Studies on the mechanism of ribonucleotide reductases

Article abstract of DOI:10.1021/ja962974q

Ribonucleotide reductases are enzymes that catalyze the conversion of ribonucleotides to 2'-deoxyribonucleotides. This important reaction is initiated by the generation of a C-3' nucleotide radical and subsequent loss of the 2'-hydroxyl group. In order to model certain steps in this mechanism, selenol ester 23 was prepared and photolyzed providing the first selective chemical access to the 3'-adenosyl radical. From product analysis it could be shown that elimination of the 2'-OH function readily takes place under general base catalysis. The rate coefficient for this reaction was determined by competition kinetics to be 1.5 · 10⁶ s^-1 in the presence of 1 M triethylammonium acetate buffer at pH 7. Without catalyst the elimination rate is about 10³ times slower. It can be concluded that a similar mechanism is also feasible for the key steps of the enzyme catalyzed reaction.

Full text of DOI:10.1021/ja962974q

2784
J. Am. Chem. Soc. 1997, 119, 2784-2794
Studies on the Mechanism of Ribonucleotide Reductases
Roman Lenz and Bernd Giese*
Contribution from the Department of Chemistry, UniVersity of Basel,
St. Johanns-Ring 19, CH-4056 Basel, Switzerland
ReceiVed August 23, 1996X
Abstract: Ribonucleotide reductases are enzymes that catalyze the conversion of ribonucleotides to 2′-deoxyribo-
nucleotides. This important reaction is initiated by the generation of a C-3′ nucleotide radical and subsequent loss
of the 2′-hydroxyl group. In order to model certain steps in this mechanism, selenol ester 23 was prepared and
photolyzed providing the first selective chemical access to the 3′-adenosyl radical. From product analysis it could
be shown that elimination of the 2′-OH function readily takes place under general base catalysis. The rate coefficient
for this reaction was determined by competition kinetics to be 1.5‚106 s-1 in the presence of 1 M triethylammonium
acetate buffer at pH 7. Without catalyst the elimination rate is about 103 times slower. It can be concluded that a
similar mechanism is also feasible for the key steps of the enzyme catalyzed reaction.
Ribonucleotide reductases are key enzymes in all living
Scheme 1
organisms. They catalyze the transformation of ribonucleotides
1 to the corresponding 2′-deoxyribonucleotides 2 providing
the monomeric precursors required for DNA biosynthesis1
(Scheme 1).
These enzymes, therefore, play a critical role in the regulation
of cell growth and present a promising target for the design of
chemotherapeutic agents in the field of antitumor and antiviral
treatment.2 Despite dramatic differences in structure and
cofactor requirements3 a common radical-based reaction mech-
anism has been postulated by Stubbe1c,4 for all ribonucleotide
reductases (Scheme 2).
The key steps involve formation of C-3′ nucleotide radical 3
via hydrogen atom abstraction by a cofactor induced sulfur based
radical, protonation of the 2′-hydroxyl group of 3 by one of
two redox-active cysteine residues, and elimination of water
leading to the 2′-deoxy-3′-ketonucleotide radical 4, which is
subsequently reduced to the C-3′ radical 5 by the thiol/thiolate
pair. In the final step, the initially abstracted hydrogen atom
Ha is transferred back to the C-3′ position forming 2 with
concomitant regeneration of the protein radical. Indirect
evidence for the involvement of radical species in this enzymatic
reaction has been obtained by numerous biological studies using
isotopically labeled substrates,5 mechanism-based inhibitors,6
and site-directed mutants.7 In spite of several attempts using
modified substrates as radical traps for both enzyme8 and
substrate radicals,9 direct observation of nucleotide radical
intermediates was not possible until very recently.10 It was the
aim of our work to find an independent chemical access to C-3′
nucleoside radicals. In mechanistic studies of these radicals,
especially with regard to a potential 2′-hydroxy elimination,
important insight could be gained with respect to the key steps
of the proposed mechanism.
In the known studies concerning the dehydration of R,â-
dihydroxyalkyl radicals, the radical generation took place via
hydrogen atom abstraction by hydroxyl radicals. While this
method was very successful with simple model systems such
as ethylene glycol,12 it proved to be less useful with more
complex compounds like carbohydrates or nucleosides due to
unselective hydrogen abstraction from all carbon atoms of the
carbohydrate moiety13 and preferred addition of hydroxyl
radicals to nucleobases.14 Hence, it was necessary to develop
a method enabling us to selectively generate C-3′ nucleoside
radicals. One possible way to achieve this task is to introduce
a suitable functionality at C-3′ which is chemically stable but
can be cleaved off homolytically under mild conditions. A
X Abstract published in AdVance ACS Abstracts, March 1, 1997.
(1) (a) Thelander, L.; Reichard, P. Annu. ReV. Biochem. 1979, 48, 133-
158. (b) Eriksson, S.; Sjo¨berg, B. M. In Allosteric Enzymes; Herve´, G.,
Ed.; CRC Press: Boca Raton, 1990; pp 189-215. (c) Stubbe, J. AdV.
Enzymol. Relat. Areas Mol. Biol. 1990, 63, 349-417. (d) Thelander, L.;
Gra¨slund, A. In Metal Ions in Biological Systems; Sigel, H., Sigel, A., Eds.;
Marcel Dekker, Inc.: New York, 1994; pp 109-129.
(7) (a) Mao, S. S.; Johnston, M. I.; Bollinger, J. M.; Stubbe, J. Proc.
Natl. Acad. Sci. U.S.A. 1989, 86, 1485-1490. (b) Mao, S. S.; Holler, T. P.;
Yu, G. X.; Bollinger, J. M.; Booker, S.; Johnston, M. I.; Stubbe, J.
Biochemistry 1992, 31, 9733-9743. (c) Mao, S. S.; Holler, T. P.; Bollinger,
J. M.; Yu, G. X.; Johnston, M. I.; Stubbe, J. Biochemistry 1992, 31, 9744-
9751. (d) Mao, S. S.; Yu, G. X.; Chalfoun, D.; Stubbe, J. Biochemistry
1992, 31, 9752-9759.
(2) (a) Cory, J. G.; Chiba, P. In Inhibitors of Ribonucleotide Diphosphate
Reductase ActiVity; Cory, J. G., Cory, A. H., Eds.; CRC Press: Boca Raton,
FL, 1989; pp 245-264. (b) Robins, M. J.; Samano, M. C.; Samano, V.
Nucleotides Nucleosides 1995, 14, 485-493.
(8) Auguste, S. P.; Young, D. W. J. Chem. Soc., Perkin Trans. 1 1995,
395-404.
(9) Samano, V.; Robins, M. J. J. Am. Chem. Soc. 1992, 114, 4007-
4008.
(3) Reichard, P. Science 1993, 260, 1773-1777.
(10) Preliminary EPR studies with a 2′-fluoromethylene substituted
substrate analog suggest the presence of a nucleotide based radical.11
(11) van der Donk, W. A.; Yu, G.; Silva, D. J.; Stubbe, J.; McCarthy, J.
R.; Jarvi, E. T.; Matthews, D. P.; Reswick, R. J.; Wagner, E. Biochemistry
1996, 35, 8381-8391.
(12) Steenken, S.; Davies, M. J.; Gilbert, B. C. J. Chem. Soc., Perkin
Trans. 2 1986, 1003-1010 and references cited therein.
(13) von Sonntag, C.; Dizdaroglu, M. Carbohydr. Res. 1977, 58, 21-
30.
(4) (a) Ashley, G. W.; Stubbe, J. Pharmac. Ther. 1985, 30, 301-329.
(b) Stubbe, J; van der Donk, W. A. Chem. Biol. 1995, 2, 793-801.
(5) (a) Stubbe, J. Molec. Cell. Biochem. 1983, 50, 25-45. (b) Ashley,
G. W.; Harris, G.; Stubbe, J. J. Biol. Chem. 1986, 261, 3958-3964.
(6) (a) Thelander, L.; Larsson, A.; Hobbs, J.; Eckstein, F. J. Biol. Chem.
1976, 251, 1398-1405. (b) Harris, G.; Ator, M. A.; Stubbe, J. Biochemistry
1984, 23, 5214-5225. (c) Ator, M. A.; Stubbe, J. Biochemistry 1985, 24,
7214-7221. (d) Salowe, S. P.; Ator, M. A.; Stubbe, J. Biochemistry 1987,
26, 3408-3416. (e) Samano, V.; Robins, M. J. J. Org. Chem. 1991, 56,
7108-7113.
(14) von Sonntag, C. The Chemical Basis of Radiation Biology; Taylor
& Francis: London, 1987; pp 116-166.
S0002-7863(96)02974-5 CCC: $14.00 © 1997 American Chemical Society

Studies on the Mechanism of Ribonucleotide Reductases
J. Am. Chem. Soc., Vol. 119, No. 12, 1997 2785
Scheme 2
functionality fulfilling these criteria is the selenol ester group.
This group is usually easy to introduce15 and stable under non-
basic conditions and in the absence of good nucleophiles. Due
to its weak Se-C bond the selenol ester moiety is cleaved either
by directly using UV light (λ > 320 nm) or in a photochemically
induced radical chain process in the presence of a mediator such
as Bu3SnH.15 The initially formed acyl radical readily decar-
bonylates16 leading to the desired nucleoside radical.
In this paper we describe the synthesis of 3′-selenocarbonyl
substituted nucleoside derivatives based on adenosine 6 and their
application as precursors of C-3′ nucleoside radicals in studies
to clarify the mechanism of the biosynthesis of 2′-deoxyribo-
nucleotides.
at the 3′-position (Scheme 3). This stereochemistry can be
rationalized by an attack of the organometallic reagent from
the R-face of the carbohydrate moiety due to the efficient
shielding of the â-face by the nucleobase. This assumption is
supported by results obtained when organometallic ethynyl
derivatives were added to 3′-ketonucleosides.23 In these cases,
selectivities were found between 1.5:1 and 50:1 in favor of the
R-attack depending on the reagent and the nucleobase.
The conversion of the vinyl grignard into the less basic cerium
reagent is decisive for the addition reaction since it prevents
the elimination of the nucleobase which occurs readily after
deprotonation at the C-2′ position of the carbohydrate residue.24
After silylation of the 3′-hydroxyl group in 10 with Horning’s
reagent,25 compound 11 was readily ozonolyzed, and subsequent
reductive workup with dimethyl sulfide26 afforded aldehyde 12
in 68% yield. Further oxidation under mild conditions using
potassium permanganate27 led to carboxylic acid 13 in 77% yield
(Scheme 3).
The coupling constant J1′,2′ of less than 0.5 Hz in the NMR
spectrum of aldehyde 12 refers to a dihedral angle between
C-1′-H-1′ and C-2′-H-2′ of nearly 90° which is characteristic
for a 3′-endo conformation (3T2 conformation) of the carbohy-
drate residue (Figure 1). Additional NOE experiments were
carried out on aldehyde 12 to ensure the proposed configuration
at C-3′. Indeed, a strong NOE effect (22% and 14%) was
observed between H-4′ and the hydrogen of the aldehyde
function which indicates a cis-arrangement of these two sub-
stituents and therefore a D-xylose configuration of the carbo-
hydrate moiety.
Results and Discussion
Synthesis of the Radical Precursors. The first step on the
way to 3′-phenylselenocarbonyl adenosine derivatives involved
protection of the 2′- and 5′-hydroxyl groups. Silylation with
tert-butyldimethylsilyl chloride in pyridine afforded the desired
product 7 in 50% yield.18 A three-step procedure allowed the
protection of the nucleobase: silylation of the 3′-hydroxyl group
was followed by dibenzoylation of the nucleobase in pyridine
and final deprotection of the hydroxyl group under acidic
conditions led to protected nucleoside 8 in quantitative yield.
The free alcohol function of 8 was then oxidized with the Dess-
Martin reagent19,20 to the corresponding ketone 9 in 99% yield.
The subsequent addition of a vinylcerium reagent,21,22 generated
in situ from vinylmagnesium bromide and cerium trichloride,
led in 82% yield to allyl alcohol 10 having the (S)-configuration
(21) Imamoto, T.; Takiyama, N.; Nakamura, K.; Hatajima, T.; Kamija,
Y. J. Am. Chem. Soc. 1989, 111, 4392-4398.
(22) Bender, S. L.; Moffett, K. K. J. Org. Chem. 1992, 57, 1646-
1647.
