Selective Generation and Reactivity of 5′-Adenosinyl and 2′-Adenosinyl Radicals

Article abstract of DOI:10.1002/chem.200305488

The reaction of hydrated electrons (e_aq^-) with 8-bromoadenosine 7 has been investigated by radiolytic methods coupled with product studies. Pulse radiolysis revealed that one-electror reductive cleavage of the C-Br bond gives the C8 radical 8 followed by a fast radical translocation to the sugar moiety. The reaction is partitioned between C5′ and C2′ positions in a 60:40 ratio leading to 5′-adenosinyl radical 9 and 2′-adenosinyl radical 11. This radical translocation from C8 to different sites of the sugar moiety has also been addressed computationally by means of DFT B3LYP calculations. In addition, ketone 21 was prepared and photolyzed providing an independent generation of C2′ radical 11. Both C5′ and C2′ radicals undergo unimolecular reactions. Radical 9 attacks adenine with a rate constant of 1.0 × 10⁴ s^-1 and gives the aromatic aminyl radical 10, whereas C2′ radical 11 liberates adenine with a rate constant of 1.1 × 10⁵ s^-1.

Full text of DOI:10.1002/chem.200305488

FULL PAPER
Selective Generation and Reactivity of 5’-Adenosinyl and
2’-Adenosinyl Radicals
Chryssostomos Chatgilialoglu,*[a] Maria Duca,[a, b] Carla Ferreri,[a] Maurizio Guerra,[a]
Marcella Ioele,[a, c] Quinto G. Mulazzani,[a] Harald Strittmatter,[d] and Bernd Giese*[d]
Abstract: The reaction of hydrated
electrons (eꢀaq)with 8-bromoadenosine
7 has been investigated by radiolytic
methods coupled with product studies.
Pulse radiolysis revealed that one-elec-
radical 9 and 2’-adenosinyl radical 11.
This radical translocation from C8 to
different sites of the sugar moiety has
also been addressed computationally
by means of DFT B3LYP calculations.
In addition, ketone 21 was prepared
and photolyzed providing an indepen-
dent generation of C2’ radical 11. Both
C5’ and C2’ radicals undergo unimolec-
ular reactions. Radical 9 attacks ade-
nine with a rate constant of 1.0î104 sꢀ1
and gives the aromatic aminyl radical
10, whereas C2’ radical 11 liberates ad-
enine with a rate constant of 1.1î
105 sꢀ1.
ꢀ
tron reductive cleavage of the C Br
bond gives the C8 radical 8 followed
by a fast radical translocation to the
sugar moiety. The reaction is parti-
tioned between C5’ and C2’ positions
in a 60:40 ratio leading to 5’-adenosinyl
Keywords:
elimination
cyclization
nucleosides
¥
¥
¥
pulse radiolysis ¥ radical reactions
Introduction
aqueous solution of adenosine, adenosine-5’-monophosphate
and polyadenylic acid at pH 7 afforded (5’R):(5’S)ratios of
1.8, 0.4, and 1.6, respectively.[4] It has been proposed that a
C5’ radical might intramolecularly attack the C8,N7 double
bond of the adenine moiety to form 5’,8-cycloadenosides
2a,b as final products (Scheme 1). The fact that molecular
oxygen inhibits these reactions, was interpreted as trapping
of the C5’ radical before its attack at the C8 position of ade-
nine.
In 1968 Keck discovered that attack of HOC radicals at ade-
nine-5’-phosphate 1 (R=PO32ꢀ)leads among other products
to the 5’,8-cycloadenosides 2a and 2b.[1] The effect of pH on
the formation of the diastereomers 2a,b has been studied in
detail,[2] and it turned out that only the (5’R)-isomer 2a is
enzymatically active.[3] Radiation-induced damages to poly-
adenylic acid, as well as to free adenosine have also been in-
vestigated in some details.[4] Depending on the substrate and
the experimental conditions, the ratio of the (5’S)- and
(5’R)-isomers changes substantially. g-Irradiation of an
[a] Dr. C. Chatgilialoglu, M. Duca, Dr. C. Ferreri, Dr. M. Guerra,
Dr. M. Ioele, Dr. Q. G. Mulazzani
ISOF, Consiglio Nazionale delle Ricerche
Via P. Gobetti 101, 40129 Bologna (Italy)
Fax : (+39)051-639-8349
E-mail: chrys@isof.cnr.it
Scheme 1. The two diastereoisomers of 5’,8-cycloadenosine derivatives.
[b] M. Duca
Present address: Laboratoire de Biophysique
Museum National d×Histoire Naturelle
43 rue Cuvier, 75231 Paris cedex 05 (France)
Similar reactions have been observed in the 2’-deoxy-ribo
series. g-Irradiation of an aqueous solution of 2’-deoxyade-
nosine afforded mainly the (5’R)-isomer, whereas the
(5’R):(5’S)ratios, using single- and double-stranded DNA,
are approximately 2.[5,6] Some of us recently investigated the
[c] Dr. M. Ioele
Present address: Istituto Centrale per il Restauro
Laboratorio Di Chimica
Piazza San Francesco di Paola 9, 00184 Roma (Italy)
reaction of hydrated electrons (eꢀaq)with 8-bromo-2 ’-deoxy-
[d] Dr. H. Strittmatter, Prof. B. Giese
Department of Chemistry, University of Basel
St. Johanns Ring 19, 4056 Basel (Switzerland)
Fax : (+41)61-267-1105
[7]
adenosine (3)by radiolytic methods.
