- Email: [email protected]
Advances in solid alanine radiolysis understanding
Available online at www.sciencedirect.com
Spectrochimica Acta Part A 69 (2008) 904–910
Advances in solid alanine radiolysis understanding J. Raffi a,∗ , S. Talbi a , J.-M. Dolo b , T. Garcia b , J. Kister c a
LRMO, Case 461, UMR 6171, CEA & Universit´e Paul C´ezanne, 13397 Marseille Cedex 20, France b Laboratoire National Henri Becquerel, CEA-Saclay, 91191 Gif sur Yvette, France c GOAE, UMR 6171, Universit´ e Paul C´ezanne, 13397 Marseille Cedex 20, France Received 19 February 2007; received in revised form 11 May 2007; accepted 17 May 2007
Abstract To better understand the composite character of irradiated alanine ESR spectra, a comparative study of few simple amino acids is performed in order to identify the different radio-induced radicals and their proportions. A dedicated spin-trapping method coupled with High Performance Liquid Chromatography (HPLC) is developed and carried out on irradiated alanine, glycine and valine; labeled or not. This study leads us to obtain different isolated trapped radical spectra where hyperfine coupling constants could be evaluated. For alanine, only two species are identified with relative proportions around 97 and 3% in contradiction with recent published articles. The main species has a particularity on its hyperfine coupling constants when labeled carbons are used. Very high hyperfine coupling constants are observed with the carboxylic acid function carbon for the three studied amino acid. © 2007 Elsevier B.V. All rights reserved. Keywords: Electron spin resonance; Spin-trapping; Alanine; Radiation; Dosimetry
1. Introduction Based on studies of Bradshaw [1], ESR alanine dosimetry was developed by Regulla [2]. At LNHB, the method had an empirical and pragmatic development for ionizing radiation metrology since 1980 [3]. Alanine is an amino acid with a particular behaviour considering its radiolysis. Indeed, irradiated alanine is stable in solid phase, and has a characteristic ESR spectrum which allows quantitative measurements [4]. The first ESR studies showed the production of a radical called SAR (“stable alanine radical”), created after departure of the amino function [5]. This reaction was questioned in 1997 by Sagstuen et al. [6], who proposed a new hypothesis with the formation of two other described radicals. Theoretical studies, using Density functional Theory (DFT) simulation, also strengthen this hypothesis, showing the potential stability in solid phase of such radicals, which depends of surroundings intact alanine molecules [7,8]. Since, different radical proportions have been proposed [9–14] with variations depending on different experimental spectra. ∗
Corresponding author at: LRMO, Case 461, UMR 6171, CEA & Universit´e Paul C´ezanne, Facult´e de Saint-J´erˆome, Avenue Escadrille Normandie Ni´emen, 13397 Marseille Cedex 20, France. Tel.: +33 4 91 28 88 66; fax: +33 4 91 28 88 66. E-mail address: [email protected] (J. Raffi). 1386-1425/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2007.05.040
However, the isolated spectra have been only obtained by simulation. To determine experimentally the proportions of these radicals, spin-trapping experiments [5,15,16] have been carried out coupled to High Performance Chromatography (HPLC) [17,18] in order to obtain isolated species and their associated ESR spectra. The objective is to transform a parent radical (stable in powder but unstable in liquid phase) in a nitroxyde radical, stable in liquid state and then recording of ESR spectra allowing chemical interpretations until the initial parent radical. This method has been developed for quantitative purposes. Obtained spectra and radical proportions may be used in metrology to improve accuracy of ESR alanine dosimetry by deconvolution [14]. 2. Experimental Alanine and 2-methyl-2-nitroso-propane (MNP or “tNB”) dimer were purchased from Sigma–Aldrich. Specifically labeled amino acids were used to determine the chemical structure of the different radicals. Irradiations were performed at room temperature in the cobalt 60 cell of CEA-Cadarache, supplying a dose rate of about 5 kGy h−1 . Samples were also stored and analysed at room temperature and shielded against UV-rays.
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
ESR spectra were recorded with a Bruker X band spectrometer (model EMS104). Selected recording parameters were: microwave power of 15.77 mW, amplitude modulation of 0.143 mT and three scans on the sweep width of 8 mT. Spin-trapping method has been described previously [15–17]. Optimisation has been performed in order to obtain maximal monomer form in the initial MNP solution and maximal trapped radicals in the final trapped solution to get the better and the reproducible yield of reaction, including a fast dissolution of the alanine powder in the aqueous trap solution. This yield is usually low as there is a competition between the trapping reaction and the destruction of parent radicals by water. To get the pure ESR spectrum of each species, experiments have been carried out by HPLC. The HPLC apparatus is an Alliance Waters 2695, with an Interchrom C18 column coupled to a Waters 2996 diode array detector (DAD). The nitroxide solution was eluted with a water flow rate was 1 ml min−1 ; the fractions were collected during 2 min each.
905
Fig. 2. Influence of the trapping reaction temperature on the ESR signal intensity of the nitroxide solution spectrum.
the room temperature generally found in literature [5,15,16], but it is not established here if this increased yield is related to the higher monomer concentration, to the faster dissolution or both together.
