Biomimetic simulation of free radical-initiated cascade reactions postulated to occur at the active site of ribonucleotide reductases

Article abstract of DOI:10.1021/ja983449p

Treatment of 5'-O-nitro esters of nucleosides with tributylstannane and AIBN at elevated temperatures caused β-scission of the resulting 5'-oxygen radical to give formaldehyde and dehomologated erythrofuranosyl nucleosides. Analogous treatment of 6'-O-nitro esters of homonucleosides [(5-deoxy-β-D- ribo-hexofuranosyl)adenine or uracil nucleosides derived from D-glucose] resulted in generation of a 6'-oxygen radical followed by abstraction of H3' by a [1,5]-hydrogen shift. Radical quenching with tributyltin deuteride gave 3'-[²H]homonucleosides. This deuterium transfer, and inversion of configuration at C3' with unprotected homonucleosides, confirmed the relay- generation of C3' free radicals. Analogous treatment of 6'-O-nitro esters of homonucleosides containing a 2'-chloro (30) or 2'-O-tosyl (40) substituent resulted in complete disappearance of starting material and generation of (R)-2-(2-hydroxyethyl)-3(2H)-furanone (33). Generation of a 6'-oxygen radical, [1,5]-hydrogen shift of H3' to give a C3' radical, and loss of the 2'-substituent would give unstable intermediates that could lose the heterocyclic base from Cl' to give 33. This radical-initiated cascade simulates reactions postulated to occur at the active site of ribonucleotide reductases. Generation of a C3' radical from 40 and loss of toluenesulfonic acid via a [1,2]-electron shift would generate a radical intermediate that could undergo deuterium transfer followed by β-elimination of the base to give the deuterated furanone 33 as observed. This is in harmony with a new mechanism for substrate reduction of nucleotides to give 2'-deoxy products. Generation of a C3' radical from 30 and loss of a chlorine atom by β- radical elimination would result in conjugate elimination of base and generation of 33 without incorporation of deuterium, as observed. Thus, one- electron elimination processes (as well as the previously postulated two- electron loss with groups from C2') must be considered with mechanism-based inactivators of ribonucleotide reductases. Biomimetic reactions and new mechanistic considerations are discussed.

Full text of DOI:10.1021/ja983449p

VOLUME 121, NUMBER 7
FEBRUARY 24, 1999
© Copyright 1999 by the
American Chemical Society
Biomimetic Simulation of Free Radical-Initiated Cascade Reactions
Postulated To Occur at the Active Site of Ribonucleotide Reductases1
Morris J. Robins,* Zhiqiang Guo,† Mirna C. Samano,‡ and Stanislaw F. Wnuk§
Contribution from the Department of Chemistry and Biochemistry, Brigham Young UniVersity,
ProVo, Utah 84602-5700
ReceiVed September 28, 1998
Abstract: Treatment of 5′-O-nitro esters of nucleosides with tributylstannane and AIBN at elevated temperatures
caused â-scission of the resulting 5′-oxygen radical to give formaldehyde and dehomologated erythrofuranosyl
nucleosides. Analogous treatment of 6′-O-nitro esters of homonucleosides [(5-deoxy-â-D-ribo-hexofuranosyl)-
adenine or uracil nucleosides derived from D-glucose] resulted in generation of a 6′-oxygen radical followed
by abstraction of H3′ by a [1,5]-hydrogen shift. Radical quenching with tributyltin deuteride gave 3′-[2H]-
homonucleosides. This deuterium transfer, and inversion of configuration at C3′ with unprotected homo-
nucleosides, confirmed the relay-generation of C3′ free radicals. Analogous treatment of 6′-O-nitro esters of
homonucleosides containing a 2′-chloro (30) or 2′-O-tosyl (40) substituent resulted in complete disappearance
of starting material and generation of (R)-2-(2-hydroxyethyl)-3(2H)-furanone (33). Generation of a 6′-oxygen
radical, [1,5]-hydrogen shift of H3′ to give a C3′ radical, and loss of the 2′-substituent would give unstable
intermediates that could lose the heterocyclic base from C1′ to give 33. This radical-initiated cascade simulates
reactions postulated to occur at the active site of ribonucleotide reductases. Generation of a C3′ radical from
40 and loss of toluenesulfonic acid via a [1,2]-electron shift would generate a radical intermediate that could
undergo deuterium transfer followed by â-elimination of the base to give the deuterated furanone 33, as observed.
This is in harmony with a new mechanism for substrate reduction of nucleotides to give 2′-deoxy products.
Generation of a C3′ radical from 30 and loss of a chlorine atom by â-radical elimination would result in
conjugate elimination of base and generation of 33 without incorporation of deuterium, as observed. Thus,
one-electron elimination processes (as well as the previously postulated two-electron loss with groups from
C2′) must be considered with mechanism-based inactivators of ribonucleotide reductases. Biomimetic reactions
and new mechanistic considerations are discussed.
Introduction
performed elegant molecular mechanistic studies that clarified
the role of free radical initiators and the resulting reaction
cascade that results in reduction of substrate ribonucleotides to
2′-deoxy analogues. This field has been reviewed frequently.2-4
The ribonucleoside diphosphate reductase (RDPR) from Es-
Ribonucleotide reductases (RNRs) are the crucial enzymes
that execute the only de noVo biosynthesis of DNA monomers.
Reichard and other investigators2 defined basic functional and
structural features of RNRs, and Stubbe and co-workers3
* To whom correspondence should be addressed.
† Present address: Neurocrine Biosciences, San Diego, CA.
‡ Present address: Glaxo Wellcome, Research Triangle Park, NC.
§ Present address: Department of Chemistry, Florida International
University, Miami, FL.
(1) Nucleic Acid Related Compounds. 106. Part 105: Robins, M. J.;
Lewandowska, E.; Wnuk, S. F. J. Org. Chem. 1988, 63, 7375-7381.
(2) (a) Thelander, L.; Reichard, P. Annu. ReV. Biochem. 1979, 48, 133-
158. (b) Reichard, P. Science 1993, 260, 1773-1777. (c) Sjo¨berg, B.-M.
Struct. Bonding (Berlin) 1997, 88, 139-173.
10.1021/ja983449p CCC: $18.00 © 1999 American Chemical Society
Published on Web 01/28/1999

1426 J. Am. Chem. Soc., Vol. 121, No. 7, 1999
Robins et al.
Scheme 1a
a Proposed mechanism for reduction of nucleoside diphosphate substrates with RDPR.8
cherichia coli (EC 1.17.4.1) has been studied most extensively
and consists of two homodimeric subunits (R1 and R2). R1
contains cysteine residues that are involved in redox chemistry,
binding sites for substrates and allosteric nucleotides that control
the enzyme velocity and substrate selection, and a cysteine
residue postulated to be the proximal radical initiator. R2
contains a diferric iron cluster associated with a tyrosyl free
radical, which constitutes the “stable” radical system postulated
to initiate an intriguing reaction cascade during each enzyme
turnover. Mammalian and certain viral RDPRs have structural
organization similar to that of E. coli.2-4
by long-range electron transfer from Tyr222) as the proximal
initiator for abstraction of H3′ from the substrate. Base-promoted
heterolytic cleavage of C2′-O2′ (hydrogen bonding of OH3′
with the carboxylate of Glu441) was proposed to assist loss of
water from C2′.3c,d Hydrogen transfer from the dithiol pair (Cys
225/462) at the R-face of C2′ and electron transfer from the
resulting disulfide radical anion to C3′ would give the C3′
radical that would regain hydrogen from Cys439 to produce
the 2′-deoxynucleotide and regenerate the thiyl radical.3c,d
Additional evidence for C3′ radical initiation as the first step
in the reduction of substrates, and also in mechanism-based
inactivation of RNRs, has been reported.7
Siegbahn’s very recent theoretical analysis of the substrate
reaction cascade8 correlated mechanistic processes with amino
acid residues identified in X-ray crystal structures of the
subunits.9 Scheme 1 illustrates this analysis,8 which is based
on the Stubbe mechanism but adds new considerations and
clarifies postulates3,5 that were inconsistent with chemical
properties. Abstraction of H3′ from substrate 1 by Cys439 is
assisted by hydrogen bonding with Glu441 and Asn437, which
promotes loss of water (H2O-2′) from the resulting C3′ radical
2 by a [1,2]-electron shift to C2′ with concomitant net transfer
Stubbe’s3,5 original mechanistic rationalization for radical-
mediated 2′-deoxygenation of ribonucleotides by RDPR was
based5a-d on a mechanism proposed for conversion of ethylene
glycol to acetaldehyde with acidic Fenton’s reagent6a (hydroxyl
radical initiation6). The enzymatic process was considered to
be initiated by abstraction of H3′ from the substrate by the
tyrosyl radical,5c or a protein residue,5d to give a C3′ radical.
Protonation of the 2′-hydroxyl by a cysteine thiol was invoked
to assist heterolytic departure of a water molecule (H2O-2′) to
produce a cation radical. Transfer of a hydride equivalent (from
a dithiol pair) to the R-face of C2′ and return of H3′ to C3′
would produce the 2′-deoxynucleotide.5d Recent refinements of
this hypothesis invoked a thiyl radical3c,d,5e-h (Cys439, generated
(5) (a) Stubbe, J.; Kozarich, J. W. J. Am. Chem. Soc. 1980, 102, 2505-
2507. (b) Stubbe, J.; Kozarich, J. W. J. Biol. Chem. 1980, 255, 5511-
5513. (c) Stubbe, J.; Ackles, D. J. Biol. Chem. 1980, 255, 8027-8030. (d)
Stubbe, J.; Ator, M.; Krenitsky, T. J. Biol. Chem. 1983, 258, 1625-1631.