It was found that 3
captures electrons and rapidly loses the bromide ion to give
E-mail: bernd.giese@unibas.ch
the corresponding C8 radical 4. Radical 4 abstracts intramo-
1249
Chem. Eur. J. 2004, 10, 1249 1255
DOI: 10.1002/chem.200305488
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

FULL PAPER
lecularly a hydrogen atom exclusively from the C5’ position
affording selectively C5’ radical 5 (Scheme 2). This allowed
for the first time to study the fate of the 2’-deoxyadenosin-
5’-yl radical 5 properly, and in particular the cyclization step
5!6, which occurs with a rate constant of 1.6î105 sꢀ1.
and with the 8-bromo-2’-deoxyadenosine
1010 mꢀ1 sꢀ1).[7]
3
(1.6î
The reaction of an aqueous solution of 7 (1 mm)and
tBuOH (0.25m)at pH ~7 with eꢀaq in the absence of O2 was
complete within approximately 300 ns. At this time, no sig-
nificant absorption was detected in the 300 750 nm region.
However, a spectrum containing two bands centered at 350
and 480 nm, respectively, developed in 100 ms (Figure 1).
The time profile for the formation of the transient with
lmax =350 nm (Figure 1, inset a)follows a first-order kinetic
Scheme 2. Chemical studies of hydrated electrons with 8-bromo-2’-deoxy-
adenosine.
Figure 1. Absorption spectrum obtained from the pulse radiolysis of an
Ar-purged solution containing 1 mm 7 and 0.25m tBuOH at pH ~7, taken
100 ms after the pulse; dose=20 Gy, optical path=2.0 cm. Insets: a)Time
dependence of absorption at 350 nm; dose=23.8 Gy. b)Dependence of
kobs (see text)from the radiation dose.
We have now found in radiation studies with 8-bromoade-
nosine 7 that in the ribo series the intramolecular hydrogen
abstraction by the initially formed C8 radical is partitioned
between two channels, generating both C5’ and C2’ radicals
with similar rate constants.[8]
Results and Discussion
with a rate constant (kobs)that is independent of 7 in the
concentration range 0.2 1 mm. But kobs increased with the
dose/pulse ratio (Figure 1, inset b). This dose dependence is
due to the mixing of the first-order growth and the second-
order decay of the species. Using an empirical expression
for the fitting of the experimental data and extrapolating at
zero dose,[13] a rate constant kc =(1.0Æ0.2)î104 sꢀ1 at 208C
is obtained. The absorbance at 350 nm is also found to vary
substantially with the dose. An apparent molar extinction
coefficient (eapp)of 6100 Æ100mꢀ1 cmꢀ1 at 350 nm was calcu-
lated by extrapolating to zero dose and assuming a radiation
chemical yield G=0.27 mmolJꢀ1, which is the G of hydrated
electrons.[9]
Reaction of hydrated electrons (eaqꢀ) with 8-bromoadeno-
sine 7: Radiolysis of neutral water leads to eꢀaq, HOC and HC
as shown in [Eq. (1)]. The values in parentheses represent
the radiation chemical yields (G values)in units of
mmolJꢀ1.[9] The reactions of eaꢀq with the substrates were
studied in O2-free solutions containing 0.25m tBuOH. With
this amount of tBuOH, HOC is scavenged efficiently
[Eq. (2), k2 =6.0î108 mꢀ1 sꢀ1], whereas HC reacts only slowly
[Eq. (3), k3 =1.7î105 mꢀ1 sꢀ1].[9,10] Therefore, the reactions of
HC may be relevant since they can account for as much as
~20% of the products (see below).
In analogy to the reactions of 8-bromo-2’-deoxyadenosine
3,[7] we assigned the transient in Figure 1 to the conjugated
aminyl radical 10, and the observed rate to the cyclization of
radical 9 (Scheme 3). Compared with the 2’-deoxyribosides,
in the ribo case the spectrum containing the two bands de-
veloped more slowly (100 vs 20 ms)and the absorbance at
the maximum (350 vs 360 nm)was about two thirds. The
cyclization rate constant for reaction 9!10 (kc =1.0î
104 sꢀ1)is 16 times slower than that of the 2 ’-deoxy-ribo ana-
logue 5!6 (Scheme 2). This is probably a consequence of
the conformational changes in going from ribo to 2’-deoxy-
ribo derivatives.
H2O R eꢀaq ð0:27Þ; HOC ð0:28Þ; HC ð0:062Þ
ð1Þ
ð2Þ
ð3Þ
k2
ƒ!
C
HOC þ tBuOH
ðCH Þ CðOHÞCH þ H O
3
2
2
2
k3
ƒ!
C
HC þ tBuOH
ðCH Þ CðOHÞCH þ H
3
2
2
2
The pseudo first-order rate constant, kobs, for the reaction of
eꢀaq with a defined amount of 8-bromoadenosine 7 was deter-
mined by measuring the rate of the decrease of the optical
density of eꢀaq at 720 nm (e=1.9î104 mꢀ1 cmꢀ1).[11] From the
slope of kobs versus [7], the bimolecular rate constant was
determined to be 1.1î1010 mꢀ1 sꢀ1, which is very similar to
the analogous reactions with adenosine (1.0î1010 mꢀ1 sꢀ1),[12]
Because the extinction coefficients of radicals 6 and 16
are similar (ca. 1î104 mꢀ1 cmꢀ1),[7,14] it is reasonable to
1250
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 1249 1255

Adenosinyl Radicals
1249 1255
assume a similar extinction coefficient also for radical 10.
This analysis suggests that about 60% of the radicals, pro-
duced by reaction of the hydrated electron with 7, leads to
radical 10 (Scheme 3).