(15) For a recent overview of methods for the preparation of selenol
ester, see: Ogawa, A.; Sonoda, N. In ComprehensiVe Organic Synthesis;
Trost, B. M., Ed.; Pergamon Press: New York, 1991; Vol. 6, pp 461-468.
(16) For examples of the use of acyl selenides in radical chemistry, see
references cited in Crich, D.; Yao, Q. J. Org. Chem. 1996, 61, 3566-
3570.
(23) (a) Huss, S.; De las Heras, F. G.; Camarasa, M. J. Tetrahedron
1991, 47, 1727-1736. (b) Jung, P. M. J.; Burger, A.; Bielmann, J.-F.
Tetrahedron Lett. 1995, 36, 1031-1034.
(17) In the case of the pivaloyl radical the rate of decarbonylation is
1.2‚105 s-1 at 25 °C, see: Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980,
13, 317-323.
(24) For addition reactions of alkylcerium reagents to easily enolizable
ketones, see: Imamoto, T.; Sugiura, Y.; Takiyama, N. Tetrahedron Lett.
1984, 25, 4233-4236.
(18) Ogilvie, K. K.; Beaucage, S. L.; Schifman, A. L.; Theriault, N. Y.;
Sadana, K. L. Can. J. Chem. 1978, 56, 2768-2780.
(25) TMSCl/TMSimidazole/N,O-bisTMSacetamide (2:3:3), commercially
available: Fluka Product No. 85432. See: Chambaz, E. M.; Horning, E.
C. Anal. Biochem. 1969, 30, 7-24.
(19) For the preparation of the Dess-Martin reagent, see: (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.
(26) Pappas, J. J.; Keaveney, W. P.; Gancher, E.; Berger, M. Tetrahedron
Lett. 1966, 4273-4278.
(20) (a) Samano, V.; Robins, M. J. J. Org. Chem. 1990, 55, 410-412.
(b) Samano, V.; Robins, M. J. J. Org. Chem. 1990, 55, 5186-5188.
(27) Abiko, A.; Roberts, J. C.; Takemasa, T.; Masamune, S. Tetrahedron
Lett. 1986, 27, 4537-4540.

2786 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Lenz and Giese
Scheme 3a
Scheme 4a
a (a) TBDMSCl, pyr, room temperature, 50%. (b) TMSCl, pyr, 0
°C f room temperature, then BzCl, 0 °C f room temperature. (c)
cat. TsOH, EtOH, room temperature, quantitative (from 7). (d) Dess-
Martin periodinane,19 CH2Cl2, room temperature, 99%. (e)
H2CdCHMgBr, CeCl3, THF, -78 °C, then 9, -78 °C, 82%. (f)
Horning’s reagent,25 CH2Cl2, room temperature (g) O3, CH2Cl2, -78
°C, then Me2S, -78 °C f room temperature, 68% (from 10). (h) 1 M
a (a) CDT, DMF, 0 °C f room temperature, then PhSeH, 0 °C,
79%. (b) BF3‚OEt2, CH2Cl2, reflux, 98%. (c) 5% HCl in EtOH, 50
°C, 79%.
With regard to our planned model studies on the 2′-
deoxygenation of ribonucleotides, we needed to synthesize a
3′-selenocarbonyl nucleoside bearing a free 2′-hydroxyl function.
Therefore, we decided to replace the tert-butyldimethylsilyl
residues by protecting groups which would be less stable toward
acidic hydrolysis, prior to the introduction of the selenol ester
function. Starting from 10, reaction with tetrabutylammonium
fluoride in THF31 at 0 °C led to the deprotected vinyl nucleoside
17 in good yield. Standard conditions (acetone/camphorsulfonic
acid (CSA)36 or 2,2-dimethoxypropane/CSA/DMF37 ) for ac-
etonide formation did not lead to product 18. However, the
use of the strong Lewis acid aluminum trichloride38 in the pres-
ence of acetone gave rise to a quick reaction, and the acetonide
18 was obtained in almost quantitative yield. Consecutive
treatment with acetic anhydride in pyridine39 afforded the 2′-
O-acetyl derivative 19 in 97% yield. The conversion of 19 into
the corresponding carboxylic acid 21 was carried out following
the two-step procedure described earlier (ozonolysis under
reductive conditions to 20 and subsequent permanganate oxida-
tion). The selenol ester 22 was obtained in 72% yield by reac-
tion of acid 21 with phenyl dichlorophosphate to the mixed
anhydride which was treated in situ with benzeneselenol.40 Com-
plete deprotection of the carbohydrate moiety and the nucleobase
took place in 78% yield in the presence of 5% HCl in ethanol.
Radical precursor 23 was obtained as crystalline hydrochloride
and could be easily separated from side products by filtration.
On the other hand, treatment of 22 with 80% trifluoroacetic
acid41 at room temperature led only to the hydrolysis of the
acetonide and afforded 24 in 86% yield (Scheme 5).
t
KMnO4, BuOH, phosphate buffer, room temperature, then saturated
Na2SO3, HCl (f pH 3), 0 °C, 77%.
Figure 1. NOE effects and coupling constant J1′,2′ in nucleoside 12 in
agreement with a xylo configuration and a 3′-endo conformation.
Activation of the carboxylic acid function in 13 to the
corresponding acid chloride by reaction with either dichloro-
methyl methyl ether28 or an R-chloroenamine29 was unsuccess-
ful. We then turned our attention to a method leading directly
from 13 to the protected selenol ester 14 using acyl 1,2,4-
triazolides as reactive intermediates.30 These activated acyl
compounds can be produced in situ and react with selenols to
give the corresponding selenol esters under mild conditions and
in good yields. Thus, compound 13 was treated with carbonyldi-
1,2,4-triazole (CDT) at room temperature to form the triazolide
which reacted subsequently with benzeneselenol to give 14 in
79% yield. The deprotection of the carbohydrate moiety in 14
was challenging since selenol esters react easily with nucleo-
philes, and consequently the choice of the reaction conditions
was limited. Thus, treatment of 14 with fluoride in aprotic me-
dium31 led only to decomposition products. On the other hand,
the use of trifluoroboron etherate32 in dichloromethane allowed
solely for the obtaining of the 5′-deprotected compound 15 in
98% yield. Other methods allowing for hydrolysis of silyl ethers
(TBAF/AcOH/THF, HFaqueous/CH3CN,33 NH4F/CH3OH,34 Dowex
50W-X8/CH3OH35) resulted also in sole deprotection at the 5′-
position. Additional cleavage of the benzoyl group at the
nucleobase was achieved by using stronger acidic conditions.
Thus, treatment of compound 14 with 5% HCl in ethanol at 50
°C gave 16 in 79% yield (Scheme 4). Interestingly, the selenol
ester function was stable under these conditions.
(33) (a) Newton, R. F.; Reynolds, D. P.; Finch, M. A. W.; Kelly, D. R.;
Roberts, S. M. Tetrahedron Lett. 1979, 3981-3984. (b) Collington, E. W.;
Finch, H.; Smith, I. J. Tetrahedron 1985, 26, 681-684.
(34) (a) White, J. D.; Amedio, J. C., Jr.; Gut, S.; Jayasinghe, L. J. Org.
Chem. 1989, 54, 4268-4271. (b) Zhang, W.; Robins, M. J. Tetrahedron
Lett. 1992, 33, 1177-1180.
(35) Corey, E. J.; Ponder, J. W.; Ulrich, P. Tetrahedron Lett. 1980, 21,
137-140.
(36) (a) Schmidt, O. Th. In Methods in Carbohydrate Chemistry;
Whistler, R. L., Wolfrom, M. L., Eds.; Academic Press Inc.: New York,
1963; Vol. 2, pp 318-325. (b) de Belder, A. N. AdV. Carbohydr. Chem.
1965, 20, 219-302.
(28) Ottenheijm, H. C. J.; DeMan J. H. M. Synthesis 1975, 163-164.
(29) Devos, A.; Remion, J.; Frisque-Hesbain, A.-M.; Colens, A.; Ghosez,
L. J. Chem. Soc., Chem. Commun. 1979, 1180-1181.
(30) Gais, H.-J. Angew. Chem. 1977, 89, 251-252.
(31) Corey, E. J.; Venkateswarlu, A. J. Am. Chem. Soc. 1972, 94, 6190-
6191.
(37) Lipshutz, B. H.; Barton, J. C. J. Org. Chem. 1988, 53, 4495-4499.
(38) Lal, B.; Gidwani, R. M.; Rupp, R. H. Synthesis 1989, 711-713.
(39) (a) Weber, H.; Khorana, H. G. J. Mol. Biol. 1972, 72, 219-249.
(b) Zhdanov, R. I.; Zhenodarova, S. Synthesis 1975, 222-245.
(40) Ireland, R. E.; Norbeck, D. W.; Mandel, G. S.; Mandel, N. S. J.
Am. Chem. Soc. 1985, 107, 3285-3294.
(32) Kelly, D. R.; Roberts, S. M.; Newton, R. F. Synth. Commun. 1979,
9, 295-299.
(41) Leblanc, Y.; Fitzsimmons, B. J.; Adams, J; Perez, F.; Rokash, J. J.
Org. Chem. 1986, 51, 789-793.

Studies on the Mechanism of Ribonucleotide Reductases
J. Am. Chem. Soc., Vol. 119, No. 12, 1997 2787
Scheme 5a
By using radical precursor 24 with acetate at C-2′ as a better
leaving group, it was possible to demonstrate that the formation
of free nucleobase is fully related to the 2′-deoxygenation
(Scheme 6 and Table 2). The first irradiation experiments in
the presence of Bu3SnH lead to N-benzoyladenine (33), enolone
34, and acetic acid in high yields. The 2′-deoxy-3′-ketonucleo-
side 31 was only detected in 0 and 15% yields, respectively
(entry 1 and 2). The amount of 31 could be increased to 52%
under milder conditions avoiding the workup step (evaporation
of the solvent for subsequent NMR analysis) by using CD3OD
as solvent (entry 3). Due to the use of Bu3SnD as trapping
reagent a selective deuteration at C-2′ of 31 (>90% as estimated
by NMR) and at C-3 of 34 (52% by GC/MS) could be observed.
Only in the presence of the more reactive radical scavenger
tBuSH was it possible to detect the formation of the direct
reduction product 27, exclusively in the xylo configuration (entry
t
4). Through H/D exchange betweeen CD3OD and BuSH a
a (a) TBAF, THF, 0 °C, 82%. (b) AlCl3, acetone/Et2O 2:1, 0 °C,
99%. (c) Ac2O, pyr, 0 °C f room temperature, 97%. (d) O3, CH2Cl2,
-78 °C, then Me2S, -78 °C f room temperature, 69%. (e) 1 M
selective deuteration at C-3′ of 27 (>80% by NMR) took place.
The results in Table 2 clearly indicate that the free nucleobase
33 is quantitatively formed from the 2′-deoxy-3′-ketonucleoside
31, since the decrease in the yield of nucleobase 33 (57 and
40%) parallels an increase in the yield of ketone 31 by 52 and
37%, respectively (entry 1-3). In addition, the free nucleobase
33 and enolone 34, are detected in nearly the same yield under
mild conditions (entry 3). This leads to further support of the
assumption that the nucleobase 33 is entirely formed from
ketone 31. The somewhat lower yields of enolone 34 in
comparison to free nucleobase 32 or 33 in the other experiments
are probably due to further reactions of 34 acting as a Michael
acceptor.43
The results listed in Tables 1 (entry 1-3) and 2 moreover
indicate that the elimination of acetic acid from the initially
formed C-3′ radical is a fast process (compared to the elimina-
tion of water) which competes successfully with the direct
trapping reaction to 27. This observation is in agreement with
pulse radiolytic studies of ethylene glycol derivatives by Schulte-
Frohlinde. He could show that elimination of acetic acid45
occurs in neutral aqueous solution about 103 times faster than
elimination of water46 (Scheme 7).