3. Results and discussion 3.1. Spin-trapping of alanine 3.1.1. Optimization of spin-trapping parameters The reaction of a radical A• upon the double bond N O of a spin trap gives a stabilizated nitroxide radical R–N(–A)–O• by the de localization of the single electron between N and O atoms. But this reaction is in competition with the reaction of the radicals with the solvent. Commercial MNP is only in solid dimer form thus, numerous authors [5,15,16] insist on the necessity to move the equilibrium towards the maximal concentration of the monomer form which is the efficient trap. As already described previously [12,14], major specie is immediately observed (Fig. 1). For the optimization, ESR signal amplitude, measured as the sum of the six lines, is used to characterize spin-trapping efficiency. In case of alanine, to increase the dissolution speed and consequently the spin-trapping yield (so to get the higher ESR signal), the optimal temperature is around 35–40 ◦ C (Fig. 2) instead of
Fig. 1. ESR spectrum of the alanine nitroxide solution.
3.1.2. Influence of pH on ESR signal intensity The ESR spectrum of the nitroxide solution shows a triplet of doublet (Fig. 1) with two hyperfine coupling constants of aN = 1.60 mT and aH = 0.50 mT. Because of the multiple acidic functions of the parent molecule, investigations were done on the ESR signal regarding the pH of the solution. As assumed, its intensity appears dependent on the pH (Fig. 3). As the parent molecule has two acidic functions thus the main radical induced in powder, got par departure of the amino function [5,9], still has an acidic function and so in solution, the nitroxide radical derived from it. There are two possible states of the nitroxide radical, Fig. 3 shows that an equilibrium exists: starting from point “a” (at approximate pH of 10, obtained by NaOH addition), there is a decrease of the signal by HCl addition until point “b” (pH 2.5), where no ESR signal still exists; but if NaOH is added in the solution (point “c” at pH 7.6), the signal reappears. To explain that the radical form in acidic pH is not observable by ESR, the existence of a second equilibrium between this nitroxide radical and a diamagnetic species (Fig. 4) must be
Fig. 3. pH influence upon the ESR signal intensity.
906
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
Fig. 6. Labeled 13-C positions in alanine.
Fig. 4. Possible diamagnetic form (another hypothesis requires a tautomeric equilibrium such as the one proposed in Fig. 10).
supposed. Such type of hydrogen bonding was also proposed for a MNP/tyrosil adduct [18]. In a previous study [12,14], a triplet of triplet was also found; but because of its very low intensity, it was difficult to check if the second triplet was a 1:1:1 form (due to a nitrogen atom) or a 1:2:1 form (due to a CH2 group), which are relative to radicals A2 or A3 (Fig. 5). Multiple experimentations were performed and it appears that this structure is the A3 with hyperfine constants of aN = 1.58 mT, aH (two protons) = 0.52 mT. The comparison of ESR intensities implies that the two species relative concentrations are about 97 and 3%. 3.2. Spin-trapping of specifically labeled alanine
Fig. 7. ESR spectrum of nitroxide radicals derived from irradiated alanine labeled on carbon 2.
different from unlabeled alanine spectrum (Fig. 1) where there is a triplet (N) of doublet (H), thus six lines owing the same intensity. If one of the parent alanine is labeled with carbon 13, carbon 1 its spin 2 transforms each line into two new ones (separated by an hyperfine constant which is the higher value, the alone electron and the labeled carbon are closer); thus 12 lines instead of 6 must be observed, which is the case for the three labeling. The hyperfine constant of the main radical is larger in case of a labeling on carbon 2 than on carbon 3 (Figs. 7 and 8 and Table 1).
In order to improve our knowledge upon radical structures, alanine specifically labeled in positions 1, 2 or 3 (Fig. 6) were irradiated. Spectra obtained after spin-trapping (Figs. 7–9) are
Fig. 5. Possible structures of nitroxide radicals [8].
Fig. 8. ESR spectrum of nitroxide radicals derived from irradiated alanine labeled on carbon 3.
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
907
Table 1 Hyperfine constants (mT) of nitroxide radicals derived from irradiated labeled alanine (T = triplet; D = doublet) Multiplicity
T × D, none (Fig. 2)
T × D × D, 1 (Fig. 13)
T × D × D, 2 (Fig. 11)
T × D × D, 3 (Fig. 12)
A(N) A(H) A(13C)
1.60 0.50
1.60 0.53 1.27
1.61 0.50 0.68
1.61 0.5.3 0.39
None, 1, 2 and 3 are labeling positions.
Fig. 10. Nitroxide radicals showing bridge: A, from [14]; B, our hypothesis, including a tautomeric equilibrium. Fig. 9. ESR spectrum of nitroxide radicals derived from irradiated alanine labeled on carbon 1.