Further insight into the reaction was gained by experi-
ments in the presence of N,N,N’,N’-tetramethyl-p-phenyl-
enediamine (TMPD).[15] Pulsing O2-free solutions of 1 mm 7
containing 0.25m tBuOH and different concentrations of
TMPD (25 100 mm)at pH ~7 led to the oxidation of
TMPD. The rate constants for this oxidation, and the associ-
ated formation of TMPDC+ were measured at 565 nm (e=
1.25î104 mꢀ1 cmꢀ1), which represents one of the absorption
+
maxima of TMPDC (Figure 2).[16,17] Because under these ex-
perimental conditions HOC radicals are trapped by tBuOH,
and the reaction of TMPD with the carbon centered radicals
derived from tBuOH is unimportant in the time scale of our
experiments,[18] we concluded that the TMPDC+ radical
cation is formed from intermediates generated by the reac-
tion of eꢀaq with 7.[19]
Figure 2. Time dependence of absorption at 565 nm obtained from the
pulse radiolysis of an Ar-purged solution containing 1 mm 7, 0.25m
tBuOH and 96 mm TMPD at pH ~7; dose=22 Gy, optical path=2.0 cm.
Insets: a)Dependence of the first component k1 from the [TMPD].
b)Dependence of the second component k2 from the [TMPD].
+
The yield of TMPDC (corrected for its decay)increased
by increasing TMPD concentration, varying between 15%
for [TMPD]=25 mm to 40% for TMPD=100 mm relatively
to the yield of eꢀaq. Figure 2 shows that an induction period is
observed for the formation of TMPDC+, which could be sat-
isfactorily fitted using a two consecutive reactions model.
The first component is independent of TMPD concentration
(Figure 2, inset a)and occurs with a rate constant of kf =
(1.1Æ0.1)î105 sꢀ1. The second component depends on the
concentration of TMPD and the rate constant kTMPD =(4.6Æ
0.3)î108 mꢀ1 sꢀ1 (Figure 2, inset b)is assigned to the species
reacting with TMPD.
(kf =1.1î105 sꢀ1)to the heterolysis of the glycosidic bond in
radical 11 producing radical cation 12 and the base anion 13.
Analogous heterolytic b-C,O-bond cleavages of tertiary,
oxygen-substituted carbon radicals are well known.[20] Radi-
cal cation 12 oxidizes TMPD giving presumably compound
14. It is worth mentioning that in the case of the deoxyribo-
sides the build up of TMPDC+ was not observed.[7] This
strengthens our mechanistic suggestion because a heterolytic
cleavage of b-C,O-bonds in secondary alkyl radicals has
never been observed.
Based on the experimental data obtained from the inde-
pendent generation of the 2’-adenosinyl radical 11 (see
below), we suggest that the initially produced radical 8 af-
fords not only the 5’-radical 9 but also the 2’-radical 11
(Scheme 3). We assigned the observed unimolecular process
Independent generation of the 2’-adenosinyl radical (11):
The suggestion that radical 11 liberates adenine in a b-bond
cleavage was proven by its independent generation in a Nor-
rish photoreaction of 2’-acetylated nucleoside derivative 21
(Scheme 4). The synthesis of 21 started from the partly pro-
Scheme 3. Proposed mechanism for the fate of radical 5 based on pulse radiolysis studies.
1251
Chem. Eur. J. 2004, 10, 1249 1255
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

C. Chatgilialoglu, B. Giese et al.
FULL PAPER
Scheme 4. Synthesis of 2’-acetylated nucleoside 21.
Figure 3. Plot obtained from the competition between fragmentation and
hydrogen abstraction from GSH by radical 11 (Scheme 5)according to
Equation (4).
tected adenosine 17, which was benzoylated at the NH2
group of adenine (17!18), and oxidized at the free 2’-OH
function of the ribose (18!19). Vinylation of the resulting
ketone 19 yielded enol ether 20 whose deprotection led to
the 2’-acetylated nucleoside derivative 21. The C,C-bond for-
mation anti to the adenine (19!20)was proven by NOE ex-
periments.
Photolysis of 21 generated the adenosin-2’-yl radical 11,
which was trapped by glutathione (GSH)(Scheme 5.) This
led to the ribo- and arabino-products 22 and 23, respectively,
in a 1:3.5 ratio. The preferred attack of the thiol from the a-
face of the C2’ radical is due to the efficient b-face shielding
by the adenine moiety. In addition, adenine 24 was set free.
The dependence of the product ratio (22+23)/24 upon the
concentration of glutathione was determined in competition
kinetic experiments (Figure 3).
reacts about 10 times slower than the analogous unsubstitut-
ed alkyl radical. A reasonable assumption for the reaction
of C2’ radical 11 with GSH is a rate constant of about
106 mꢀ1 sꢀ1,[21] which suggests a rate constant kf of about 2î
105 sꢀ1 for the b-elimination. This is in good agreement with
the rate observed in the radiolysis studies.
DFT calculations: Theoretical calculations with the DFT
method were performed at the B3LYP6-31G* level[23,24] to
determine the factors that affect the activation energies of
radical translocation from the adenine to the C5’, C3’ and
C2’ positions of the ribose.[25] All three reactions are calcu-
lated to be strongly exothermic as expected on the thermo-
chemical grounds.[26] Radical translocation to the C5’ posi-
tion has the lowest activation energy (3.1 kcalmolꢀ1), which
is similar to that of the 2’-deoxyribose analogue (3.2
kcalmolꢀ1).[7] Hydrogen abstraction from the C2’ position
occurs with a barrier of 9.6 kcalmolꢀ1, whereas the analo-
gous reaction in the deoxyribonucleoside has a much higher
activation energy (16.2 kcalmolꢀ1). This difference of
6.6 kcalmolꢀ1 can be attributed to the stabilization of the
C2’ radical by the hydroxyl group in the ribose series. Radi-
cal translocation to the C3’ positions has the highest activa-
tion barrier (11.1 kcalmolꢀ1)in the ribonucleoside radical 8.
