Effect of Na+ and Ca2+ ions on the photochemical formation of hydrogen peroxide in frozen aqueous solutions of adenine derivatives

Article abstract of DOI:10.1134/S0036024417120160

The content of hydrogen peroxide is determined in NaCl-containing (0.001–0.1 M) aqueous solutions of adenosine-5'-diphosphate (ADP) and adenosine (Ado) irradiated with near-UV light at 77 K. The obtained data are compared to estimates of the integral intensity (Int) of the EPR signals from the irradiated solutions prior to their thawing and the contents of different components in these spectra (including peroxyl free radicals), calculated by analyzing these spectra from their modeling. It is found that at a low content of photosensitizers (2 × 10^?4 M), the relationship between [H₂O₂] and Int differs for ADP and Ado (a positive linear correlation in the case of ADP and a decreasing nonlinear dependence for Ado). It is assumed that these differences are due to the combination of two main factors: (1) the difference between the self-association constants (K_sa) of ADP and Ado and (2) the effect of the freezing process, during which added NaCl exerts a disaggregating action of the formation of ADP and Ado associates. The relation between [H₂O₂] and the EPR signal characteristics differs for high (1 × 10^?3 M) and low (2 × 10^?4 M) concentrations of photosensitizers, suggesting there are two paths of hydrogen peroxide formation in the investigated systems. According to the first path, H₂O₂ is formed during thawing as a result of interaction between peroxyl radicals O₂ ^-· and HO₂ ^· that accumulate during irradiation. According to the second path, the formation of H₂O₂ is direct (without releasing peroxyl radicals into the medium). This probably occurs inside aggregates of adenine derivatives, which form upon freezing the solution. In both cases, adding CaCl₂ to test solutions of Ado and ADP ([CaCl₂] = 0.1–0.2 × [NaCl]) results in a substantial drop in the yield of H₂O₂.

Full text of DOI:10.1134/S0036024417120160

ISSN 0036-0244, Russian Journal of Physical Chemistry A, 2017, Vol. 91, No. 12, pp. 2466–2474. © Pleiades Publishing, Ltd., 2017.
Original Russian Text © T.A. Lozinova, A.V. Lobanov, A.V. Lander, 2017, published in Zhurnal Fizicheskoi Khimii, 2017, Vol. 91, No. 12, pp. 2146–2154.
PHOTOCHEMISTRY
AND MAGNETOCHEMISTRY
+
2+
Effect of Na and Ca Ions on the Photochemical Formation
of Hydrogen Peroxide in Frozen Aqueous Solutions
of Adenine Derivatives
a,
b
c
T. A. Lozinova *, A. V. Lobanov , and A. V. Lander
a
Emanuel Institute of Biochemical Physics, Russian Academy of Sciences, Moscow, 119334 Russia
b
Semenov Institute of Chemical Physics, Russian Academy of Sciences, Moscow, 119991 Russia
c
Institute of Earthquake Prediction Theory and Mathematical Geophysics,
Russian Academy of Sciences, Moscow, 113556 Russia
*
е-mail: fepr@sky.chph.ras.ru
Received January 18, 2017
Abstract—The content of hydrogen peroxide is determined in NaCl-containing (0.001–0.1 M) aqueous solu-
tions of adenosine-5'-diphosphate (ADP) and adenosine (Ado) irradiated with near-UV light at 77 K. The
obtained data are compared to estimates of the integral intensity (Int) of the EPR signals from the irradiated
solutions prior to their thawing and the contents of different components in these spectra (including peroxyl
free radicals), calculated by analyzing these spectra from their modeling. It is found that at a low content of
photosensitizers (2 × 10−4 M), the relationship between [H O ] and Int differs for ADP and Ado (a positive
2
2
linear correlation in the case of ADP and a decreasing nonlinear dependence for Ado). It is assumed that
these differences are due to the combination of two main factors: (1) the difference between the self-associa-
tion constants (K ) of ADP and Ado and (2) the effect of the freezing process, during which added NaCl
sa
exerts a disaggregating action of the formation of ADP and Ado associates. The relation between [H O ] and
2
2
−
3
−4
the EPR signal characteristics differs for high (1 × 10 M) and low (2 × 10 M) concentrations of photo-
sensitizers, suggesting there are two paths of hydrogen peroxide formation in the investigated systems.
According to the first path, H O is formed during thawing as a result of interaction between peroxyl radicals
2
2
−i
i
2
O2 and HO that accumulate during irradiation. According to the second path, the formation of H O is
2
2
direct (without releasing peroxyl radicals into the medium). This probably occurs inside aggregates of adenine
derivatives, which form upon freezing the solution. In both cases, adding CaCl to test solutions of Ado and
2
ADP ([CaCl ] = 0.1–0.2 × [NaCl]) results in a substantial drop in the yield of H O .
2
2
2
Keywords: adenine derivatives, photolysis, hydrogen peroxide, EPR, peroxyl radicals
DOI: 10.1134/S0036024417120160
INTRODUCTION
[2]. Based on a comparison of the obtained results, it
was assumed that the observed uniqueness of AX
derivatives in terms of the photoproduction of H O
It was shown in [1, 2] that hydrogen peroxide can
be found in samples obtained by thawing aqueous
solutions of adenine derivatives AX (where X is H for
adenine (A), ribose for adenosine (Ado), and ribose-
2
2
could be due to their associative properties [2].