(e) Mao, S. S.; Holler, T. P.; Yu, G. X.; Bollinger, J. M., Jr.; Booker, S.;
Johnston, M. I.; Stubbe, J. Biochemistry 1992, 31, 9733-9743. (f) Mao, S.
S.; Holler, T. P.; Bollinger, J. M., Jr.; Yu, G. X.; Johnston, M. I.; Stubbe,
J. Biochemistry 1992, 31, 9744-9751. (g) Mao, S. S.; Yu G. X.; Chalfoun,
D.; Stubbe, J. Biochemistry 1992, 31, 9752-9759. (h) van der Donk, W.
A.; Zeng, C.; Biemann, K.; Stubbe, J. Biochemistry 1996, 35, 10058-10067.
(6) (a) Walling, C.; Johnson, R. A. J. Am. Chem. Soc. 1975, 97, 2405-
2407. (b) Walling, C. Acc. Chem. Res. 1998, 31, 155-157. (c) MacFaul,
P. A.; Wayner, D. D. M.; Ingold, K. U. Acc. Chem. Res. 1998, 31, 159-
162.
(3) (a) Ashley, G. W.; Stubbe, J. In Inhibitors of Ribonucleoside
Diphosphate Reductase ActiVity; Cory, J. G., Cory, A. H., Eds.; Pergamon
Press: New York, 1989; pp 55-87. (b) Stubbe, J. AdV. Enzymol. Relat.
Areas Mol. Biol. 1990, 63, 349-419. (c) Stubbe, J.; van der Donk, W. A.
Chem. Biol. 1995, 2, 793-801. (d) Stubbe, J.; van der Donk, W. A. Chem.
ReV. 1998, 98, 705-762.
(4) (a) Lammers, M.; Follmann, H. Struct. Bonding (Berlin) 1983, 54,
27-91. (b) Stubbe, J. J. Biol. Chem. 1990, 265, 5329-5332. (c) Robins,
M. J.; Samano, M. C.; Samano, V. Nucleosides Nucleotides 1995, 14, 485-
493. (d) Sjo¨berg, B.-M. In Nucleic Acids and Molecular Biology; Eckstein,
F., Lilley, D. M. J., Eds.; Springer-Verlag: Berlin, 1995; Vol. 9, pp 192-
221. (e) Robins, M. J. Nucleosides Nucleotides, in press.

Cascade Reactions at the ActiVe Site of RNRs
J. Am. Chem. Soc., Vol. 121, No. 7, 1999 1427
of the 3′-hydroxyl hydrogen to O2′.8 The hydrogen-bonded
transition state for 2 f 3a has minimal charge separation in
harmony with chemical reactions at C2′. We had demonstrated
that free radical reactions proceed readily at C2′,10 whereas
generation of cationic character3,5 is energetically prohibitive
(C2′ is bonded to the electron-deficient anomeric center).10a
Hydrogen transfer from Cys225 to C2′ of the ketone radical 3b
is exothermic to produce the most stable intermediate in the
sequence, the 2′-deoxy-3′-ketone 4.8 Stubbe’s mechanism is
vague regarding electron/proton transfer to 4 from a disulfide
anion radical,3,5 and a recent alternative pathway was suggested
without theoretical support.9f A chemically plausible postulate,
with a kinetically competent activation energy barrier, invokes
a protein residue-assisted addition of the thiol moiety of Cys225
(generated by hydrogen transfer from the proximal HSCys462
Scheme 2a
•
to SCys225) to the ketone function in the hydrogen-bonded
•
complex 4.8 Attack of the SCys462 radical at the hydrogen-
bonded sulfur of the resulting thiohemiacetal complex 5 then
generates the C3′ radical 6 and a hydrogen-bonded cystine
disulfide complex. Hydrogen transfer from Cys439 to C3′ is
exothermic to give the 2′-deoxynucleotide product 7 and
regenerate •SCys439. Reduction of the disulfide to the HSCys225/
HSCys462 dithiol pair is then required for the next catalytic
cycle. This elegant theoretical analysis8 preserves the funda-
mental concepts of the Stubbe mechanism,3,5 corrects10 the
postulated3a,b,5a-d generation of cation radical character at C2′,
and offers for the first time a chemically and theoretically
plausible route for the overall rate-limiting reduction of the 2′-
deoxy-3′-ketone intermediate 4.
a Proposed mechanism for inactivation of RDPR by 2′-chloro-2′-
deoxy-NDPs.3c
Thelander et al.13 Inhibition of RNRs with 2′-chloro analogues
was proposed to involve analogous initiation by abstraction of
H3′3,14 followed by spontaneous heterolytic cleavage of the C2′-
Cl bond to release chloride and produce a cation radical.14
Abstraction of hydrogen from a Cys by C2′14b and â-elimination
of H2′/base and H4′/inorganic pyrophosphate from the 2′-deoxy-
3′-ketonucleotide were postulated to produce the 2-methylene-
3(2H)-furanone Michael acceptor, which effected covalent
alkylation/inactivation of the enzyme.
Lenz and Giese performed photochemical studies with
selenoester models that fragment to generate nucleoside mimics
of the natural nucleotide C3′ radical 2.11 Photolysis rates were
pH dependent, and addition of acetate buffer enhanced rates in
harmony with base-promoted assistance of the cleavage of water
from C2′ (2 f 3a, Scheme 1), rather than acid-catalyzed
generation of a radical cation.3a,b,5a-d The photolysis results11
are compatible with studies of Schulte-Frohlinde and co-workers
on radiolytic generation and decomposition of oxygen-contain-
ing radicals.12
Scheme 2 illustrates Stubbe’s recent hypothesis for this
mechanism-based inactivation.3c Abstraction of H3′ from 8 gives
C3′ radical 9. Loss of chloride and the 3′-hydroxyl proton gives
ketone radical 10. Hydrogen transfer from Cys439 to 10 at the
â-face gives 11a (with regeneration of •SCys439) and that from
Cys225/462 to the R-face gives 11b (without regeneration of
•SCys439). Dissociation of 11 from the enzyme active site
followed by H2′/base and H4′/iPP â-eliminations would produce
the Michael inactivator,3c,14 2-methylene-3(2H)-furanone (12).
Over a decade ago, we15 began biomimetic studies to simulate
the initiation/elimination cascade that occurs during reductions
and mechanism-based inactivations mediated by RNRs. Our
rudimentary models generated 3′-deoxy C3′ radicals with
substitutents at C2′.16 Treatment of 2′-(azido, bromo, chloro,
iodo, and methylthio)nucleoside 3′-thionocarbonates with Bu3SnH/
AIBN produced 3′-deoxy C3′ radicals that underwent loss of
the 2′-substitutent to give 2′,3′-didehydro-2′,3′-dideoxynucleo-
sides. In contrast, analogous 3′-thionocarbonates with 2′-fluoro
or 2′-O-(mesyl or tosyl) substituents (anionic leaving groups)
underwent hydrogen transfer to the C3′ radical to give the 3′-
deoxy-2′-[fluoro or O-(mesyl or tosyl)] derivatives.16 A radical
relay system has now been constructed17 with homologated
nucleoside analogues to allow generation of 6′-oxyl radicals
[from 6′-O-nitro-2′-(substituted)homonucleosides] that are po-
sitioned to abstract H3′ and produce 3′-hydroxyl-containing C3′
radicals. Chlorine atom17a or toluenesulfonic acid17b loss from
Mechanism-based inactivation of RDPR with 2′-(azido and
chloro)-2′-deoxynucleoside 5′-diphosphates was discovered by
(7) (a) Coves, J.; Le Hir de Fallois, L.; Le Pape, L.; De´cout, J.-L.;
Fontecave, M. Biochemistry 1996, 35, 8595-8602. (b) Le Hir de Fallois,
L.; De´cout, J.-L.; Fontecave, M. J. Chem. Soc., Perkin Trans. 1 1997, 2587-
2595. (c) Persson, A. L.; Eriksson, M.; Katterle, B.; Po¨tsch, S.; Sahlin, M.;
Sjo¨berg, B.-M. J. Biol. Chem. 1997, 272, 31533-31541. (d) Gerfen, G. J.;
van der Donk, W. A.; Yu, G.; McCarthy, J. R.; Jarvi, E. T.; Matthews, D.
P.; Farrar, C.; Griffin, R. G.; Stubbe, J. J. Am. Chem. Soc. 1998, 120, 3823-
3835. (e) van der Donk, W. A.; Gerfen, G. J.; Stubbe, J. J. Am. Chem. Soc.
1998, 120, 4252-4253. (f) van der Donk, W. A.; Yu, G.; Pe´rez, L.; Sanchez,
R. J.; Stubbe, J.; Samano, V.; Robins, M. J. Biochemistry 1998, 37, 6419-
6426.
(8) Siegbahn, P. E. M. J. Am. Chem. Soc. 1998, 120, 8417-8429.
(9) (a) Nordlund, P.; Sjo¨berg, B.-M.; Eklund, H. Nature 1990, 345, 593-
598. (b) Uhlin, U.; Eklund, H. Nature 1994, 370, 533-539. (c) Uhlin, U.;
Eklund, H. J. Mol. Biol. 1996, 262, 358-369. (d) Kauppi, B.; Nielsen, B.