However, the nitroxide radical derived from alanine labeled in position 1 leads to a very strong value of the A(13C) hyperfine constant (Fig. 9), almost twice of the value of position 2 where the parent radical is centered, which is in contradiction with this radical structure. An equivalent phenomena in case of valine (1.4 mT) and glycine (1.25 mT) exists; the irradiated unlabeled glycine gives a main nitroxyde radical (triplet 1:1:1 by triplet 1:2:1) which can only correspond to the departure of the amino function; the irradiated 13C-1 labeled glycine leads to a triplet (N) of triplet (CH2 ) of doublet (13-C) with a large hyperfine constant (1.25 mT instead of 0.71 mT for a labeling in position 2) (Table 2). This value must be due to an important overlapping of the electronic orbitals of the alone electron (between N and O atoms of the nitroxide function) and of the electron of the cation carboxylic acid function. Intra or intermolecular associations have already been found elsewhere; for instance, in case of nitroxide-sugar radicals, the lines are broadened out as the radical concentration is increased by slow water evaporation at room temperature due the great number of possible hydrogen bridges; in the same way, we proved [10] that an H bridge obstructs the trapping of a galactose radical; we already
supposed in this paper such a bridge to explain the diamagnetic species in acid medium (Fig. 4). But such structures, involving H bridges were proved elsewhere; see for instance Fig. 10-A [19]. A structure such as the one shown on Fig. 10B is thus plausible and the two hypotheses (Figs. 4 and 10B) grow stronger, even if put on to explain different phenomena. As already said, the same idea was proposed in case of a MNP/tyrosil adduct [18]. 3.3. HPLC of alanine nitroxyde radical solutions Fig. 11 shows the evolution of UV spectra during HPLC of an irradiated alanine solution, without trap, i.e. also without nitroxide radicals. Peaks P1(A) of alanine and P2(A) of its major radiolytic product are clearly visible (Table 3). The maximal absorption of alanine is around 217.8 nm (Fig. 14).
Table 2 Hyperfine constants (mT) of nitroxide radicals derived from irradiated amino acids labeled on carbon 1 Glycine
A(N) A(H) A(13C-1)
1.60 0.84 (2H) 1.25
Alanine
Valine
Main radical
Minor radical
1.60 0.50 1.2.7
1.58 0.52 (2H) n.d.
1.57 0.38 1.40
Fig. 11. Evolution of UV spectra during HPLC of an aqueous solution of irradiated alanine (without spin trap).
908
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
Table 3 Peaks of chromatograms of irradiated alanine Pi(A) and pure trap Pi(ST) aqueous solutions
Table 4 Peaks Pi(N) of chromatogram of the nitroxide solution and comparison with those of irradiated alanine Pi(A) or trap solutions Pi(ST)
Peaks
Observations
Peaks
tR (min)
Radiolytic product derived from alanine Alanine; UV peak: maximum at 217.8 nm Impurity Impurity Trap (monomer form); UV peak: maximum at 243.7 nm Trap (dimer in equilibrium with monomer)
P1(N)
1.5–2
tR (min)
P1(A) P2(A) P1(ST) P2(ST) P3(ST)
4.4 6 4.6 6.1 10
P4(ST)
15–35
Peaks
P2(N)
4–4.5
P1(A) + P1(ST)
P3(N)
5.5
P2(A) + P2(ST)
P4(N)
6.7
P5(N)
7.9
P6(N)
10–11
P3(ST)
P7(N)
15–35
P4(ST)
Conclusions Not always observed, not attributed Alanine radiolytic product overlapping trap impurity Alanine overlapping trap impurity Non attributed; UV peak max: 237.8 nm (a) Main radical (b); UV peak max: 269.7 nm Trap (monomer form); UV peak: maximum at 243.7 nm Trap (dimer in equilibrium with monomer)
(a) may be the minor radical, (b) ESR spectrum got from the relative fraction collected after HPLC.
Fig. 12. Evolution of UV spectra during HPLC of an aqueous solution of MNP trap (without alanine and nitroxide radicals).
Fig. 12 shows the evolution of UV spectra during HPLC of a MNP trap solution alone. Four peaks exist (Table 3). Peaks P1(ST) and P2(ST) are not to the “parasite” radical derived from the spin trap, the di tertiobutyl nitroso radical because of its retention time (tR ) is longer for this type of column [17]; they are probably due to impurities. Peak P3(ST) is due to the monomer with a maximal absorption around 243.7 nm (Fig. 15). Peak P4(ST), with a retention time from 15 to approximately 35 mn, is due to a dynamic equilibrium between the monomer and the dimer form of the trap. Fig. 13 shows the evolution of the UV spectra during HPLC of irradiated alanine trapped with MNP, i.e. also including nitrox-
Fig. 14. UV spectrum of the alanine fraction: peak P2(A).
ide radicals derived from alanine. It shows six peaks, followed by a smaller one (Table 4), with a long retention time range easily attributable to the monomer–dimer equilibrium (Fig. 15). The attribution of the other peaks were done by analyzing retention times (tR ) and UV absorptions of the relative fractions (Figs. 14–17); some of them can be easily attributed to the peaks already observed in alanine and trap solutions (Table 3): • Peak P7(N) corresponds to the dimer peak P4(ST). • Peak P6(N) corresponds to the monomer peak P3(ST).