A comparison of the results from determining
2 2
their absorption bands (λ = 240–400 nm) and the
results from analyzing the EPR spectra recorded for
5
'-diphosphate for adenosine-5'-diphosphate (ADP)) H O in AX samples irradiated at the wavelengths of
irradiated with near-UV light at 77 K. The formation
of H O in Ado samples intensifies in the presence of
2
2
NaCl [1]. There is no H O in similarly irradiated the irradiated samples before thawing suggests that
2
2
samples of guanosine-5'-monophosphate and thy- H O in the investigated systems could form via the
2
2
mine, but trace amounts of it can be found in cytidine reaction
solutions. The content of hydrogen peroxide has been
−
2
i
i
2
–
determined in solutions of the derivatives of different
nucleic acid bases at a relatively high content of them
O + HO + H O → H O + O + OH .
(1)
2
2
2
2
−i
i
–
3
(
1 × 10 M) after irradiating these solutions under The formation of free radicals O2 and HO in the
2
identical conditions in the presence of NaCl (0.1 M) investigated systems is confirmed by their EPR spectra
466
2

+
2+
EFFECT OF Na AND Ca IONS
2467
−
•
O
2
•
AOH
−
•
•
HO
Cl
2
2
50 G
•
−
E
A
•
•
Ri
NO2
Fig. 1. Main basic spectra used in modeling the EPR signals of adenine derivatives.
obtained at 77 K and analyzed via their modeling [3, immersion in liquid nitrogen and released from the
4
], which is similar to the method used in this work.
The aims of this work are
cartridges immediately before irradiation. Irradiation
was performed using a DPSh-1000 high-pressure
mercury lamp with UFS-5 light filter (transmission,
(
a) to compare the results from determining the
2
40–400 nm). In [2], we observed no dependence of
photoinduced formation of hydrogen peroxide and
analyzing the EPR spectra of NaCl-containing solu-
tions of ADP and Ado to data obtained earlier, and to
analyze the effect of NaCl on the correlation between
the photoproduction of H O and the signal inten-
sity/composition of the EPR spectra of irradiated
solutions of Ado and ADP before thawing;
the yield of H O on the intensity of irradiation (I ) in
2
2
hν
the range of 0.2–1 at low concentrations of A and Ado,
despite the considerable difference between the inte-
gral intensities (Int) of the corresponding EPR signals.
The intensity of irradiation was therefore limited using
2
2
gratings calibrated on a spectrophotometer. The I
hν
value was ≈0.6 of the maximum intensity of irradia-
(
b) to analyze the effect of CaCl (a presumable tion. The samples were irradiated at 77 K (in liquid
2
complexing agent) on the possible relationship nitrogen) in a quartz Dewar flask. EPR spectra were
between these values.
recorded at 77 K (frequency, 9.5 GHz) on an EPR
spectrometer manufactured at our laboratory. The
microwave radiation power was W ≈ 200 μW and W ≈
EXPERIMENTAL
2
mW. The amplitude of modulation was ~2 G.
Our ADP (trisodium salt) came from Sigma; our
Ado, from Serva (Germany). Other reagents were of
reagent grade. Solutions were prepared using water
from a Millipore purification system. All experiments
were performed in a weakly acidic medium (pH 6.6).
The pH of the solutions was adjusted to this value
before freezing by adding HCl and NaOH. The con-
A quantitative estimate of the contribution from
different components to the total EPR spectra (signals
S) recorded at g ≈ 2.00 was obtained by constructing a
model signal using the DECO_2016 computer simula-
tion software, which is a modification of the versions
used earlier. The calculation procedure was described
in [4]. The individual (basic) signals from the compo-
nents that contributed the most to the recorded spec-
tra of irradiated solutions of AX are shown in Fig. 1.
−4
tents of ADP and Ado were 2 × 10 M. Control mea-
surements were performed at the higher concentration
−3
used in earlier studies, [AX] = 1 × 10 M. The content
−i
These are the signals from (1) peroxyl radicals O ; (2)
–4
2
of [NaCl] was varied in the range of (2–1000) × 10 M;
i
CaCl was added in amounts of 0.1–0.2 × [NaCl]. peroxyl radicals HO ; (3) radical cations of adenine,
2
2
Samples were frozen in Teflon cartridges via rapid most likely in their deprotonated form (pK
< 4.2 [5]);
a
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 91 No. 12 2017

2468
LOZINOVA et al.
Table 1. Results from determining H O and analyzing the EPR spectra for samples AX + NaCl and AX + NaCl + CaCl
2
2
Composition of solutions,
М × 104
2
Relative content of components
Comp-
ound
[H O ],
2 2
μM
Int, c.u.