All irradiation experiments discussed so far were conducted
in methanol. In order to simulate the reaction conditions found
in an enzyme pocket more accurately, mixtures of acetonitrile
and aqueous buffer solutions were employed. We chose the
triethylammonium acetate buffer (TEAA buffer, pH 7) which
is volatile and therefore easy to remove. Since Bu3SnH is not
soluble under these conditions, it had to be replaced by the water
soluble tris[3-(2-methoxyethoxy)propyl]stannane (35) described
by Breslow.47 The irradiation of selenol ester 23 using tin
hydride 35 and 0.1 M TEAA buffer led to an astonishing result
(Table 1, entry 4). Despite the presence of the efficient H-donor
35 mainly the products following the 2′-OH elimination were
formed.49 This indicates that the use of a buffered aqueous
solution strongly accelerates the elimination of the 2′-hydroxy
group from the C-3′ nucleoside radical.
t
KMnO4, BuOH, phosphate buffer, room temperature, then saturated
Na2SO3, HCl (f pH 3), 0 °C, 91%. (f) PhOP(O)Cl2, NEt3, THF, 0 °C,
then PhSeH, NEt3, 0 °C, 72%. (g) 5% HCl in EtOH, 50 °C, 78%. (h)
80% CF3COOH, room temperature, 86%.
Photolysis of the Radical Precursors. All irradiation experi-
ments were carried out at 20 °C in methanol or acetonitrile/
water (1:1) and, in most cases, in the presence of an H-donor
such as Bu3SnH or tBuSH as radical scavenger. The composi-
tion of the crude product mixture was quantitatively analyzed
by 1H-NMR spectroscopy using either pentachloroethane or the
total integral of the two adenine protons as internal standards.
Photolysis reactions of radical precursor 23 in methanol in
t
the presence of Bu3SnH and BuSH led exclusively to the
formation of the direct reduction products 6 in good yields
(Scheme 6 and Table 1, entries 1 and 2). The preferred attack
of the H-donor from the R-face of the C-3′ nucleoside radical
25 is due to the efficient â-face shielding by the nucleobase.
Through H/D exchange between the deuterated solvent and the
thiol, a selective deuteration at C-3′ (>80% as estimated by
1H-NMR) could be observed (entry 2). These results clearly
indicate that the selenol ester 23 is an excellent radical precursor
and that the selective C-3′ radical generation can either be
achieved by a photoinduced chain reaction (entry 1) or direct
photolysis (entry 2). The fact that in the presence of efficient
t
radical traps like Bu3SnH or BuSH the formation of 2′-
deoxygenated products could not compete with direct reduction
led to the use of Bu6Sn2 as radical mediator (entry 3). In this
case due to the much slower radical trapping reaction by H
abstraction from the solvent, 57% of adenine (32), 10% of 2,3-
dihydro-2-hydroxymethylfuran-3-one (34), and only 35% of the
reduction products 6 were formed.
Although the appearance of the free nucleobase in nucleoside
reactions is often indicative for unspecific degradation, the
presence of enolone 34 suggests that adenine (32) might be
formed from 2′-deoxy-3′-ketoadenosine (30) by an elimination
reaction. The elimination of the nucleobase from 2′-deoxy-3′-
ketonucleosides is a well-known reaction42,43 and is also
discussed by Stubbe as an important step in the ribonucleotide
reductase mechanism when substrate-based inhibitors6c,44 or site-
directed mutated enzymes7c were employed. The absence of
free nucleobase 32 under the conditions of entry 1 shows that
adenine (32) is not formed directly from the radical 25 or the
radical precursor 23.
Competition Kinetic Studies. To gain a better understanding
of the factors controlling the radical induced dehydration,
(43) Schreiber, S. L.; Ikemoto, N. Tetrahedron Lett. 1988, 29, 3211-
3214.
(44) Ator, M. A.; Stubbe, J. J. Biol. Chem. 1986, 261, 3595-3599.
(45) Matsushige, T.; Koltzenburg, G.; Schulte-Frohlinde, D. Ber. Bun-
senges. Phys. Chem. 1975, 79, 657-661.
(46) Burchill, C. E.; Perron, K. M. Can. J. Chem. 1971, 49, 2382-2389.
(47) (a) Light, J.; Breslow, R. Tetrahedron Lett. 1990, 31, 2957-2958.
(b) Light, J.; Breslow, R. Org. Synth. 1995, 72, 199-207.48
(48) The boiling point of the tin bromide is not given correctly in this
reference. In order to purify this compound, we applied Kugelrohr
destillation at 240-250 °C and 0.05 mbar.
(42) (a) Sugiyama, H.; Fujimoto, K.; Saito, I. J. Am. Chem. Soc. 1995,
117, 2945-2946. (b) Hansske, F.; Madej, D.; Robins, M. J. Tetrahedron
1984, 40, 125-135. (c) Binkley, R. W.; Hehemann, D. G.; Binkley, W.
W. J. Org. Chem. 1978, 43, 2573-2576.
(49) In addition, the irratiation time could be reduced from over 2 h
(Table 1, entry 1-3) to less than 5 min (entry 4).

2788 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Lenz and Giese
Scheme 6
Table 1. Results of Irradiation Reactions with Radical Precursor
23
Table 2. Results of Irradiation Reactions with Radical Precursor
24
entry
reaction conditions
6 (xylo/ribo) 32 (%) 34 (%)
27
31
33
34
AcOH
(%)
entry
reaction conditions
(%) (%) (%) (%)
1
2
3
4
Bu3SnH, CH3OH, 2.5 h
tBuSH, CD3OD, 2.5 h
Bu6Sn2, CH3OH, 2.5 h
94% (3:1)
78%a (4:1)
1
2
3
4
Bu3SnH, CH3OH, 3 ha
Bu3SnH, CH3OH, 3 ha
Bu3SnD, CD3OD, 3 h
tBuSH, CD3OD, 3 h
84
67
27
70
78
90
88
92
50
15
41
35% (3:1)
57
80
10
50
52b
25c
54f
R3SnH (35)b, 0.1 M TEAAc 14% (4:1)
9d,e
CH3CN/H2O 1:1, 5 min
a Workup: evaporation of solvent (reduced pressure, 40 °C). b >90%
deuterated at C-2′ (NMR). c 52% deuterated at C-3 (GC/MS). d Ex-
clusively with xylo configuration. e >80% deuterated at C-3′ (NMR).
a
>80% deuteration at C-3′ (NMR). b R ) 3-(2-methoxyethoxy)-
propyl. c Triethylammonium acetate.
f
∼40% deuterated at C-3 (NMR).
competition kinetic experiments were accomplished. As il-
lustrated below, the unimolecular dehydration process (25 f
28, with rate coefficient kobs) competes with the bimolecular
direct reduction (25 f 6, with rate coefficient kH).
Scheme 7
a factor of 5, probably due to stabilization of the polar transition
state.53 Moreover, results of Beckwith54 and Newcomb55,56
show that kH decreases by about the same amount on going
from alkyl to R-methoxyalkyl and R,â-dimethoxyalkyl radicals,
the latter being more accurate models for the C-3′ nucleoside
radical 25. This behavior is rationalized by the higher stability
of R-oxyalkyl and especially R,â-dioxyalkyl radicals and by the
unfavorable transition state between the electron rich radical
and the nucleophilic H-atom donor. A combination of both
solvent and substituent effect allowed us to set kH in the present
When the tin hydride 35 involved in the reduction step is
used in large excess with respect to the radical precursor, this
second-order reaction can be treated as a pseudo-first-order
process. Under this constraint and provided that the competing
reactions are irreversible and the products observed are stable
under the reaction conditions, the ratio of the rate coefficients
can be calculated directly from the product distribution and the
initial radical trap concentration according to eq 1. Since ketone
30 was not detected by NMR, the concentration of adenine (32),
being the only stable product of the dehydration sequence, was
used as a measure for the amount of originally formed 28.50
The product ratio [32]/[6] was determined using reversed-phase
HPLC.
case to approximately 1‚106 M-1 s-1 57
.
The results of a series of competition reactions with radical
precursor 23 (3 mM in acetonitrile/water 1:1) in the presence
of 0.1 M TEAA buffer (pH 7) and various concentration of tin
hydride 35 (17-155 mM) are shown in Figure 2.
The excellent linear dependence of the product ratio [32]/[6]
on the reciprocal tin hydride concentration 1/[35] provides
support for the applicability of eq 1. From the slope a ratio of
kobs/kH ) 0.142 M is obtained, which leads to a rate coefficient
kobs for the elimination of water from radical 25 of 1.4‚105 s-1
at 20 °C. In the absence of buffer or in the presence of 0.5 M
sodium chloride, values for kobs of only 2‚103 s-1 were measured
in single kinetic determinations. These results clearly show that
kobs
kH
[32]
[6]
1
[35]
)
‚
(1)
In order to come to an absolute value of kobs the rate coefficient
kH for the H-atom transfer from tin hydride 35 to the C-3′
nucleoside radical 25 has to be accurately determined. Using
a calibrated radical clock,51 kH for the H-transfer from 35 to a
tertiary alkyl radical at 20 °C in acetonitrile was measured to
be 1.2‚106 M-1 s-1 in good accord to the corresponding value52
for Bu3SnH (kH ) 1.6‚106 M-1 s-1). Changing the solvent to
acetonitrile/water 1:1 led to an acceleration of the reaction by
(53) For an analogous explanation using thiols as H-donors see ref 51.
(54) Beckwith, A. L. J.; Glover, S. A. Aust. J. Chem. 1987, 40, 157-
173.
(55) Johnson, C. C.; Horner, J. H.; Tronche, Ch.; Newcomb, M. J. Am.
Chem. Soc. 1995, 117, 1684-1687.
(50) This assumption is further supported by the fact that the combined
yield of 6 and 32 (as judged by HPLC) exceeded 80% in every kinetic
experiment.
(51) Tronche, Ch.; Martinez, F. N.; Horner, J. H.; Newcomb, M.; Senn,
M.; Giese, B. Tetrahedron Lett. 1996, 37, 5845-5848.
(52) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc.
1981, 103, 7739-7742.
(56) Dhanabalasingham, B.; Martinez, F. N.; Simakov, P. A.; Newcomb,
M. Unpublished results.
(57) Recent results by Newcomb56 support this value. He determined a
preliminary value of 1.3‚106 M-1 s-1 for the H-atom transfer from the tin
hydride 35 to a secondary R,â-dimethoxyalkyl radical in acetonitrile/water
(1:1) at 34 °C.

Studies on the Mechanism of Ribonucleotide Reductases
J. Am. Chem. Soc., Vol. 119, No. 12, 1997 2789
1
Figure 4. pH dependence of the relative rate coefficients kobs/kH from
reactions of radical 25 in acetonitrile/water (1:1) in the presence of tin
hydride 35.
Figure 2. Ratios of [32] to [6] formed from reactions of radical 25 in
acetonitrile/water (1:1) in the presence of 0.1 M triethylammonium
acetate and various concentrations of tin hydride 35.
Scheme 8
Table 3. Catalytic Effect of Various Buffers on the Dehydration
Reaction (25 f 28)
buffer
pH
kobs/kH
acceleration ratea
10 mM phosphate
10 mM phosphate
25 mM imidazole
100 mM piperazine
100 mM TEAFb
20 mM citrate
8
7
7
6
6
5
0.244 M
0.082 M
0.024 M
0.023 M
0.016 M
0.006 M
162
55
16
16
11
4
a Compared with the reaction in the absence of buffer. b Triethyl-
ammonium formate.
was determined in the range from pH 3-6 (10 mM citrate
buffer) and from pH 5-7.5 (0.1 M phosphate buffer).
The two plots in Figure 4 clearly indicate that with increasing
proton concentration the rate of elimination steadily decreases.
Over the whole pH range studied, the elimination of water from
the C-3′ nucleoside radical 25 is therefore subject to general
base catalysis.59 Consequently in the case of the TEAA buffer
the catalytically active species is acetate. For the observed rate
coefficient we can therefore write kobs ) 1.5‚106 M-1 s-1‚[AcO-].
General Base Catalysis in the Enzyme Reaction Mecha-
nism. In general base catalyzed reactions the rate determining
step is a proton transfer from the substrate to the base which is
followed by a rapid transformation of the intermediate to the
product. That means for the reaction investigated in this study
that the C-3′ nucleoside radical 25 is deprotonated by acetate
to the ketyl 36. In a subsequent, faster step the 2′-hydroxy group
of 36 is eliminated as OH-, leading to the carbonyl conjugated
radical 28 (Scheme 8, top). The rate of the overall reaction is
solely dependent on the first step and is thus directly proportional
Figure 3. Relative rate coefficients kobs/kH from reactions of radical
25 in acetonitrile/water (1:1) in the presence of tin hydride 35 and
various concentrations of triethylammonium acetate (TEAA).
the acceleration of the elimination reaction is due to the TEAA
buffer and that its influence is not a salt effect. As the results
in Table 3 illustrate, the catalytic effect on the dehydration is
not restricted to TEAA. The acceleration factor in the last
column refers to the competition kinetic in the absence of buffer
(kobs/kH ) 1.5‚10-3 M).