These calculations show that the difference of the activa-
tion energies between H-atom abstraction from C5’ (8!9)
and C2’ (8!11)is 6.5 kcalmol ꢀ1, whereas experimental ob-
servations indicate that the processes are competitive.
Maybe the activation entropies of these two pathways are
different and favor the hydrogen abstraction from C2’. Ac-
tually, in the ground state of 8 the distance between the rad-
ical site at C8 is considerably longer (3.84 ä)to the hydro-
gen at C5’ than to that at C2’ (2.70 ä). In order to reach the
transition state for the H-abstraction from C5’, one looses
not only the freedom of rotation around the C4’,C5’-bond
but one also decreases conformational freedom in the sugar
ring. Thus, the activation entropy favors the hydrogen ab-
straction from C2’ compared with the radical translocation
to C5’.
Scheme 5. Competition kinetics from the selective generation of C2’ radi-
cal 11.
Because the measurements were carried out with an at
least tenfold excess of glutathione, the H-abstraction could
be analyzed by a pseudo first-order kinetic. Thus, the ratio
between the rate of the hydrogen abstraction kH and the
rate of elimination kf are described by [Equation (4)].
½22Š þ ½23Š
½24Š
kH
kf
ð4Þ
¼
½GSHŠ
From the slope of the straight line in Figure 3 a kH/kf ratio
of 4.3mꢀ1 is obtained. The rates of H-transfer from alkane-
thiols to alkyl radicals depend strongly upon the radical sub-
stituents.[21,22] For example, the a,b-dialkoxyalkyl radical
Product studies from continuous radiolysis of 8-bromoade-
nosine 7: In order to further strengthen our mechanistic sug-
1252
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 1249 1255

Adenosinyl Radicals
1249 1255
gestions derived from the pulse radiolysis of 7, we also car-
ried out product studies from the g-radiolysis experiments.
Samples containing 3 mL of aqueous solution of 7 (ca.
1.5 mm)and tBuOH (0.25m)at pH ~7, were irradiated
under stationary-state conditions with a total dose up to
1.5 kGy at a dose rate of about 20 Gyminꢀ1, followed by
HPLC analysis. Although complex chromatograms were ob-
tained, the main reaction prod-
Fe(CN)46ꢀ, since its concentration is several orders higher
than that of Fe(CN)36ꢀ. So that, reduction of radical 28 fol-
lowed by fast protonation gives compound 26, which can
readily dehydrate (the OH group in the 5’ position is anti to
the H atom in the 8 position), and yields cyclonucleoside 27.
The formation of the hydrated aldehyde 25 (5%)should be
due to the oxidation of radical 9.
uct was the free base (adenine)
that accounts for 40 45% of
the reacted bromide. These re-
sults are in excellent agreement
with the pulse radiolysis obser-
vations reported above. It is
worth mentioning that adenine
was a minor reaction product in
the analogous experiment with
3,[7] which further motivates our
conclusion that the free base
can derive from the C2’ radical
chemistry.
In the case of 3, both radicals
5 and 6 are readily oxidized by
Fe(CN)36ꢀ, the rate constants
being
4.2î109
and
8.3î
108 mꢀ1 sꢀ1 respectively, whereas
radical 6 can also be reduced
by strong reductants.[7] It was
suggested that under g-irradia-
tion of aqueous solutions of
4 mm K4Fe(CN)6 (reductant), a
continuous generation of micro- Scheme 6. Proposed mechanism for the formation of nucleosides derived from C5’ radical 9.
molar levels of the oxidant
Fe(CN)36ꢀ could be obtained.
Such a low concentration of ox-
Conclusion
idant should allow radical 9 before being oxidized to cyclize
to radical 10.
The results described herein demonstrate that the reaction
of 8-bromoadenosine 7 with eꢀaq at pH ~7 leads to 5’-adeno-
sinyl radical 9 and 2’-adenosinyl radical 11 in a ratio of
about 60:40. In addition, radical 11 was selectively generat-
ed by photolyzing the precursor 21. Using time-resolved
spectroscopy and competition kinetic methods we could
show that these radicals undergo unimolecular reactions. We
found that the C5’ radical adds intramolecularly to the
C8,N7 double bond of the adenine moiety with a rate con-
stant of 1.0î104 sꢀ1 affording (5’R)and (5 ’S)-isomers in a
ratio of about 3.5:1, whereas the C2’ radical liberates ade-
nine with a rate constant of 1.1î105 sꢀ1 (Scheme 3). Our
findings can furnish a molecular basis for forthcoming ex-
periments with RNA radicals, as well as with adenosine de-
rivatives that play crucial roles in biological processes.
A deareated aqueous solution (1 L)containing 1.5 m m of
7 (520 mg), 0.25m tBuOH and 4 mm K4Fe(CN)6 at pH ~7
was g-irradiated with a total dose up to 3 kGy at a dose rate
of 18 Gyminꢀ1. After work-up, the following compounds
were eluted in the order (yields in parenthesis): adenosine
5’-carboxyaldehyde 25 (5%), (5’R)-5’,8-cycloadenosine 29
(14%), adenosine (10%), 5’,8-cyclo-5’-deoxyadenosine 27
(3%), (5’S)-5’,8-cycloadenosine 31 (4%), and adenine
(50%).