−
i
i
•
A
•
•
−i
Cl2
[
AX]
[NaCl] [CaCl2]
O2
HO
Ri
AOH
2
ADP
ADP
ADP
ADP
ADP*
ADP*
ADP
ADP
ADP
ADP
ADP
ADP
Ado
Ado
Ado*
Ado
Ado
Ado
Ado
Ado
Ado
10
10
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1000
1000
1000
1000
1000
1000
10
12
10
–
2
–
100
–
100
–
100
–
–
2
10
–
2
–
–
–
100
–
–
2
5
10
250
33
190
26
240
43
30
39
10
1
1115
969
625
805
962
721
169
122
94
0.212
0.244
0.252
0.342
0.167
0.313
0.064
0.107
0.060
0.087
0.103
0.067
0.237
0.143
0.202
0.312
0.000
0.030
0.085
0.065
0.109
0.155
0.168
0.088
0.038
0.213
0.000
0.282
0.138
0.121
0.111
0.111
0.127
0.242
0.210
0.166
0.079
0.274
0.258
0.178
0.228
0.215
0.197
0.110
0.202
0.161
0.184
0.157
0.156
0.176
0.359
0.285
0.333
0.297
0.235
0.126
0.226
0.177
0.298
0.239
0.236
0.228
0.198
0.207
0.287
0.106
0.195
0.144
0.267
0.314
0.300
0.243
0.254
0.079
0.261
0.140
0.270
0.141
0.149
0.191
0.222
0.245
0.296
0.401
0.129
0.094
0.104
0.087
0.108
0.074
0.087
0.173
0.124
0.174
0.250
0.204
0.044
0.046
0.169
0.091
0.160
0.151
0.161
0.084
0.000
a
0.066
0.061
0.110
0.121
0.121
0.061
0.014
0.004
0.016
0.000
0.000
0.015
0.069
0.130
0.059
0.133
0.004
0.000
0.000
0.006
0.020
113
48
41
15
3
–
а
а
1000
1000
1000
1000
10
12
10
–
–
81
110
140
1151
b
335b
413b
787
157
167
177
113
b
38
145
195
72
45
48
167
Light filter, UFS-5. *Duration of irradiation t = 32 min; for other samples, t = 16 min. Intensity of irradiation I = 0.6; from [2], I = 1;
from [2], Ihν = 0.2. Shown are the average values of the results obtained in parallel experiments.
hν
hν
b
•
After the EPR spectra were recorded, the irradiated
(
4) ribose radicals Ri in Ado and ADP (these radicals
are mainly localized on C5'); (5) C8OH adducts of samples were stored for 24 h in liquid nitrogen before
•
−i
2
determining the H O content. In each experiment,
adenine AOH ; (6) Cl , which is present in NaCl-
containing solutions; (7) electrons E stabilized in the
2
2
–
the determination of H O was performed in parallel
2
2
i
2
with the irradiated samples for control samples that
matrix; and (8) NO radicals. The presence of the lat-
ter signal in the spectra was probably due to nitrate were not irradiated but were stored for the same time
admixtures in the components of the solutions. at 77 K.
Although this signal was clearly visible at the high
The quantitative determination of Н О was per-
microwave power (W), the contribution from the sig-
2
2
i
2
formed by spectral-iodometric means [6]. To accom-
plish this, the samples were thawed at room tempera-
nals of NO radicals to signals S at W ≈ 200 μW nor-
mally does not exceed ~5%. The parameters and con-
ditions for obtaining the basic spectra were given in [3, ture after photolysis. After measuring their volume, they
4
] and in the works cited there. In addition to the were supplied with 1 mL Н SO (0.2 М). Dissolved
2 4
abovementioned signals, the signal from the Dewar
flask was also added to the basic systems.
oxygen was displaced by blowing with СО . The sam-
2
ples were then mixed with 2 mL a deaerated 5% potas-
The yield from different paramagnetic products was
determined from the spectra recorded under conditions
in which the dependence of the integral intensity of all
the components on W was close to linear (W =
00 μW), similarly the same way as in earlier studies.
sium iodide water solution and purged with СО once
2
−
again. The emission of iodine, which forms complex I3
1
/2
anions with excess iodide anions, was detected via spec-
−1
−1
2
trophotometry (λmax = 351 nm, ε = 26400 M cm ).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 91
No. 12
2017

+
2+
EFFECT OF Na AND Ca IONS
2469
Int
Int
−
•
O
•
2
Ri
Cl
•
−•
HO
200
1
3
6
8
200
0
2
4
7
9
2
2
•
•
A
AOH
0
Int
k = 2
k = 2
k = 2
100
100
50
0
100
5
50
50
0
0
5
0 G
2
00
0
200
0
k = 2
k = 2
k = 2
100
100
50
0
100
50
0
10
50
0
Fig. 2. Examples of EPR spectra registered in irradiated solutions of (1–5) ADP and (6–10) Ado; [AX] = 2 × 10− M. Irradiation
4
was performed with a UFS-5 light filter; I = 0.6. The duration of irradiation was 16 min. The content of additives of inorganic
hν
salts was (1, 6) [NaCl] = 0.1 M; (3) [NaCl] = 0.001 M; (8) [NaCl] = 0.0012 M; (2, 7) [NaCl] = 0.1 M + [CaCl ] = 0.01 M;
2
(
4, 9) [NaCl] = 0.001 M + [CaCl ] = 0.0002 M; and (5, 10) [CaCl ] = 0.001 M. The experimental EPR spectra are represented
2
2
by gray lines; the model, by black lines. The results from estimating the contribution from the main components of signals
(in arbitrary units) are presented under the spectra. W = 200 μW.