In order to obtain a better understanding of the dependence
of the elimination step on the TEAA buffer, the ratio of kobs/kH
was measured by single kinetic determination for buffer
concentrations ranging from 10 to 500 mM.
From the plot in Figure 3 it is evident that the rate of the
elimination reaction is linearly dependent on the buffer con-
centration according to the relation kobs ) 1.47‚kH‚[TEAA] or
kobs ) 1.5‚106 M-1 s-1‚[TEAA], respectively. Such a linear
rate dependence on the buffer concentration is characteristic of
reactions following general acid or base catalysis.58 Studying
the pH dependence of the dehydration reaction should allow us
to specify the general catalysis. Therefore, the ratio of kobs/kH
to the acetate concentration (with kobs ) kcat‚[AcO-] and kcat
)
1.5‚106 M-1 s-1).
Although an analogous base catalyzed mechanism has already
been suggested1c,4a for the dehydration step in the ribonucleotide
reductase mediated reaction based on model studies using simple
R,â-dihydroxyalkyl radicals under specific hydroxide catalysis,60
it never became as popular as the alternative acid catalyzed
(58) (a) Jencks, W. P. Chem. ReV. 1972, 72, 705-718. (b) Jencks, W.
P. Acc. Chem. Res. 1976, 9, 425-432. (c) Hoffmann, R. W. Aufkla¨rung
Von Reaktionsmechanismen; Thieme Verlag: Stuttgart, 1976; pp 231-243.
(59) General base catalysis has also been observed in the radical-induced
dehydration reaction of carbohydrates, see: (a) Tamba, M.; Quintiliani, M.
Radiat. Phys. Chem. 1984, 23, 259-263. (b) Akhlaq, M. S.; Al-Baghdadi,
S.; von Sonntag, C. Carbohydr. Res. 1987, 16, 71-83.

2790 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Lenz and Giese
Scheme 9
Figure 5. Schematic drawing of the active-site region of the E. coli
ribonucleotide reductase with a plausible position for a model-built
substrate molecule.
mechanism shown in Scheme 2. According to our findings,
this acid catalyzed mechanism, however, seems not to be
operative under reaction conditions where weak acids are used.
This can be rationalized by a much more unfavored equilibrium
in the case of the general acid catalyzed reaction compared with
the corresponding general base catalysis (Scheme 8, bottom).
Using the known values for 2-deoxyribose radicals12,61 the
pKa value for the 3′-hydroxy group of radical 25 and the
protonated 2′-hydroxy group of radical 37 can be estimated to
9.861 and 1.4,12 respectively.62 Thus, under neutral conditions
using triethylammonium acetate as buffer this leads to a more
advantageous equilibrium for the general base catalyzed process
(K1 ≈ 10-5) than for the reaction under general acid catalysis
(K2 ≈ 3‚10-10). Because the difference of the pKa values is
large, one can assume for both reactions that the faster proton
transfer steps are diffusion-controlled63 (k-1 ≈ k-2 ≈ 1010 M-1
s-1). The rate coefficients k1 and k2 therefore become 105 and
3 M-1 s-1, respectively. This estimation leads to the conclusion
that the deprotonation step 25 f 36 is much faster than the
protonation step 25 f 37. The fair agreement between the
measured value for kcat (≈106 M-1 s-1) and k1 supports this
interpretation.
For a general base catalysis to appear in the ribonucleotide
reductase mediated process, an appropriate basic functionality
has to be present at a suitable position in the active cavity of
the enzyme. That this is indeed the case has been shown only
recently by Uhlin and Eklund.64 They crystallized the R1
subunit of the E. coli reductase thus providing the first detailed
structure of the active center of a ribonucleotide reductase
(Figure 5).
In this model the three polar amino acid side chains from
serine (S224), asparagine (N437), and glutamate (E441), all of
them highly conserved in other reductases, are forming hydrogen
bonds with O-4′, O-2′, and O-3′, respectively. Supposing this
model to be correct, E441 would thus be perfectly positioned
to act as a general base able to deprotonate the 3′-hydroxy group
after generation of the C-3′ nucleotide radical.
reactivity, the acetate group and the glutamate side chain behave
very similarly having pKa values of 4.76 and 4.4, respectively.
There is, however, a fundamental difference between enzyme
and solution chemistry. While catalysis in solution is intermo-
lecular showing second-order kinetics, reactions in an enzyme-
substrate complex are better described as intramolecular pro-
cesses with first-order kinetics. An appropriate model for an
enzyme reaction is therefore the corresponding reaction in
solution where the catalyst is an integral part of the molecule.65
Furthermore, it is well-known that in the case of general acid-
base catalysis intramolecularly catalyzed processes are about
as efficient as their intermolecular counterparts at concentrations
for the general catalyst of 1-10 M.66 Using such a value for
the acetate concentration in the present study, the rate coefficient
for the elimination of water in the ribonucleotide reductase
mediated process can be estimated to 106-107 s-1, which
signifies a rate enhancement of at least 103 as compared to the
uncatalyzed dehydration reaction. Such a strongly accelerated
deoxygenation step is absolutely essential for the enzyme
mechanism.
Since it is known from model studies by von Sonntag67 and
Asmus68 that hydrogen atom abstraction from hydroxyalkyl
compounds by thiyl radicals (kH ≈ 103-104 M-1 s-1) is 104-
105 times slower than the reverse reaction, an unfavorable
f
equilibrium for the corresponding enzyme reaction (1
3,
r
Scheme 2) can be expected. In order to drive this equilibrium
to the products, it has to be followed by a fast and irreversible
step such as the proposed general base catalyzed â-elimination,
which is able to compete successfully with the hydrogen atom
transfer from cysteine 439 to the C-3′ radical 3.
The analogy between our model reaction and the enzyme
mechanism ends with the reduction of 4 to the 2′-deoxy-3′-
ketonucleotide 38 (Scheme 9). To prevent elimination of the
nucleobase from 38, as observed in the present study, a rapid
reduction of the carbonyl group by a disulfide radical anion
has been postulated.1c,4 This reaction is thermodynamically
not feasible in the absence of acid, since the reduction of a
ketone to the corresponding ketyl (E0 ) -2.1 V for the isopropyl
ketyl69) by a disulfide radical anion (E0 ) -1.6 V for the
The question now arises to what extent the results of the
general catalysis by acetate in solution can be compared with
an enzymatic catalysis by glutamate. As to the chemical
(60) (a) Bansal, K. M.; Gra¨tzel, M.; Henglein, A.; Janata, E. J. Phys.
Chem. 1973, 77, 16-19. (b) Steenken, S. J. Phys. Chem. 1979, 83, 595-
599.
(61) Hayon, E.; Simic, M. Acc. Chem. Res. 1974, 7, 114-121.
(62) Compared with alcohols (pKa(ROH) ≈ 17, pKa(ROH2+) ≈ -2),
the presence of the carbon-centered radical in 25 therefore increases the
aciditiy of the R-hydroxy group by more than 7 orders of magnitude, while
the basicity of the â-hydroxy group is only enhanced by 3-4 orders of
magnitude.
(65) (a) Jencks, W. P. Catalysis in Chemistry and Enzymology;
McGraw-Hill: New York, 1969. (b) Page, M. I. In The Chemistry of Enzyme
Action; Page, M. I., Ed.; Elsevier: Amsterdam, 1984, pp 229-269.
(66) Page, M. I. Chem. Soc. ReV. 1973, 2, 295-323.
(67) Akhlaq, M. S.; Schuchmann, H.-P.; von Sonntag, C. Int. J. Radiat.
Biol. 1987, 51, 91-102.
(63) Eigen, M. Angew. Chem., Int. Ed. Engl. 1964, 3, 1-19.
(64) Uhlin, U.; Eklund, H. Nature 1994, 370, 533-539.
(68) Scho¨neich, C.; Bonifacic, M.; Asmus, K.-D. Free Rad. Res.
Commun. 1989, 6, 393-405.

Studies on the Mechanism of Ribonucleotide Reductases
J. Am. Chem. Soc., Vol. 119, No. 12, 1997 2791
as a white foam: Rf ) 0.28 (EtOAc/pentane 1:3); 1H NMR (300 MHz,
CDCl3) δ 8.65 (s, 1H), 8.39 (s, 1H), 7.87-7.84 (m, 4H), 7.51-7.45
(m, 2H), 7.37-7.32 (m, 4H), 6.15 (d, J ) 5.3 Hz, 1H), 4.65 (t, J ) 5.1
Hz, 1H), 4.26 (m, 1H), 4.23 (m, 1H), 3.99 (dd, J ) 12.5, 2.6 Hz, 1H),
3.86 (dd, J ) 12.5, 2.7 Hz, 1H), 2.70 (d, J ) 3.8 Hz, 1H), 0.93 (s,
9H), 0.82 (s, 9H), 0.12 (s, 3H), 0.115 (s, 3H), -0.08 (s, 3H), -0.22 (s,
3H); 13C NMR (75.5 MHz, CDCl3) δ 172.1, 153.0, 152.2, 151.8, 143.1,
134.5, 134.1, 132.8, 130.5, 129.4, 128.8, 128.6, 127.9, 88.0, 85.4, 76.6,
71.3, 63.1, 26.0, 25.5, 18.5, 17.8, -5.10, -5.30, -5.66, -5.75; FAB-
MS (NBA) m/z 706 [M + 3]+, 705 [M + 2]+, 704 [M + 1]+, 600 [M
+ 1 - Bz]+, 344 [ABz2 + 1]+, 240 [ABz + 1]+, 136 [A + 1]+, 105
Bz+; FAB-MS (NBA + KCl) m/z 743 [M + 1 + K]+, 742 [M +
K]+. Anal. Calcd for C36H49N5O6Si2 [703.99]: C, 61.42; H, 7.02; N,
9.95. Found: C, 61.66; H, 6.81; N, 9.59.
disulfide radical anion from alkyl thiols70) is highly endothermic.
In the presence of an acid catalyst (such as the protonated
glutamate residue), however, the reduction leads via the
protonated ketone 39 to the R-hydroxy radical 40 which has a
much less negative reduction potential (E0 ) -1.4 V for
2-hydroxy-2-propyl radical69). According to this suggestion,
the glutamate residue, therefore, not only acts as a base in the
elimination step but also as an acid in the reduction sequence
(38 f 40).
Conclusions
The synthesis of a novel 3′-selenocarbonyl nucleoside allows
for the first time a selective generation of the 3′-adenosyl
radical.71 Using competition kinetic methods we could show
that these radicals give rise to a rapid elimination of the 2′-
hydroxy group. This observation lends additional support to
the mechanism for the important enzymatic 2′-deoxygenation
reaction of ribonucleotides proposed by Stubbe1c,4 in which C-3′
nucleotide radicals represents key intermediates. However, in
contrast to this mechanism where acid catalysis by a cysteine
residue is postulated, the observed elimination reaction in
solution is exclusively subject to general base catalysis. This
finding, together with the recent discovery of a glutamate residue
in the active site of the E. coli reductase,64 favors general base
catalysis in the enzyme mediated reaction.
6-N,N-Dibenzoyl-9-(2,5-di-O-(tert-butyldimethylsilyl)-â-D-erythro-
pentofuran-3-ulosyl)adenine (9). A sample of Dess-Martin perio-
dinane reagent19 (17.0 g, 40.0 mmol) was dissolved under argon in
anhydrous CH2Cl2 (200 mL) at 0 °C. Then, a solution of nucleoside
8 (14.1 g, 20.0 mmol) in anhydrous CH2Cl2 (100 mL) was added under
stirring. The yellowish mixture was allowed to warm up slowly to
room temperature and was stirred in the dark. The reaction was
complete after 3 days (TLC control, eluent: CH2Cl2/acetone 40:1). The
mixture was treated with saturated aqueous NaHCO3 (300 mL)
containing Na2S2O3 (30 g), stirred for 15 min, and extracted with Et2O
(500 mL). The organic phase was then washed with saturated aqueous
NaHCO3 (2 × 200 mL), H2O (200 mL), and brine (100 mL) and dried
over Na2SO4. After evaporation of the solvent under reduced pressure,
13.8 g (99%) of ketone 9 were obtained as a pale yellow foam: Rf )
1
0.45 (CH2Cl2/acetone 40:1); H NMR (300 MHz, CDCl3) δ 8.66 (s,
Experimental Section
1H), 8.37 (s, 1H), 7.88-7.85 (m, 4H), 7.52-7.46 (m, 2H), 7.38-7.33
(m, 4H), 6.20 (d, J ) 8.3 Hz, 1H), 4.91 (dd, J ) 8.3, 0.8 Hz, 1H), 4.32
(dd, J ) 2.1, 1.3 Hz, 1H), 3.98-3.97 (m, 2H), 0.90 (s, 9H), 0.70 (s,
9H), 0.08 (s, 3H), 0.06 (s, 3H), -0.03 (s, 3H), -0.27 (s, 3H); 13C NMR
(75.5 MHz, CDCl3) δ 208.0, 172.1, 153.2, 152.5, 152.1, 142.7, 134.0,
132.9, 129.4, 128.7, 127.6, 85.2, 82.6, 77.8, 62.3, 25.8, 25.2, 18.2, 17.9,
-4.85, -5.47, -5.66, -5.75; FAB-MS (NBA) m/z 703 [M + 2]+,
702 [M + 1]+, 344 [ABz2 + 1]+, 240 [ABz + 1]+, 105 Bz+; FAB-
MS (NBA + KCl) m/z 741 [M + 1 + K]+, 740 [M + K]+. Anal.