Also under these conditions, adenine was found once
again as the major reaction product (50%)and this further
suggests that heterolytic cleavage of the C2’ radical is faster
than oxidation by mm levels of Fe(CN)36ꢀ generated in situ.
Regarding the formation of products that derived from C5’
radical chemistry, Scheme 6 shows our mechanistic propos-
al.[27] Cyclization of C5’ radical 9 should afford mainly two
aminyl radicals 28 and 30, which are in the chair conforma-
tion. The fate of these radicals strongly depends on the con-
centration of the iron species. We suggest that the mm level
of Fe(CN)36ꢀ generated in situ can easily oxidize radicals 28
and 30 to give products 29 and 31, respectively. But radicals
28 and 30 can also be reduced to a minor extent by
Experimental Section
Pulse radiolysis: Pulse radiolysis with optical absorption detection was
performed by using the 12 MeV linear accelerator, which delivered 20
200 ns electron pulses with doses between 5 and 50 Gy, by which HOC, HC,
and eꢀaq are generated with 1 20 mm concentrations. The pulse irradiations
1253
Chem. Eur. J. 2004, 10, 1249 1255
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

C. Chatgilialoglu, B. Giese et al.
FULL PAPER
were performed at room temperature (22Æ28C)on samples contained in
Spectrosil quartz cells of 2 cm optical path length. Solutions were protect-
ed from the analyzing light by means of a shutter and appropriate cut-off
filters. The bandwith used throughout the pulse radiolysis experiments
was 5 nm. The radiation dose per pulse was monitored by means of a
charge collector placed behind the irradiation cell and calibrated with a
C(CHaHb)OCH3),4.09 (sb, 1H, HO), 3.97 (d, J=11.1, 1H, H-5’), 3.83 (d,
J=11.1 Hz, 1H, H-5’), 3.48 (s, 3H, OCH3), 0.95 (s, 9H, tBuSi), 0.91 (s,
9H, tBuSi), 0.15 (s, 3H, MeSi), 0.14 (s, 3H, MeSi), 0.09 (s, 3H, MeSi),
0.07 (s, 3H, MeSi); 13C NMR (75.5 MHz, CDCl3): d=172.3, 157.7, 152.0,
153.7, 151.4, 145.6, 134.1, 132.8, 129.4, 128.6, 126.9, 86.0, 85.4, 84.1, 81.7,
79.5, 63.1, 54.8, 25.9, 25.6, 18.5, 17.9, ꢀ4.7, ꢀ4.9, ꢀ5.4, ꢀ5.6. Because of
the low yield we have directly deprotected the compound.
ꢀ
N2O-saturated solution containing 0.1m HCO2 and 0.5mm methyl violo-
gen, using Ge=9.66î10ꢀ4 m2 Jꢀ1 at 602 nm.[28] G(X)represents the
number of moles of species X formed or consumed per Joule of energy
absorbed by the system.
(2-Acetyl-b-d-arabinofuranosyl)-adenine (21): Trimethylsilyl chloride
(33.5 mg, 0.31 mmol)was added at 20 8C to an acetonitrile solution
(3 mL)of enol ether 20 (40.9 mg, 0.54 mmol)and sodium iodide (512 mg,
0.34 mmol). After 1 h the reaction was quenched with a saturated solu-
tion of NH4Cl (60 mL)and the mixture extracted with dichloromethane
(3î60 mL). After drying over MgSO4 the solvent was removed under re-
duced pressure. This led to a yellow foam that was treated for 2 h with a
40% methylamine solution (2 mL)and ethanol (0.6 mL.) Evaporation
under reduced pressure and coevaporation with toluene (two times)af-
forded a residue that was solved in tetrahydrofuran (2 mL). A tetrahy-
drofuran solution of tetrabutylammonium fluoride (180mL of a 1m solu-
tion, 0.18 mmol)was added and, after 1 h, the reaction mixture was
quenched with trimethylsilyl chloride (85.9 mg, 0.79 mmol). After addi-
tion of water (1 mL), the mixture was co-evaporated twice under reduced
pressure with toluene and the residue separated by reversed phase
HPLC [tetraethylammonium acetate solution (20 mL, pH 7):CH3CN 98:2
(2 min)! 84:16 within 13 min]. This led to the deprotected ketone 21
(9.7 mg. 31.3 mmol, 61.1%)as a colourless powder. 1H NMR (300 MHz),
D2O): d=8.32 and 8.13 (s each, 1H each, H-2 and H-8), 6.57 (s, 1H, H-
1’), 4.58 (d, J=7.0 Hz, 1H, H-3’), 4.09 (dd, J=12.8, 2.0 Hz, 1H, Hb-5’),
3.90 (dd, J=12.8, 4.4 Hz, 1H, Ha-5’), 2.39 (s, 3H, COCH3); 13C NMR
(75.5 MHz, D2O): d=212.5, 156.0, 153.2, 149.2, 142.0, 118.7, 87.6, 85.9,
82.8, 77.8, 60.7, 27.9; MS (ESI): m/z: 310 [M+H]+ ; HR-MS (FAB): calcd
for C12H15N5O5: 310.1151; found: 310.1151 [M+H]+.