RESULTS AND DISCUSSION
determining the integral intensities of the EPR spectra
in these samples before thawing (Int) and the results
Table 1 presents the results from determining the
hydrogen peroxide content in thawed AX samples irra- from assessing the contribution from the main compo-
diated at 77 K. The same table shows the results from nents to the registered total spectra. The content of
−
3
Table 2. Comparison of the H O yield and the Int of the EPR signals for the solutions of AX (1 × 10 М) + NaCl after
2
2
irradiation with a UFS-5 light filter (16 min) and data obtained earlier, along with data from calculating the concentrations
Ado
sa
ADP
Ado
Ado
saN
ADP
sa2
ADP
of dimeric associates [AX ] using the values of K
and Ksa in [11, 12] (*) and of Ksa2 , K
, K
, and KsaN in
2
a
[8, 9] ( )
[
H O ], μM/Int, c. u.
[AX ], μM
2
2
2
[
NaCl], M
Ihν
Reference
[2]
Ado
ADP
A
Ado
ADP
A
а
аа
0.1
1
440/1144
−
145/900
350/900
110/79
−
250/1115
90/598
−
84/1116
−
14*−24**;
2*−4**; 387 −400
а
аа
а
аа
0
.6
0.4
.2
~0.4
408 −423
359 −386
~
−
[1]
[2]
[1]
0
45/294
0
−
Double indices denote the results from calculating [AX ] while assuming that all values of K grow by a factor of 1.8 when the solutions
2
sa
freeze (this value was estimated from the data in [14] for purine solutions).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 91 No. 12 2017

2470
LOZINOVA et al.
−
4
Table 3. Comparison of the H O yield and the Int of the EPR signals in the solutions of AX (2 × 10 М) + NaCl after irra-
2
2
diation with a UFS-5 light filter (16 min) and data obtained earlier, plus the data from calculating the concentrations of dimeric
Ado
sa
ADP
Ado
Ado
saN
ADP
, K
sa2
, and KsaN in [8, 9] (a)
ADP
associates [AX ] using the values of K
and Ksa in [11, 12] (*) and of Ksa2 , K
2
[
H O ], μM/Int, c. u.
Reference
[AX ], μM
2
2
2
[
NaCl], M
.1
Ihν
Ado
ADP
A
Ado
ADP
A
0
1
81/1151
−
110/335
195/167
145/157
−
−
34/1285
[2]
[2]
0.6*−1.1**; 0.1*−0.2**;
а
аа
а
аа
а
аа
6
7 −74
50 −59
62 −69
0
0
.6
.2
190/625
−
39/122
30/169
15/48
−
32/239
−
0
0
.0012
.0010
0.6
−
−
0.0002
Double indices denote the results from calculating [AX ] while assuming that all values of K grow by a factor of 1.8 when the solutions
2
sa
freeze (this value was estimated from the data in [14] for purine solutions).
H O in the control samples (the unirradiated samples mental methods. For AX, UV spectroscopy has
2
2
revealed the existence of two hypochromic effects that
indicate the association of molecules [8, 9]. The first
that were stored for the same time intervals at 77 K)
did not exceed 1 μM. Examples of the EPR spectra
registered for the irradiated samples at 77 K and the
results from their modeling are shown in Fig. 2. It
should be noted that the tables show the concentration
of added NaCl. Since we used a trisodium salt of ADP,
–3
effect is apparent when [AX] < 1 × 10 М; the second
effect, at higher [AX]. The authors attribute the
observed effect to a substantial (up to two orders of
magnitude) increase in the formation constant of the
dimeric associates (K ), compared to the corre-
+
sa2
the actual [Na ] value at low contents of NaCl there-
fore exceeds those given in the table.
sponding value for the formation of associates of
higher orders, which is assumed to be the same (K ).
saN
(
I) The effect of NaCl. Tables 2 and 3 show the A more complex model that in the case of noncovalent
results from comparing measurements of the H O
bonds considers the loss of the rotational and transla-
2
2
yield and the Int of the signals in AX + NaCl systems tional degrees of freedom of monomeric aromatic sin-
to data obtained earlier. The yield of H O in ADP gle molecules upon their association also shows a
2
2
solutions exceeds that of H O in Ado solutions at the sharp drop in Ksa when N ≥ 3 [10].