Fig. 13. Evolution of UV spectra during HPLC of the aqueous solution of nitroxide radicals.
Fig. 15. UV spectrum of the MNP trap (monomer form) fraction: peak P3(ST).
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
909
• Peak P1(N) was not attributed; it is sporadically observed and it was impossible to record an ESR spectrum from the relative collected fraction. • Even if it always exists, Peak P4(N) is not attributed. Maybe it is due to the minor radical but it was impossible to record an ESR spectrum from the relative collected fraction. • Peak P5(N) is attributed to the major radical as it was possible to record an ESR spectrum from the relative collected fraction; it should be noticed that this fraction was the only one belonging an ESR spectrum. 4. Conclusions and perspectives
Fig. 16. UV spectrum of the major radical nitroxide fraction: peak P5(N).
• Peak P3(N) corresponds to alanine, P2(A), but not only; it hides one impurity of the trap P2(ST). • Peak P2(N) corresponds to the radiolytic product of alanine, P1 (A), also hiding one impurity of the trap P1(ST). The three other peaks cannot be attributed to the ones described in Table 3:
This work proved that the ESR spectrum of solid irradiated alanine is not attributed only to the radical A1 but also to another: A3 , even if it is produced in a very weak proportion (≈3%). The chemical attributions of these two species were achieved and, using HPLC, for the first time, isolated trapped radicals were obtained. The spin-trapping method combined with HPLC, applied to the three studied amino acids, labeled or not, leads us to state about specificity of alanine radiolysis. If hydrogen bridges are known as contributor of orthorhombic crystal edifice of this organic solid, it seems that, as the high value of the hyperfine coupling constant in the liquid phase of the main trapped radical shows, it is probably also the case for the radical in the solid. Lot of hydrogen bridges may exist between radicals or with intact alanine. This hypothesis also suggested by Pauwels et al. [8] by simulation, is experimentally confirmed. It is suggested that intra or intermolecular associations may be found between alanine main radicals that implies solid various conformational positions. Thus, differences observed in solid irradiated alanine spectra may be due to the presence of radical chemically identical (i.e. leading to one main trapped radical), but with different surroundings (close carboxylic or amino functions) so with conformational differences in the alanine crystal. Acknowledgments Thanks to J. Vicente (CEA-Cadarache, DSV/DEVM) for carrying out irradiations and Laboratoire National de M´etrologie et d’Essais (LNE) for financial support. References
Fig. 17. UV spectrum of the unknown product of the nitroxide fraction: peak P4(N).
[1] W. Bradshaw, D. Cadena, G. Crawford, H. Spetzler, Radiat. Res. 17 (1962) 11. [2] D.F. Regulla, U. Deffner, Int. J. Appl. Isot. 33 (1982) 1101. [3] D. Mosse, Radiat. Phys. 31 (1988) 413. [4] ISO/ASTM, no. 51607, Use of an alanine-electron paramagnetic resonance dosimetry system, 2002. [5] M. Kuwabara, Y. Lion, P. Riesz, Int. J. Radiat. Biol. 39 (1981) 490. [6] E. Sagstuen, E. Hole, S. Haugedal, W. Nelson, J. Phys. Chem. 101 (1997) 9763. [7] F. Ban, S.T. Wetmore, R.J. Boyd, J. Phys. Chem. A103 (1999) 6650. [8] E. Pauwels, V. Van Speybroek, P. Lahorte, M. Waroquier, J. Phys. Chem. A105 (2001) 8794. [9] M. Heydari, E. Malinen, E. Hole, E. Sagstuen, Phys. Chem. A 106 (2002) 8971.
910
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
[10] E. Malinen, E.A. Hult, E. Hole, Radiat. Res. 159 (2003) 149. [11] J.-M. Dolo, E. Pichot, V. Feaugas, Appl. Magn. Reson. 15 (1998) 269. [12] S. Talbi, J. Raffi, S. Ar´ena, P. Piccerelle, P. Prinderre, J.-M. Dolo, Spectrochim. Acta 60 (2004) 1335. [13] J.-M. Dolo, F. Moignau, Appl. Radiat. Isot. 62 (2005) 281. [14] J.-M. Dolo, J. Raffi, P. Piccerelle, Rev. Franc. Metrol. 4 (2005) 5. [15] A. Minegishi, R. Bergene, P. Riesz, Int. J. Radiat. Biol. 38 (1980) 627.
[16] C. Thi´ery, J.-P. Agnel, C. Fr´ejaville, J. Raffi, J. Phys. Chem. 87 (1983) 4485. [17] J.-P. Agnel, C. Thi´ery, C. Battesti, P. Vincent, J. Raffi, Anal. Lett. 18 (1985) 1013. [18] S. Quian, Y.-R. Chen, L.J. Deterding, Y.C. Fann, C.F. Chignell, Biochem. J. 363 (2002) 281. [19] E. Janzen, I. Lopp, J. Magn. Res. 7 (1972) 107.