6-N,N-Dibenzoyl-3’,5’-di-O-(tert-butyldimethylsilyl)-adenosine (18): Tri-
methylsilyl chloride (4.70 g, 43.4 mmol)was added at 20 8C to a solution
of 17 (10.2 g, 20.6 mmol)in piperidine (80 mL.) After 80 min benzoyl
chloride (14.4 g, 102 mmol)was added and quenched after 90 min with a
saturated solution of NaHCO3 (1000 mL). The solution was extracted
with dichloromethane (4î100 mL), the extracts dried over MgSO4 and
the solvent removed under reduced pressure. The residue was solved in
ethanol (30 mL)and p-toluenesulfonic acid (160 mg, 0.94 mmol)added at
08C. After 2 h, the reaction was quenched with NaHCO3 (1 g,
11.9 mmol)and the ethanol removed under reduced pressure. The resi-
due was solved in water (100 mL), extracted with dichloromethane (4î
100 mL), dried over MgSO4 and the solvent removed under reduced
pressure. Chromatography over silica gel (dichloromethane/ethyl acetate
10:1)yielded nucleoside 18 as a yellow foam (9.40 g, 13.4 mmol, 67%).
1H NMR (300 MHz, CDCl3): d=8.65, 8.29 (s, H-2 and H-8, 1H each),
7.87 7.84 (m, 4H, Ph), 7.51 7.46 (m, 2H, Ph), 7.38 7.33 (m, 4H, Ph),
6.03 (d, J=4.7 Hz, 1H, H-1’), 4.57 4.63 (m, 2H, H-2’, H-3’), 4.14 (dd, J=
3.4, 3.5 Hz, 1H, H-4’), 3.92 (dd, J=3.6, 11.5 Hz, 1H, H-5’), 3.77 (dd, J=
3.2, 11.5 Hz, H-5’), 0.96, 0.88 (s each, 9H each, tBu), 0.18, 0.06, 0.03 0.01
(s each, 3H, CH3); 13C NMR (75.5 MHz, CDCl3): d=172.2, 152.8, 152.1,
151.7, 143.5, 134.0, 132.9, 129.4, 128.7, 127.8, 89.0, 85.5, 74.8, 71.5, 62.3,
25.9, 25.7, 18.4, 18.0, ꢀ4.6, ꢀ4.9, ꢀ5.4, ꢀ5.6; MS (ESI): m/z: 705
[M+H]+ ; HR-MS (FAB): calcd for C36H49N5O6Si2: 705.0163; found:
705.0161 [M+H]+.
Kinetic experiments: The modified adenosine 21 (0.62 mg, 0.20 mmol)
and a large excess of glutathione (20 mmol to 420 mmol)were solved in
aqueous tetraethylammonium chloride (TEAA, 200 mL of a 100 mm solu-
tion, pH 7.0)in a polymethyl methacrylate cuvette. The solution was de-
oxygenated with argon (30 min), thermostated at 208C, and irradiated
(500 W, Hg high pressure lamp, 320 nm cut-off filter)for 10 min. Then,
water (0.5 mL)was added and the solution analyzed by RP-HPLC
(20 mm teaa:CH3CN 98:2 to 84:16 within 13 min). The compounds 22:23
and 24 were quantified using known material. A plot of [22]+[23]/[24]
against the glutathione concentration gave a straight line with a slope of
4.3mꢀ1 (r=0.998).
6-N,N-Dibenzoyl-3’,5’-di-O-(tert-butyldimethylsilyl)-2’-keto-adenosine
(19): A solution of Dess Martin periodinane (6.90 g, 16.3 mmol)and pyr-
idine (6.39 g, 80.8 mmol)in dichloromethane (30 mL)was added at 20 8C
to nucleoside 18 (5.70 g, 8.10 mmol)in dichloromethane (35 mL.) After
23 h the solution was quenched at 08C with sodium thiosulfate (25.0 g,
100 mmol)and a saturated solution of NaHCO (200 mL). The mixture
3
was extracted with dichloromethane, dried with MgSO4 and the solvent
removed under reduced pressure. Chromatography on silica gel (di-
chloromethane/ethyl acetate 10:1)led to 19 as a colorless foam (3.60 g,
5.13 mmol, 63%). 1H NMR (300 MHz, CDCl3): d = 8.55, 8.08 (s each,
1H each, H-8 and H-2), 7.86 7.83 (m, 4H, Ph), 7.51 7.46 (m, 2H, Ph),
7.38 7.33 (m, 4H, Ph), 5.84 (s, 1H, H-1’), 5.20 (d, J=8.6 Hz, 1H, H-3’),
4.03 (dd, J=8.8, 3.8 Hz, 1H, H-4’), 4.03 (d, J=12.7 Hz, 1H, Hb-5’), 3.92
(dd, J=12.4, 3.6 Hz, 1H, Ha-5’), 0.95 (s, 9H, tBuSi), 0.77 (s, 9H, tBuSi),
0.23 (s, 3H, MeSi), 0.18 (s, 3H, MeSi), 0.01 (s, 3H, MeSi), ꢀ012 (s, 3H,
MeSi); 13C NMR (75.5 MHz, CDCl3): d=206.8, 172.0, 153.2, 152.2, 151.9,
144.5, 133.7, 133.6, 133.1, 129.3, 128.6, 126.8, 80.1, 79.9, 71.4, 60.7, 25.5,
18.1, ꢀ4.5, ꢀ5.3, ꢀ5.5, ꢀ5.6. Because of the lability of the substance
during chromatographic purification, a high resolution mass spectrum
could not be performed.
Continuous radiolysis: Continuous radiolyses were performed at room
temperature (22Æ28C)on 3 mL or 1 L samples using a 60Co-Gammacell,
with dose rates between 18 ꢀ20 Gyminꢀ1. The absorbed radiation dose
was determined with the Fricke chemical dosimeter, by taking G(Fe3+)=
[29]
ꢀ
1.61 mmolJꢀ1
.