2
2
−4
relatively low value of [AX] = 2 × 10 М and the fairly
The values of K obtained in [8, 9] approximately
saN
high value of [NaCl] = 0.1 М. In other cases, the yield correspond to the K values calculated using the
sa
of H O proves to be higher in Ado solutions than in
2
2
isodesmic model only in the close range of [AX]. The
ADP solutions. We assume that the observed ratios of following K values were obtained for Ado and ADP
sa
the H O yields in Ado and ADP solutions can be
2
2
via calculations using the isodesmic model and the
explained by a combination of two main causes:
NMR data (in D O and NaNO , ionic strength I of
2
3
Ado
The first of these is the difference between the self-
= 15 М−1 (pD = 7) [11] and
solutions is 1): K
sa
association constants (K ) of Ado and ADP. The litera-
sa
ADP
−1
Ksa = 1.8 М (pD = 8.9) [12].
ture data on determining the self-association con-
A reduction in pH raises the K value for nucleo-
AX
sa
sa
stants of AX (K ) in liquid solutions (these constants
tides, which is ~40% of the one obtained at more alka-
line pH values [13]. The corresponding values
obtained in [8, 9] (0.05 M phosphate buffer, pH 6.9)
are usually calculated in accordance with the widely
accepted isodesmic model, which assumes the identity
of all stages of the formation of multidimensional
Ado
ADP
−
1
−1
associates) are characterized by an extremely wide were KsaN = 120 М and KsaN = 51 М ; for second-
Ado
sa2
ADP
spread. (The results from determining the self-associ-
ation constant for adenosine-5'-triphosphate (ATP),
4
−1
order associates, K
= 1.6 × 10 М and Ksa2 =
4
−1
0
.5 × 10 М .
Tables 2 and 3 show the results from calculating the
concentrations of dimeric AX associates (AX ) in liq-
ATP
Ksa , differ by a factor of more than 60; for adenos-
AMP
ine-5'-monophosphate (AMP), the determined K
2
sa
uid solutions using the K values given in the works
values differ by a factor of 7 [7]). It was shown in [8, 9]
sa
that, in addition to the differences between possible cited above. In our calculations, we considered the
theoretical models chosen to analyze association, an probability of the formation of dimeric and trimeric
essential role is played by the working interval of AX associates. In both cases, the possibility that Ksa will
concentrations that is available to different experi- rise as the temperature falls during freezing was also
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 91
No. 12
2017

+
2+
EFFECT OF Na AND Ca IONS
2471
[
H
2 2
O ], µM
hold for other compounds, the use of the results from
[8, 9] is completely justified.
2
2
1
1
50
00
50
00
It should be noted that the determined K values
sa
depend largely on the content of inorganic salts in the
solutions. There are experimental data indicating that
in liquid solutions, the complexation of AX with met-
als promotes stacking interaction between molecules
[
7, 11, 12, 15–17]. These data apply to both adenine
nucleotides and to adenosine [17]. However, the K in
sa
Ado + NaCl (1 М) solutions is only ~1.2 times higher
than in pure aqueous solutions of Ado [17]. For ADP,
K increases by a factor of 1.3 when a 1 M solution of
sa
NaCl is used instead of a 0.1 M NaCl solution. This
effect is therefore unlikely to be important in the salt
−3
−1
5
0
concentration range that we uses (10 –10 М).
The second cause is the effect of freezing. This pro-
cess is apparently the most important (and the most
difficult to define) factor for a possible relationship
between the yield of H O and the association of AX.
0
2
2
0
500
1000 Int
It is known that the freezing of aqueous solutions of
aromatic organic molecules leads to their aggregation.
In the presence of inorganic salts, however, these
compounds disaggregate [18, 19]. It is assumed that
both vertical stacking and horizontal interactions
occur during the aggregation of molecules. The degree
of aggregation depends on the type of AX molecules,
their concentration, and the composition of the solu-
tion. The presence of NaCl in the frozen solutions
could thus be an important factor in the considered
experiments.
Fig. 3. Correlation between the yield of H O and the inte-
2
2
gral intensity of the EPR signals registered in irradiated
solutions of AX + NaCl before thawing. [AX] = 2 ×
‒
4
1
0
M; [NaCl] = 0.001–0.1 M. In the experiments with
CaCl , [CaCl ] = 0.1–0.2 × [NaCl] (Table 1). All samples
2
2
were irradiated with a UFS-5 light filter. ○ ADP, t =
1
6 min; ⊖ ADP, t = 32 min;
△
Ado, t = 16 min; • ADP +
CaCl , t = 16 min; ▲ Ado + CaCl , t = 16 min; I = 0.6.
2
2
hν
From [2]:
▽ Ado, t = 16 min, I = 0.2; , Ado, t =
hν
hν hν
3
2 min, I = 0.2; n| Ado, t = 16 min, I = 1.
A possible relation between the yield of H O and
2
2
the Int of EPR signals from AX + NaCl solutions with
rising [NaCl] is shown in Fig. 3. In the case of ADP,
the respective correlation was close to linear (coeffi-
taken into account. For purine solutions (Pu), this rise
is as high as ~1.8 as we go from room temperature to
4.9°C [14].
cient of linear correlation K was 0.986). In the case
cor
Calculations of [AX ] upon such a rise in K show of Ado, however, the dependence of the yield of H O
2
sa
2
2
a significant increase in [AX ] only when we use the on Int is apparently described by a falling curve.
2
K values from [11, 12]. In this case, however, the pos- (It should be noted that the values provided in this fig-
sa
sible [AX ] turns out to be much lower than the ure for Ado refer to the results obtained at different I .)