Spectrochimica Acta Part A 69 (2008) 904–910
Advances in solid alanine radiolysis understanding J. Raffi a,∗ , S. Talbi a , J.-M. Dolo b , T. Garcia b , J. Kister c a
LRMO, Case 461, UMR 6171, CEA & Universit´e Paul C´ezanne, 13397 Marseille Cedex 20, France b Laboratoire National Henri Becquerel, CEA-Saclay, 91191 Gif sur Yvette, France c GOAE, UMR 6171, Universit´ e Paul C´ezanne, 13397 Marseille Cedex 20, France Received 19 February 2007; received in revised form 11 May 2007; accepted 17 May 2007
Abstract To better understand the composite character of irradiated alanine ESR spectra, a comparative study of few simple amino acids is performed in order to identify the different radio-induced radicals and their proportions. A dedicated spin-trapping method coupled with High Performance Liquid Chromatography (HPLC) is developed and carried out on irradiated alanine, glycine and valine; labeled or not. This study leads us to obtain different isolated trapped radical spectra where hyperfine coupling constants could be evaluated. For alanine, only two species are identified with relative proportions around 97 and 3% in contradiction with recent published articles. The main species has a particularity on its hyperfine coupling constants when labeled carbons are used. Very high hyperfine coupling constants are observed with the carboxylic acid function carbon for the three studied amino acid. © 2007 Elsevier B.V. All rights reserved. Keywords: Electron spin resonance; Spin-trapping; Alanine; Radiation; Dosimetry
1. Introduction Based on studies of Bradshaw [1], ESR alanine dosimetry was developed by Regulla [2]. At LNHB, the method had an empirical and pragmatic development for ionizing radiation metrology since 1980 [3]. Alanine is an amino acid with a particular behaviour considering its radiolysis. Indeed, irradiated alanine is stable in solid phase, and has a characteristic ESR spectrum which allows quantitative measurements [4]. The first ESR studies showed the production of a radical called SAR (“stable alanine radical”), created after departure of the amino function [5]. This reaction was questioned in 1997 by Sagstuen et al. [6], who proposed a new hypothesis with the formation of two other described radicals. Theoretical studies, using Density functional Theory (DFT) simulation, also strengthen this hypothesis, showing the potential stability in solid phase of such radicals, which depends of surroundings intact alanine molecules [7,8]. Since, different radical proportions have been proposed [9–14] with variations depending on different experimental spectra. ∗
Corresponding author at: LRMO, Case 461, UMR 6171, CEA & Universit´e Paul C´ezanne, Facult´e de Saint-J´erˆome, Avenue Escadrille Normandie Ni´emen, 13397 Marseille Cedex 20, France. Tel.: +33 4 91 28 88 66; fax: +33 4 91 28 88 66. E-mail address: [email protected] (J. Raffi). 1386-1425/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2007.05.040
However, the isolated spectra have been only obtained by simulation. To determine experimentally the proportions of these radicals, spin-trapping experiments [5,15,16] have been carried out coupled to High Performance Chromatography (HPLC) [17,18] in order to obtain isolated species and their associated ESR spectra. The objective is to transform a parent radical (stable in powder but unstable in liquid phase) in a nitroxyde radical, stable in liquid state and then recording of ESR spectra allowing chemical interpretations until the initial parent radical. This method has been developed for quantitative purposes. Obtained spectra and radical proportions may be used in metrology to improve accuracy of ESR alanine dosimetry by deconvolution [14]. 2. Experimental Alanine and 2-methyl-2-nitroso-propane (MNP or “tNB”) dimer were purchased from Sigma–Aldrich. Specifically labeled amino acids were used to determine the chemical structure of the different radicals. Irradiations were performed at room temperature in the cobalt 60 cell of CEA-Cadarache, supplying a dose rate of about 5 kGy h−1 . Samples were also stored and analysed at room temperature and shielded against UV-rays.
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
ESR spectra were recorded with a Bruker X band spectrometer (model EMS104). Selected recording parameters were: microwave power of 15.77 mW, amplitude modulation of 0.143 mT and three scans on the sweep width of 8 mT. Spin-trapping method has been described previously [15–17]. Optimisation has been performed in order to obtain maximal monomer form in the initial MNP solution and maximal trapped radicals in the final trapped solution to get the better and the reproducible yield of reaction, including a fast dissolution of the alanine powder in the aqueous trap solution. This yield is usually low as there is a competition between the trapping reaction and the destruction of parent radicals by water. To get the pure ESR spectrum of each species, experiments have been carried out by HPLC. The HPLC apparatus is an Alliance Waters 2695, with an Interchrom C18 column coupled to a Waters 2996 diode array detector (DAD). The nitroxide solution was eluted with a water flow rate was 1 ml min−1 ; the fractions were collected during 2 min each.
905
Fig. 2. Influence of the trapping reaction temperature on the ESR signal intensity of the nitroxide solution spectrum.
the room temperature generally found in literature [5,15,16], but it is not established here if this increased yield is related to the higher monomer concentration, to the faster dissolution or both together.