Calcd for C36H47N5O6Si2 [701.98]: C, 61.60; H, 6.75; N, 9.98.
Found: C, 61.46; H, 6.75; N, 9.98.
General Methods. 1H NMR and 13C NMR spectra were recorded
on a Varian Gemini 300 spectrometer (1H at 300 MHz, 13C at 75.5
MHz). Ultraviolet spectra were recorded on a Perkin-Elmer Lambda
2 spectrometer. Combustion analyses were performed at the Microana-
lytical Laboratory at the University of Basel. Mass spectra were
obtained on a VG 70-250 spectrometer. Analytical and preparative
TLC were performed on E. Merck silica gel 60 F254 plates. Unless
otherwise indicated, all separations were carried out under flash
chromatography (FC) conditions on silica gel (Chemische Fabrik
Uetikon, 230-400 mesh, particle size 35-70 µm). Analytical HPLC
was performed on Hewlett Packard 1050 series chromatographs. All
solvents were predried by standard methods. All other commercially
obtained reagents were used as received.
Synthesis of the Radical Precursors. 6-N,N-Dibenzoyl-2′,5′-di-
O-(tert-butyldimethylsilyl)adenosine (8). Disilylated nucleoside 718
(9.60 g, 19.4 mmol) was coevaporated twice with anhydrous pyridine
(50 mL each) and then dissolved in anhydrous pyridine (150 mL), and
the solution was cooled to 0 °C. Trimethylsilyl chloride (4.9 mL, 38.7
mmol) was added under stirring, and the mixture was allowed to warm
to room temperature. After 1 h, benzoyl chloride (11.2 mL, 96.8 mmol)
was slowly added dropwise at 0 °C, and the reaction mixture was again
brought to room temperature. After 2 additional h, it was carefully
hydrolyzed at 0 °C with saturated aqueous NaHCO3 (250 mL) and
extracted three times with EtOAc (400 mL, 200 mL, 100 mL). The
combined organic phases were washed with 2 N HCl and with H2O
and then dried over Na2SO4. After evaporation of the solvent in vacuo
and coevaporation with toluene, the residue was dissolved in anhydrous
EtOH (150 mL), cooled to 0 °C, and treated under stirring with p-tolu-
enesulfonic acid (150 mg). The reaction was complete after 1 h (TLC
control, eluent: EtOAc/pentane 1:3). Two spatulas of solid NaHCO3
were added, the solvent was evaporated under reduced pressure, and
the residue was dissolved in EtOAc (250 mL). The organic phase was
washed with H2O (2 × 150 mL) and brine (100 mL), dried over Na2-
SO4, and concentrated in vacuo to give 13.6 g (100%) of nucleoside 8
6-N-Benzoyl-9-(2,5-di-O-(tert-butyldimethylsilyl)-3-C-vinyl-â-D-
xylofuranosyl)adenine (10). CeCl3‚7H2O (18.5 g, 49.6 mmol) was
dried at 140 °C (oil bath temperature) in vacuo (4‚10-2 mbar) for 8 h.
The resulting white powder was then cooled under argon to 0 °C (ice
bath) and treated under vigorous stirring with cold anhydrous THF (100
mL), and the mixture was stirred overnight at room temperature. The
suspension was cooled to -78 °C, and vinylmagnesium bromide (50
mL, 1 M solution in THF) was added. After stirring for 4 h at -78
°C, a solution of ketone 9 (5.61 g, 7.99 mmol) in anhydrous THF (20
mL) was added slowly with a motor-driven syringe pump for 1 h.
After 2 additional h, the orange suspension was treated with AcOH
(6 mL) and allowed to warm up to room temperature. The reaction
mixture was partitioned between diluted aqueous NaHCO3 (300 mL)
and EtOAc (300 mL), the phases were separated, and the aqueous
phase was extracted twice with EtOAc (2 × 150 mL). The com-
bined organic phases were washed with diluted aqueous NaHCO3 (200
mL), H2O (150 mL), and brine (100 mL), dried over MgSO4, and
concentrated under reduced pressure. FC (eluent: EtOAc/CH2Cl2/
pentane 1:1:2) afforded 4.10 g (82%) of nucleoside 10 as a yellowish
foam: Rf ) 0.27 (EtOAc/CH2Cl2/pentane 1:1:2); 1H NMR (300 MHz,
CDCl3) δ 8.94 (broad s, 1H), 8.83 (s, 1H), 8.55 (s, 1H), 8.05-8.01
(m, 2H), 7.65-7.57 (m, 1H), 7.57-7.52 (m, 2H), 6.16 (s, 1H), 6.08
(dd, J ) 17.2, 10.8 Hz, 1H), 5.78 (s, 1H), 5.68 (dd, J ) 17.3, 1.7 Hz,
1H), 5.44 (dd, J ) 10.8, 1.7 Hz, 1H), 4.25 (s, 1H), 4.19-4.11 (m,
2H), 4.03 (dd, J ) 11.7, 2.8 Hz, 1H), 0.92 (s, 9H), 0.91 (s, 9H), 0.09
(s, 6H), 0.08 (s, 3H), -0.04 (s, 3H); 13C NMR (75.5 MHz, CDCl3) δ
164.6, 152.4, 151.3, 149.4, 142.6, 135.1, 133.7, 132.6, 128.9, 127.8,
123.0, 117.3, 91.3, 85.2, 82.7, 82.5, 61.6, 25.7, 18.1, 17.9, -4.6, -4.9,
-5.5, -5.8; FAB-MS (NBA) m/z 627 [M + 2]+, 626 [M + 1]+, 240
[ABz + 1]+, 136 [A + 1]+, 105 Bz+; FAB-MS (NBA + KCl) m/z
665 [M + 1 + K]+, 664 [M + K]+. Anal. Calcd for C31H47N5O5Si2
[625.92]: C, 59.49; H, 7.57; N, 11.19. Found: C, 59.32; H, 7.41; N,
11.01.
(69) Schwarz, H. A.; Dodson, R. W. J. Phys. Chem. 1989, 93, 409-
414.
(70) Surdhar, P. S.; Armstrong, D. A. J. Phys. Chem. 1986, 90, 5915-
5917.
(71) For the generation of unnatural derivatives, see: (a) Robins, M. J.;
Guo, Z.; Samano, M. C.; Wnuk, S. F. J. Am. Chem. Soc. 1996, 118, 11317-
11318. (b) Lehmann, T. E.; Berkessel, A. J. Org. Chem. 1997, 62, 302-
309.
(72) Xi, C.; Jun-Dong, Z.; Li-He, Z. Synthesis 1989, 383-384.
(73) Newcomb, M. Tetrahedron 1991, 49, 1151-1176.

2792 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Lenz and Giese
6-N-Benzoyl-9-(2,5-di-O-(tert-butyldimethylsilyl)-3-O-(trimethyl-
silyl)-3-C-vinyl-â-D-xylofuranosyl)adenine (11). Nucleoside 10 (4.10
g, 6.55 mmol) was dissolved under argon in anhydrous CH2Cl2 (50
mL), treated with Horning’s silylation reagent25 (6 mL), and stirred
for 72 h at room temperature. The reaction mixture was partitioned
between H2O (150 mL) and Et2O (150 mL), the phases were separated,
and the organic phase was washed with H2O (2 × 150 mL) and brine
(100 mL). After drying over Na2SO4 and concentration in vacuo, 4.66
g (6.68 mmol) of crude silyl ether 11 was obtained as a white foam
which was employed directly in the next step without further purifica-
tion: Rf ) 0.32 (EtOAc/pentane 1:2), Rf ) 0.46 (EtOAc/CH2Cl2/pentane
1:1:2); 1H NMR (300 MHz, CDCl3) δ 9.04 (broad s, 1H), 8.82 (s, 1H),
8.36 (s, 1H), 8.05-8.02 (m, 2H), 7.64-7.51 (m, 3H), 6.07 (dd, J )
17.7, 11.3 Hz, 1H), 6.01 (d, J ) 1.7 Hz, 1H), 5.44 (d, J ) 11.3 Hz,
1H), 5.41 (d, J ) 17.8 Hz, 1H), 4.54 (d, J ) 1.7 Hz, 1H), 4.51 (t, J )
5.3 Hz, 1H), 4.03 (d, J ) 5.3 Hz, 2H), 0.95 (s, 9H), 0.90 (s, 9H),
0.155 (s, 3H), 0.145 (s, 3H), 0.12 (s, 3H), -0.04 (s, 12H); 13C NMR
(75.5 MHz, CDCl3) δ 164.7, 152.7, 152.4, 149.2, 142.0, 134.9, 133.8,
132.6, 128.7, 127.9, 122.9, 119.0, 90.8, 85.3, 83.9, 82.7, 62.9, 26.0,
25.7, 18.5, 17.9, 2.14, -4.47, -4.75, -5.21, -5.28.
6-N-Benzoyl-9-(2,5-di-O-(tert-butyldimethylsilyl)-3-O-(trimethyl-
silyl)-3-C-formyl-â-D-xylofuranosyl)adenine (12). Crude nucleoside
11 (4.66 g, 6.68 mmol) was dissolved in CH2Cl2 (70 mL) and the
solution was cooled to -78 °C. The reaction flask was then connected
to an ozonolysis apparatus, and ozone (3% O3 in O2, flow 15 mL/min)
was passed through the solution for 1 h. The reaction mixture was
then treated with Me2S (1.47 mL, 20 mmol) and stirred for 2 h at room
temperature. After removal of the solvent in vacuo, the crude product
was purified by FC (eluent: pentane/CH2Cl2/tBuOMe 3:1:1), and 3.10
g (68% from 10) of aldehyde 12 were obtained as a yellowish foam:
Rf ) 0.25 (pentane/CH2Cl2/tBuOMe 3:1:1); 1H NMR (300 MHz, CDCl3)
δ 9.81 (s, 1H), 8.93 (broad s, 1H), 8.80 (s, 1H), 8.23 (s, 1H), 8.03-
8.00 (m, 2H), 7.62-7.50 (m, 3H), 6.16 (s, 1H), 4.98 (dd, J ) 5.3, 8.1
Hz, 1H), 4.67 (s, 1H), 3.93 (m, 2H), 0.945 (s, 9H), 0.87 (s, 9H), 0.27
(s, 3H), 0.105 (s, 3H), 0.10 (s, 3H), 0.07 (s, 3H), 0.003 (s, 9H).