2
hν
observed yield of H O . Calculations of [AX ] based
2
2
2
Assuming there is a relation between the yield of
2
on the values of K and K in [8, 9] give values con- H O and the association of AX, the dependence
sa2
saN
2
sistent with the yield of H O (Tables 2 and 3).
observed for Ado can apparently be explained by the
disaggregating effect of NaCl. This effect is not appar-
ent in ADP solutions, due probably to stronger com-
2
2
In [2], it was assumed associates play an important
role in the formation of H O , based on a comparison
2
2
+
plexation of Na with the hydroxyls of the phosphate
of our experimental data and the K values reported in
sa
groups of ADP (the formation of ionic bonds between
the literature obtained via NMR for the derivatives of
+
–
Na and O ). In [18], luminescent measurements
showed that in the case of Ado, the disaggregating effect
various nucleic bases. Since the K values given in [8,
sa2
9
] exceed these by more than three orders of magni-
−3
of NaCl is very apparent even at [NaCl] = 10 M. Such
tude, we must estimate their ratios for different com-
pounds. This is apparently essential when dealing with
frozen solutions as well.
changes in aggregation cannot be achieved in the
nucleotide (AMP) even at [NaCl] = 1 М. The disag-
gregating effect of NaCl in solutions of A is apparent at
intermediate values of [NaCl].
Ado
Pu
According to [11, 15] the K /K ratio obtained
sa
sa
It is obvious that the formation of aggregates in the
via NMR for Ado and Pu at pH 7 is ~6.5. The
investigated systems is determined by a combination of
Ado
sa2
Pu
sa2
Ado
Pu
saN
K
/K
andK /K
ratios are ~5.3 and 7.5, two main factors: the self-association of AX and the
saN
respectively. These ratios are quite similar, and if they disaggregating action of NaCl. The assumption that
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 91 No. 12 2017

2472
LOZINOVA et al.
Table 4. The coefficients of linear correlation (K ) of the yield of hydrogen peroxide ([H O ]) with EPR signal intensities
cor
2
2
i
2
−i
2
i
2
S (Int), the intensities of signals of radicals [HO ] and product P = [O ] × [HO ], calculated using the linear relation
y = A + B × x
2
2
Kcor
R
R
lin
allo
Refe-
rence
Sample group
n
i
i
2
i
2
Int
[HO ]
P
Int
[HO ]
P
Int
[HO ]
P
2
Ado + A + ADP 16 0.822
0.924
0.927
0.676
0.854
0.859
0.857
0.930
0.869
[2]
[2]
[2]
[2]
1
2
0.874* 0.936* 0.950* 0.764* 0.876* 0.902* 0.944* 0.958* 0.939*
14 0.865 0.924 0.931 0.748 0.853 0.867 0.860 0.939 0.872
0.913* 0.935* 0.954* 0.834* 0.874* 0.910* 0.948* 0.971* 0.949*
10 0.825 0.944 0.930 0.681 0.892 0.865 0.861 0.937 0.869
0.906* 0.952* 0.959* 0.821* 0.906* 0.920* 0.950* 0.961* 0.942*
0.888 0.941 0.934 0.788 0.886 0.872 0.859 0.940 0.873
Ado + A
Ado + ADP
Ado
1
0
8
8
6
0.946* 0.949* 0.960* 0.895* 0.901* 0.922* 0.959* 0.975* 0.953*
2
2
value of linear correlation. R2 is the coefficient of determina-
R
is the coefficient of determination, which coincides with the K
lin
cor
allo
2
tion for allometric dependence y = C + D × x , n is the number of points in a group. * The values of the same parameters after excluding
samples with [AX] = 2 × 10−4 M; [NaCl] = 0.1 M. The samples were irradiated with a UFS-5 light filter.
the formation of H O is related to the association of these paths seems to be largely determined by the
2
2
AX [2] thus does not seem to contradict the literature structure of the frozen solutions.
data.
(
II) The effect of CaCl . When setting the aims of
2
2+
−3
this work, we assumed that adding Ca ions to the
investigated solutions could greatly increase the yield
of hydrogen peroxide through the abovementioned
promotion of stacking interaction between molecules
by metal ions [7, 11, 12, 15–17]. The constants of the
In [2], it was established for AX solutions (1 × 10 М)
containing 0.1 М NaCl that the yield of H O deter-
mined upon thawing the irradiated samples correlated
2
2
i
2
with the content of free radicals HO and the product
i
2
−i
2
of [HO ] × [O ] (P), both of which were calculated
2+
ADP association with Ca ions exceed the corre-
from the EPR spectra. Most of the data (>70%) pro-
+
sponding values for the bonds with Na ions by almost
−
3
vided in [2] were obtained at [AX] = 1 × 10 M, which
is relatively high. Removing the data corresponding to
two orders of magnitude. (For the equilibrium con-
3−
(3 − n)−
stants of the reactions M + ADP ↔ MADP
,
−
4
the low [AX] of 2 × 10 M from the results provided
in [2] considerably raises the coefficients of linear cor-
relation and determination (Table 4). In contrast to
where M is a metal ion and n is its valence, the values
of logK are 3.08 and 1.12, respectively [20].) However,
the results obtained for the photosystems in this work
flatly contradicted the assumption of a possible
increase in the yield of H O .
these results, there is no correlation between the yield
i
2
of H O and the content of radicals HO or the P value
2
2
2
2
−
4
(
Kcor < 0.900) at [AX] = 2 × 10 M (the concentration
In all of the analyzed cases, adding CaCl to AX +
NaCl solutions leads to a considerable drop in the
2
used in this work), not even in the case of ADP.