B.; Ramaswamy, S.; Larsen, I. K.; Thelander, M.; Thelander, L.; Eklund,
H. J. Mol. Biol. 1996, 262, 706-720. (e) Logan, D. T.; Su, X.-D.; Åberg,
A.; Regnstro¨m, K.; Hajdu, J.; Eklund, H.; Nordlund, P. Structure 1996, 4,
1053-1064. (f) Eriksson, M.; Uhlin, U.; Ramaswamy, S.; Ekberg, M.;
Regnstro¨m, K.; Sjo¨berg, B.-M.; Eklund, H. Structure 1997, 5, 1077-1092.
(10) (a) Robins, M. J.; Sporns, P.; Muhs, W. H. Can. J. Chem. 1979,
57, 274-282. (b) Robins, M. J.; Wilson, J. S.; Hansske, F. J. Am. Chem.
Soc. 1983, 105, 4059-4065.
(13) Thelander, L.; Larsson, B.; Hobbs, J.; Eckstein, F. J. Biol. Chem.
1976, 251, 1398-1405.
(14) (a) Harris, G.; Ator, M.; Stubbe, J. Biochemistry 1984, 23, 5214-
5225. (b) Ator, M. A.; Stubbe, J. Biochemistry 1985, 24, 7214-7221.
(15) Samano, M. C. Ph.D. Dissertation, Brigham Young University, 1992.
(16) Robins, M. J.; Wnuk, S. F.; Herna´ndez-Thirring, A. E.; Samano,
M. C. J. Am. Chem. Soc. 1996, 118, 11341-11348.
(11) Lenz, R.; Giese, B. J. Am. Chem. Soc. 1997, 119, 2784-2794.
(12) (a) Behrens, G.; Koltzenburg, G.; Ritter, A.; Schulte-Frohlinde, D.
Int. J. Radiat. Biol. 1978, 33, 163-171. (b) Behrens, G.; Koltzenburg, G.;
Schulte-Frohlinde, D. Z. Naturforsch. 1982, 37C, 1205-1227. (c) Koltzen-
burg, G.; Behrens, G.; Schulte-Frohlinde, D. J. Am. Chem. Soc. 1982, 104,
7311-7312.
(17) (a) Robins, M. J.; Guo, Z.; Samano, M. C.; Wnuk, S. F. J. Am.
Chem. Soc. 1996, 118, 11317-11318. (b) Robins, M. J.; Guo, Z.; Wnuk,
S. F. J. Am. Chem. Soc. 1997, 119, 3637-3638.

1428 J. Am. Chem. Soc., Vol. 121, No. 7, 1999
Robins et al.
Scheme 3
Scheme 5a
Scheme 4a
a Substituents included X ) Cl, Br, I, SR, SOR, and SO2R.22
these R-hydroxy radicals provided the first simulation of the
radical cascade processes postulated to occur at the active site
of RNRs.
a (a) Bu3SnD/AIBN/benzene (or benzene/DMAC)/∆.
Results and Discussion
multistep procedures (for each base) and often give low overall
yields.25 Deprotection of 2′,3′-O-isopropylidene-6′-O-nitro-
homouridine17a (18a) gave 6′-O-nitrohomouridine (18b). Homo-
adenosine17b (28b) and 6′-O-nitrohomoadenosine17b (27b) were
converted into their 2′,3′-O-isopropylidene derivatives 28a and
27a, respectively, by standard procedures. Regioselective to-
sylation26 of 6′-O-nitrohomoadenosine (27b) gave 6′-O-nitro-
2′-O-tosylhomoadenosine17b (40). Nitration18 of 3′-O-TBDMS-
2′-O-tosylhomoadenosine (prepared by bis-silylation of 2′-O-
tosylhomoadenosine17b and selective primary desilylation27) and
deprotection also gave 40, whereas attempted nitration of 28a
resulted in glycosyl cleavage. The stabilizing effect of a 2′-O-
tosyl ester against acid-catalyzed hydrolysis of the glycosyl bond
of adenosine was known.28
An initial approach15 involved generation of 5′-oxyl radicals
from readily available 5′-O-nitro esters of protected ribonucleo-
sides.18 However, treatment of 5′-O-nitro-2′,3′-O-isopropylide-
neuridine18 (13a) with Bu3SnD/AIBN/benzene/∆19 resulted in
â-scission (loss of formaldehyde from 15a) to give the 4′-
glycosyl radical (Scheme 3). Deuterium transfer occurred
stereoselectively at the â-face to give the dehomologated
erythrose derivative 16a (R/S, ∼7:3). Deuterium transfer to the
5′-oxyl radical 15a also occurred (to give 14a, after aqueous
workup), but no deuterium exchange was detected at C3′ (1H
NMR). Treatment of 13a with Bu3SnD/AIBN/xylenes/∆ gave
a higher ratio of 16a/14a (∼1:1). Deprotection of 16a gave 9-(â-
D-erythrofuranosyl)uracil20 (17a). Similar treatment of the
adenine analogue 13b gave 17b, and analogous dehomologation
was observed with 2′-O-(tert-butyldimethylsilyl)-3′-deoxy-5′-
O-nitroadenosine.15 This provides a convenient new route to
tetrofuranosyl nucleosides.
Barton’s nitrite21 and Wagner’s δ-substituted aryl ketone22
(Scheme 4) photolysis studies had shown that a six-membered
transition state is favorable for abstraction of hydrogen by an
oxyl radical. A [1,5]-hydrogen shift was observed with oxyl
radicals generated from nitrate esters with Bu3SnH.19b Carbo-
hydrates with benzoyl groups tethered at C4′ were recently
synthesized for proposed generation of C3′ radicals,23 but
irridiation would generate oxygen radicals with ꢀ (rather than
the favored δ) separation from H3′. Wagner demonstrated that
photoactivated elimination (even with iodine) proceeded poorly
if a seven-membered transition state was required.22b
Treatment of 18a (Scheme 5) with Bu3SnD/AIBN/benzene/∆
1
gave 21a and 3′-[2H]21a (86%, ∼1:4; H NMR, HRMS). The
1
decrease (∼80%) in the H NMR signal at δ 4.76 (H3′) is
consistent with generation of 6′-oxyl radical 19a, [1,5]-
shift19b,21,22 of H3′ (19a f 20a), and quenching of the C3′
radical by deuterium transfer from the stannane (20a f 3′-[2H]-
21a), in competition with deuterium transfer to the 6′-oxyl
(25) (a) Hollmann, J.; Schlimme, E. Liebigs Ann. Chem. 1984, 98-107.
(b) Lassota, P.; Kusmierek, J. T.; Stolarski, R.; Shugar, D. Z. Naturfosch.
1987, 42C, 589-598. (c) Rawson, T. E.; Webb, T. R. Nucleosides
Nucleotides 1990, 9, 89-96. (d) Kappler, F.; Hampton, A. In Nucleic Acid
Chemistry: ImproVed and New Synthetic Procedures, Methods, and
Techniques; Townsend, L. B., Tipson, R. S., Eds.; Wiley: New York, 1991;
Vol. 4, pp 240-244.
(26) Wagner, D.; Verheyden, J. P. H.; Moffatt, J. G. J. Org. Chem. 1974,
39, 24-30.
(27) Robins, M. J.; Samano, V.; Johnson, M. D. J. Org. Chem. 1990,
55, 410-412.
Our syntheses of 6′-O-nitrohomonucleosides utilized sugar
precursors15,17,24 (full experimental details are in the Supporting
Information), because 5′-homologations of nucleosides require
(28) Brown, D. M.; Fasman, G. D.; Magrath, D. I.; Todd, A. R. J. Chem.
Soc. 1954, 1448-1455.
(29) (a) Quantitative estimations of unimolecular rate constants for the
C4′-C5′ â-scission to release formaldehyde from 15 (Scheme 3, k1 ≈ 107
s-1), for [1, 5]-hydrogen shifts from C3′ to O6′ (19 f 20, Scheme 5, k1 ≈
107-108 s-1), (41 f 42, Scheme 9, k1 ≈ 107 s-1), and for the corresponding
[1, 5]-hydrogen shifts that result in generation of C3′ radicals in Schemes
7 (k1 ≈ 106 s-1) and 8 (k1 ≈ 108 s-1) (in competition with deuterium transfer
(18) Lichtenthaler, F. W.; Mu¨ller, H. J. Synthesis 1974, 199-201.
(19) (a) Vite, G. D.; Fraser-Reid, B. Synth. Commun. 1988, 18, 1339-
1342. (b) Lopez, J. C.; Alonso, R.; Fraser-Reid, B. J. Am. Chem. Soc. 1989,
111, 6471-6473.
(20) Kline, P. C.; Serianni, A. S. J. Org. Chem. 1992, 57, 1772-1777.
(21) Barton, D. H. R.; Beaton, J. M.; Geller, L. E.; Pechet, M. M. J.
Am. Chem. Soc. 1961, 83, 4076-4083.
(22) (a) Wagner, P. J.; Sedon, J. H.; Lindstrom, M. J. J. Am. Chem. Soc.
1978, 100, 2579-2580. (b) Wagner, P. J.; Lindstrom, M. J.; Sedon, J. H.;
Ward, D. R. J. Am. Chem. Soc. 1981, 103, 3842-3849.
(23) Lehmann, T. E.; Berkessel, A. J. Org. Chem. 1997, 62, 302-309.
(24) (a) Hiebl, J.; Zbiral, E. Tetrahedron Lett. 1990, 31, 4007-4010.