(s, 1H), 8.37 (s, 1H), 8.02-8.00 (m, 2H), 7.59-7.54 (m, 1H), 7.50-
7.44 (m, 2H), 7.38-7.34 (m, 5H), 6.99 (s, 1H), 6.06 (d, J ) 2.2 Hz,
1H), 4.72 (d, J ) 2.2 Hz, 1H), 4.55 (t, J ) 3.7 Hz, 1H), 4.20 (dd, J )
11.8, 3.7 Hz, 1H), 4.04 (dd, J ) 11.8, 3.6 Hz, 1H), 0.92 (s, 9H), 0.80
(s, 9H), 0.12 (s, 3H), 0.10 (s, 3H), 0.04 (s, 3H), -0.02 (s, 3H); 13C
NMR (75.5 MHz, CDCl3) δ 201.2, 164.7, 152.4, 149.7, 149.2, 142.2,
135.7, 133.4, 132.6, 129.0, 128.6, 128.6, 127.8, 126.1, 123.3, 90.6,
90.0, 83.6, 80.9, 62.0, 25.7, 25.4, 18.0, 17.6, -4.78, -5.00, -5.59,
-5.82; FAB-MS (NBA) m/z 787 [M + 4]+, 786 [M + 3]+, 785 [M
+ 2]+, 784 [M + 1]+, 783 M+, 782 [M - 1]+, 781 [M - 2]+, 780 [M
- 3]+ 599 [M + 1 - PhSeCO]+, 598 [M - PhSeCO]+, 240 [ABz +
1]+, 136 [A + 1]+, 105 Bz+; FAB-MS (NBA + KCl) m/z 825 [M +
3 + K]+, 824 [M + 2 + K]+, 823 [M + 1 + K]+, 822 [M + K]+, 821
[M - 1 + K]+, 820 [M - 2 + K]+, 819 [M - 3 + K]+, 818 [M - 4
+ K]+.
6-N-Benzoyl-9-(2-O-(tert-butyldimethylsilyl)-3-C-phenylseleno-
carbonyl-â-D-xylofuranosyl)adenine (15). A solution of selenol ester
14 (50 mg, 64 µmol) and BF3‚Et2O (0.24 mL, 1.92 mmol) in anhydrous
CH2Cl2 (3 mL) was heated under reflux. After 2 h, the mixture was
treated with EtOAc (20 mL) and washed with H2O (3 × 20 mL) and
brine (2 × 10 mL). After removal of the solvent in vacuo, 42 mg
(98%) of selenol ester 15 were obtained as a white solid: 1H NMR
(300 MHz, CDCl3) δ 9.17 (broad s, 1H), 8.81 (s, 1H), 8.01 (s, 1H),
8.04-8.00 (m, 2H), 7.65-7.50 (m, 1H), 7.48-7.39 (m, 7H), 6.66-
6.62 (m, 1H), 5.94 (d, J ) 6.8 Hz, 1H), 5.21 (d, J ) 2.2 Hz, 1H), 5.04
(d, J ) 6.7 Hz, 1H), 4.38 (broad s, 1H), 4.10-4.06 (m, 1H), 3.99-
3.94 (m, 1H), 0.77 (s, 9H), 0.05 (s, 3H), -0.47 (s, 3H).
9-(2-O-(tert-Butyldimethylsilyl)-3-C-phenylselenocarbonyl-â-D-xy-
lofuranosyl)adenine (16). Nucleoside 14 (160 mg, 204 µmol) was
treated with 5% HCl in EtOH (6 mL) for 5 h at 50 °C, then EtOAc (60
mL) was added, and the reaction mixture was washed with H2O (4 ×
30 mL). The combined aqueous phases were extracted with EtOAc
(20 mL), and the organic phases were washed with brine (2 × 20 mL),
dried over Na2SO4, and concentrated under reduced pressure. The crude
product was purified by FC (eluent: CH2Cl2/EtOH 15:1) and 91 mg
(79%) of selenol ester 16 were obtained as a white precipitate: Rf )
0.27 (CH2Cl2/EtOH 15:1); 1H NMR (300 MHz, CDCl3) δ 8.36 (s, 1H),
7.74 (s, 1H), 7.54-7.50 (m, 2H), 7.44-7.39 (m, 3H), 7.24 (dd, J )
10.8, 3.0 Hz, 1H), 5.83 (d, J ) 7.1 Hz, 1H), 5.71 (broad s, 2H), 5.23
(s, 1H), 5.07 (d, J ) 7.1 Hz, 1H), 4.34 (t, J ) 1.6 Hz, 1H), 4.33 (ddd,
J ) 13.4, 2.9, 1.5 Hz, 1H), 3.93 (ddd, J ) 13.3, 11.1, 2.0 Hz, 1H),
0.77 (s, 9H), 0.04 (s, 3H), -0.49 (s, 3H);); 1H NMR (300 MHz, DMSO-
d6) δ 8.39 (s, 1H), 8.17 (s, 1H), 7.80 (s, 1H), 7.47 (broad s, 2H), 7.44-
7.40 (m, 5H), 5.88 (d, J ) 5.3 Hz, 1H), 5.44 (dd, J ) 6.5, 5.0 Hz, 1H),
4.98 (d, J ) 5.3 Hz, 1H), 4.22 (dd, J ) 4.7, 3.8 Hz, 1H), 3.73-3.68
(m, 2H), 0.71 (s, 9H), -0.02 (s, 3H), -0.35 (s, 3H); 13C NMR (75.5
MHz, CDCl3) δ 204.6, 156.0, 148.2, 152.5, 140.8, 136.0, 129.3, 129.0,
128.2, 125.8, 89.0, 88.5, 82.7, 80.9, 62.7, 25.7, 17.7, -4.47, -5.71;
FAB-MS (NBA) m/z 569 [M + 4]+, 568 [M + 3]+, 567 [M + 2]+,
566 [M + 1]+, 565 M+, 564 [M - 1]+, 563 [M - 2]+, 562 [M - 3]+,
381 [M + 1 - PhSeCO]+, 380 [M - PhSeCO]+, 136 [A + 1]+; FAB-
MS (NBA + KCl) m/z 607 [M + 3 + K]+, 606 [M + 2 + K]+, 605
[M + 1 + K]+, 604 [M + K]+, 603 [M - 1 + K]+, 602 [M - 2 +
K]+, 601 [M - 3 + K]+, 600 [M - 4 + K]+.
6-N-Benzoyl-9-(3-C-vinyl-â-D-xylofuranosyl)adenine (17). Nu-
cleoside 10 (5.11 g, 8.16 mmol) was dissolved in anhydrous THF (100
mL) under argon and cooled to 0 °C. The solution was then treated
with TBAF (20.4 mL, 1 M solution in THF) and stirred for 3 h. The
solvent was removed in vacuo, and the residue was purified by FC
(short column with large diameter, eluent: CH2Cl2/MeOH gradient 9:1
f 2:1). A brown oil (3.0 g) was obtained which still contained TBA
salts and which crystallized on standing. Dissolving this mixture in
EtOH and precipitating the product with a 1:1 Et2O/CH2Cl2 mixture
afforded after filtration 2.78 g (82%) of deprotected nucleoside 17 as
a white precipitate: Rf ) 0.41 (CH2Cl2/EtOH 8:1); 1H NMR (300 MHz,
DMSO-d6) δ 11.2 (broad s, 1H), 8.76 (s, 1H), 8.60 (s, 1H), 8.06-8.03
(m, 2H), 7.67-7.61 (m, 1H), 7.57-7.52 (m, 2H), 6.23 (d, J ) 5.4 Hz,
1H), 6.14 (dd, J ) 17.5, 10.9 Hz, 1H), 6.11 (broad s, 1H), 5.51 (s,
1H), 5.43 (dd, J ) 17.5, 1.9 Hz, 1H), 5.31 (dd, J ) 10.9, 1.9 Hz, 1H),
4.81 (t, J ) 5.6 Hz, 1H), 4.18 (dd, J ) 5.4, 1.0 Hz, 1H), 4.12 (dd, J
) 6.8, 2.7 Hz, 1H), 3.75-3.58 (m, 2H); 13C NMR (75.5 MHz, DMSO-
d6) δ 165.6, 151.8, 150.2, 151.5, 143.0, 136.6, 133.4, 132.4, 128.5,
125.6, 115.9, 90.6, 86.0, 82.9, 80.8, 59.5; FAB-MS (NBA) m/z 399
6-N-Benzoyl-9-(2,5-di-O-(tert-butyldimethylsilyl)-3-C-carboxy-â-
D-xylofuranosyl)adenine (13). Aldehyde 12 (75 mg, 107 µmol) was
t
dissolved in a mixture of BuOH (1.5 mL) and phosphate buffer (pH
7, 1 mL). The solution was then stirred with KMnO4 (1 mL of 1 M
aqueous solution) at room temperature. After 2 h, the excess of
permanganate was reduced by addition of saturated aqueous Na2SO3
(2 mL). The reaction mixture was then cooled to 0 °C, carefully
acidified to pH 3 with diluted HCl and extracted with EtOAc (3 × 10
mL). The combined organic phases were dried over MgSO4, treated
with silica gel (0.5 g), and concentrated in vacuo. The residue was
purified by FC (eluent: CH2Cl2/CH3OH/AcOH 100:10:1), and 53 mg
(77%) of carboxylic acid 13 were obtained as a fine white precipitate:
Rf ) 0.30 (CH2Cl2/CH3OH/AcOH 100:10:1); 1H NMR (300 MHz,
acetone-d6) δ 8.70 (s, 1H), 8.68 (s, 1H), 8.15-8.12 (m, 2H), 7.67-
7.53 (m, 3H), 6.30 (d, J ) 3.4 Hz, 1H), 4.87 (d, J ) 3.4 Hz, 1H), 4.69
(t, J ) 4.5 Hz, 1H), 4.11 (d, J ) 4.4 Hz, 2H), 0.92 (s, 9H), 0.83 (s,
1
9H), 0.13 (s, 3H), 0.11 (s, 3H), 0.04 (s, 3H), -0.04 (s, 3H); H NMR
(300 MHz, DMSO-d6) δ 11.2 (s, 1H), 8.76 (s, 1H), 8.60 (s, 1H), 8.05-
8.03 (m, 2H), 7.67-7.61 (m, 1H), 7.57-7.52 (m, 2H), 6.15 (d, J )
3.5 Hz, 1H), 6.12-5.92 (broad s, 1H), 4.69 (d, J ) 3.5 Hz, 1H), 4.50
(t, J ) 4.8 Hz, 1H), 3.94 (d, J ) 4.7 Hz, 2H), 0.86 (s, 9H), 0.77 (s,
9H), 0.05 (s, 3H), 0.03 (s, 3H), -0.02 (s, 3H), -0.10 (s, 3H); 13C NMR
(75.5 MHz, acetone-d6) δ 172.1, 166.2, 152.8, 152.5, 151.1, 143.3,
134.9, 133.2, 129.3, 129.2, 125.4, 89.9, 84.2, 83.9, 83.3, 63.3, 26.3,
25.9, 19.0, 18.3, -4.67, -4.94, -5.05, -5.29.
6-N-Benzoyl-9-(2,5-di-O-(tert-butyldimethylsilyl)-3-C-phenylsele-
nocarbonyl-â-D-xylofuranosyl)adenine (14). A sample of carboxylic
acid 13 (230 mg, 357 µmol) was dissolved in anhydrous DMF (3 mL)
under argon and cooled to 0 °C. 1,1′-Carbonyldi-1,2,4-triazole (147
mg, 893 µmol) was added, and the solution was stirred overnight at
room temperature. The reaction mixture was then cooled to 0 °C and
treated with PhSeH (96.0 µL, 893 µmol). After stirring for 1 h at 0
°C, EtOAc (50 mL) was added, and the mixture was washed with
brine (4 × 30 mL), dried over Na2SO4, and concentrated under re-
duced pressure. The crude product was purified by FC (eluent: EtOAc/
CH2Cl2/pentane 1:1:3), and 220 mg (79%) of selenol ester 14 were
obtained as a fine white precipitate: Rf ) 0.38 (EtOAc/CH2Cl2/pen-
1
tane 1:1:3); H NMR (300 MHz, CDCl3) δ 9.44 (broad s, 1H), 8.78

Studies on the Mechanism of Ribonucleotide Reductases
J. Am. Chem. Soc., Vol. 119, No. 12, 1997 2793
[M + 2]+, 398 [M + 1]+, 282 [M - Bz]+, 240 [ABz + 1]+, 136 [A
+ 1]+, 105 Bz+, 77 Ph+; FAB-MS (NBA + KCl) m/z 437 [M + 1 +
K]+, 436 [M + K]+. Anal. Calcd for C19H19N5O5 [397.40]: C, 57.43;
H, 4.82; N, 17.62. Found: C, 57.09; H, 4.86; N, 17.40.
room temperature. After 30 min the excess of oxidant was reduced
by adding saturated aqueous Na2SO3 (10 mL). The reaction mixture
was cooled to 0 °C, carefully acidified with HCl to pH 3, and extracted
with CH2Cl2 (4 × 30 mL) whereby the aqueous phase was gradually
brought to pH 2 by addition of HCl. The combined organic phases
were dried over MgSO4 and concentrated in vacuo, and the residue
was coevaporated consecutively with toluene, CHCl3, and CH2Cl2
yielding 1.16 g (91%) of carboxylic acid 21 as a fine white
precipitate: 1H NMR (300 MHz, DMSO-d6) δ 11.25 (s, 1H), 8.77 (s,
1H), 8.63 (s, 1H), 8.07-8.05 (m, 2H), 7.68-7.53 (m, 3H), 6.32 (s,
1H), 5.59 (s, 1H), 4.63 (broad s, 1H), 4.20 (broad d, J ) 13.9 Hz, 1H),
6-N-Benzoyl-9-(3,5-O-isopropylidene-3-C-vinyl-â-D-xylofurano-
syl)adenine (18). A suspension of nucleoside 17 (737 mg, 1.85 mmol)
in dry acetone (20 mL) was prepared under argon and cooled to 0 °C.