We may therefore assume the existence of two yield of H O (Table 1, Fig. 3). There is no correlation
2
2
paths of hydrogen peroxide formation in the analyzed between the amount of the produced hydrogen perox-
systems: (1) the formation of H O via the interaction ide and any of the components of the recorded EPR
2
2
−
i
i
spectra. However, there does seem to be a correlation
between the total Int of the EPR signals and the yield
of O and HO radicals (that accumulated during the
2
2
irradiation) in the process of thawing the solutions and
of H O (Fig. 3). In solutions of Ado + NaCl, the yield
(
2) the immediate formation of H O during irradia-
2
2
2
2
of H O in both the presence and the absence of CaCl
2
tion (possibly inside the aggregates of AX). The second
2
2
falls greatly upon an increase in the Int of the EPR sig-
nals. This in turn is due to an increase in the salt con-
tent (Table 1, Fig. 3). At the same content of AX,
path is also supported by the formation of H O in Ado
2
2
−
4
and A solutions (2 × 10 М) containing 0.1 M NaCl
during the irradiation of samples at λ = 290–450 nm,
which encompasses only the long-wavelength absorp-
tion region of AX [2]. For the samples irradiated at the
low I of 0.2 for 16 min, the yield of H O was 23–
NaCl, and the CaCl additive, the drop in the yield of
2
H O in the presence of CaCl is 1.6–3.3 times smaller
2
2
2
in the Ado solutions than in the ADP solutions. The
hν
2
2
2+
yield of H O in Ado solutions containing Ca
2
9 μM and there were virtually no EPR signals. The
2
2
quantitative ratio of the peroxide formed in each of exceeds the yield of H
O in similar ADP solutions by
2
2
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 91
No. 12
2017

+
2+
EFFECT OF Na AND Ca IONS
2473
a factor of 1.5–7.5, depending on the content of salts in [AX] visibly leads to greater transparency of the
in the solutions: 0.1 M NaCl, 0.01 M CaCl and 1 × samples, which probably means the impact of this fac-
2
−3
−4
tor becomes smaller. On the other hand, it was pro-
posed in [2–4] that the formation of hydroxide ions in
the weakly acidic pH range of 4–6.5 is possible with
the protonation of photoinduced triplet states of AX by
H O molecules, since the pK values for purines in
1
0 М NaCl, 2 × 10 М CaCl , respectively.
2
The drop in the formation of H O in the AX solu-
2
2
tions containing CaCl could obviously be due to the
2
2+
particular nature of the bonding between Ca ions
and Ado. In addition to the phosphate groups of
nucleotides, which are the main centers for bonding
metal cations [21], the hydroxyl groups of ribose and
endocyclic nitrogen atoms N7, N1, and N3 of the base
are also possible centers of their bonding (according to
crystallographic data) [21]. It is assumed that the
donor nitrogen atoms of the base and hydroxyl groups
of ribose play no notable role (according to NMR) in
2
a
the ground state are lower than those for purines in the
triplet state [26, 27]. Unfortunately, since we do not
know the data on the changes in the characteristics of
the photoexcited triplet states of AX complexes with
divalent metals, we cannot draw conclusions on the
possibility of this path of the formation of hydroxide
2+
ions in ADP complexes with Ca .
The conformational changes that affect the glyco-
2+
bonding nucleotides with Mg ions, the properties of
2
+
sidic bonds of ATP molecules in the presence of Mg
2+
which are similar to those of Ca ions [22]. Neverthe-
are shown in [28]. In principle, such changes in the
conformation of the investigated AX in the presence of
1
15
less, it has been shown by means of H– N HMBC
that N1 atoms of the adenine fragment of ATP interact
2+
Ca ions could considerably affect the structure of the
with Mg2 ions in the pH range of ~2.5–5.0 [23]. P
+
31
2+
frozen aggregates. The ionic radius of Ca greatly
2+
NMR shows that phosphate groups also bond Mg .
2+
exceeding that of Mg is also likely to be a significant
1
Diffusion measurements of H indicate that only one
factor in this effect.
2+
ligand and one Mg ion participate in ATP-Mg com-
plexes. Alkalizing the solution to pH ~ 7 reduces (but
In [1, 2], it was suggested that the unique ability of
AX to photoproduce H O could play an important
15
2
2
does not completely eliminate) the N chemical shift
role in the processes of evolutionary biochemistry.
2+
observed in the presence of Mg . This indicates the
participation of N1 atoms in the interaction with metal
ions. It should be noted that these experiments were
performed at very high concentrations of ATP (0.2 M)
where the concentration of ATP associates was high.
The data on the inhibition of the H O photoproduc-
2
2
2
+
tion by Ca ions generally do not contradict the sug-
2+
+
gested role of AX, since the [Ca ]/[Na ] ratio of 0.1–
.2 used in this work is much higher that the
0
2+
+
2+
−2
[Ca ]/[Na ] ratio in seawater: ~0.02 [29].