(b) Gautier, C.; Leroy, R.; Monneret, C.; Roger, P. Tetrahedron Lett. 1991,
32, 3361-3364.
from Bu3SnD to nucleoside oxyl radicals) were calculated with a rate
29b
constant (k2 ) 2 × 108 M-1 s-1
)
for the reaction t-BuO• + Βu3SnH f
t-BuOH + Bu3Sn•. Major assumptions include using a “constant” concentra-
tion for Bu3SnD (used in 5-40 molar excess), using the known rate
constant29b for our “related” deuterium transfer to primary oxyl radicals,
ignoring competing processes that generate minor byproducts, and using
our detection limit (∼2%)29c for the nucleoside bimolecular deuterium
transfer byproduct. (b) Scaiano, J. C. J. Am. Chem. Soc. 1980, 102, 5399-
5400. (c) Robins, M. J.; Sarker, S.; Wnuk, S. F. Nucleosides Nucleotides
1998, 17, 785-790.

Cascade Reactions at the ActiVe Site of RNRs
J. Am. Chem. Soc., Vol. 121, No. 7, 1999 1429
Scheme 6a
Scheme 8a
a (a) AcCl/collidine/CH2Cl2/-78 °C. (b) (i) PTCCl/DMAP/CH3CN;
(ii) Bu3SnH/AIBN/toluene/∆. (c) (i) TFA/H2O; (ii) Ac2O/pyridine; (iii)
(persilylated)uracil/SnCl4/CH3CN. (d) NH3/MeOH.
Scheme 7a
a (a) Bu3SnD/AIBN/benzene/∆. (b) (i) (Bu3Sn)2O/CHCl3/∆; (ii) Br2.
(c) TBDMSCl/pyridine. (d) (i) TsNHNH2/MeOH; (ii) NaBH4/MeOH/
∆. (e) CrO3/pyridine/Ac2O. (f) TEA/MeOH.
O-nitrohomouridine (30) (Scheme 8) with Bu3SnD/AIBN/
benzene/∆ resulted in decomposition of 30 with concomitant
generation of uracil and (R)-2-(2-hydroxyethyl)-3(2H)-furanone
(33). NMR and HRMS spectra, and an independent synthesis
of the silyl ether 34, confirmed the structure of the rather
unstable enone 33. Incubation of 2′-chloro-2′-deoxynucleotides
with RNRs is known to produce the 2-methylene-3(2H)-
furanone14 (12) analogue of 33.
A plausible mechanism for conversion of 30 into 33 involves
generation of the 6′-oxyl radical 31 and relay [1,5]-H3′
abstraction. Loss of a chlorine atom,17 but not a chloride anion,14
would produce enol 32. Radical chains would be propagated
by deuterium transfer from Bu3SnD to chlorine atoms. Conjugate
elimination (or tautomerization of 32 into the 2′-deoxy-3′-ketone
and â-elimination) of uracil would give 33. During in vivo
inactivation of RNRs by 2′-chloro-2′-deoxynucleotides, released
chlorine atoms might be reduced to chloride by proximal thiol
groups. Hydrogen bonding of the 3′-hydroxyl proton to Glu441
would enhance negative character at C2′. The incipient enolate
could remove the proton (Ha) from Cys439 to give 11a (some
migration of [3H]3′ to C2′ at the â-face is observed3b,14b), or
accept a proton at the R-face to produce 11b (Scheme 2), with
a one-electron alteration in the oxidation state(s) of the thiol(s).
a (a) Bu3SnD/AIBN/benzene/DMAC/∆.
radical (∼20%) (and exchange during workup). Analogous
treatment
[Bu3SnD/AIBN/benzene/N,N-dimethylacetamide
(DMAC)/∆)] of 6′-O-nitrohomouridine (18b), with a more
conformationally flexible sugar ring, gave 3′-[2H]21b and its
xylo epimer 3′-[2H]22 (87%, ∼1.3:1) with >98% deuterium
incorporation at C3′ (1H NMR) of both compounds.29 Formation
of the ribo and xylo C3′ epimers, as well as the complete 3′-
deuteration, confirmed the intermediacy of a 3′-radical. The xylo
epimer, 22, of homouridine was synthesized independently from
3-O-acetyl-1,2-O-isopropylidene-R-D-glucofuranose30 (23) by a
route (Scheme 6) parallel to that used17a for 21b.
Treatment of 27a (Scheme 7) with Bu3SnD/AIBN/benzene/∆
gave 28a and 3′-[2H]28a (83%, ∼1:1; H NMR, HRMS). The
1
unprotected 27b (more conformationally flexible sugar ring) was
converted into a mixture of homoadenosine (28b, no 2H
exchange at C3′) and the completely deuterated xylo epimer
3′-[2H]29 (∼81%, ∼1.3:1). Adenine (∼5%) and an unidentified
product (∼10%, possibly a 3′-ketone) also were isolated.
Nucleoside 3′-ketones decompose readily by â-elimination (H2′/
base), especially in the presence of bases. Addition of tetrabu-
tylammonium fluoride (TBAF) to a solution of the unidentified
product resulted in its immediate decomposition with release
of adenine. Repetition of parallel treatments of 18b, and HPLC
of the reaction mixture, indicated a correspondingly unstable
compound plus uracil.
Methyl 2-deoxy-R-D-arabino-hexofuranoside35 (35) was pre-
pared from 2-deoxyglucose. Oxidation36 of 35, silylation of the
5-oxo derivative 36, and deoxygenation37 of 37 gave the 2,5-
dideoxy sugar 38. Oxidation31 of 38 gave the 3-ketone 39 which
underwent â-elimination to give the TBDMS-protected 3(2H)-
furanone derivative 34 upon treatment with triethylamine (TEA)/
MeOH. Formation of 34 is in harmony with results on C3′
oxidations of 5′-O-tritylthymidine, during which the 3′-ketone
Deuterium transfer from the stannane to the 6′-oxyl radical
(to give 28b, after workup) competes successfully with intramo-
lecular [1,5]-abstraction of H3′ with adenine homonucleoside
27b. When a C3′ radical is generated, it is quenched by
deuterium transfer at the R-face to produce 3′-[2H]29 (more
hindered â-face than uracil analogues31). Parallel treatment of
(32) Szarek, W. A.; Ritchie, R. G. S.; Vyas, D. M. Carbohydr. Res. 1978,
62, 89-103.
(33) The 2′-chloro-2′-deoxy-6′-O-nitrohomouridine model was chosen
to mimic inactivation of RNRs by 2′-chloro-2′-deoxyuridine 5′-phos-
phates.13,14 Our first studies15 with 2′-(azido and chloro)-5′-O-nitrouridine
derivatives had shown that generation of a 5′-oxyl radical was faster than
cleavage of the C2′-chlorine bond, but reduction of an azido group to
amino34 was competitive with generation of a 5′-oxyl radical.
(34) (a) Samano, M. C.; Robins, M. J. Tetrahedron Lett. 1991, 44, 6293-
6296. (b) Poopeiko, N. E.; Pricota, T. I.; Mikhailopulo, I. A. Synlett 1991,
342-342.
1
27b with Bu3SnH gave a similar product distribution. The H
NMR signal at δ 3.87 (H3′) in the spectrum of 9-(5-deoxy-â-
D-xylo-hexofuranosyl)adenine32 (29) was absent in the spectrum
of 3′-[2H]29.
With relay generation of C3′ radicals clearly established, we
proceeded with biomimetic modeling of radical cascade de-
composition reactions.17,33 Treatment of 2′-chloro-2′-deoxy-6′-
(35) Walker, T. E.; Ehler, D. S.; Unkefer, C. J. Carbohydr. Res. 1988,
181, 125-134.
(30) Gramera, R. E.; Ingle, T. R.; Whistler, R. L. J. Org. Chem. 1964,
(36) Tsuda, Y.; Hanajima, M.; Matsuhira, N.; Okuno, Y.; Kanemitsu,
K. Chem. Pharm. Bull. 1989, 37, 2344-2350.
29, 2074-2075.
(31) Hansske, F.; Madej, D.; Robins, M. J. Tetrahedron 1984, 40, 125-
(37) Caglioti, L. Organic Syntheses; Wiley: New York, 1988; Collect.
Vol. VI, pp 62-63.
135.

1430 J. Am. Chem. Soc., Vol. 121, No. 7, 1999
Robins et al.
Scheme 9a
Scheme 10a
a (a) (i) TBDPSCl/pyridine; (ii) Dess-Martin periodinane/CH2Cl2.
(b) Bu3SnD/AIBN/benzene/∆.
that case, a 2′-deoxy-3′-ketone intermediate is formed by a [1,2]-
hydride shift (H3′ from C3′ to C2′) with loss of tosylate.41 The
present concerted [1,2]-electron shift with generation of a
carbonyl group at C3′ would provide the driving force for
expulsion of toluenesulfonic acid (42 f 43). Zipse performed
a theoretical study on a C3′ radical species in which a seven-
membered ring (connecting OH3′ and OH2′ by hydrogen
bonding with an “XH” species) intermediate was found to
undergo concerted loss of water from C2′.43 Formation of the
seven-membered hydrogen-bonded intermediate 42 and con-
certed elimination of toluenesulfonic acid are not unreasonable
in benzene solution. The calculations of Siegbahn assumed an
average dielectric constant of ꢀ ) 4 for the aqueous-surrounded
protein environment of the active site of RDPR.8 His study
indicated that transition state energies were affected to a very
minor extent by changing parameters for the dielectric constant
(vacuum versus ꢀ ) 4). Therefore, benzene (ꢀ ) 2.3) should
be a better model for the RDPR active site environment than
polar media.11 Model reactions in more polar, especially
aqueous, media would favor more polarized transition states,
which Nature apparently has selected against in protein environ-
ments at the active sites in which nonpolarized, concerted
processes are energetically favored.8
a (a) Bu3SnD/AIBN/benzene/∆.
derivative undergoes â-elimination under mild conditions to give
(R)-2-[(trityloxy)methyl]-3(2H)-furanone.31,38 Such â-elimina-
tions with other nucleoside39,40 and 2′-deoxy-3′-oxopento-
furanose38a derivatives have been reported. Spectral data for 34
were compatible with those for the radical cascade product 33.