It dissolved instantaneously upon addition of a solution of AlCl3 (740
mg, 5.56 mmol) in dry Et2O (11 mL). After stirring for 1 h at room
temperature, the reaction mixture was treated with diluted aqueous
NaHCO3 (20 mL) and extracted with CHCl3 (40 mL) and CH2Cl2 (2
× 30 mL). The combined organic phases were dried over MgSO4 and
concentrated under reduced pressure. Acetonide 18 (800 mg, 99%)
was obtained as a white foam: Rf ) 0.40 (CH2Cl2/EtOH 10:1); 1H NMR
(300 MHz, CDCl3) δ 9.04 (broad s, 1H), 8.82 (s, 1H), 8.77 (s, 1H),
8.04-8.01 (m, 2H), 7.61-7.50 (m, 3H), 6.27 (s, 1H), 6.03 (dd, J )
18.1, 11.4 Hz, 1H), 5.58 (d, J ) 11.3 Hz, 1H), 5.39 (d, J ) 18.1 Hz,
1H), 4.37 (broad s, 1H), 4.34 (broad s, 1H), 4.24 (dd, J ) 13.5, 0.8
Hz, 1H), 4.16 (dd, J ) 13.5, 1.5 Hz, 1H), 3.04 (broad s, 1H), 1.44 (s,
3H), 1.25 (s, 3H); 13C NMR (75.5 MHz, CDCl3) δ 164.6, 152.4, 150.8,
149.2, 142.3, 136.7, 133.8, 132.7, 128.8, 127.8, 123.2, 119.7, 99.1,
91.5, 82.4, 81.7, 75.8, 58.9, 29.8, 23.2; FAB-MS (NBA) m/z 439 [M
+ 2]+, 438 [M + 1]+, 240 [ABz + 1]+, 105 Bz+; FAB-MS (NBA +
KCl) m/z 477 [M + 1 + K]+, 476 [M + K]+. Anal. Calcd for
C22H23N5O5 [437.46]: C, 60.40; H, 5.30; N 16.01. Found: C, 59.94;
H, 5.53; N 15.02.
4.06 (d, J ) 13.8 Hz, 1H), 2.12 (s, 3H), 1.40 (s, 3H), 1.31 (s, 3H); 13
C
NMR (75.5 MHz, DMSO-d6) δ 169.9, 168.4, 165.6, 151.9, 151.7, 150.4,
141.9, 133.2, 132.4, 128.5, 125.2, 99.1, 87.0, 81.7, 79.9, 73.8, 58.4,
28.8, 20.4.
6-N-Benzoyl-9-(2-O-acetyl-3,5-O-isopropylidene-3-C-phenylsele-
nocarbonyl-â-D-xylofuranosyl)adenine (22). A solution of carboxylic
acid 21 (805 mg, 1.62 mmol) in anhydrous THF (30 mL) was cooled
to 0 °C under argon. NEt3 (680 µL, 4.85 mmol) and PhOPOCl2 (480
µL, 3.24 mmol) were added under stirring, the mixture was agitated
for 1 h, and then addition of NEt3 (1.16 mL, 7.90 mmol) and PhSeH
(520 µL, 4.90 mmol) was followed by stirring for 30 min at 0 °C and
subsequent treatment with H2O and EtOAc (50 mL each). The phases
were separated, and the aqueous phase was extracted with EtOAc (2
× 30 mL). The combined organic phases were washed with brine (2
× 30 mL), dried over MgSO4, and concentrated under reduced pressure.
Purification of the residue by FC (eluent: CH2Cl2/acetone 5:1) afforded
739 mg (72%) of selenol ester 22 as a white precipitate: Rf ) 0.22
6-N-Benzoyl-9-(2-O-acetyl-3,5-O-isopropylidene-3-C-vinyl-â-D-xy-
lofuranosyl)adenine (19). Acetonide 18 (1.87 g, 3.85 mmol) was
coevaporated twice with anhydrous pyridine (10 mL each), then
dissolved in anhydrous pyridine (25 mL) under argon, and cooled to 0
°C. AcOH (400 µL, 4.24 mmol) was added, and the reaction mixture
was stirred at room temperature overnight. The solvent was removed
in vacuo, and the residue was coevaporated twice with toluene (10 mL
each). The crude product was purified by FC (eluent: CH2Cl2/acetone
2:1) yielding 1.79 g (97%) of acetate 19 as a white foam: Rf ) 0.34
1
(CH2Cl2/acetone 5:1); H NMR (300 MHz, CDCl3) δ 9.04 (broad s,
1H), 8.84 (s, 1H), 8.72 (s, 1H), 8.05-8.03 (m, 2H), 7.64-7.50 (m,
3H), 7.45-7.39 (m, 5H), 6.25 (d, J ) 1.3 Hz, 1H), 5.83 (d, J ) 1.3
Hz, 1H), 4.70 (t, J ) 1.6 Hz, 1H), 4.19 (m, 2H), 2.13 (s, 3H), 1.63 (s,
3H), 1.54 (s, 3H); 13C NMR (75.5 MHz, CDCl3) δ 200.6, 168.1, 164.5,
152.9, 151.5, 149.4, 141.6, 135.9, 133.7, 132.7, 129.5, 129.4, 128.8,
127.8, 124.7, 123.1, 101.2, 87.8, 87.5, 82.5, 73.1, 59.2, 29.2, 22.8, 20.5;
FAB-MS (NBA) m/z 640 [M + 3]+, 639 [M + 2]+, 638 [M + 1]+,
637 M+, 636 [M - 1]+, 635 [M - 2]+, 394 [M - 3]+, 240 [ABz +
1]+, 105 Bz+; FAB-MS (NBA + KCl) m/z 678 [M + 2 + K]+, 677
[M + 1 + K]+, 676 [M + K]+, 675 [M - 1 + K]+, 674 [M - 2 +
K]+, 673 [M - 3 + K]+, 672 [M - 4 + K]+. Anal. Calcd for
C29H27N5O7Se [636.47]: C, 54.73; H, 4.28; N 11.00. Found: C, 55.20;
H, 4.42; N 10.43.
1
(CH2Cl2/acetone 2:1); H NMR (300 MHz, CDCl3) δ 8.93 (broad s,
1H), 8.82 (s, 1H), 8.76 (s, 1H), 8.04-8.02 (m, 2H), 7.61-7.50 (m,
3H), 6.33 (s, 1H), 5.96 (dd, J ) 18.1, 11.4 Hz, 1H), 5.51 (s, 1H), 5.41
(d, J ) 11.4 Hz, 1H), 5.25 (d, J ) 18.1 Hz, 1H), 4.34 (broad s, 1H),
4.23-4.12 (m, 2H), 2.12 (s, 3H), 1.47 (s, 3H), 1.36 (s, 3H); 13C NMR
(75.5 MHz, CDCl3) δ 168.5, 164.5, 152.7, 151.3, 149.2, 142.0, 136.5,
133.7, 132.6, 128.8, 127.8, 123.2, 118.1, 99.4, 88.4, 82.2, 81.0, 75.4,
58.4, 29.8, 22.9, 21.1; FAB-MS (NBA) m/z 481 [M + 2]+, 480 [M +
1]+, 479 M+, 242 [M + 2 - ABz]+, 241 [M + 1 - ABz]+, 240 [M -
ABz]+ and [ABz + 1]+, 136 [A + 1]+, 105 Bz+, 77 Ph+; FAB-MS
(NBA + KCl) m/z 520 [M + 1 + K]+, 519 [M + K]+. Anal. Calcd
for C24H25N5O6 [479.50] C, 60.12; H, 5.26; N 14.61. Found: C, 59.87;
H, 5.37; N 13.91.
9-(3-C-Phenylselenocarbonyl-â-D-xylofuranosyl)adenine (23). Se-
lenol ester 22 (366 mg, 575 µmol) was dissolved in a solution of 5%
HCl in EtOH (10 mL) and stirred for 5 h at 50 °C. During this time,
a white precipitate appeared. The solvent was removed under reduced
pressure, and the residue was dissolved in MeOH (8 mL) and treated
with Et2O (20 mL). The white precipitate formed was filtered off and
washed several times with cold Et2O. The filtrate was concentrated
under reduced pressure and treated under the same conditions to obtain
more precipitate. On the whole, 218 mg (78%) of selenol ester 23
was obtained in its hydrochloride form: Rf ) 0.15 (CH2Cl2/EtOH 8:1);
1H NMR (300 MHz, CD3OD) (free amine) δ 8.31 (s, 1H), 8.24 (s,
1H), 7.53-7.37 (m, 5H), 5.97 (d, J ) 3.9 Hz, 1H), 4.79 (d, J ) 3.9
Hz, 1H), 4.44 (t, J ) 4.4 Hz, 1H), 3.88 (d, J ) 4.4 Hz, 2H); 1H NMR
(300 MHz, CD3OD) (hydrochloride) δ 8.65 (s, 1H), 8.43 (s, 1H), 7.53-
7.37 (m, 5H), 6.16 (d, J ) 3.6 Hz, 1H), 4.71 (d, J ) 3.5 Hz, 1H), 4.51
(t, J ) 4.1 Hz, 1H), 3.96 (d, J ) 4.1 Hz, 2H); 13C NMR (75.5 MHz,
CD3OD) (hydrochloride) δ 205.1, 152.0, 149.7, 145.7, 143.9, 137.2,
130.2, 129.8, 127.9, 120.3, 91.1, 90.7, 84.8, 83.5, 61.5; FAB-MS
(NBA) (free amine with M+ ) 451) m/z 454 [M + 3]+, 453 [M + 2]+,
452 [M + 1]+, 451 M+, 450 [M - 1]+, 449 [M - 2]+, 448 [M - 3]+,
266 [M - PhSeCO]+, 137 [A + 2]+, 136 [A + 1]+; FAB-MS (NBA
+ KCl) m/z 492 [M + 2 + K]+, 490 [M + K]+, 488 [M - 2 + K]+;
UV(MeOH) λ (ꢀ) 206 (44 000), 261 (25 000), 300 (1500). Anal. Calcd
for C17H17N5O5Se‚HCl‚H2O [504.78]: C, 40.45; H, 4.00; N, 13.87.
Found: C, 40.37; H, 4.03; N, 14.13.
6-N-Benzoyl-9-(2-O-acetyl-3-C-formyl-3,5-O-isopropylidene-â-D-
xylofuranosyl)adenine (20). Nucleoside 19 (1.78 g, 3.71 mmol) was
dissolved in CH2Cl2 (30 mL), and the solution was cooled to -78 °C.
The reaction flask was then connected to an ozonolysis apparatus and
ozone (3% O3 in O2, flow 15 mL/min) was passed through the solution
for 1 h. The reaction mixture was then treated with Me2S (816 µL,
11.1 mmol) and stirred for 1 h at room temperature. The solvent was
removed in vacuo, and the residue was purified by FC (eluent: CH2-
Cl2/acetone 3:1) to yield 1.23 g (69%) of aldehyde 20 as a yellowish
precipitate: Rf ) 0.32 (CH2Cl2/acetone 3:1); 1H NMR (300 MHz,
CDCl3) δ 9.66 (s, 1H), 8.96 (broad s, 1H), 8.81 (s, 1H), 8.71 (s, 1H),
8.05-8.02 (m, 2H), 7.62-7.51 (m, 3H), 6.39 (s, 1H), 5.52 (s, 1H),
4.67 (broad s, 1H), 4.22 (d, J ) 14.0 Hz, 1H), 4.06 (dd, J ) 14.0, 1.9
Hz, 1H), 2.15 (s, 3H), 1.43 (s, 3H), 1.39 (s, 3H); 13C NMR (75.5 MHz,
CDCl3) δ 198.0, 169.5, 164.8, 152.8, 151.3, 149.4, 141.6, 133.6, 132.8,
128.8, 127.9, 123.3, 99.7, 88.5, 84.6, 82.3, 73.8, 58.9, 29.1, 22.0, 20.7;
FAB-MS (NBA) m/z 483 [M + 2]+, 482 [M + 1]+, 243 [M + 1 -
ABz]+, 240 [ABz + 1]+, 136 [A + 1]+, 105 Bz+, 77 Ph+; FAB-MS
(NBA + KCl) m/z 521 [M + 1 + K]+, 520 [M + K]+. Anal. Calcd
for C23H23N5O7 [481.47]: C, 57.38; H, 4.82; N 14.55. Found: C, 57.42;
H, 5.24; N 13.25.
