
1H); 13C NMR (125 MHz, CDCl3) δ (ppm) -5.7, -5.6, -5.3,
-4.9,-4.8, 18.2, 18.3, 25.6, 25.8, 62.3, 70.3, 70.4, 71.8, 84.2, 85.8,
86.4, 88.1, 171.1; HRMS (ESI) calcd for C17H35FO4Si2
378.2058, found 378.2059.
2-Deoxy-2,2-difluoro-3,5-di-O-(triisopropylsilyl)-D-ribofuranose
(14). Compound 14 was obtained according to the reduction
procedure to synthesize 12, using 13 (160 mg, 0.33 mmol) as the
starting material. Column chromatography (hexanes/ethyl acet-
ate 10:1) provided 14 (146 mg, 0.3 mmol, 91%) as a colorless oil:
1H NMR (CDCl3, 400 MHz) δ (ppm) 1.04 (stack, 78H), 3.48 (d,
J=11.3 Hz, 1H), 3.67 (m, 1.8H), 3.81 (m, 1.8H), 3.88 (dt, J=2.1,
11.2 Hz, 1H), 4.02 (m, 0.8H), 4.24 (m, 1H), 4.39 (dt, J = 2.0,
10.7 Hz, 1H), 4.67 (m, 0.8H), 5.02 (dd, J=5.1, 9.8 Hz, 0.8H), 5.11
(dd, J=6.1, 11.3 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ (ppm)
11.87, 11.90, 12.1, 12.3, 17.72, 17.73, 17.78, 17.83, 17.86, 17.89,
62.2, 62.34, 62.36, 69.6, 69.7, 69.8, 70.0, 71.9, 72.0, 72.1, 72.3,
83.95, 84.02, 85.3, 95.3, 95.5, 95.6, 95.8, 96.0, 96.2, 96.3, 96.5,
119.5, 120.1, 121.5, 121.6, 122.1, 123.6, 124.2; HRMS (ESI) calcd
for C23H48F2O4Si2 482.3059, found 482.3054.
2-Deoxy-2-bromo-3,5-di-O-(tert-butyldimethylsilyl)-D-ribono,
arabino-lactones (15). To a solution of 7 (180 mg, 0.5 mmol) and
triethylamine (303 mg, 3 mmol) in 6 mL of CH2Cl2 at 0 °C was
added TMSOTf (333 mg, 1.5 mmol), and the solution was
stirred at this temperature for 30 min. A solution of NBS
(134 mg, 0.75 mmol) in 1.5 mL of CH2Cl2 was added, and
stirring was continued for 1 h at 0 °C. The reaction mixture was
poured into saturated NaHCO3 solution and extracted with
CH2Cl2 (3 ꢀ 5 mL). The combined organic layer was washed
with brine, dried over anhydrous Na2SO4, and concentrated
in vacuo. Flash chromatography (hexanes/ethyl acetate 35:1)
afforded a mixture of two isomers of 15 (120 mg, 0.27 mmol,
55%, 1:1.4, arabino/ribono) as pure pale yellow liquid. A small
amount of this mixture was separated to obtain the pure
compounds. The stereochemistry was determined by an ob-
served NOE between the 2- and 4-protons of the arabino isomer.