Treatment of 40 (Scheme 9) with Bu3SnD/AIBN/benzene/∆
resulted in its decomposition into 2′-O-tosylhomoadenosine
2
[28%, no H at C3′ (1H NMR)] plus adenine and (R)-2-(2-
1
hydroxyethyl)-3(2H)-furanone (33, 62%). H NMR spectra of
this 33 had ∼30% reduction in the signal at δ 5.71 (H4)
(corresponds to H2′ of 40). HRMS peaks at m/z 129.0545 (100,
MH+ [C6H9O3] ) 129.0552) and 130.0619 (41, MH+ [C6H8-
DO3] ) 130.0614) confirmed the incorporation of deuterium.
Thus, radical-induced decompositions of these 2′-chloro 30 and
2′-tosylate 40 analogues proceed by different mechanisms
(Schemes 8 and 9, respectively).
Additional evidence for the mechanism in Scheme 9 was
obtained by parallel treatment of 2′-deoxy-3′-ketone44 derivatives
46 (Scheme 10). Silylation of 2′-deoxy-2′-deuterioadenosine10b
[45; C2′(R/S), ∼85:15) and Dess-Martin oxidation45 gave 46.
Downfield shifts of H2′,2′′ peaks, reduction in their intensities,
Generation of the 6′-oxyl radical (40 f 41) (Scheme 9)
followed by [1,5]-shift of H3′ would give the C3′ radical (41
f 42). Loss of toluenesulfonic acid from 42, with a concerted
[1,2]-electron shift, would produce the C2′-radical intermediate
43. Deuterium transfer from the stannane to 43 would occur
selectively at the less hindered R face10b to give the 2′-deoxy-
2′-deuterio-3′-oxohomoadenosines [44; C2′(R/S), ∼30:70] which
would undergo anti â-elimination to give 33 (with ∼30%
deuterium remaining at C4). Deuterium transfer from Bu3SnD
to 43 would propagate radical chains (40 f 43). In contrast,
the decomposition of 2′-chloro analogue 30 (Scheme 8) (with
no deuterium incorporation into 33) is analogous to our
elimination reactions in which generation of a 3′-deoxy C3′
radical was followed by loss of (azido, bromo, chloro, iodo, or
methylthio) radicals from C2′ to give the 2′,3′-olefin.16 In that
series, generation of the 3′-deoxy C3′ radical with 2′-(fluoro,
mesylate, or tosylate) substituents resulted in hydrogen transfer
from the stannane to C3′ and retention of the 2′-substituent.
The radical-induced loss of toluenesulfonic acid (42 f 43)
(Scheme 9) is analogous to the [1,2]-hydride shift rearrange-
ment,40-42 which converts 2′-O-tosyladenosine into 9-(2-deoxy-
â-D-threo-pentofuranosyl)adenine (LiEt3BH/THF/DMSO).41 In
1
and splitting simplifications44 in the H NMR spectra were
consistent with 46. Treatment of 46 with Bu3SnD/AIBN/
benzene/∆ (identical control reaction mixture; however, the
thermal â-elimination is not dependent on Bu3SnD/AIBN) gave
(R)-2-{[(tert-butyldiphenylsilyl)oxy]methyl}-3(2H)-furanone (47),
a dehomologated analogue of 33. Reduction (∼15%) of the 1H
NMR signal at δ 5.75 (H4) is in harmony with an anti-
stereospecific â-elimination (2HS/adenine) from 46 (∼85% (S)-
[2H]). Such decompositions of 2′-deoxy-3′-ketonucleosides with
elimination of the base are well-known.31,38-40
The mechanism illustrated in Scheme 9 simulates substrate
reactions postulated to occur at the active site of E. coli RDPR
(Scheme 1), except for reduction of the 3′-ketone 44 and return
of a hydrogen atom to C3′. The external tributylstannyl radical
generates internal 6′-oxyl radical 41 analogous to generation
of the proximal •SCys439 by long-range electron transfer from
•OTyr122 in RDPR. The [1,5]-shift of H3′ to O6′ generates the
C3′ radical 42 analogous to the conversion of 1 f 2 by RDPR.
The [1,2]-electron shift with loss of toluenesulfonic acid converts
(38) (a) Binkley, R. W.; Hehemann, D. G.; Binkley, W. W. J. Org. Chem.
1978, 43, 2573-2575. (b) Schreiber, S. L.; Ikemoto, N. Tetrahedron Lett.
1988, 29, 3211-3214.
(42) (a) Kawana, M.; Kuzuhara, H. Tetrahedron Lett. 1987, 28, 4075-
4078. (b) Kawana, M.; Kuzuhara, H. J. Chem. Soc., Perkin Trans. 1 1992,
469-478.
(43) Zipse, H. J. Am. Chem. Soc. 1995, 117, 11798-11806.
(44) Samano, V.; Robins, M. J. J. Org. Chem. 1990, 55, 5186-5188.
(45) (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155-4156.
(b) Ireland, R. E.; Liu, L. J. Org. Chem. 1993, 58, 2899.
(39) Sugiyama, H.; Fujimoto, K.; Saito, I. J. Am. Chem. Soc. 1995, 117,
2945-2946.
(40) Kawana, M.; Takeuchi, K.; Ohba, T.; Kuzuhara, H. Nucleic Acids
Res., Symp. Ser. No. 17, 1986, 37-40.
(41) Hansske, F.; Robins, M. J. J. Am. Chem. Soc. 1983, 105, 6736-
6737.

Cascade Reactions at the ActiVe Site of RNRs
J. Am. Chem. Soc., Vol. 121, No. 7, 1999 1431
evaporated, and the residue was chromatographed (CHCl3 f 1.5%
MeOH/CHCl3) to give 16a (8 mg, 41%; 4′R/S, ∼7:3): 1H NMR δ
1.30, 1.44 (2s, 2 × 3H), 4.17 (s, 0.7H), 4.33 (d, J ) 3.8 Hz, 0.3H),
5.04-5.08 (m, 1H), 5.17 (d, J ) 5.8 Hz, 1H), 5.38 (s, 1H), 5.71 (dd,
J ) 8.1, 2.2 Hz, 1H), 7.21 (d, J ) 8.0 Hz, 1H), 8.71 (br s, 1H, ex);
HRMS (CI) m/z 256.1043 (100, MH+ [C11H14DN2O5] ) 256.1044).
Further elution gave 2′,3′-O-isopropylideneuridine (10 mg; 46%).
Procedure A. A solution of 16a (8 mg, 0.031 mmol) in TFA/H2O
(9:1, 1 mL) was stirred at 0 °C for 1 h. Volatiles were evaporated,
EtOH was added and evaporated, and the residue was chromatographed
(CHCl3 f 7% MeOH/CHCl3) to give 17a (5 mg, 75%; 4′R/S, ∼7:3)
42 f 43 analogous to the conversion of 2 f 3a by RDPR.
Stereoselective (∼70%) transfer of deuterium at the R-face of
43 gives 44, whereas analogous transfer of hydrogen occurs
with complete stereoselectivity at the R-face of 3b to give 4
with RDPR. It is remarkable that our biomimetic system fully
simulates reactions performed by an E. coli RDPR Glu441 f
Gln441 site-directed mutant.7c The mutant enzyme also was
unable to perform the final reduction step (Gln441 lacks the
acidic hydrogen-bonding capability of Glu441), and the 3′-
ketonucleotide (analogous to our 44) was the end product of
that mutant enzyme sequence.7c
2
with data as reported20 except for H effects: 1H NMR (Me2SO-d6) δ
3.65 (d, J ) 2.2 Hz, 0.7H), 4.14-4.22 (m, 1.3H); HRMS (CI) m/z
216.0738 (45, MH+ [C8H10DN2O5] ) 216.0731).
Summary and Conclusions
9-(â-D-Erythrofuranosyl)adenine (17b). Treatment of 13b18 [60
mg, 0.17 mmol; prepared (20%) by nitration of 2′,3′-O-isopropylide-
neadenosine with N-nitropyrazole/triflic acid/CH3CN47] with Bu3SnH
(1.05 mL, 1.14 g, 3.9 mmol) as described for 17a [with chromatography
(EtOAc f 20% Me2CO/EtOAc)] gave 16b (24 mg, 50%): MS m/z
277 (38, M+), 164 (100), 136 (61). Deprotection of 16b (30 mg, 0.1
mmol) by procedure A, chromatography [Dowex 1 × 2 (OH-), MeOH/
H2O (1:1)], and crystallization (MeOH/Et2O) gave 17b (19 mg, 75%):
mp 235-237 °C (lit.20 mp 230-232 °C dec).