(
1
With no Mg , it is more than 6 × 10 M when K
.3 М [11] is used in calculations, and ~0.2 M when
using Ksa2 = 8000 М and K = 50 М in [9].) How-
=
sa2
−1
−1
−1
ACKNOWLEDGMENTS
The authors thank E.N. Degtyarev for his great
help in this work.
saN
15
ever, we believe that when a change in the N chemi-
cal shift is induced by the self-association of the nucle-
otide, the changes cannot be selective for the N1 atom.
Complexes of nucleotides with metals are thought to
be outer-sphere complexes in which the bonds
between metal ions and the nitrogen of heterocycles
REFERENCES
1
. T. A. Lozinova, A. V. Lobanov, and A. V. Lander, Russ.
J. Phys. Chem. A 89, 1492 (2015).
are formed through molecules of H O [24].
2
Analysis of the EPR spectra reveals no systematic
2. T. A. Lozinova, A. V. Lobanov, and A. V. Lander, Russ.
J. Phys. Chem. A 90, 2297 (2016).
change in the Int of the signals, or in the relative con-
tent of their components when CaCl is added to the
3. T. A. Lozinova and A. V. Lander, Biophysics 58, 341
2
(
2013).
irradiated solutions before they are frozen (Table 1). In
4
. T. A. Lozinova and A. V. Lander, Russ. J. Phys. Chem.
A 88, 163 (2014).
the presence of CaCl , however, there is always a drop
2
i
2
•
in the total relative content of the HO and AOH rad-
5
. K. Kobayashi, J. Phys. Chem. B 114, 5600 (2010).
icals. (With the pairwise summation of the relative
6. A. V. Lobanov, N. A. Rubtsova, Yu. A. Vedeneeva, and
content of other different components of the S signals,
G. G. Komissarov, Dokl. Chem. 421, 190 (2008).
no unidirectional change is observed when CaCl is
2
7
8
. L. Soccorsi and M. Cignitti, Bioelectrochem. Bioen-
added to the solutions).
erg. 9, 219 (1982).
The possibility of hydroxide ions being near pho-
. J. Morcillo, E. Gallego, and F. Peral, J. Mol. Struct.
toinduced reaction centers is important for the forma-
1
57, 353 (1987).
i
•
tion of HO and AOH radicals [1, 2]. At a relatively
2
9. F. Peral and E. Gallego, Biophys. Chem. 85, 79 (2000).
–3
high content of AX (1 × 10 M), interaction between
electron-deficient centers and the OH that accumu-
1
0. D. A. Beshnova, A. O. Lantushenko, D. B. Davies, and
–
M. P. Evstigneev, J. Chem. Phys. 130, 165105 (2009).
lates on a sample’s surface is possible, due to the opac- 11. K. H. Scheller, F. Hofstetter, P. R. Mitchell, et al.,
ity of the samples [25]. However, a fivefold reduction
J. Am. Chem. Soc. 103, 247 (1981).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 91 No. 12 2017

2474
LOZINOVA et al.
1
2. K. H. Scheller and H. Sigel, J. Am. Chem. Soc. 105, 21. W. Saenger, Principles of Nucleic Acid Structure
5
891 (1983).
3. R. Tribolet and H. Sigel, Eur. J. Biochem. 170, 617
1988).
4. K. E. van Holde and G. P. Rossetti, Biochemistry 6,
189 (1967).
5. R. B. Martin, Chem. Rev. 96, 3043 (1996).
6. H. Sigel, R. Tribolet, R. Malini-Balakrishnan, and
R. B. Martin, Inorg. Chem. 26, 2149 (1987).
(Springer, New York, 1987).
2
2. Z. Szabó, Coord. Chem. Rev. 252, 2362 (2008).
1
(
23. L. Jiang and X.-A. Mao, Spectrochim. Acta A 57, 1711
(2001).
1
2
24. E. M. Bianchi, S. A. A. Sajadi, and H. Sigel, Chem.
Eur. J. 9, 881 (2003).
1
1
2
2
2
2
5. S.-C. Park, E.-S. Moon, and H. Kang, Phys. Chem.
Chem. Phys. 12, 12000 (2010).
6. J. J. Aaron and J. D. Winefordner, Photochem. Photo-
17. B. I. Sukhorukov, I. Ya. Gukovsky, A. I. Petrov, et al.,
biol. 18, 97 (1973).
Int. J. Quantum Chem. 17, 339 (1980).
7. J. W. Longworth, R. O. Rahn, and R. G. Shulman,
18. V. Kleinwächter, Collect. Czech. Chem. Commun.
J. Chem. Phys. 45, 2930 (1966).
3
71, 1622 (1972).
8. L. Jiang and X.-A. Mao, Polyhedron. 21, 435 (2002).
19. Physico-Chemical Properties of Nucleic Acids, Ed. by 29. H. Trinks, W. Schroder, and C. K. Biebricher, Origins
J. Duchesne (Academic, London, 1973), p. 119.
Life Evol. Biosphere 35, 429 (2005).
Translated by Yu. Modestova
20. R. M. Smith, A. E. Martell, and Y. Chen, Pure Appl.
Chem. 63, 1015 (1991).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 91
No. 12
2017