The reactions of 7 (purchased from Sigma)with e and
aq
HC were investigated using deareated aqueous solutions containing
1.5 mm substrate and 0.25m tBuOH in the presence or absence of 4 mm
K4Fe(CN)6 at pH ~7. The 1 L solution in the presence of 4 mm
K4Fe(CN)6 was g-irradiated with a total dose up to 3 kGy. The crude re-
action mixture was passed through ion-exchange resin (Amberlite IRA-
400)in order to eliminate the iron salts. The reaction crude was lyophi-
lized and then separated by chromatography on RP silica gel (water/ace-
tonitrile 7:3). The following compounds were eluting in the order (yields
are based on the recovered starting material): 25[30] (Rf =0.91; 11 mg,
0.043 mmol; 5%), 29[31] (Rf =0.83; 31 mg; 0.12 mmol; 14%), adenosine
(Rf =0.62; 23 mg; 0.086 mmol; 10%), 27[32] (Rf =0.56; 6.4 mg;
0.026 mmol; 3%), 31[31] (Rf =0.56; 9.1 mg; 0.034 mmol; 4%), adenine
(Rf =0.47; 58 mg; 0.43 mmol; 50%), 7 (Rf =0.34; 221 mg; 0.64 mmol;
43% recovery).
6-N,N-Dibenzoyl-9-[3’,5’-di-O-(tert-butyldimethylsilyl)-2’-(vinyl-1’’-
methyl enol ether)-b-d-arabinofuranosyl]-adenine (20): At ꢀ788C tert-
butyllithium (0.7 mL of a 1.6m solution in pentane, 1.12 mmol)was
added slowly to a solution of methyl vinyl ether (620 mg, 10.7 mmol)in
tetrahydrofuran (1.6 mL). The mixture was warmed up to 08C for 3 min
and then cooled down again to ꢀ788C. During this procedure the yellow
solution became pale yellow-green. After 5 min the reaction mixture was
added to a cooled (08C)solution of ketone 19 (213 mg, 0.3 mmol)in tet-
rahydrofuran (5 mL). The mixture was quenched after 10 min with a sat-
urated NH4Cl solution (50 mL), water was added (50 mL) and extracted
with dichloromethane (3î50 mL). From the dried solution (MgSO4)the
solvent was evaporated under reduced pressure and the residue purified
by chromatography over silica gel (pentane/methyl acetate 3:1). This led
to enol ether 20 (65.4 mg, 28%). 1H NMR (300 MHz, CDCl3): d = 8.68
(s, 2H, H-8 and H-2), 7.87 7.84 (m, 4H, Ph), 7.50 7.45 (m, 2H, Ph),
7.37 7.32 (m, 4H, Ph), 6.68 (s, 1H, H-1’), 5.00 (1H, H-3’), 4.44 (d, J=
3.3 Hz, 1H, C(CHaHb)OCH3), 4.19 (s, 1H, H-4’), 4.13 (d, J=3.1 Hz, 1H,
Acknowledgments
Work supported in part by the European Community×s Human Potential
Programme under contract HPRN-CT-2002-00184 [SULFRAD] and by
the Swiss National Science Foundation. We thank Angelo Monti and
Alessandro Martelli for assistance with pulse radiolysis experiments.
1254
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 1249 1255

Adenosinyl Radicals
1249 1255
[1] K. Keck, Z. Naturforsch. B 1968, 23, 1034 1043.
[20] For example, see: A. Gugger, R. Batra, P. Rzadek, G. Rist, B. Giese,
J. Am. Chem. Soc. 1997, 119, 8740 8741; H. Strittmatter, A. Dussy,
U. Schwitter, B. Giese, Angew. Chem. 1999, 111, 238; Angew. Chem.
Int. Ed. 1999, 38, 135 137.
[2] J. A. Raleigh, W. Kremers, R. Whitehouse, Radiat. Res. 1976, 65,
414 422; J. A. Raleigh, A. F. Fuciarelli, Radiat. Res. 1985, 102, 165
175.
[3] A. Hampton, P. J. Harper, J. Sasaki, Biochemistry 1972, 11, 4965
4969; T. P. Haromy, J. Raleigh, M. Sundaralingam, Biochemistry
1980, 19, 1718 1722.
[4] A. F. Fuciarelli, F. Y. Shum, J. A. Raleigh, Radiat. Res. 1987, 110,
35 44; A. J. Alexander, P. Kebarle, A. F. Fuciarelli, J. A. Raleigh,
Anal. Chem. 1987, 59, 2484 2491.
[5] M. Dizdaroglu, P. Jaruga, H. Rodriguez, Free Radical Biol. Med.
2001, 30, 774 784; for early work, see: N. Marriagi, J. Cadet, R.
Teoule, Tetrahedron 1976, 32, 2385 2387.
[6] M.-L. Dirksen, W. F. Brakely, E. Holwitt, M. Dizdaroglu, Int. J.
Radiat. Biol. 1988, 54, 195 204; M. Dizdaroglu, M.-L. Dirksen, H.
Jiang, J. H. Robbins, Biochem. J. 1987, 241, 929 932.
[7] C. Chatgilialoglu, M. Guerra, Q. G. Mulazzani, J. Am. Chem. Soc.
2003, 125, 3839 3848; R. Flyunt, R. Bazzanini, C. Chatgilialoglu,
Q. G. Mulazzani, J. Am. Chem. Soc. 2000, 122, 4225 4226.
[8] It is worth mentioning that the analogous reaction of hydrated elec-
trons (eꢀaq)with 8-bromoguanosine behaves in a completely different
manner. See: M. Ioele, R. Bazzanini, C. Chatgilialoglu, Q. G. Mulaz-
zani, J. Am. Chem. Soc. 2000, 122, 1900 1907.