Treatment of 5′-O-nitropentofuranosyl nucleosides with
Bu3SnD/AIBN/benzene/∆ resulted in â-scission of the C4′-
C5′ bond of the O5′ radical intermediate, rather than [1,4]-
abstraction of H3′ (five-membered transition state). This pro-
vides access to tetrofuranosyl homologues (Scheme 3). We have
constructed 6′-O-nitrohomonucleosides and demonstrated free
radical relay exchange of H3′ for D3′ with Bu3SnD. Treatment
of 2′-chloro-2′-deoxy-6′-O-nitrohomouridine (30) under analo-
gous conditions resulted in decomposition of 30 to give uracil
plus (R)-2-(2-hydroxyethyl)-3(2H)-furanone (33) with no in-
corporation of deuterium into 33 (Scheme 8). In contrast,
analogous treatment of 6′-O-nitro-2′-O-tosylhomoadenosine (40)
resulted in decomposition of 40 to give adenine and 33 with
∼30% deuterium incorporation at C4 (Scheme 9). Thus, the
results with 40 (Scheme 9) are in harmony with the Stubbe/
Siegbahn mechanism (Scheme 1) for reduction of substrates,
and provide a biomimetic model for free radical-induced relay
reaction cascades postulated to occur at the active site of
ribonucleotide reductases. However, the results with 30 (Scheme
8) are consistent with â-elimination of a chlorine atom in
harmony with photochemical studies of Wagner (Scheme 4),
and are incompatible with loss of chloride anion (Scheme 2).
Our studies have provided experimental evidence for dif-
ferentiation between one-electron (Scheme 8) and two-electron
(Scheme 9) loss with 2′-substitutents upon generation of O3′-
containing C3′ radicals. Studies are in progress to evaluate
solvent (benzene is known to stabilize chlorine atoms46), ionic,
and general base effects on such reactions.
1-(5-Deoxy-â-D-ribo-hexofuranosyl)uracil (21b). A solution of
1-(2,6-di-O-acetyl-3-O-benzoyl-5-deoxy-â-D-ribo-hexofuranosyl)-
uracil17a (446 mg, 1 mmol) in NH3/MeOH (10 mL) was stirred in a
sealed flask for 18 h at ambient temperature. Volatiles were evaporated,
and the residue was chromatographed (10% MeOH/CH2Cl2) and
recrystallized (EtOH) to give 21b (211 mg, 82%): mp 151-152 °C
1
(lit.25b 154-157 °C); UV max 262 nm (ꢀ 10 100); H NMR (Me2SO-
d6) δ 1.76 (2 × dd, J ) 13.6, 7.0 Hz, 2H), 3.42-3.52 (m, 2H), 3.79
(“q”, J ) 5.0 Hz, 1H), 3.86 (dd, J ) 7.0, 5.0 Hz, 1H), 4.05 (“q”, J )
5.0 Hz, 1H), 4.51 (t, J ) 5.0 Hz, 1H, ex), 5.08 (d, J ) 5.0 Hz, 1H, ex),
5.32 (d, J ) 5.0 Hz, 1H, ex), 5.63 (d, J ) 8.1 Hz, 1H), 5.69 (d, J )
5.0 Hz, 1H), 7.60 (d, J ) 8.1 Hz, 1H), 11.30 (br s, 1H, ex); 13C NMR
(Me2SO-d6) δ 36.61, 57.75, 72.93, 73.33, 80.67, 88.77, 102.27, 141.39,
150.92, 163.30; MS (CI) m/z 259 (12, MH+), 129 (25), 113 (100).
1-(5-Deoxy-2,3-O-isopropylidene-â-D-ribo-hexofuranosyl)uracil
(21a). TsOH‚H2O (19 mg, 0.1 mmol) and triethyl orthoformate (0.498
mL, 444 mg, 3 mmol) were added to a suspension of 21b (258 mg, 1
mmol) in dried Me2CO (10 mL), and stirring was continued for 3 h at
ambient temperature. NaHCO3 (84 mg, 1 mmol) was added, and stirring
was continued for 30 min. The mixture was diluted (EtOAc) and
filtered, and the filtrate was evaporated. The residue was chromato-
graphed (5% MeOH/CH2Cl2) to give 21a as a white foam (244 mg,
82%): UV max 259 nm; 1H NMR δ 1.35, 1.57 (2s, 2 × 3H), 2.00 (dd,
J ) 12.0, 6.5 Hz, 2H), 2.70 (br s, 1H, ex), 3.73-3.85 (m, 2H), 4.20
(td, J ) 6.5, 5.0 Hz, 1H), 4.76 (“t”, J ) 6.5, 5.0 Hz, 1H), 4.99 (dd, J
) 6.5, 2.0 Hz, 1H), 5.59 (d, J ) 2.0 Hz, 1H), 5.75 (d, J ) 8.0 Hz,
1H), 7.25 (d, J ) 8.0 Hz, 1H), 9.83 (br s, 1H, ex); 13C NMR δ 25.88,
27.72, 35.92, 60.22, 84.11, 84.64, 85.91, 94.70, 103.27, 115.37, 142.92,
150.41, 163.52; MS m/z 298 (1, M+), 283 (40, M - Me). Anal. Calcd
for C13H18N2O6 (298.3): C, 49.81; H, 5.70; N, 26.40. Found: C, 49.76;
H, 5.81; N, 26.20.
Experimental Section
A capillary apparatus was used for uncorrected melting points. UV
1
spectra are of solutions in MeOH. H (200 or 500 MHz) and 13C (50
or 125 MHz) NMR spectra are of solutions in Me4Si/CDCl3 unless
otherwise specified. Mass spectra (MS and HRMS) were obtained by
electron impact (20 eV), chemical ionization (CI; CH4), or fast atom
bombardment (FAB; thioglycerol matrix) techniques. Reagent grade
chemicals were used, and solvents were dried by reflux and distillation
from CaH2 (except acetone/P2O5) under an argon atmosphere. TLC was
performed with Merck kieselgel 60-F254 sheets, and products were
detected with 254 nm light or by color development (I2 or 10% H2SO4/
MeOH). Merck kieselgel 60 (230-400 mesh) was used for column
chromatography. RP-HPLC was performed with a SpectraPhysics P200
pump system with an Apex Prepsil column (25 cm). NH3/MeOH was
saturated at -10 °C. Elemental analyses were by M-H-W Laboratories,
Phoenix, AZ. Only key experiments are described in this paper. Full
experimental details, spectral data, and characterization for all com-
pounds are available in the Supporting Information.
1-(5-Deoxy-2,3-O-isopropylidene-6-O-nitro-â-D-ribo-hexofurano-
syl)uracil (18a). Procedure B. Cold fuming nitric acid (3 mL; d )
1.5 g/mL) in Ac2O (3 mL) was added to 21a (60 mg, 0.2 mmol) in
Ac2O (5 mL) at -60 °C, stirring was continued for 20 min, and the
solution was poured into ice-cold saturated NaHCO3/H2O. The mixture
was extracted (EtOAc), and the combined organic phase was washed
(brine) and dried (Na2SO4). Volatiles were evaporated, and the residue
was chromatographed (3% MeOH/CH2Cl2) and recrystallized (EtOH)
1
to give 18a (63 mg, 92%): mp 146-147 °C; UV max 260 nm; H
NMR δ 1.35, 1.58 (2 × s, 2 × 3H), 2.18 (dd, J ) 13.0, 6.6 Hz, 2H),
4.15 (td, J ) 6.6, 4.8 Hz, 1H), 4.56 (“td”, J ) 6.5, 2.5 Hz, 2H), 4.73
(dd, J ) 6.5, 4.8 Hz, 1H), 5.09 (dd, J ) 6.5, 1.8 Hz, 1H), 5.47 (d, J
) 1.8 Hz, 1H), 5.75 (dd, J ) 8.0 Hz, 2.2 Hz, 1H), 7.19 (d, J ) 8.0 Hz,
1H), 8.75 (br s 1H, ex); 13C NMR δ 25.30, 27.15, 30.60, 69.67, 83.85,
1-(4-Deuterio-â-D-erythrofuranosyl)uracil (17a). A solution of
13a18 (25 mg, 0.076 mmol), AIBN (5 mg, 0.03 mmol), and Bu3SnD
(0.103 mL, 111 mg, 0.38 mmol) in dried xylene (5 mL) was
deoxygenated (Ar, 45 min) and refluxed for 1 h. Volatiles were
(47) (a) Olah, G. A.; Narang, S. C.; Fung, A. P. J. Org. Chem. 1981,
46, 2706-2709. (b) Gizeiwicz, J., Wnuk, S. F., Robins, M. J. J. Org. Chem.,
in press.
(46) Russell, G. A. J. Am. Chem. Soc. 1958, 80, 4987-4996.

1432 J. Am. Chem. Soc., Vol. 121, No. 7, 1999
Robins et al.
84.15, 84.28, 95.77, 102.70, 114.83, 143.10, 149.59, 162.71; HRMS
(FAB) m/z 344.1097 (48, MH+ ) 344.1094). Anal. Calcd for
C13H17N3O8 (343.3): C, 45.48; H, 4.99; N, 12.24. Found: C, 45.64;
H, 5.06; N, 12.15.
9-(5-Deoxy-3-deuterio-2,3-O-isopropylidene-â-D-ribo-hexofurano-
syl)adenine (3′-[2H]28a). Treatment of 27a (0.03 mmol, 11 mg) by
procedure C [with chromatography (8% MeOH/CH2Cl2)] gave 28a/
1
3′-[2H]28a (∼1:1; 8 mg, 83%): UV max 259 nm; H NMR (Me2SO-
d6) δ 4.20 (t, J ) 7.1 Hz, 1, H4′), 4.87 (dd, J ) 6.2, 3.3 Hz, 0.50,
H3′), 5.47 (d, J ) 2.7 Hz, 1, H2′), other peaks same as in the spectrum
of 28a; the 13C NMR (Me2SO-d6) peak at δ 83.71 (C3′) was reduced
to ∼50% intensity; HRMS (FAB) m/z 322.1511 (100, MH+ [C14H20N5O4]
) 322.1515), 323.1583 (98, MH+ [C14H19DN5O4] ) 323.1578).