6-N-Benzoyl-9-(2-O-acetyl-3-C-phenylselenocarbonyl-â-D-xylo-
furanosyl)-adenine (24). Selenol ester 22 (355 mg, 558 µmol) was
dissolved in 80% aqueous trifluoroacetic acid (5 mL) and stirred for
30 min at room temperature. After evaporation of the solvent under
reduced pressure, the residue was coevaporated with toluene (2 × 10
mL) and then purified by FC (eluent: CH2Cl2/EtOH 20:1) to give 286
6-N-Benzoyl-9-(2-O-acetyl-3-C-carboxy-3,5-O-isopropylidene-â-
D-xylofuranosyl)adenine (21). A solution of aldehyde 20 (1.23 g, 2.55
t
mmol) in BuOH (30 mL) and phosphate buffer (pH 7, 15 mL) was
treated with KMnO4 (10 mL, 1 M aqueous solution) under stirring at

2794 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Lenz and Giese
mg (86%) of selenol ester 24 as a yellowish precipitate: Rf ) 0.17
evaporated in vacuo, and the residue was coevaporated with 5 mL of
toluene. NMR analysis showed the formation of 14% of 6 (xylo/ribo
4:1), 80% of adenine (32), and 50% of enolone 34.
1
(CH2Cl2/EtOH 20:1); H NMR (300 MHz, CDCl3) δ 9.11 (broad s,
1H), 8.80 (s, 1H), 8.10 (s, 1H), 8.06-8.02 (m, 2H), 7.65-7.40 (m,
8H), 6.69 (broad s, 1H), 6.11 (d, J ) 5.2 Hz, 1H), 5.98 (d, J ) 5.2 Hz,
1H), 5.70-5.55 (broad s, 1H), 4.41 (t, J ) 2.6 Hz, 1H), 4.10 (dd, J )
13.2, 2.5 Hz, 1H), 4.00 (dd, J ) 13.0, 1.9 Hz, 1H), 2.09 (s, 3H); 13C
NMR (75.5 MHz, CDCl3) δ 204.3, 169.6, 164.9, 152.2, 150.5, 150.0,
142.5, 135.9, 133.2, 132.9, 129.3, 129.1, 128.8, 128.0, 125.8, 123.4,
88.2, 87.0, 83.0, 80.9, 61.1, 20.5; FAB-MS (NBA) m/z 601 [M +
4]+, 600 [M + 3]+, 599 [M + 2]+, 598 [M + 1]+, 597 M+, 596 [M -
1]+, 595 [M - 2]+, 594 [M - 3]+, 412 [M - PhSeCO]+, 240 [ABz +
1]+, 136 [A + 1]+, 105 Bz+; FAB-MS (NBA + KCl) m/z 639 [M +
3 + K]+, 638 [M + 2 + K]+, 637 [M + 1 + K]+, 636 [M + K]+, 635
[M - 1 + K]+, 634 [M - 2 + K]+, 633 [M - 3 + K]+, 632 [M - 4
+ K]+.
Photolyses of Radical Precursor 24. A solution of selenol ester
24 (11.2 mg, 18.8 µmol) and Bu3SnH (12.5 µL, 47.2 µmol) in 3 mL of
CH3OH was purged with argon for 10 min and irradiated with lamp A
for 3 h. After evaporation of the solvent in vacuo, the crude reaction
mixture was analyzed by 1H-NMR. Besides 7% of the starting material,
84% of 6-N-benzoyladenine (33), 78% of enolone 34, and 90% of
AcOH could be detected. A second NMR spectrum, recorded after 2
h, revealed only 40% of 34; the yields of the other products were
unchanged. Compound 33: Rf ) 0.31 (CH2Cl2/EtOH 12:1); 1H NMR
(300 MHz, CD3OD) δ 8.70 (s, 1H), 8.46 (s, 1H), 8.12-8.09 (m, 2H),
7.69-7.64 (m, 1H), 7.60-7.54 (m, 2H).
A repetition of the photolysis under the same conditions led to the
formation of 67% of 33, 41% of 34, 88% of AcOH, and 15% of the
2′-deoxy-3′-ketonucleoside 31. Compound 31:20b 1H NMR (300 MHz,
CD3OD) δ 8.70 (s, 1H), 8.66 (s, 1H), 8.11-8.06 (m, 2H), 7.67-7.51
(m, 3H), 6.81 (t, J ) 7.0 Hz, 1H), 4.29 (t, J ) 3.0 Hz, 1H), 3.89-3.86
(m, 2H), 3.30 (m, 2H).
Photolyses of the Radical Precursors. Irradiation experiments were
carried out at 20 °C using either a Heraeus TQ-150 (150 W) Hg high-
pressure lamp with Pyrex coating (lamp A) or an Osram (500 W) Hg
high-pressure lamp with 320 nm filter (lamp B). The reaction mixtures
1
were analyzed by H-NMR; products were identified by comparison
with authentic material or the known data for analogous compounds
(references cited). Yields were determined by product isolation or from
NMR spectra using internal standards (pentachloroethane or total
integral of adenine protons).
A solution of 24 (10.0 mg, 16.7 µmol) and Bu3SnD (11.1 µL, 41.9
µmol) in 0.7 mL of CD3OD was purged with argon for 10 min and
irradiated with lamp A for 3 h. The following products could be
detected by 1H-NMR: 27% of 6-N-benzoyladenine (33), 25% of
enolone 34 (52% deuterated at C-3 as judged by GC/MS), 92% of
AcOH, and 52% of the 2′-deoxy-3′-ketonucleoside 31 (>90% mono-
deuterated at C-2′).
A solution of 24 (13.2 mg, 22.1 µmol) in 0.7 mL of CD3OD was
purged for 10 min with argon. tBuSH (5.0 µL, 45 µmol) was added,
and the mixture was irradiated for 3 h using lamp B. The NMR analysis
showed the formation of 70% of 6-N-benzoyladenine (33), 54% of
enolone 34 (ca. 40% deuterated at C-3), 50% AcOH, and 9% of direct
reduction product 27, exclusively with xylo configuration and > 80%
deuterated at C-3′. Compound 27: 1H NMR (300 MHz, CD3OD) δ
8.68 (s, 1H), 8.66 (s, 1H), 8.09-8.06 (m, 2H), 7.67-7.51 (m, 3H),
6.29 (d, J ) 1.7 Hz, 1H), 5.39 (t, J ) 1.8 Hz, 1H), 4.32 (dd, J ) 5.9,
4.8 Hz, 1H), 3.99 (dd, J ) 11.8, 4.9 Hz, 1H), 3,94 (dd, J ) 11.9, 5.9
Hz, 1H), 2.15 (s, 3H).
Photolyses of Radical Precursor 23. A solution of selenol ester
23 (14.1 mg, 29.0 µmol) and Bu3SnH (17 µL, 64 µmol) in 4 mL of
CH3OH was purged with argon for 10 min and subsequently irradiated
for 2.5 h with lamp A. The solvent was removed in vacuo, and the
products were purified by preparative TLC (eluent: CH2Cl2/EtOH 4:1).
7.3 mg (94%) of the reduction products 6 (xylo/ribo 3:1) were obtained
1
as a white precipitate: Rf ) 0.19 (CH2Cl2/EtOH 4:1); H NMR (300
MHz, CD3OD) (xylo-isomer71) δ 8.33 (s, 1H), 8.20 (s, 1H), 5.96 (d, J
) 1.9 Hz, 1H), 4.46 (t, J ) 1.9 Hz, 1H), 4.36-4.31 (m, 1H), 4.20 (dd,
J ) 3.8, 1.6 Hz, 1H), 3.96 (dd, J ) 11.9, 4.6 Hz, 1H), 3.90 (dd, J )
11.9, 6.1 Hz, 1H); (ribo-isomer) δ 8.30 (s, 1H), 8.17 (s, 1H), 5.96 (d,
J ) 6.4 Hz, 1H), 4.73 (dd, J ) 6.4, 5.1 Hz, 1H), 4.31 (dd, J ) 5.2, 2.5
Hz, 1H), 4.16 (q, J ) 2.5 Hz, 1H), 3.88 (dd, J ) 12.5, 2.5 Hz, 1H),
3.73 (dd, J ) 12.5, 2.8 Hz, 1H).
A solution of selenol ester 23 (8.3 mg, 17.1 µmol) in 0.7 mL of
CD3OD was purged for 10 min with argon. tBuSH (2.9 µL, 26 µmol)
was added, and the mixture was irradiated for 2.5 h using lamp B. The
yield of reduction products 6 (selectively deuterated at C-3′) determined
Competition Kinetic Experiments with Radical Precursor 23.
Indirect kinetic studies involved conventional competition experi-
ments.72 They were performed at 20 °C with CH3CN/H2O 1:1 as
solvent and with tin hydride 35 as radical trap. The product ratio [32]/
[6] was determined by analytical HPLC (Merck Lichrospher 100, RP
18 (5 µm), 250 mm × 4 mm, 25 °C) at 254 nm. Solvents: A: 20
mM TEAA buffer, B: CH3CN. Gradient: A/B: 98/2(2 min)s13
mins85/15s10 mins0/100(10 min). Retention times: 9.2 min (32,
correspondence factor: 13.32 mM/AU), 11.3 min (6 ribo isomer,
correspondence factor: 14.62 mM/AU), 10.5 min (6 xylo isomer).
In a typical experiment, selenol ester 23 (3 mM) and an excess of
tin hydride 35 (20-150 mM, freshly prepared from the corresponding
tin bromide47) were dissolved under argon in 1.0 mL of degassed CH3-
CN/H2O 1:1, containing the buffer stated, and subsequently irradiated
for 5 min with lamp A (see photolyses). The reaction mixture (0.1
mL) was diluted with 0.9 mL of H2O and directly analyzed by HPLC.
1
by H-NMR analysis was 78% (xylo/ribo 4:1).
A solution of selenolester 23 (11.6 mg, 23.8 µmol) and Bu6Sn2 (13
µL, 25 µmol) in 4 mL of CH3OH was purged with argon for 5 min
and then irradiated with lamp A for 3 h. The solvent was removed in
1
vacuo, and the residue analyzed by H-NMR. Thirty-five percent of
the reduction products 6 (xylo/ribo 3:1), 55% of adenine (32), and 10%
of enolone 34 could be detected. Preparative TLC (eluent: CH2Cl2/
EtOH 4:1) yielded 1.8 mg (57%) of adenine (32) and 2.2 mg (35%) of
reduction products 6. Compound 32: Rf ) 0.22 (CH2Cl2/EtOH 4:1);
1H NMR (300 MHz, CD3OD) δ 8.13 (s, 1H), 8.04 (s, 1H). Compound
34:42a,c Rf ) 0.41 (CH2Cl2/EtOH 12:1); 1H NMR (300 MHz, CD3OD)
δ 8.54 (d, J ) 2.5 Hz, 1H), 5.71 (d, J ) 2.5 Hz, 1H), 4.53 (ddd, J )
4.6, 3.0, 0.8 Hz, 1H), 3.97 (dd, J ) 12.5, 2.9 Hz, 1H), 3.83 (dd, J )
12.5, 4.5 Hz, 1H).
A solution of selenol ester 23 (8.0 mg, 16.4 µmol) and tin hydride
35 (40 µL, 85 µmol) in 5 mL of CH3CN/H2O (1:1) containing 0.1 M
triethylammonium acetate buffer (TEAA buffer) was purged with argon
for 10 min and irradiated for 5 min using lamp A. The solvent was
Acknowledgment. This work was supported by the Swiss
National Science Foundation. R.L. thanks the Swiss Society
of Chemical Industry for a grant.
JA962974Q