Arabino-15: 1H NMR (CDCl3, 500 MHz) δ (ppm) 0.041 (s, 3H),
0.053 (s, 3H), 0.10 (s, 3H), 0.13 (s, 3H), 0.85 (s, 9H), 0.90 (s, 9H),
3.76 (dd, J=2.0, 12.2 Hz, 1H), 3.93 (dd, J=2.2, 12.2 Hz, 1H),
2-Deoxy-2-fluoro-3,5-di-O-(tert-butyldimethylsilyl)-D-ribofur-
anose (19). Compound 19 was obtained according to the reduc-
tion procedure to synthesize 12, using 18 (100 mg, 0.22 mmol) as
the starting material. Column chromatography (hexanes/ethyl
acetate 15:1) provided 19 (95 mg, 0.205 mmol, 95%) as a
4.31 (m, 1H), 4.39 (t, J=5.0 Hz, 1H), 4.47 (d, J=5.7 Hz, 1H); 13
C
NMR (125 MHz, CDCl3) δ (ppm) -5.6, -5.5, -5.0, -4.8, 18.1,
18.2, 25.6, 25.8, 46.1, 60.3, 68.8, 85.0, 170.9; HRMS (ESI) calcd
J. Org. Chem. Vol. 74, No. 16, 2009 5787

Cen and Sauve
JOCFeatured Article
colorless oil: 1H NMR (CDCl3, 500 MHz) δ (ppm) 1.02 (stack,
96 H), 3.28 (d, J=7.3 Hz, 2.2H), 3.68 (dd, J=3.7, 11 Hz, 1H),
3.76 (dd, J=2.7, 11 Hz, 1H), 3.80 (dd, J=1.8, 11 Hz, 2.2H), 3.92
(dd, J=2.2, 11 Hz, 2.2H), 4.09 (dt, J=1.9, 6.7 Hz, 2.2H), 4.14
(dd, J=1.2, 12.3 Hz, 1H), 4.23 (s, 1H), 4.49 (dd, J=1.7, 2.8 Hz,
1H), 4.59 (dd, J=3.7, 53.4 Hz, 2.2H), 4.71 (ddd, J=3.8, 10.4,
23.5 Hz, 2.2H), 4.85 (dt, J=4.4, 51.7 Hz, 1H), 5.24 (dd, J=4.2,
8 Hz, 1H), 5.27 (t, J = 6.1 Hz, 2.2H); 13C NMR (125 MHz,
CDCl3) δ (ppm) 11.83, 11.85, 11.9, 12.2, 17.71, 17.72, 17.81,
17.85, 17.87, 17.88, 61.8, 63.5, 69.8, 69.9, 71.9, 72.0, 84.0, 85.67,
85.70, 87.6, 89.2, 93.5, 95.0, 95.7, 95.9, 99.1, 99.3; HRMS (ESI)
calcd for C17H37FO4Si2 380.2214, found 380.2212.
1-Chloro-2-deoxy-2-fluoro-3,5-di-O-(triisopropylsilyl)-D-ara-
binofuranose (20). Compound 12 (40 mg, 0.086 mmol) was
dissolved in 0.5 mL of CH2Cl2 and triethylamine (12.2 mg,
0.12 mmol). To this solution was added at 0 °C methanesulfonyl
chloride (11.5 mg, 0.1 mmol). After 3 h of stirring under argon
at room temperature, the mixture was evaporated in vacuo, and
the residue was taken up in ethyl acetate. The solution was
washed with saturated NaHCO3, followed by 1 M HCl, water,
and brine. Solvent was concentrated under reduced pressure to
afford 20 (40 mg, 0.086 mmol, almost quantitatively) as a
yellow liquid: 1H NMR (CDCl3, 400 MHz) δ (ppm) 1.08 (stack,
42H), 3.89 (dd, J=3.7, 11.7 Hz, 1H), 3.96 (dd, J=2.9, 11.7 Hz,
1H), 4.30 (dd, J=3.4, 8.3 Hz, 1H), 4.58 (dd, J=5.1, 14.8 Hz,
1H), 5.12 (d, J=51.7 Hz, 1H), 6.15 (d, J=12.5 Hz, 1H); 13C
NMR (125 MHz, CDCl3) δ (ppm) 11.9, 12.0, 17.80, 17.84, 17.9,
61.4, 75.1, 75.4, 88.61, 88.64, 95.3, 95.6, 103.8, 105.3; HRMS
(ESI) calcd for C23H48ClFO3Si2 482.2815, found 482.2822.
1-(20-Deoxy-20-fluoro-30,50-di-O-(triisopropylsilyl)arabinofur-
anosyl)nicotinamide (21). Compound 20 (25 mg, 0.052 mmol)
and nicotinamide (15 mg, 0.12 mmol) were dissolved in 1 mL
of CH2Cl2. To this solution at 0 °C was added an ice-cold
solution of nicotinamide (15 mg, 0.12 mmol) and AgSbF6
(36 mg, 0.104 mmol) in 1.5 mL of acetonitrile. The reaction
mixture was kept at room temperature overnight. Solvent was
evaporated under reduce pressure, and the residue was redis-
solved in methanol and passed through a short pad of Celite.
Concentrated crude product (which contained a mixture of
R and β isomers) was examined by NMR and used for the next
step without further purification.
1-(20-Deoxy-20-fluoroarabinofuranosyl)nicotinamide (22). To
a solution of 21 (25 mg, 0.044 mmol) in 1 mL of DMF were
added acetic acid (10.6 mg, 0.176 mmol) and tetramethylam-
monnium fluoride (16.4 mg, 0.176 mmol). The reaction was
stirred at room temperature overnight and then was concen-
trated and purified by HPLC on a Waters RP-18 XBridge
PrepShield 19 ꢀ 50 mm column (solvent was 20 mM ammonium
acetate, compounds were eluted at a flow rate of 2 mL/min) to
afford 22 (β-isomer, tR=8 min, 7 mg, 0.027 mmol, 62%) and
R-isomer (tR = 6.7 min, 2 mg, 0.008 mmol, 18%). β-Isomer:
1H NMR (D2O, 400 MHz) δ (ppm) 3.88 (dd, J=4.8, 13.0 Hz,
1H), 4.0 (dd, J=1.9, 12.9 Hz, 1H), 4.28 (dd, J=4.7, 8.3 Hz, 1H),
4.51 (dt, J=5.5, 17.6 Hz, 1H), 5.5 (dt, J=4.6, 51.3 Hz, 1H), 6.68
(dd, J=4.7, 9.8 Hz, 1H), 8.23, (t, J=6.6 Hz, 1H), 8.95 (d, J=8.1 Hz,
1H), 9.18 (d, J = 6.2 Hz, 1H), 9.57 (s, 1H); 13C NMR
(125 MHz, D2O) δ (ppm) 59.4, 71.1, 71.3, 85.07, 85.11, 93.9,
94.0, 94.1, 95.6, 128.1, 133.8, 141.5, 143.8, 146.1, 165.7; HRMS
(ESI) calcd for C11H13FN2O4 256.0859, found 256.0865.
(1R,2R)-1-Chloro-2-deoxy-2-fluoro-3,5-di-O-(tert-butyldimethyl-
silyl)-D-ribofuranose (23). Compound 23 was obtained according to
the chlorination procedure to synthesize 20 using 19 (10 mg, 0.021
mmol) as the starting material to afford 23 (10 mg, 0.021 mmol,
almost quantitatively) as yellow liquid: 1H NMR (CDCl3,
500 MHz) δ (ppm) 1.02 (stack, 51 H), 3.86 (stack, 3.4H), 4.02
(dd, J=1.8, 11.8 Hz, 0.7H), 4.10 (m, 0.7H), 4.31 (d, J=2.1 Hz, 1H),
4.55 (quintet, J=3 Hz, 1H), 4.82 (dt, J=4.7, 49 Hz, 1H), 4.95 (dd,
J=3.3, 45 Hz, 0.7H), 6.06 (d, J=11.1 Hz, 0.7H), 6.21 (d, J=4.1 Hz,
1H); 13C NMR (125 MHz, CDCl3) δ (ppm) 11.8, 11.9, 12.1, 12.2,
17.77, 17.81, 17.84, 17.86, 17.95, 17.96, 31.5, 52.5, 61.6, 62.3, 68.6,
68.8, 85.8, 87.9, 88.7, 89.6, 92.8, 93.0, 93.4, 93.7, 95.6, 97.2; HRMS
(ESI) calcd for C17H36ClFO3Si2 398.1876, found 398.1880.
1-(20-Deoxy-20-fluoro-30,50-di-O-(tert-butyldimethylsilyl)-D-ri-
bofuranosyl)nicotinamide (24). Compound 23 (11 mg, 0.026
mmol) and nicotinamide (8 mg, 0.065 mmol) were dissolved in
1 mL of CH2Cl2. To this solution at 0 °C was added an ice-
cold solution of nicotinamide (8 mg, 0.065 mmol) and AgSbF6
(8.9 mg, 0.026 mmol) in 1.5 mL of acetonitrile. The reaction
mixture was kept at room temperature overnight. Solvent was
evaporated under reduced pressure, and the residue was redis-
solved in methanol and passed through a short pad of Celite.
Concentrated crude product was examined by NMR and used
for the next step without further purification.
1-(2-Deoxy-2-fluoro-D-ribofuranosyl)nicotinamide (25). Com-
pound 25 was obtained according to the deprotection procedure
to synthesize 22, using 24 (12.6 mg, 0.026 mmol) as the starting
material. Crude product was purified by HPLC on a Waters RP-
18 XBridge PrepShield 19 ꢀ 50 mm column (solvent was 20 mM
ammonium acetate, compounds were eluted at a flow rate of
2 mL/min) to afford 25 (β-isomer, tR=14.5 min, 3 mg, 0.012
mmol, 45%) and R-isomer (tR=10.6 min, 3.1 mg, 0.012 mmol,
46%). β-Isomer: 1H NMR (D2O, 600 MHz) δ (ppm) 3.75 (dd, J=
2.4, 13.2 Hz, 1H), 3.97 (dd, J=2.4, 13.8 Hz, 1H), 4.31 (m, 1H),
4.35 (m, 1H), 5.20 (dd, J=4.2, 49.2 Hz, 1H), 6.47 (d, J=14.4 Hz,
1H), 8.10 (t, J=7.8 Hz, 1H), 8.80 (dd, J=1.2, 7.8 Hz, 1H), 9.16
(d, J=6.6 Hz, 1H), 9.54 (s, 1H); 13C NMR (125 MHz, D2O)
δ (ppm) 55.15, 55.18, 55.22, 58.8, 67.3, 67.4, 85.8, 94.2, 95.7,
97.2, 97.5, 128.5, 134.1, 140.8, 142.9, 145.9, 165.2; NOESY NOE
0
correlation between sugar H3 and nicotinamide H2; HRMS
(ESI) calcd for C11H13FN2O4 256.0859, found 256.0863.
1
r-Isomer: H NMR (D2O, 500 MHz) δ (ppm) 3.82 (dd, J =
4.5, 10.5 Hz, 1H), 4.0 (dd, J=2.0, 10.5 Hz, 1H), 4.61 (m, 1H), 4.81
(m, 1H), 5.65 (dt, J=4.5, 43.5 Hz, 1H), 6.79 (dd, J=3.5, 8.5 Hz,
1H), 8.29, (dd, J=5.5, 7.0 Hz, 1H), 9.01 (d, J=6.5 Hz, 1H), 9.17
(d, J = 5.5 Hz, 1H), 9.40 (s, 1H); NOESY NOE correlations
0
between sugar H4 and nicotinamide H2, sugar H4 and nicotina-
0
0
mide H4, and sugar H1 and H3 .
0
1-Methylsulfonyl-2-deoxy-2,2-difluoro-3,5-di-O-(triisopropyl-
silyl)-D-ribofuranose (26). Compound 14 (146 mg, 0.3 mmol)
was dissolved in 1.1 mL of CH2Cl2 and triethylamine (42 mg,
0.42 mmol). To this solution was added at 0 °C methanesulfonyl
chloride (41 mg, 0.35 mmol). After 3 h of stirring under argon at
room temperature, the mixture was evaporated in vacuo, and the
residue was taken up in ethyl acetate. The solution was washed
with saturated NaHCO3, followed by 1 M HCl, water, and brine.
Solvent was concentrated under reduced pressure to afford 26
(165 mg, 0.3 mmol, almost quantitatively) as a yellow liquid:
1H NMR (CDCl3, 400 MHz) δ (ppm) 1.07 (stack, 74H), 3.07 (s,
1.9H), 3.08 (s, 3H), 3.83 (dd, J=3.8, 11.4 Hz, 0.64H), 3.89 (m, 2H),
4.00 (m, 1.6H), 4.26 (dd, J=4.0, 8.1 Hz, 1H), 4.47 (dd, J=4.7,
16.5 Hz, 1H), 4.59 (m, 0.64H), 5.83 (d, J=7.0 Hz, 0.64H), 5.92 (d,
J=6.8 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ (ppm) 11.84,
11.85, 12.1, 12.2, 17.67, 17.73, 17.76, 17.79, 17.81, 17.87, 17.91,
40.0, 40.2, 46.3, 61.5, 61.8, 69.0, 69.2, 69.4, 70.98, 71.12, 71.2, 71.4,
84.7, 84.8, 88.0, 99.4, 99.6, 99.9, 100.1, 100.3, 100.5; HRMS (ESI)
calcd for C24H50F2O6SSi2 560.2835, found 560.284.
1-(20-Deoxy-20,20-difluoro-30,50-di-O-(triisopropylsilyl)ribofura-
nosyl)nicotinamide (27). Compound 26 (420 mg, 0.75 mmol) and
nicotinamide (732 mg, 6 mmol) were dissolved in 20 mL of
CH3CN/CH2Cl2 (1:1). To this solution was added TMSOTf
(167 mg, 0.75 mmol) under argon, and the reaction mixture was
allowed to reflux overnight. Solvent was evaporated in vacuo,
andconcentrated crudeproduct was examinedbyNMRand used
for the next step without further purification.
1-(20-Deoxy-20,20-difluororibofuranosyl)nicotinamide (28). Com-
pound 28 was obtained according to the deprotection procedure to
5788 J. Org. Chem. Vol. 74, No. 16, 2009

Cen and Sauve
JOCFeatured Article
synthesize 22, using 27 (440 mg, 0.75 mmol) as the starting
material. Crude product was purified by HPLC on a Waters RP-
18 XBridge PrepShield 19 ꢀ 50 mm column (solvent was 20 mM
ammonium acetate, compounds were eluted at a flow rate of
2 mL/min) to afford 28 (β-isomer, tR = 17.8 min, 78 mg, 0.28
mmol, 38%) and the R-isomer (tR=14.8 min, 103 mg, 0.37 mmol,
50%). β-Isomer: 1H NMR (D2O, 600 MHz) δ (ppm) 3.83 (d, J=
11.4 Hz, 1H), 4.01 (d, J=12.6 Hz, 1H), 4.24 (d, J=7.8 Hz, 1H), 4.46
(dd, J=10.8, 20.4 Hz, 1H), 6.56 (d, J=8.4 Hz, 1H), 8.20 (t, J=6.6
Hz, 1H), 8.92 (d, J=7.8 Hz, 1H), 9.20 (d, J=6.6 Hz, 1H), 9.61 (s,
1H); 13C NMR (125 MHz, D2O) δ (ppm) 58.4, 67.3, 67.5, 67.6,
83.2, 83.3, 93.3, 93.4, 93.5, 93.6, 119.4, 121.7, 124.0, 128.6, 134.3,
141.4, 143.6, 146.9, 165.4; HRMS (ESI) calcd for C11H12F2N2O4
274.0765, found 274.0771. r-Isomer: 1H NMR (D2O, 500 MHz)
δ(ppm) 3.78 (dd, J=4.5, 11.0 Hz, 1H), 3.91 (dd, J=1.5, 11 Hz, 1H),
4.59 (m, 1H), 6.75 (t, J=5.0 Hz, 1H), 8.24, (dd, J=5.0, 6.5 Hz, 1H),
8.98 (d, J=6.5 Hz, 1H), 9.11 (d, J=4.5 Hz, 1H), 9.34 (s, 1H).
20-Deoxy-20,20-difluororibonicotinamide Mononucleotide (29).
To a flame-dried round-bottom flask were added 28 (5 mg, 0.018
mmol), 6-methylnicotinamide (12.4 mg, 0.091 mmol), and
0.5 mL of trimethyl phosphate. At 0 °C, 13.9 mg (0.091 mmol)
of phosphorus oxychloride was added to the reaction mixture.
This solution was stirred at 0 °C for another 2 h. Ice was added to
quench the reaction, and the pH was adjusted to 7 by adding
NaOH solution and phosphate buffer. The crude product
was concentrated and purified by HPLC on a Waters RP-18
XBridge PrepShield 19 ꢀ 50 mm column (solvent was 20 mM
ammonium acetate, compound was eluted at a flow rate of
2 mL/min) to afford 29 (tR=6.2 min, 5 mg, 0.014 mmol, 78%):
1H NMR (D2O, 500 MHz) δ (ppm) 4.02 (dd, J=3.2, 13.4 Hz,
1H), 4.19 (dt, J=2.4, 13.4 Hz, 1H), 4.43 (d, J=8.4 Hz, 1H), 4.65
(stack, 2H), 6.75 (dd, J=2.3, 8.9 Hz, 1H), 8.38 (t, J=6.5 Hz, 1H),
9.10 (d, J=8.2 Hz, 1H), 9.36 (d, J=5.8 Hz, 1H), 9.78 (s, 1H);
13C NMR (125 MHz, D2O) δ (ppm) 59.7, 69.4, 69.5, 69.6, 69.8,
86.80, 86.84, 94.0, 94.2, 94.4, 94.5, 115.9, 118.4, 120.9, 123.4,
128.5, 134.1, 140.9, 143.2, 146.8, 165.4; HRMS (ESI) calcd for
C11H14F2N2O7P 355.0501, found 355.0503.
98.5, 98.8, 99.2, 100.7, 128.3, 133.9, 140.3, 142.6, 145.8, 165.6; MS
(Mþ) calcd 337.06, found 337.45.
20-Deoxy-20-fluoroarabino-NADþ (32). A single reaction (50 μL)
containing 6 mM 31, 2 mM ATP, 10 mM MgCl2, 1 μL of
pyrophosphatase (1 unit), 5 μL of NMNAT-1 (13.5 μM), and
50 mM phosphate buffer (pH ∼7.4) was incubated at 37 °C for 1 h.
The reaction was terminated by addition of 3 μL of 10% TFA and
purified by HPLC on a Waters RP-18 XBridge PrepShield 19ꢀ
50 mm column (solvent was 20 mM ammonium acetate, com-
pound was eluted at a flow rate of 2 mL/min, tR=14.7 min, 95%
versus ATP with recovery of unreacted 32): 1H NMR (D2O,
600 MHz) δ (ppm) 4.41 (m, 1H), 4.46 (stack, 2H), 4.57 (stack,
3H), 4.67 (dd, J=3.6, 5.4 Hz, 1H), 4.80 (dt, J=4.8, 17.4 Hz, 1H),
4.90 (t, J=6 Hz, 1H), 5.71 (dt, J=4.8, 51 Hz, 1H), 6.20 (d, J=6 Hz,
1H), 6.81, (dd, J=4.8, 9.6 Hz, 1H), 8.37, (s, 1H), 8.40 (dd, J=6.6,
7.8 Hz, 1H), 9.05, (d, J=7.8 Hz, 1H), 9.38 (d, J=6.6 Hz, 1H), 9.51
(s, 1H); 13C NMR (125 MHz, D2O) δ (ppm) 63.4, 65.4, 70.3, 70.8,
71.0, 74.1, 83.5, 83.8, 83.9, 86.8, 93.68, 93.73, 93.9, 95.3, 118.4,
128.3, 133.2, 140.0, 141.2, 143.4, 146.03, 148.9, 152.2, 165.1; MS
(Mþ) calcd 666.11, found 666.60.
20-Deoxy-20-fluororibonicotinamide Mononucleotide (33). Com-
pound 33 was obtained according to the phosphorylation proce-
dure to synthesize 29 using 25 (3.5 mg, 0.014 mmol) as the starting
material. Crude product was concentrated and purified by HPLC
on a Waters RP-18 XBridge PrepShield 19ꢀ50 mm column
(solvent was 20 mM ammonium acetate, compound was eluted
at a flow rate of 2 mL/min) to afford 33 (tR=5.9 min, 3.8 mg,
0.011 mmol, 83%): 1H NMR (D2O, 600 MHz) δ (ppm) 3.95 (d,
J=12.6 Hz, 1H), 4.16 (d, J=13.2 Hz, 1H), 4.52 (stack, 2H), 5.40
(d, J=50.4 Hz, 1H), 6.67 (d, J=13.8 Hz, 1H), 8.31 (s, 1H), 9.00 (s,
1H), 9.35 (s, 1H), 9.74 (s, 1H); 13C NMR (150 MHz, D2O) δ (ppm)
55.16, 55.20, 60.1, 60.6, 69.3, 69.4, 86.97, 86.98, 90.0, 91.5, 94.6,
94.8, 127.8, 133.4, 141.0, 143.4, 145.9, 165.7; HRMS (ESI) calcd
for C11H15FN2O7P 337.0595, found 337.0599.
20-Deoxy-20-fluororibo-NADþ (34). A single reaction mixture
(50 μL) containing 3.8 mM 33, 10 mM ATP, 10 mM MgCl2, 1 μL
of pyrophosphatase (1 unit), 10 μL of NMNAT-1 (27 μM), and
50 mM phosphate buffer (pH ∼7.4) was incubated at 37 °C for
1 h. The reaction was terminated by addition of 3 μL of 10%
TFA and purified by HPLC on a Waters RP-18 XBridge
PrepShield 19ꢀ50 mm column (solvent was 20 mM ammonium
acetate, compound was eluted at a flow rate of 2 mL/min, tR=
18.6 min, 94% versus 33): 1H NMR (D2O, 600 MHz) δ (ppm)
3.44 (dd, J=7.2, 12 Hz, 1H), 3.54 (dd, J=4.2, 11.4 Hz, 1H), 4.10
(m, 1H), 4.15 (dd, J= 4.8, 11.4 Hz, 2H), 4.27 (t, J=2.4 Hz, 1H),
4.37 (dd, J=1.8, 14.4 Hz, 1H), 4.40 (t, J=3.6 Hz, 1H), 4.49 (stack,
2H), 5.30, (dt, J=3.0, 51 Hz, 1H), 5.95, (d, J=6.0 Hz, 1H), 6.41
(dd, J=2.4, 13.8 Hz, 1H), 8.11, (s, 1H), 8.13 (dd, J=6.6, 7.8 Hz,
1H), 8.74 (d, J=8.4 Hz, 1H), 9.13 (d, J=6.6 Hz, 1H), 9.32 (s, 1H);
13C NMR (125 MHz, D2O) δ (ppm) 62.5, 63.45, 63.49, 65.36,
65.40, 67.7, 67.8, 70.4, 72.1, 73.9, 83.8, 83.9, 85.1, 85.2, 86.6, 94.1,
95.7, 97.2, 97.5, 102.4, 118.4, 128.7, 133.8, 139.2, 139.8, 140.3,
142.5, 145.99, 146.02, 149.0, 152.8, 155.4, 156.5, 165.0; HRMS
(ESI) calcd for C21H27FN7O13P2 666.1121, found 666.1132.