1-(5-Deoxy-3-deuterio-â-D-ribo-hexofuranosyl)uracil (3′-[2H]21b)
and 1-(5-Deoxy-3-deuterio-â-D-xylo-hexofuranosyl)uracil (3′-[2H]22).
Treatment of 18b (12 mg, 0.04 mmol) by procedure C [DMAC (1 mL)
added for solubility] [with chromatography (15% MeOH/CH2Cl2)]
followed by RP-HPLC (15 f 40% CH3CN/H2O; 2.8 mL/min, 60 min)
gave 3′-[2H]21b and 3′-[2H]22 (9 mg, 87%; ∼1.3:1). Data for 3′-[2H]-
1-(5-Deoxy-3-deuterio-2,3-O-isopropylidene-â-D-ribo-hexofurano-
syl)uracil (3′-[2H]21a). Procedure C. A solution of 18a (10 mg, 0.03
mmol), Bu3SnD (40 µL, 44 mg, 0.15 mmol), and AIBN (∼2 mg) in
dried benzene (5 mL) was deoxygenated (Ar, 20 min) and then heated
for 1 h at reflux. Volatiles were evaporated, and the residue was
chromatographed (5% MeOH/CH2Cl2) to give 21a/3′-[2H]21a (∼1:4;
1
7.5 mg, 86%): UV max 259 nm; H NMR δ 4.20 (t, J ) 6.5 Hz, 1,
H4′), 4.76 (“t”, J ) 6.5, 5.0 Ηz, ∼0.2, H3′), 4.99 (d, J ) 2.0 Hz, 1,
H2′), other peaks same as for 21a; MS (CI) m/z 300 (100, MH+ [3′-
[2H]21a]), 299 (22, MH+ [21a]).
1-(5-Deoxy-6-O-nitro-â-D-ribo-hexofuranosyl)uracil (18b). A solu-
tion of the residue (EtOAc) from deprotection of 18a (68 mg, 0.2 mmol)
by procedure A was washed (saturated NaHCO3/H2O, brine) and dried
(Na2SO4). Volatiles were evaporated, and the residue was chromato-
graphed (7% MeOH/CH2Cl2) to give 18b (55 mg, 90%) as a white
solid: mp 161-162.5 °C; UV (MeOH) max 261 nm; 1H NMR (Me2SO-
d6) δ 2.06-2.14 (m, 2H), 3.82-3.88 (m, 2H), 4.12 (“q”, J = 5.0 Hz,
1H), 4.62 (dd, J ) 10.6, 6.4 Hz, 2H), 5.23 (d, J ) 5.2 Hz, 1H, ex),
5.45 (d, J ) 5.5 Hz, 1H, ex), 5.64 (d, J ) 8.0 Hz, 1H), 5.72 (d, J )
4.9 Hz, 1H), 7.63 (d, J ) 8.0 Hz, 1H), 11.38 (br s, 1H, ex); 13C NMR
(Me2SO-d6) δ 30.00, 70.78, 72.43, 72.86, 79.68, 89.24, 102.03, 141.44,
150.62, 163.05; HMRS (CI) m/z 304.0789 (6, MH+ [C10H14N3O8] )
304.0781).
9-(5-Deoxy-â-D-ribo-hexofuranosyl)adenine (28b). A solution of
9-(2-O-acetyl-3,6-di-O-benzoyl-5-deoxy-â-D-ribo-hexofuranosyl)-
adenine17b (531 mg, 1 mmol) in NH3/MeOH (10 mL) was stirred in a
sealed flask overnight at ambient temperature. Volatiles were evapo-
rated, and the residue was dissolved (H2O) and washed (CH2Cl2, 3×).
The aqueous phase was evaporated, and the residue was recrystallized
(H2O) to give 28b (234 mg, 83%): mp 230-232 °C (lit.48 mp 231.5-
232.5 °C); UV max 260 (ꢀ 15 200); 1H NMR (Me2SO-d6) δ 1.79 (dd,
J ) 12.2, 7.1 Hz, 2H), 3.45 (dd, J ) 10.8, 5.4 Hz, 2H), 4.00 (td, J )
7.1, 4.8 Hz, 1H), 4.09 (“q”, J = 4.8 Hz, 1H), 4.48 (t, J ) 5.0 Hz, 1H,
ex), 4.66 (q, J ) 5.2 Hz, 1H), 5.17 (d, J ) 4.8 Hz, 1H, ex), 5.41 (d,
J ) 5.2 Hz, 1H, ex), 5.82 (d, J ) 5.2 Hz, 1H), 7.28 (br s, 2H, ex), 8.12
(s, 1H), 8.30 (s, 1H); 13C NMR (Me2SO-d6) δ 36.57, 57.57, 72.98,
73.48, 81.08, 87.50, 119.28, 140.03, 149.59, 152.81, 156.22; MS (CI)
m/z 282 (42, MH+), 136 (100).
1
21b: UV max 261 nm; H NMR (Me2SO-d6) no peak at δ 3.79 (H3′
2
f H3′), 3.86 (dd, J ) 7.0, 5.0 Hz, 1, H4′), 4.05 (t, J ) 5.0 Hz, 1,
H2′), other peaks same as those for 21b; 13C NMR (Me2SO-d6) peaks
same as those for 21b except no peak at δ 72.93 (C3′); HRMS (FAB)
m/z 259.0900 (23, M+ [C10H13DN2O6] ) 259.0915). Data for 3′-[2H]-
22: UV max 260 nm; 1H NMR (Me2SO-d6) no peak at δ 3.79 (H3′ f
2H3′), 3.96 (t, J ) 4.0 Hz, 1, H2′), 4.24 (t, J ) 6.6 Hz, 1, H4′), other
peaks same as those for 22; 13C NMR (Me2SO-d6) peaks same as those
for 22 except no peak at δ 75.44 (C3′); HRMS (FAB) m/z 259.0933
(10, M+ [C10H13DN2O6] ) 259.0915).
A sample of 18b (17 mg) was treated as described, except the first
evaporation residue was partitioned (CH2Cl2/H2O). The aqueous phase
was evaporated, and the residue was subjected to RP-HPLC as described
to give uracil (∼5%; tR ) 24.4 min), 3′-[2H]21b (∼50%; tR ) 27.7
min), 3′-[2H]22 (∼36%; tR ) 31.4 min), and an unidentified product
(∼7%; tR ) 34.3 min). This unidentified product decomposed im-
mediately to give uracil upon addition of TBAF/THF.
1-(5-Deoxy-â-D-xylo-hexofuranosyl)uracil (22). A solution of
1-(2,3,6-tri-O-acetyl-5-deoxy-â-D-xylo-hexofuranosyl)uracil (26; 384
mg, 1 mmol) in NH3/MeOH (10 mL) was stirred in a sealed flask
overnight at ambient temperature. Volatiles were evaporated, and the
residue was chromatographed (15% MeOH/CH2Cl2) and recrystallized
(EtOH) to give 22 (215 mg, 83%): mp 159-160.5 °C; UV max 262
1
nm (ꢀ 10 200); H NMR (Me2SO-d6) δ 1.84 (dd, J ) 13.2, 6.6 Hz,
2H), 3.53 (dd, J ) 11.3, 5.1 Hz, 2H), 3.79 (dd, J ) 3.4, 2.8 Hz, 1H),
3.96 (t, J ) 4.0 Hz, 1H), 4.24 (td, J ) 6.6, 2.8 Hz, 1H), 4.57 (t, J )
5.1 Hz, 1H, ex), 5.39 (d, J ) 3.4 Hz, 1H, ex), 5.60 (d, J ) 4.0 Hz, 1H,
ex), 5.65 (d, J ) 8.1 Hz, 1H), 5.78 (d, J ) 4.0 Hz, 1H), 7.72 (d, J )
8.1 Hz, 1H), 11.29 (br s, 1H, ex); 13C NMR (Me2SO-d6) δ 31.68, 58.11,
75.44, 80.56, 81.35, 91.19, 101.03, 141.69, 150.75, 163.54; HRMS
(FAB) m/z 259.0935 (100, MH+ [C10H15N2O6] ) 259.0930). Anal.
Calcd for C10H14N2O6 (258): C, 46.51; H, 5.46; N, 10.85. Found: C,
46.40; H, 5.69; N, 10.69.
9-(5-Deoxy-2,3-O-isopropylidene-â-D-ribo-hexofuranosyl)ade-
nine (28a). Treatment of 28b (100 mg, 0.35 mmol) as described for
21b f 21a [with concentrated NH3/H2O (2 mL) in place of NaHCO3]
and evaporation of volatiles gave a residue that was slurried with
EtOAc/acetone (1:1). Filtration of the suspension, evaporation of the
filtrate, and chromatography (8% MeOH/CH2Cl2) and recrystallization
(MeOH) of the residue gave 28a (94 mg, 82%): mp 269-271 °C dec
9-(5-Deoxy-â-D-xylo-hexofuranosyl)adenine (29). Bu3SnH (41 µL,
44.6 mg, 0.15 mmol) and AIBN (∼2 mg) were added to 27b (10 mg,
0.03 mmol) in DMAC (1 mL) and dried benzene (9 mL). The solution
was deoxygenated (Ar, 20 min) and heated at reflux for 2 h [AIBN
(∼2 mg) was added after 1 h]. Volatiles were evaporated, the residue
was dissolved (H2O), and the aqueous solution was washed (CH2Cl2,
3×). The aqueous phase was evaporated, and the residue was subjected
to RP-HPLC (10 f 40% CH3CN/H2O; 2.8 mL/min, 80 min) to give
adenine (∼4%, tR ) 38.0 min), homoadenosine (28b, ∼47%; tR ) 40.3
min), 29 (∼37%, tR ) 47.6 min), and an unidentified product (∼12%,
tR ) 61.8 min). Data for compound 29:32 UV 260 nm; 1H NMR
(Me2SO-d6) δ 1.82 (dd, J ) 12.8, 6.4 Hz, 2H), 3.49 (m, t after D2O
ex, J ) 6.4 Hz, 2H), 3.87 (m, 1H), 4.25 (m, 2H), 4.53 (t, J ) 4.8 Hz,
1H, ex), 5.79 (d, J ) 1.2 Hz, 1H), 5.88 (m, 2H, ex), 7.33 (br s, 2H,
ex), 8.13 (s, 1H), 8.21 (s, 1H); 13C NMR (Me2SO-d6) δ 31.73, 57.83,
75.89, 79.94, 81.46, 89.48, 118.73, 139.72, 148.58, 152.30, 156.01;
HRMS (FAB) m/z 282.1201 (5, MH+ [C11H16N5O4] ) 282.1202).