3. Results and discussion 3.1. Spin-trapping of alanine 3.1.1. Optimization of spin-trapping parameters The reaction of a radical A• upon the double bond N O of a spin trap gives a stabilizated nitroxide radical R–N(–A)–O• by the de localization of the single electron between N and O atoms. But this reaction is in competition with the reaction of the radicals with the solvent. Commercial MNP is only in solid dimer form thus, numerous authors [5,15,16] insist on the necessity to move the equilibrium towards the maximal concentration of the monomer form which is the efficient trap. As already described previously [12,14], major specie is immediately observed (Fig. 1). For the optimization, ESR signal amplitude, measured as the sum of the six lines, is used to characterize spin-trapping efficiency. In case of alanine, to increase the dissolution speed and consequently the spin-trapping yield (so to get the higher ESR signal), the optimal temperature is around 35–40 ◦ C (Fig. 2) instead of
Fig. 1. ESR spectrum of the alanine nitroxide solution.
3.1.2. Influence of pH on ESR signal intensity The ESR spectrum of the nitroxide solution shows a triplet of doublet (Fig. 1) with two hyperfine coupling constants of aN = 1.60 mT and aH = 0.50 mT. Because of the multiple acidic functions of the parent molecule, investigations were done on the ESR signal regarding the pH of the solution. As assumed, its intensity appears dependent on the pH (Fig. 3). As the parent molecule has two acidic functions thus the main radical induced in powder, got par departure of the amino function [5,9], still has an acidic function and so in solution, the nitroxide radical derived from it. There are two possible states of the nitroxide radical, Fig. 3 shows that an equilibrium exists: starting from point “a” (at approximate pH of 10, obtained by NaOH addition), there is a decrease of the signal by HCl addition until point “b” (pH 2.5), where no ESR signal still exists; but if NaOH is added in the solution (point “c” at pH 7.6), the signal reappears. To explain that the radical form in acidic pH is not observable by ESR, the existence of a second equilibrium between this nitroxide radical and a diamagnetic species (Fig. 4) must be
Fig. 3. pH influence upon the ESR signal intensity.
906
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
Fig. 6. Labeled 13-C positions in alanine.
Fig. 4. Possible diamagnetic form (another hypothesis requires a tautomeric equilibrium such as the one proposed in Fig. 10).
supposed. Such type of hydrogen bonding was also proposed for a MNP/tyrosil adduct [18]. In a previous study [12,14], a triplet of triplet was also found; but because of its very low intensity, it was difficult to check if the second triplet was a 1:1:1 form (due to a nitrogen atom) or a 1:2:1 form (due to a CH2 group), which are relative to radicals A2 or A3 (Fig. 5). Multiple experimentations were performed and it appears that this structure is the A3 with hyperfine constants of aN = 1.58 mT, aH (two protons) = 0.52 mT. The comparison of ESR intensities implies that the two species relative concentrations are about 97 and 3%. 3.2. Spin-trapping of specifically labeled alanine
Fig. 7. ESR spectrum of nitroxide radicals derived from irradiated alanine labeled on carbon 2.
different from unlabeled alanine spectrum (Fig. 1) where there is a triplet (N) of doublet (H), thus six lines owing the same intensity. If one of the parent alanine is labeled with carbon 13, carbon 1 its spin 2 transforms each line into two new ones (separated by an hyperfine constant which is the higher value, the alone electron and the labeled carbon are closer); thus 12 lines instead of 6 must be observed, which is the case for the three labeling. The hyperfine constant of the main radical is larger in case of a labeling on carbon 2 than on carbon 3 (Figs. 7 and 8 and Table 1).
In order to improve our knowledge upon radical structures, alanine specifically labeled in positions 1, 2 or 3 (Fig. 6) were irradiated. Spectra obtained after spin-trapping (Figs. 7–9) are
Fig. 5. Possible structures of nitroxide radicals [8].
Fig. 8. ESR spectrum of nitroxide radicals derived from irradiated alanine labeled on carbon 3.
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
907
Table 1 Hyperfine constants (mT) of nitroxide radicals derived from irradiated labeled alanine (T = triplet; D = doublet) Multiplicity
T × D, none (Fig. 2)
T × D × D, 1 (Fig. 13)
T × D × D, 2 (Fig. 11)
T × D × D, 3 (Fig. 12)
A(N) A(H) A(13C)
1.60 0.50
1.60 0.53 1.27
1.61 0.50 0.68
1.61 0.5.3 0.39
None, 1, 2 and 3 are labeling positions.
Fig. 10. Nitroxide radicals showing bridge: A, from [14]; B, our hypothesis, including a tautomeric equilibrium. Fig. 9. ESR spectrum of nitroxide radicals derived from irradiated alanine labeled on carbon 1.