20-Deoxy-20,20-difluoro-NADþ (30). A single reaction (50 μL)
containing 6 mM 29, 2 mM ATP, 10 mM MgCl2, 1 μL of
pyrophosphatase (1 unit), 5 μL of NMNAT-1 (13.5 μM), and
50 mM phosphate buffer (pH ∼7.4) was incubated at 37 °C for 1 h.
The reaction was terminated by addition of 3 μL of 10% TFA and
purified by HPLC on a Waters RP-18 XBridge PrepShield 19 ꢀ
50 mm column (solvent was 20 mM ammonium acetate, com-
pound was eluted at a flow rate of 2 mL/min, tR=12.7 min, 90%
versus ATP with recovery of unreacted 30): 1H NMR (500 MHz,
D2O) δ (ppm) 4.28 (stack, 2H), 4.36 (m, 1H), 4.42 (s, 1H), 4.55
(stack, 3H), 4.73 (stack, 2H), 6.16 (d, J=5.6 Hz, 1H), 6.69 (d, J=
9.3 Hz, 1H), 8.38 (dd, J=6.5, 7.9Hz, 1H), 8.41(s, 1H), 8.60(s, 1H),
9.04 (d, J=8.2 Hz, 1H), 9.38 (d, J=6.3 Hz, 1H), 9.54 (s, 1H); 13
C
NMR (125 MHz, D2O) δ (ppm) 59.7, 65.18, 65.22, 69.3, 69.49,
69.55, 69.7, 70.3, 74.3, 83.9, 84.0, 86.76, 86.81, 94.0, 94.2, 94.3,
94.5, 118.4, 119.7, 121.8, 123.8, 128.4, 133.9, 140.2, 140.8, 143.1,
146.7, 148.9, 151.6, 154.7, 165.2; HRMS (ESI) calcd for
C21H26F2N7O13P2 684.1026, found 684.1019.
20-Deoxy-20-fluoroarabinonicotinamide Mononucleotide (31).
Compound 31 was obtained according to the phosphorylation
procedure to synthesize 29 using 22 (7 mg, 0.027 mmol) as the
starting material. Crude product was concentrated and purified
by HPLC on a Waters RP-18 XBridge PrepShield 19 ꢀ 50 mm
column (solvent was 20 mM ammonium acetate, compound
was eluted at a flow rate of 2 mL/min) to afford 31 (tR=5.4 min,
Acknowledgment. We thank Ellison Medical Foundation
for a New Scholar in Aging Award 2007 to A.A.S. We also
thank Cornell University for funding. This work was sup-
ported by NIH funds for purchase of a new NMR spectro-
meter to Weill Cornell Medical College. We thank Cliff Soll
of Hunter College of New York City for HRMS measure-
ments. We thank Rockefeller Proteomics Resource for use of
MS facilities.
1
6 mg, 0.018 mmol, 67%): H NMR (D2O, 400 MHz) δ (ppm)
4.04 (m, 1H), 4.19 (m, 1H), 4.34 (m, 1H), 4.56 (dt, J = 5.0,
17.8 Hz, 1H), 5.51 (dt, J=4.6, 51.4 Hz, 1H), 6.67 (dd, J=4.8,
8.8 Hz, 1H), 8.24 (t, J=7.1 Hz, 1H), 8.92 (d, J=7.9 Hz, 1H), 9.31
(d, J=6.1 Hz, 1H), 9.39 (s, 1H); 13C NMR (125 MHz, D2O) δ
(ppm) 55.17, 55.20, 55.23, 60.60, 60.62, 72.6, 72.8, 90.17, 90.19,
Supporting Information Available: 1H and 13C NMR spec-
tra for all new compounds described herein. This material is
available free of charge via the Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 16, 2009 5789