9-(5-Deoxy-3-deuterio-â-D-xylo-hexofuranosyl)adenine (3′-[2H]-
29). Treatment of 27b (10 mg, 0.03 mmol) with Bu3SnD (as described
for 29) and RP-HPLC gave adenine, homoadenosine (28b), 3′-[2H]29,
and an unidentified product with yield ratios and retention times similar
to those described for 29. Data for 3′-[2H]29 (3 mg, 35%): UV max
260 nm; 1H NMR (Me2SO-d6) all peaks such as those for 29 except no
1
(lit.25d 265-268 °C dec); UV max 259 nm; H NMR (Me2SO-d6) δ
1.29, 1.50 (2 × s, 2 × 3H), 1.61-1.70 (dd, J ) 13.7, 6.0 Hz, 1H),
1.72-1.81 (dd, J ) 13.7, 7.1 Hz, 1H), 3.38 (dd, J ) 11.2, 5.6 Hz,
2H), 4.20 (td, J ) 7.1, 3.3 Hz, 1H), 4.50 (t, J ) 5.0 Hz, 1H, ex), 4.87
(dd, J ) 6.2, 3.3 Hz, 1H), 5.47 (dd, J ) 6.2, 2.7 Hz, 1H), 6.07 (d, J
) 2.7 Hz, 1H), 7.32 (br s, 2H, ex), 8.14 (s, 1H), 8.30 (s, 1H); 13C
NMR (Me2SO-d6) δ 25.27, 27.03, 36.26, 57.14, 82.95, 83.03, 83.71,
88.46, 113.36, 119.07, 139.92, 148.94, 152.76, 156.12; HRMS (FAB)
m/z 322.1512 (17, MH+ [C14H20N5O4] ) 322.1515).
9-(5-Deoxy-2,3-O-isopropylidene-6-O-nitro-â-D-ribo-hexofurano-
syl)adenine (27a). Protection of 27b17b (326 mg, 1 mmol) (as described
for 28b f 28a) [with chromatography (5% MeOH/CH2Cl2)] gave 27a
1
(293 mg, 80%): UV max 259 nm; H NMR δ 1.37, 1.60 (2 × s, 2 ×
3H), 2.15-2.24 (dd, J ) 11.7, 6.0 Hz, 2H), 4.26-4.35 (m, 1H), 4.42-
4.52 (m, 2H), 5.01 (dd, J ) 6.3, 3.9 Hz, 1H), 5.53 (dd, J ) 6.3, 2.2
Hz, 1H), 5.83 (br s, 2H, ex), 6.01 (d, J ) 2.2 Hz, 1H), 7.87 (s, 1H),
8.32 (s, 1H); 13C NMR δ 25.36, 27.15, 30.79, 69.55, 83.44, 83.85,
84.12, 90.52, 114.73, 118.82, 140.25, 149.57, 153.11, 155.60; HRMS
(FAB) m/z 367.1381 (27, MH+ [C14H19N6O6] ) 367.1366).
(48) Ryan, K. J.; Arzoumanian, H.; Acton, E. M.; Goodman, L. J. Am.
Chem. Soc. 1964, 86, 2503-2508.

Cascade Reactions at the ActiVe Site of RNRs
J. Am. Chem. Soc., Vol. 121, No. 7, 1999 1433
2
peak at δ 3.87 (H3′ f H3′); 13C NMR (Me2SO-d6) peaks same as
6H), 0.90 (s, 9H), 1.79 (ddd, J ) 18.1, 8.9, 4.8 Hz, 1H), 2.09 (dtd, J
) 18.1, 4.0, 7.1 Hz, 1H), 3.78-3.84 (m, 2H), 4.62 (dd, J ) 8.9, 4.0
Hz, 1H), 5.70 (d, J ) 2.5 Hz, 1H), 8.49 (d, J ) 2.5 Hz, 1H); 13C NMR
(MeOH-d4) δ -5.24, -5.16, 19.26, 26.49, 35.59, 59.74, 83.34, 107.65,
181.17, 208.46; HRMS (CI) m/z 243.1413 (100, MH+ [C12H23O3Si] )
243.1416).
those for 29 except no peak at δ 75.89 (C3′); HRMS (CI) m/z 283.1277
(4, MH+ [C11H15DN5O4] ) 283.1265).
(R)-2-(2-Hydroxyethyl)-3(2H)-furanone (33). Treatment of 3017a
(10 mg, 0.031 mmol) by procedure C (45 min) [with preparative HPLC
(6 f 8% MeOH/CH2Cl2)] gave uracil and 33 (3 mg, 75%): UV max
1
259 nm; H NMR (MeOH-d4) δ 2.01-2.17 (m, 2H), 3.73 (dd, J )
(R)-2-{[(tert-Butyldiphenylsilyl)oxy]methyl}-4-deuterio-3(2H)-
furanone (47). Treatment of 9-[5-O-(tert-butyldiphenylsilyl)-2-deoxy-
2-deuterio-â-D-glycero-pentofuranos-3-ulosyl]adenine (46; 49 mg, 0.1
mmol) by procedure C [with chromatography (15 f 30% EtOAc/
7.5, 5.5 Hz, 2H), 4.63 (dd, J ) 9.3, 4.0 Hz, 1H), 5.71 (d, J ) 2.5 Hz,
1H), 8.50 (d, J ) 2.5 Hz, 1H); 13C NMR (MeOH-d4) δ 35.48, 58.68,
83.67, 107.55, 181.25, 208.42; HRMS (CI) m/z 129.0555 (100, MH+
[C6H9O3] ) 129.0552).
2
hexanes)] gave 47 (26 mg, 74%; ∼15% H4): 1H NMR δ 0.99 (s,
(R)-4-Deuterio-2-(2-hydroxyethyl)-3(2H)-furanone (4-[2H]33) from
40. Treatment of 4017b (12 mg, 0.025 mmol) by procedure C [2 h, AIBN
(∼2 mg) added after 1 h] [with preparative HPLC (6 f 10% MeOH/
CH2Cl2)] gave 2′-O-tosylhomoadenosine17b (∼3 mg, 28%; spectral data
as listed17b with no 2H exchange for H3′), adenine (∼2 mg, 62%), and
9H), 4.02 (dd, J ) 11.5, 4.0 Hz, 1H), 4.10 (dd, J ) 11.5, 2.7 Hz, 1H),
4.45 (dd, J ) 4.0, 2.7 Hz, 1H), 5.75 (d, J ) 2.5 Hz, ∼0.85H), 7.39-
7.70 (m, 10H), 8.31 (d, J ) 2.5 Hz, 1H); 13C NMR δ (19.27, 26.61),
62.69, 85.34, 107.88, (127.74, 129.82, 132.68, 132.94, 135.57, 135.63),
178.77, 202.72; HRMS (FAB) m/z 375.1417 (100, MNa+ [C21H24O3-
SiNa] ) 375.1392), 376.1437 (20, MNa+ [C21H23DO3SiNa] ) 376.1455).
4-[2H]33 (∼2 mg, 62%; with ∼30% H at C4): UV max 259 nm; H
NMR (MeOH-d4) δ 5.71 (d, J ) 2.5 Hz, ∼0.7H), all other peaks same
as for 33; 13C NMR (MeOH-d4) δ 107.55 (∼30% diminished), all other
peaks same as for 33; HRMS (CI) m/z 129.0545 (100, MH+ [C6H9O3]
) 129.0552), 130.0619 (41, MH+ [C6H8DO3] ) 130.0614).
2
1
Acknowledgment. We thank the American Cancer Society
(Grant DHP-34) and Brigham Young University development
funds for support, and Mrs. Jeanny K. Gordon for assistance
with the manuscript.
(R)-2-{2-[(tert-Butyldimethylsilyl)oxy]ethyl}-3(2H)-furanone (34).
Methyl 6-O-(tert-butyldimethylsilyl)-2,5-dideoxy-R-D-glycero-hexo-
furanosid-3-ulose (39; 137 mg, 0.5 mmol) in 20% Et3N/MeOH (10
mL) was stirred for 2 h at ambient temperature. Volatiles were
evaporated, and the residue was dissolved (EtOAc). The solution was
washed (H2O, brine) and dried (Na2SO4). Volatiles were evaporated,
and the residue was chromatographed (15% EtOAc/hexanes) to give
Supporting Information Available: Full experimental
details, spectral data, and characterization for all compounds
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
1
34 (42 mg, 34%): UV max 260 nm; H NMR (MeOH-d4) δ 0.07 (s,
JA983449P