However, the nitroxide radical derived from alanine labeled in position 1 leads to a very strong value of the A(13C) hyperfine constant (Fig. 9), almost twice of the value of position 2 where the parent radical is centered, which is in contradiction with this radical structure. An equivalent phenomena in case of valine (1.4 mT) and glycine (1.25 mT) exists; the irradiated unlabeled glycine gives a main nitroxyde radical (triplet 1:1:1 by triplet 1:2:1) which can only correspond to the departure of the amino function; the irradiated 13C-1 labeled glycine leads to a triplet (N) of triplet (CH2 ) of doublet (13-C) with a large hyperfine constant (1.25 mT instead of 0.71 mT for a labeling in position 2) (Table 2). This value must be due to an important overlapping of the electronic orbitals of the alone electron (between N and O atoms of the nitroxide function) and of the electron of the cation carboxylic acid function. Intra or intermolecular associations have already been found elsewhere; for instance, in case of nitroxide-sugar radicals, the lines are broadened out as the radical concentration is increased by slow water evaporation at room temperature due the great number of possible hydrogen bridges; in the same way, we proved [10] that an H bridge obstructs the trapping of a galactose radical; we already
supposed in this paper such a bridge to explain the diamagnetic species in acid medium (Fig. 4). But such structures, involving H bridges were proved elsewhere; see for instance Fig. 10-A [19]. A structure such as the one shown on Fig. 10B is thus plausible and the two hypotheses (Figs. 4 and 10B) grow stronger, even if put on to explain different phenomena. As already said, the same idea was proposed in case of a MNP/tyrosil adduct [18]. 3.3. HPLC of alanine nitroxyde radical solutions Fig. 11 shows the evolution of UV spectra during HPLC of an irradiated alanine solution, without trap, i.e. also without nitroxide radicals. Peaks P1(A) of alanine and P2(A) of its major radiolytic product are clearly visible (Table 3). The maximal absorption of alanine is around 217.8 nm (Fig. 14).
Table 2 Hyperfine constants (mT) of nitroxide radicals derived from irradiated amino acids labeled on carbon 1 Glycine
A(N) A(H) A(13C-1)
1.60 0.84 (2H) 1.25
Alanine
Valine
Main radical
Minor radical
1.60 0.50 1.2.7
1.58 0.52 (2H) n.d.
1.57 0.38 1.40
Fig. 11. Evolution of UV spectra during HPLC of an aqueous solution of irradiated alanine (without spin trap).
908
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
Table 3 Peaks of chromatograms of irradiated alanine Pi(A) and pure trap Pi(ST) aqueous solutions
Table 4 Peaks Pi(N) of chromatogram of the nitroxide solution and comparison with those of irradiated alanine Pi(A) or trap solutions Pi(ST)
Peaks
Observations
Peaks
tR (min)
Radiolytic product derived from alanine Alanine; UV peak: maximum at 217.8 nm Impurity Impurity Trap (monomer form); UV peak: maximum at 243.7 nm Trap (dimer in equilibrium with monomer)
P1(N)
1.5–2
tR (min)
P1(A) P2(A) P1(ST) P2(ST) P3(ST)
4.4 6 4.6 6.1 10
P4(ST)
15–35
Peaks
P2(N)
4–4.5
P1(A) + P1(ST)
P3(N)
5.5
P2(A) + P2(ST)
P4(N)
6.7
P5(N)
7.9
P6(N)
10–11
P3(ST)
P7(N)
15–35
P4(ST)
Conclusions Not always observed, not attributed Alanine radiolytic product overlapping trap impurity Alanine overlapping trap impurity Non attributed; UV peak max: 237.8 nm (a) Main radical (b); UV peak max: 269.7 nm Trap (monomer form); UV peak: maximum at 243.7 nm Trap (dimer in equilibrium with monomer)
(a) may be the minor radical, (b) ESR spectrum got from the relative fraction collected after HPLC.
Fig. 12. Evolution of UV spectra during HPLC of an aqueous solution of MNP trap (without alanine and nitroxide radicals).
Fig. 12 shows the evolution of UV spectra during HPLC of a MNP trap solution alone. Four peaks exist (Table 3). Peaks P1(ST) and P2(ST) are not to the “parasite” radical derived from the spin trap, the di tertiobutyl nitroso radical because of its retention time (tR ) is longer for this type of column [17]; they are probably due to impurities. Peak P3(ST) is due to the monomer with a maximal absorption around 243.7 nm (Fig. 15). Peak P4(ST), with a retention time from 15 to approximately 35 mn, is due to a dynamic equilibrium between the monomer and the dimer form of the trap. Fig. 13 shows the evolution of the UV spectra during HPLC of irradiated alanine trapped with MNP, i.e. also including nitrox-
Fig. 14. UV spectrum of the alanine fraction: peak P2(A).
ide radicals derived from alanine. It shows six peaks, followed by a smaller one (Table 4), with a long retention time range easily attributable to the monomer–dimer equilibrium (Fig. 15). The attribution of the other peaks were done by analyzing retention times (tR ) and UV absorptions of the relative fractions (Figs. 14–17); some of them can be easily attributed to the peaks already observed in alanine and trap solutions (Table 3): • Peak P7(N) corresponds to the dimer peak P4(ST). • Peak P6(N) corresponds to the monomer peak P3(ST).
Fig. 13. Evolution of UV spectra during HPLC of the aqueous solution of nitroxide radicals.
Fig. 15. UV spectrum of the MNP trap (monomer form) fraction: peak P3(ST).
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
909
• Peak P1(N) was not attributed; it is sporadically observed and it was impossible to record an ESR spectrum from the relative collected fraction. • Even if it always exists, Peak P4(N) is not attributed. Maybe it is due to the minor radical but it was impossible to record an ESR spectrum from the relative collected fraction. • Peak P5(N) is attributed to the major radical as it was possible to record an ESR spectrum from the relative collected fraction; it should be noticed that this fraction was the only one belonging an ESR spectrum. 4. Conclusions and perspectives
Fig. 16. UV spectrum of the major radical nitroxide fraction: peak P5(N).
• Peak P3(N) corresponds to alanine, P2(A), but not only; it hides one impurity of the trap P2(ST). • Peak P2(N) corresponds to the radiolytic product of alanine, P1 (A), also hiding one impurity of the trap P1(ST). The three other peaks cannot be attributed to the ones described in Table 3:
This work proved that the ESR spectrum of solid irradiated alanine is not attributed only to the radical A1 but also to another: A3 , even if it is produced in a very weak proportion (≈3%). The chemical attributions of these two species were achieved and, using HPLC, for the first time, isolated trapped radicals were obtained. The spin-trapping method combined with HPLC, applied to the three studied amino acids, labeled or not, leads us to state about specificity of alanine radiolysis. If hydrogen bridges are known as contributor of orthorhombic crystal edifice of this organic solid, it seems that, as the high value of the hyperfine coupling constant in the liquid phase of the main trapped radical shows, it is probably also the case for the radical in the solid. Lot of hydrogen bridges may exist between radicals or with intact alanine. This hypothesis also suggested by Pauwels et al. [8] by simulation, is experimentally confirmed. It is suggested that intra or intermolecular associations may be found between alanine main radicals that implies solid various conformational positions. Thus, differences observed in solid irradiated alanine spectra may be due to the presence of radical chemically identical (i.e. leading to one main trapped radical), but with different surroundings (close carboxylic or amino functions) so with conformational differences in the alanine crystal. Acknowledgments Thanks to J. Vicente (CEA-Cadarache, DSV/DEVM) for carrying out irradiations and Laboratoire National de M´etrologie et d’Essais (LNE) for financial support. References
Fig. 17. UV spectrum of the unknown product of the nitroxide fraction: peak P4(N).
[1] W. Bradshaw, D. Cadena, G. Crawford, H. Spetzler, Radiat. Res. 17 (1962) 11. [2] D.F. Regulla, U. Deffner, Int. J. Appl. Isot. 33 (1982) 1101. [3] D. Mosse, Radiat. Phys. 31 (1988) 413. [4] ISO/ASTM, no. 51607, Use of an alanine-electron paramagnetic resonance dosimetry system, 2002. [5] M. Kuwabara, Y. Lion, P. Riesz, Int. J. Radiat. Biol. 39 (1981) 490. [6] E. Sagstuen, E. Hole, S. Haugedal, W. Nelson, J. Phys. Chem. 101 (1997) 9763. [7] F. Ban, S.T. Wetmore, R.J. Boyd, J. Phys. Chem. A103 (1999) 6650. [8] E. Pauwels, V. Van Speybroek, P. Lahorte, M. Waroquier, J. Phys. Chem. A105 (2001) 8794. [9] M. Heydari, E. Malinen, E. Hole, E. Sagstuen, Phys. Chem. A 106 (2002) 8971.
910
J. Raffi et al. / Spectrochimica Acta Part A 69 (2008) 904–910
[10] E. Malinen, E.A. Hult, E. Hole, Radiat. Res. 159 (2003) 149. [11] J.-M. Dolo, E. Pichot, V. Feaugas, Appl. Magn. Reson. 15 (1998) 269. [12] S. Talbi, J. Raffi, S. Ar´ena, P. Piccerelle, P. Prinderre, J.-M. Dolo, Spectrochim. Acta 60 (2004) 1335. [13] J.-M. Dolo, F. Moignau, Appl. Radiat. Isot. 62 (2005) 281. [14] J.-M. Dolo, J. Raffi, P. Piccerelle, Rev. Franc. Metrol. 4 (2005) 5. [15] A. Minegishi, R. Bergene, P. Riesz, Int. J. Radiat. Biol. 38 (1980) 627.
[16] C. Thi´ery, J.-P. Agnel, C. Fr´ejaville, J. Raffi, J. Phys. Chem. 87 (1983) 4485. [17] J.-P. Agnel, C. Thi´ery, C. Battesti, P. Vincent, J. Raffi, Anal. Lett. 18 (1985) 1013. [18] S. Quian, Y.-R. Chen, L.J. Deterding, Y.C. Fann, C.F. Chignell, Biochem. J. 363 (2002) 281. [19] E. Janzen, I. Lopp, J. Magn. Res. 7 (1972) 107.