[21] C. Tronche, F. N. Martinez, J. H. Horner, M. Newcomb, M. Senn, B.
Giese, Tetrahedron Lett. 1996, 37, 5845 5848.
[22] An absolute rate constant of 4.4î106 mꢀ1 sꢀ1 for the reaction of 2’-
deoxyuridin-1’-yl radical with GSH was directly measured by laser
flash photolysis, see: C. Chatgilialoglu, C. Ferreri, R. Bazzanini, M.
Guerra, S.-Y. Choi, C. J. Emanuel, J. H. Horner, M. Newcomb, J.
Am. Chem. Soc. 2000, 122, 9525 9533.
[23] DFT calculations were carried out with the GAUSSIAN 98 system
of programs[24] employing the B3LYP functional and the valence
double-z basis set supplemented with polarization d-function on
heavy atoms (B3LYP6-31G*). Unrestricted wavefunction was used
for radical species (UB3LYP6-31G*). The nature of the transition
states were verified by frequency calculations (one imaginary fre-
quency)and energy values were corrected for the zero-point vibra-
tional energy (ZPVE).
[24] Gaussian 98, Revision A.7, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski,
J. A. Montgomery Jr., R. E. Stratmann, B J. C. urant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q.
Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Re-
plogle, J. A. Pople, Gaussian, Inc., Pittsburgh, PA, 1998.
[9] G. V. Buxton, C. L. Greenstock, W. P. Helman, A. B. Ross, J. Phys.
Chem. Ref. Data 1988, 17, 513 886, and references therein. A. B.
Ross, W. G. Mallard, W. P. Helman, G. V. Buxton, R. E. Huie, P.
Neta, NDRL-NIST Solution KineticDatabase, Ver. 3 , Notre Dame
Radiation Laboratory, Notre Dame, IN and NIST Standard Refer-
ence Data, Gaithersburg, MD, 1998 and references therein.
[10] The rate constant of HOC radical with 7 was found to be 3.8î
109 mꢀ1 sꢀ1 (unpublished work). Therefore, more than 97% of HOC
radicals are scavenged by tBuOH when [7]=1ꢁ10ꢀ3 m.
[11] G. L. Hug, Natl. Stand. Ref. Data Ser. (US Natl. Bur. Stand.) 1981,
69, 1 159.
[12] K. J. Visscher, M. P. de Haas, H. Loman, B. Vojnovic, J. M. Warman,
Int. J. Radiat. Biol. 1987, 52, 745 753.
[25] The anti conformer 8, which undergoes radical translocation, is sepa-
rated from the more stable syn conformer (DE=0.9 kcalmolꢀ1)by a
rotational barrier around the N-glycosidic bond of 2.9 kcalmolꢀ1
calculated here at the UB3LYP6-31G* level.
;
[13] For example, see: G. V. Buxton, Q. G. Mulazzani, A. B. Ross, J.
Phys. Chem. Ref. Data 1995, 24, 1055 1349.
[26] J. Berkowitz, G. B. Ellison, D. Gutman, J. Phys. Chem. 1994, 98,
2744 2765.
[14] L. P. Candeias, S. Steenken, J. Phys. Chem. 1992, 96, 937 944.
[15] TMPD, E8(TMPDC+/TMPD)=0.27 V. See: P. Wardman, J. Phys.
Chem. Ref. Data 1989, 18, 1637 1755, and references therein.
[16] S. Fujita, S. Steenken, J. Am. Chem. Soc. 1981, 103, 2540 2545; S.
Steenken, A. J. S. C. Vieira, Angew. Chem. 2001, 113, 581; Angew.
Chem. Int. Ed. 2001, 40, 571 573.
[27] For possible mechanistic paths for the formation of adenosine via
the reaction of HC with 7, see ref. [7].
[28] C. A. Kelly, E. L. Blinn, N. Camaioni, M. D×Angelantonio, Q. G.
Mulazzani, Inorg. Chem. 1999, 38, 1579 1584 (Corrigenda p. 2756).
[29] J. W. T. Spinks, R. J. Woods, An Introduction to Radiation Chemistry,
3rd ed., Wiley, New York, 1990, p. 100.
[17] P. S. Rao, E. Hayon, J. Phys. Chem. 1975, 79, 1063 1066.
[18] Generally electrophilic-type alkyl radicals like a-acyl derivatives
react with TMPD with a rate constant of ca. 108 mꢀ1 sꢀ1. The rate
constant of HC atom with TMPD is reported to be 2.2î108 mꢀ1 sꢀ1 at
pH 1 where TMPD is doubly protonated.[9]
[30] S. Liu, S. F. Wnuk, C. Yuan, M. J. Robins, R. T. Borchradt, J. Med.
Chem. 1993, 36, 883 887.
[31] A. Matsuda, M. Tezuka, K. Niizuma, E. Sugiyama, T. Ueda, Tetrahe-
dron 1978, 34, 2633 2637.
[32] A. Matsuda, K. Muneyama, T. Nishida, T. Sato, T. Ueda, Nucleic
Acids Res. 1976, 3, 3349 3357.
[19] Under the conditions employed: i)the competition between 7 and
TMPD for eꢀaq is well in favor of 7 since k(eaꢀq + TMPD)=9.1î
Received: August 28, 2003
Revised: November 13, 2003 [F5488]
107 mꢀ1 sꢀ1 [17]
,
and [7]ꢂ10î[TMPD], and ii)although the rate con-
stant of HOC with TMPD is 1.0î1010 mꢀ1 sꢀ1, (ref. [16])the HO C radi-
cals are still scavenged efficiently by tBuOH (ref. [10]).
1255
Chem. Eur. J. 2004, 10, 1249 1255
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim