68
J. Phys. Chem. 1981, 85,68-75
since we are comparing a monosubstituted nitrogen atom (trien) with a trisubstituted nitrogen (I and 11)a difference in ASo of about 10 cal mor1 K-l was expected. Our results (Table 111)are in perfect agreement. With the approach of a second proton, the order around another H30+ion is lost; however, this loss is now partly compensated by the new order which is created by the two charges on the diprotonated molecule. In the case of diamines when the distance between the two nitrogen atoms exceeds four methylene groups, ASoz becomes equal to AS01,9J4 taking into account the correction due to the statistical effect. On the other hand, in the case of t r i a m i n e ~ l ~ l l or ~ ~tetramines ’ ~ ’ ~ ~ ~ ~(see Table 111) ASo2 remains much lower than ASol also when the distance between the two terminal nitrogen atoms (first to be p r o t ~ n a t e d exceeds )~~ four methylene groups. This is due to the ability of inner amino groups to coordinate further water molecules even in the unprotonated form. As a consequence ASoz is much lower than ASol due to an increased order of water molecules around the differently charged (+ or -) amino group^.^ Obviously, the ASo values are also affected by the substitution degree of
neighboring nitrogens. Therefore, both ASo, and ASoz of I1 are lower than the corresponding entropy values of I (Table 111). The further reduction in entropy values on going to the third and fourth protonation may be explained in a similar way. The last two entropy values of I are more similar to those of trien than to those of I1 because these last protonations always refer to inner nitrogens which are disubstituted in the case of I and trien, but are trisubstituted in the case of 11.
Acknowledgment. The financial support of the Italian Research Council (CNR) is gratefully acknowledged. Supplementary Material Available: Individual data points for emf measurements including milliliters of OHadded, emf, total concentration of amine, total concentration of acid, and residual on the total concentrations of amine and acid (i.e., the difference between the analytical values and those computed by the program MINIQUAD 76A) (7 pages). Ordering information is available on any current masthead page.
Oxidation of Amino-Containing Disulfides by B r p and OH. A Pulse-Radiolysis Study A. John Elllot, Roderlck J. McEachern, and David A. Armstrong” Department of Chemistry, University of Calgary, Calgaty, Alberta, Canada, T2N 1N4 (Received; June 12, 1980)
The rate constant for reaction of B r p with dithiodipropionic acid (-4.2 X lo8 mol-l dm3s-l) was independent of pH in the range 6.611.0 and was -4.5 times smaller than those for the neutral dimethyl and diethyl disulfides. Bra. only reacted with the disulfides of cysteamine, cysteine, and penicillamine with an appreciable rate (>lo8 mol-l dm3s-’) when one or both of the amino groups were unprotonated. Homocystine was less sensitive to the degree of protonation. While the reaction of Brz-. with dithiodipropionic acid yielded a transient (A, N 450 nm) which possessed the characteristicsof a disulfide cation, the amino-containingdisulfides cited above produced transients which adsorbed with a ,A, near 380 nm. The latter transients decayed by second-order kinetics over the pH range studied, and they gave no evidence of reaction with OH-. These species have been tentatively identified as perthiyl (RSS.) radicals. Hydroxyl radicals produced composite spectra consisting of the 380-nm species and other transients which absorbed below 350 nm. Oxidized glutathione gave only weak absorptions on reaction with BrZ; and OH. The yields of sulfydryl molecules have been measured for cystine and dithiodipropionic acid. The mechanisms of radiolysis are discussed in the light of present results and earlier studies.
Introduction The 1960s and early 1970s saw a great deal of research on the radiation chemistry of aqueous solutions of sulfur-containing amino acids and related comp~unds.l-~ This arose because of the “in vivo” biological radiationprotection potential of these compounds. Although most research centered on the sulfydryl molecules, disulfides were also in~estigated.~~~ In deaerated solutions containing disulfides, the principal products were the corresponding thiol, sulfinic acid, trisulfide, and products arising from C-S ~leavage.~ Furthermore, the hydroxyl radical ap(1) J. E. Packer in “The Chemistry and the Thiol Group”, Part 2, S. Patai, Ed., Wiley, London, 1974, p 481 and references within. (2) G. C. Goyal and D. A. Armstrong, Can. J. Chem., 53,1475 (1975); J.Phys. Chem., 80, 1848 (1976). (3) T. C. Owen, A. C. Wilbraham, J. A. G. Roach, and D. R. Ellis, Radiat. Res., 50, 234 (1972). (4) J. W. Purdie, J. Am. Chem. Soc., 89, 226 (1967); Can. J . Chem., 47, 1029, 1037 (1969); 49, 725 (1971). 0022-3654/81/2085-0068$01 .OO/O
peared to be implicated to some extent in the formation of all of these product^.^ Until relatively recently, pulse-radiolysis studies of disulfides were concentrated on the one-electron reduction by e,; and COz-. because the disulfide anion absorption, A, = 400-450 nm,5,6was often observed when disulfidecontaining proteins were reacted with e,; and COz-..7-10 However, utilizing pulse radiolysis with both optical and (5) M. Hoffman and E. Hayon, J. Am. Chem. Soc., 94, 7950 (1972). (6) J. W. Purdie, H. A. Gillis, and N. V. Klassen, Can. J. Chem., 51, 3132 (1973). (7) J. R. Clement, D. A. Armstrong, N. V. Klassen, and H. A. Gillis, Can. J. Chem., 50, 2833 (1972). (8) R. H. Bisby, R. B. Cundall, J. L. Redpath, and G. E. Adams, J. Chem. SOC.,Faraday Trans. 1, 72, 51 (1976). (9) M. Faraggi, M. H. Klapper, and L. M. Dorfman, J. Phys. Chem., 82, 508 (1978). (10) A. J. Elliot, F. Wilkinson, and D. A. Armstrong, Int. J. Radiat. B i d . Relat. Stud. Phys., Chem. Med., 38, 1 (1980).
@ 1981 American Chemical Society
The Journal of Physical Chemistry, Vol. 85, No. 1, 1981 69
Oxidation of Amino-Containing Disulfides
conductometric detection, Asmus and c o - ~ o r k e r shave ~~-~~ studied the cations formed from dialkyl disulfides by one-electron oxidizing radicals such as OH, Br2--,Ag2+*, SO4-., etc. The cation yield was greater than 50% in all cases studied. In rapidrmix ESR measurements, where chemically generated OH radicals were reacted with disulfides, radicals which could not be detected in the pulse-radiolysis experiments, such as the sulfinyl radical (RSO) and those formed by hydrogen abstraction, were 0b~erved.l~ The literature cited above indicates that the disulfide bond is vulnerable to attack by one-electron oxidizing radicals such as OH and Br2--. Furthermore, although these two radicals are often used as “probes” in studying the radiation chemistry of proteins which contain disulfide l i n k a g e ~ , ~ J ~little J ~ - is l ~known about their reaction with disufide-containing amino acids. In this paper we report our findings on the reactions of OH and Br2-. with a number of disulfide-containing amino acids and related compounds. Experimental Section The organosulfur chemicals were purchased from the following suppliers: Calbiochem (dithiodipropionic acid), Schwarz and Mann (cystine), Sigma (cystaminePHC1, glutathione disulfide, homocystine), Aldrich (penicillamine disulfide, 5’-methylcysteine), Eastman Organic Chemicals (di-tert-butyl disulfide). All chemicals were used as supplied except di-tert-butyl disulfide, which was vacuum distilled before use. The solutions were prepared from triply distilled water, and then the pH was adjusted by adding perchloric acid or potassium hydroxide with the exception that for pHs near 7 and 9 a phosphate buffer (-5 X mol dm-3) and a borate buffer (-4 X mol dm-3) were used, respectively. The pulse-radiolysis experiments were performed by utilizing 2-5-ys pulses of electrons from a 1.5-MeV Van de Graaff generator as described earlier.1° A 1-cm path length optical cell was used with a fresh solution for each pulse. Before the sample was illuminated, the analyzing light from a 150-W xenon lamp was passed through a Pyrex filter to prevent photolysis of the disulfides. Dosimetry was based on a N20-saturated mol dm-3 potassium ferrocyanide solution ( G ~ 4 2 0 = 6.4 X Steady-state radiolyses were performed by using an AECL 6oCoGamma cell (dose rate -1530 rd/min). All solutions were bubbled with nitrous oxide to remove oxygen, so that on irradiation most of the solvated electrons were converted to OH (eq 1).la To form Br,-. poeaq- + N 2 0 + H20 OH + N2 + OH(1)
-
tassium bromide (-0.05 mol dm-3unless otherwise stated) was added to the N20-saturated solution (eq 2). For most OH 2Br- Br2-. OH(2)
+
+
kinetic experiments, the disulfide concentrations were 0.9 ~
~~~
(11) H. Mockel, M. BonifaEiE, and K. D. Asmus, J. Phys. Chem., 78, 282 (1974). (12) M. BonifaEiC, K. Schafer, H. Mockel, and K. D. Asmus, J . Phys. Chem., 79, 1496 (1975). (13) M. BonifaCie and G. D. Asmus, J. Phys. Chem., 80, 2426 (1976). (14) B. C. Gilbert, H. A. H. Laue, R. 0. C. Norman, and R. C. Sealy, J. Chem. SOC.,Perkin Trans. 2, 892 (1975). (15) G. E. Adams, R. H. Bisby, R. B. Cundall, J. L. Redpath, and R. L. Willson, Radiat. Res., 49, 290 (1972). (16) N. V. Klassen, J. W. Purdie, K. R. Lynn, and M. D’Iorio, Int. J. Radiat. Biol. Relat. Stud. Phys., Chem. Med., 26, 127 (1974). (17) G. E. Adams and J. L. Redpath, Int. J. Radiat. Bid. Relat. Stud. Phys., Chem. Med., 25, 129 (1974). (18) G. E. Adams and P. Wardman in “Free Radicals in Biology”, Vol. 111, W. A. Pryor, Ed., 1977, p 53.
X 10-4-6.0 X mol dm-3, which was sufficiently low to prevent any significant scavenging of the solvated electron in competition with reaction 1. However, for determining the rate of reaction of Br2-. with penicillamine disulfide at pH 6.7 and 4.2, the disulfide concentrations were 1.3 X and 3.2 X mol dm-3, respectively. In the latter solution, 10-15% of the solvated electrons would have been scavenged by the d i ~ u l f i d e . ~The , ~ concentration of tertbutyl alcohol was -1 mol dm-3 when this chemical was added to solutions to scavenge the hydroxyl radicals. In some experiments, sodium perchlorate was added to the solution to control the ionic strength. These are noted in the text. The bimolecular rate constant for reaction of B r p with the different disulfides was determined from the pseudofirst-order decay of the intense absorption (Am= = 360 n r n 9 of Br2-. (eq 3).
Br2-. + RSSR
ka
products
(3) Sulfhydryl analyses were carried out by the method of Ellmanlgt2Ousing 5,5’-dithiobis(2-nitrobenzoicacid). Trial runs using standard solutions of the thiols confirmed that the pH of the solutions and the bromide ion concentration did not interfere with the assay. A limited number of analyses for serine were performed for us on a Beckman 121 amino acid analyzer by Dr. K. J. Stevenson. Product yields are reported in terms of G values, which are yields in molecules per 100 eV of radiation energy absorbed.18 Results Dithiodipropionic Acid. The reaction of Big’ with dithiodipropionic acid formed a transient which had a broad spectrum with a, ,A = 440-460 nm (Figure la). The rate constant for this reaction, kB,did not vary significantly over the pH range studied (pH 6.6-11.0) as can be seen in Table I. Below pH 9.0, this transient decayed relatively slowly, but at pH 10.8 and 11.0 it decayed rapidly with a dose-independent half-life of 20 and 12 ps, respectively. The difference in the decay rates at pH 6.7 and 10.8 can be seen in the oscillograph traces in Figure lb. In pulse-radiolysis experiments where OH was reacted directly with dithiodipropionic acid at pH 4.0,7.0, and 10.8, weak end-of-pulsespectra were recorded, all of which were similar (Figure la). If the OH scavenger tert-butyl alcohol was present, no significant spectrum was recorded. While the spectra observed at pH 4.0 and 7.0 decayed fairly uniformly over the spectral range, for pH 10.8 the region X > 400 nm decayed faster with a dose-independent half-life of 28 ys. The spectral changes are shown in Figure l a for pH 10.8. Sulfhydryl assays were performed on y-irradiated aqueous solutions containing dithiodipropionic acid. These are listed in Table 11. Irrespective of the pH of the solution, the 3-mercaptopropionic acid yield was low when the reacting radical was Br2-. but much higher when OH was the reacting species. In the latter case, the thiol yield was 33-40% of G(0H). Cystamine. The transient that was observed following the reaction of Br2-. with cystamine at pH 9.0, 10.9, and 12.0 absorbed with a, A N 380 nm (Figure 2a). The rate constant, k3, for the reaction of Br2-. with cystamine increased with pH (Tables I and 111),but for a given pH it was not dependent on the bromide ion concentration or the ionic strength of the solution (Table 111). However, (19) G. L. Ellman, Arch. Biochem. Biophys., 82, 70 (1959). (20) E. Beutler, 0. Duron, and B. M. Kelly, J. Lab. Clin. Med., 61,882 (1961).
70
Elliot et al.
The Journal of Physical Chemistry, Vol. 85, No. 1, 1981
I 120 HOOC-CHZ-CH~-S-S-CH,- CHZ -COOH
100 “2
“2
CH2-CH2-S-S-CH2-CHZ D
a eo
z \
W 0
z a 60
5: m
WAVELENGTH (nm)
U
Flgure 1. (a) The transient spectra formed in an aqueous solution containing dithiodipropionic acid (1.0 X 10-4-10.0 X mol dm-3): by Br,-. at pH 7.0, 80 ys after a 700-rd pulse (0); by OH at pH 10.8, end-of-pulse (0)and 40 ps (A) after a 1080-rd pulse; at pH 7 with the OH scavenger red-butyl alcohol present, 20 ps after a 980-rd pulse (V). (b) The time profile of the absorption at X = 450 nm when Br,-. reacted with dithiodipropionic acid (4 X mol d d ) : (upper) at pH 6.7 following a 690-rd pulse; (lower) at pH 10.8 following a 1000-rd pulse. Sweep = 20 ps/division.
TABLE I: RSSR
+ Br,-.
k3 ----i’
’
*
40
20
Products
0:
~~
lO-’k,, mol-’ dm3 s-’
RSSR
pH-7
pH-9 pH-11
0
400
350
450
WAVELENGTH (nm)
PIC,,;
cystine 4 X lo8 mol-' dm3 s-l. As with the reaction of Br2-s, it probably reflects the Coulombic effect of the two carboxylic groups. The extremely low yield of 3-mercaptopropionic acid found (Table 11) after reactions 6 and 7 have occurred indicates that reaction 8 is not an important process. RSSR'.
+ OH-+
RSH
+ RSO
(8)
However, reaction 9 cannot be eliminated. In neutral RSSR+*+ OH-+ RS
+ RSOH
(9)
solutions, BonifaEiE et alaL2 suggested that the disulfide may decay by reactions 10 and 11, and the absence of RSH 2RSSR'. RSSR2++ RSSR (10) +
(22) G. E. Adams, J. E. Aldrich, R. H. Bisby, R. B. Cundall, J. L. Redpath, and R. L. Willson, Radzat. Res., 49, 278 (1972).
+
RSSR2+ 2H20
-
2RSOH + 2H+
(11)
formation at pH 6.5 (Table 11) is consistent with this suggestion. In contrast to dithiodipropionic acid, the absorption maxima are shifted to 380 nm for the transient species formed from cystamine, cystine, and penicillamine disulfide. No evidence for reaction with OH- above pH 10 was observed as they decayed by second-order kinetics even at pH 12 (Figure 5a). Thus they behave quite differently from the disulfide cations produced in this study from dithiodipropionicacid and di-tert-butyl disulfide, and from those investigated earlier.l1-l3 As one possible explanation, it should be noted that the commonest organosulfur radical detected by EPR spectroscopy in solid-state studies, which is characterized by a g tensor 2.058, 2.025, and 2.001,23-28 is known to absorb with a ,A, 370-380 nm. Symons and co-workersZ5produced this radical from penicillamine and found that the diffuse reflectance spectrum was dominated by a band at 375 nm. Falle et formed similar radicals from cystine and homocystine in sulfuric acid glasses, with A,, of 372 and 377 nm, respectively. The tert-butyl analogue of this radical in a 3MP glass has a A, at 380 nmn as can be seen by the spectrum shown in Figure 4a (solid squares). In early studies, this radical was mistakenly identified as the thiyl r a d i ~ a l ,but ~ ~ it> ~has ~ since been shown to contain two sulfur atoms by the use of 33Slabeling.29 The radical has now been identified as the perthiyl radical (RSS) by a number of laborat0ries~~,~~2*~' although Symons and ~ o - w o r k e r s have ~ ~ - ~postulated ~ the following structure R\ 1
,s-s
/R
R"
as a possible alternative. As noted by Symons, it is difficult to find the definitive experiment to distinguish between the two possibilities. However, the recent work by Giles and Roberts,32 where both the perthiyl radical and the alternative radical appeared to be produced, tends to argue against Symons' postulated structure, and hence the authors favor the perthiyl assignment. The close agreement between the spectrum of the tert-butylperthiyl radical and the analogous penicillamine species can be seen in Figure 4a. Production of the perthiyl radical requires C-S bond scission, and this is likely to involve an intermediate from the initial reaction with Br2--. Clear-cut evidence for a two-step process exists in the case of cystamine, where there was a discernible time lag between Brz--disappearance and 380-nm growth (Figure 2b) under certain experimental conditions. The nature of the intermediate involved here is at present a matter for speculation. (23) A. J. Elliot and F. C. Adam, Can. J. Chem., 52, 102 (1974) and references therein. (24) M. C. R. Symons, J. Chem. SOC.,Perkin Trans. 2,1618 (1974) and references therein. (25) D. J. Nelson, R. L. Petersen, and M. C. R. Symons, J. Chem. SOC., Perkin Trans. 2, 2005 (1977). (26) R. L. Petersen, D. J. Nelson, and M. C. R. Symons,J. Chem. SOC., Perkin Trans. 2, 225 (1978). (27) F. C. Adam and A. J. Elliot, Can. J . Chem., 55, 1546 (1977). (28) H. R. Falle, F. S. Dainton, and G. A. Salmon, J. Chem. SOC., Faraday Trans. 1, 2014 (1976). (29) J. H. Hadley and W. Gordy, Proc. Natl. Acad. Sci. U.S.A., 71, 3106 (1974); 72, 3486 (1975). (30) E. Sagstuen and C. Alexander, J. Chem. Phys., 68, 762 (1978). (31) J. E. Bennett and G . Brunton, J. Chem. Soc., Chem. Commun., 62 (1979). (32) J. R. M. Giles and B. P. Roberta, J. Chem. SOC.,Chem. Commun., 62 (1980).
The Journal of Physical Chemistry, Vol. 85,No. 1, 1981
74
Elliot et al.
However, it does not absorb appreciably in the range 330-700 nm. I t may be a cation-bromide ion pair or perhaps a bromine atom adduct, viz: Br2;
+ RSSR * Br- + RSSR(Br) or RSSR+Br-
(12)
The back-reaction cannot be important in either case, since the overall disappearance of B r p was uninfluenced by the bromide ion concentration (Table 111). Direct evidence for C-S cleavage in the case of cystine comes from an observation of a significant yield (-0.7 molecules/100 eV) of serine, which may, for example, be formed via reactions 13 and 14. The reaction sequences for cystamine, cystine, RSSR+Br- or RSSR(Br) OH-(H20)RSSR(0H) Br- (H+) (13) I I RSS + ROH (14) and penicillamine disulfide could be the same. Indeed the reason that no growth was observed at 380 nm for cystine and penicillamine disulfide after the Br2--absorption decayed may simply reflect the fact that the half-life of decay of the “colorless” intermediate into ita 380-nm absorption in these cases was far shorter than with cystamine. The possibility that the species absorbing at 380 nm was RSSR(0H) is in our view unlikely, since the hydroxyl group would be expected to have a pK within the pH range 4.2-12.0 of the present experiments, and this should have had profound effect on A-, Thiyl (RS) and sulfinyl (RSO) radicals formed in reactions 15 and 16 cannot acI RS + RSOH (15) I -,RSH + RSO (16) count for the 380-nm absorption for the following reasons. Thiyl radicals have been observed at pH 5-6 and have A, N 330 nm.6!33 Also their absorptions are too weak (ern, = 300-1200 M cm-1)596p33 to explain the size of the 380-nm peaks, though they may contribute to the absorbance below 340 nm. A significant contribution from RSO radicals from reaction 16 is also unlikely since BonifaEiS:et al.12 did not report an absorption maximum at 380 nm although the found a large yield of RSH from the overall reaction of H with dialkyl disulfides at high pH, viz: OH + RSSR RSO. + RSH (17) To sum up, the perthiyl radical seems to provide a very plausible explanation for the 380-nm transient produced by Brp. Further discussion of this species is given below. Mechanisms and Transients for OH. The reactions of hydroxyl radicals with dithiodipropionic acid, cystamine, cystine, penicillamine disulfide and homocystine form, in part, the same transient species as Br2-.. However, this does not imply that the reaction pathways are necessarily the same for both radicals. With dithiodipropionic acid the 450-nm absorbance, contributed by the species which decayed at pH 10.8 with a dose-independent half-life of 28 IS,indicated a cation yield for OH which was -20% of that produced by Brz--. This could be formed in two ways:l’-l3 by an electrontransfer reaction (eq 18) or by an addition reaction followed RSSR OH RSSR+* OH(18) by dissociation (eq 19 and 20).
+
+
+
-+
+
B
+
+
RSSR
+
-
R
HO’
(19)
‘S-”S-R 1
(33) M. Z. Hoffman and E. Hayon, J. Phys. Chern., 77, 990 (1973).
-
-
-00CCH2CHSSCH2CH2C00(22) radicals formed in reaction 15 would contribute to the long-lived absorption observed below 360 nm5,33in the pulse-radiolysis experiments (Figure la). Absorption of RSO in this spectral region cannot be eliminated either. For the disulfides which contain amino groups, above the pK of the NH3+ group, OH produced the same transient as Br2-., albeit in lower yield. The spectra observed were composite in the sense that, in addition to the species with ,A, at 380 nm, there was an extra absorption increasing in intensity at wavelength below 340 nm. This can be seen in Figure 2a for cystamine and, in particular, in the oscillographs in Figure 6c for homocystine. It is more prominent at lower pHs. The weaker 380-nm peaks from OH imply that the B r p mechanisms are more specific, whereas OH was capable of reacting in a number of ways. For cystine, the 380-nm peak with OH reacting at pH 7 was 40% of that from Br2--. However, the serine yield was reduced by an order of magnitude from the G value of 0.7 obtained with Br2-., showing that reaction 14 was relatively unimportant. This does not, however, eliminate RSS as the species with A, a t 380 nm, since it can be produced by alternative reactions. Above the pK of the NH3+group, OH can abstract hydrogen from the a ~ a r b o n : ~ ~ ~ ~ ~
Nr
NYz I N12 -~~c-~-cH~-s-s-cH~-cH-coo-
+ b~
-OOC-CH-CH~-S-S-CH~-CH-COO-
-
”2
+
H ~ O( 2 3 )
I1
The radical I1 would eliminate perthiyl radical
+
-+
bH
As noted earlier the low 3-mercaptopropionic acid yield from the alkaline solutions, where Br2; formed the cation, showed that reaction 8 did not occur. The high sulfhydryl yield obtained when OH was the reacting species (Table 11) indicates that reaction 17 must be viable in this system. It probably occurs through a sequence of reactions 19 and 16. Dithiodipropionic acid with its comparable thiol yield from OH at both pH 6.2 and 10.8 is different from the dialkyl disulfides where G(RSH) was negligible between pH 4.0 and 11.0 and increased both above and below these pHs.12 This was attributed to reaction 15 occurring predominantly in “neutral solutions” and reaction 16 occurring via acid or base catalysis in acidic and basic solutions.12 On that basis, with dithiodipropionic acid, one would have to conclude that reaction 16 is self-catalyzed by virtue of the two carboxylic groups. As the 3-mercaptopropionicacid yield was only 30-40% of the total 6 H yield (G(0H) = 5.6), the remaining OH radicals must be accounted for by reaction 15 and the hydrogen-abstraction reactions (eq 21 and 22). The thiyl -OOCCH2CH2SSCH2CH2C00-+ OH -00CCHCH2SSCHzCH2C00-(21)
I1
-
‘ “” -OOC-C=CH2
NT2
4- SS-CH2-CH-COO-
(24)
and the resulting olefin would be expected to form pyruvic acid. (34)M.SimiE, P.Neta, and E. Hayon, Int. J. Radiat. Phys. Chern., 3, 309 (1971). (36) N. Getoff and F. Schworer, Int. J. Radiat. Phys. Chern., 6,101 (1973).
The Journal of Physical Chemistry, Vol. 85, No. 7, 7887 75
Oxidation of Amino-Containing Disulfides
At pH 10.8 (cysteine) G(sulfhydry1)was 0.83 (Table 11), which suggests that reaction 17 is also important above the pK of the NH3+group and competes with the processes producing the 380-nm transient. Reaction of OH with the p CH2,which is adjacent to the disulfide, may also occur. Indeed below the pK of the NH3+group, OH will be directed away from the a carbon or the disulfide bond itself. This will to the 0 CH234935 reduce the yield of perthiyl radical from reactions 23 and 24, which agrees with the reduction in the 380-nm absorbance observed at lower pHs (see Figure 3). At pH 5, the cysteine yield was relatively small (0.19 molecule/100 eV, Table 11), implying that reaction 17 was unimportant for those conditions. This is consistent with another study,% in which the papain SH group was used as a trap for cysteine RSOH. The results implied that reaction 19 followed by reaction 15 gave an almost quantitative yield of RSOH near pH 6 for cysteine. The 380-nm peak for OH reacting with penicillamine above the pKs of the NH3+ groups was 80-90% of that formed by Brp. This must reflect a channelling of the OH radicals to reaction with the a carbon, because the p carbon no longer has C-H bonds available for hydrogen abstraction. The perthiyl radical would then be formed through reactions similar to reactions 23 and 24. Any hydrogen abstraction from the methyl groups attached to the carbons adjacent to the disulfide group could also lead to perthiyl formation: y 2
fH2
-
OOC-CH-C-S-S-R
1
CH3
-
OOC-CH
C ,-
4CH2 CH3
+
RSS
(25)
At pHs below the pK of the penicillamine disulfide NH3+ groups, where OH attack should be directed away from the a carbon, reaction 25 would be the mechanism for the formation of the perthiyl radical. If reactions 19 and 15 also occur in this system, the spectrum obtained from 6 H reaction at these pHs (Figure 4a) would largely be a composite of the perthiyl radical (A380 nm) and the thiyl radical6,A( 330, ern= 1200 mol-’ dm3cm-’). Production of these absorptions would account for the large trisulfide yield (G 1.5) (reaction 26) found for RSS RS RSSSR (26) N20-saturatedsolutions of penicillamine disulfide at pH ~ 5 Further . ~ evidence for reaction 15 occurring is that Purdie4 found that the G values of penicillamine sulfinic acid and penicillamine disulfide S-monoxide were 0.41 and 0.2, respectively, for NzO-saturated solutions. He concluded that both were formed from the sulfinic acid as shown in reactions 27 and 28. This would imply a min-
-
N
+ -
+ RSH RS(0)SR + RSH
2RSOH --.* RSOzH RSOH
+ RSSR
-
-
(27) (28)
imum G value of 1for RSOH from the irradiation and hence a minimum G of 1for RS. The “vinyl” compound formed in reaction 25 has a G value of 0.29: which is a minimum because this compound can undergo free-radical polymerization reactions. The OH-produced transients in Figures 2a and 3a both exhibit lower absorbance with decreasing pH and an increased ratio of the absorbance below 350 nm to that ~~
above. Thus it would appear that the processes occurring with cystamine are fairly similar to those discussed above for cystine. For homocystine the OH transient has greater absorbance at and above 400 nm (Figure 6a), as was the case for the Br2-. transient. Furthermore, it is clear from the different rates of decay at 330 and 390 nm in Figure 6c that more than one species was present, indicating that it was also a composite spectrum. Here abstraction of hydrogen from the /3 carbon and reaction 29 would provide
-
OOC-C
i”‘H-CH2-CH2-S -S -CH2-C* I YH2 YH2 +
~OOC-CH-CH~-CH~-S-S
”2
-.
CH2=CH
H -CH-COO
-CH-COO-
(29)
a pathway to perthiyl which should be less sensitive to pH than reactions 23 and 24. As well as being in agreement with the work in ref 23-32 reviewed above, the assignment of the 380-nm transient to perthiyl radical is supported by Purdie’s observations of trisulfide yields in disulfide rad io lyse^.^ Penicillamine disulfide, with the weakest C-S bond, has the largest trisulfide yield4and $so gave the largest 380-nm transient from reaction with OH in this study. One further alternative is that the 380-nm absorption was due to a disulfide cation species with the lone pairs of the amino groups coordinated into the positive disulfide center. This model can explain many of the features of the 380-nm absorptions. For example, as with the dialkyl disulfides, lower yields were observed with OH than Br2-.. Secondly the electronic energy levels and absorption maxima would be sensitive to the orientation of the electron lone pairs with respect to the S-S+. center. Thus the shift in ,A, from 380 to 400 nm for homocystine could be a natural consequence of the extra CHz in the aliphatic chain. The major difficulty is that the coordination must be so strong that reaction with OH- at pH 1 2 was not possible. While we strongly favor the perthiyl assignment, we do not categorically exclude the coordinated cation at this time. Indeed, even if perthiyl is the “380-nm transient” in all of the systems studied here, the cation may still be an undetected intermediate in perthiyl formation. In fact a reason for the decrease in k3 when both amino groups are protonated may be the difficulty of developing three positive centers on the same molecule. Oxidized Glutathione. The results in Figure 7 demonstrate that, although some 380-nm absorption may be formed by Br2-., it is relatively weak. The greatest intensity occurs below 350 nm, where thiyl radicals and species such as RSOH may be expected to absorb. The weak absorption caused by 6 H may have a significant component due to aliphatic and peptidyl radical^.^' Thus, based purely on the results for this peptide, it would appear that disulfide bonds of proteins are unlikely to produce significant absorption above 300 nm on reactions with OH or Br2-.. This supports the continued use of Br2-. as a diagnostic probe for other amino acid residues. Acknowledgment. We are indebted to Dr. K. J. Stevenson for the serine analyses and to Drs. F. C. Adam, T. S. Sorenson, D. F. Tavares, and T. Back of this department for useful discussions. We are also indebted to Professor K. D. Asmus for his helpful comments on the manuscript. The continued technical support from D. Malinsky is appreciated. Financial assistance from N.S.E.R.C.C. (Grant No. A3571) and the University of Calgary is gratefully acknowledged.
~~
(36) M. Lal, W. S. Lin, G. M. Gaucher, and D. A. Armstrong, Znt. J. Radiat. Bid. Relat. Stud. Phys., Chem. Med., 28, 549 (1975).
(37) M. SimiE, P. Neta, and E. Hayon, J. Am. Chem. Soc., 92,4763 (1970).
