J . Phys. Chem. 1993,97,3783-3190
3783
Peptide Peroxyl Radicals: Base-Induced 02*Elimination versus Bimolecular Decay. A Pulse Radiolysis and Product Study Oliver J. Mieden, Man Nien Schuchmann, and Clemens von Sonntag' Max- Planck-institut fur Strahlenchemie, Stiftstrasse 34-36, 0-4541 3 Miilheim a. d. Ruhr, Germany Received: December 7, I992
Radiolytically generated O H radicals react with the cyclic dipeptides glycine anhydride (1) and alanine anhydride (2), forming a single type of peptide radical in each case by abstracting a carbon-bound H atom at the ring. In the case of sarcosine anhydride (3), besides the C(3) or C(6) H atoms (78%), the H atoms at the N-methyl groups are also targets of the O H radical attack (22%). In N2O/O2 (4:l v/v) saturated solutions these peptide radicals add oxygen (k= 2 X lo9dm3 mol-) s-I) to form the corresponding peroxyl radicals 6 (from l ) , 7 (from 2), and 12 and 13 (from 3). The kinetics of 0 2 ' - elimination from the radicals 6 and 7 has been monitored by pulse radiolysis techniques. The pKa values of the peroxyl radicals 6 and 7 have been determined to be 10.8 and 11-2, respectively. The anions of these peroxyl radicals (6a and 7a) rapidly eliminate 02'- with the rate constants 1.6 X lo5and 3.7 X lo6 SKI, respectively. In contrast, the spontaneous HO2' elimination reactions of the peroxyl radicals 6 and 7 are very slow, with rate constants of 10, all of the peroxyl radicals 6 decay into 02'(G(NF-) = 5.8 X 10-7 mol J-I). Alternatively, one can also make use of the change in conductance since the formation of H+/Oz*-from the uncharged
3786 The Journal of Physical Chemistry, Vol. 97, No. I S . 19193
4
Mieden et al.
I
1
1
2
3
4
- lo3 [OH-] I mol dm-'
2
-10' [OH-]
Figure 3. [OH-] dependence of the observed rate constant of 02.formation (A,monitored at 350 nm in the presence of 6 X mol dm-3 TNM; 0 , determined by conductance decrease) of N2O/O2 (4:1, v/v) saturated glycine anhydride solutions (IO-' mol dm-') irradiated by an electron pulse of -5 Gy. Inset: plot of kobs-l versus [OH-]-'. radical 6 should lead to the consumption of OH- and thus to a reduction in conductance of the glycine anhydride solution as shown in Figure 2 (bottom) at pH 9.8. The same observed first-order rate constant is obtained for the formation of 02'-from these two methods under the same pH conditions. This rate constant, kobsrincreases with increasing pH (Figure 3). It approaches a plateau value of 1.6 X 105 s-I in strongly alkaline solution. Equation 16 represents the kinetics of the 02*elimination according to the mechanism outlined by reactions 8-1 1. Here k-8 denotes the pseudo-first-orderrate constant of the reaction of the deprotonated peroxyl radical 6a (or 7a, see below) with water. 1
k-s+k,,
kobs
k&ii
I -
1
+-
1
[OH-] kii
6
4
(16)
According to eq 16, kobs approaches kl I at very high [OH-]; thus, it follows that kl I = 1.6 X lo5 s-I. The same value results from the intercept when ko,-l is plotted versus [OH-]-' (inset in Figure 3). As to be expected, the value of kl I coincides with kob obtained by direct monitoring of the absorption buildup of 02.at pH 11.7 (Figure 1). The slope of the linear plot in the inset of Figure 3 (3.0 X s mol dm-3) is equal to the term (k-8 kll)/k8kllaccording to eq 16. Hence, the value of k-8 can be calculated from the slope, if for k8 a value of 10'0 dm3 mol-' s-l is assumed, which is a typical rate constant for the deprotonation of an N H group by OH-.1°+22With this assumption we calculate k-8 = 4.6 X 106 S-1. Although these two methods of monitoring 02*formation give identical kinetic data (cf. Figures 2 and 3), the yield of acid formation calculated from the conductance change at pH > 10.5 exceeds that obtained from optical measurement of NF- buildup (G(H+) = 9.0 X mol J-I at pH 11-6). It is thereforesuggested that in this pH range the product 8 is partly deprotonated (pK, < 11.6). However, we have not observed any changes in absorbance in the accessible UV region associated with this dissociation. In the alanine anhydride system results similar to those for the glycine anhydridesystem have been obtainedfor the OH--induced elimination of 0 2 ' - from the peroxyl radical 7. However, the rate constant of this reaction for radical 7 is considerably higher than that of radical 6 (Figure 4). At pH > 10.6 the 02'elimination reaction (reaction 11) becomes so fast that the rate of oxygen addition to radical 5 (cf. reaction 6, shown as a dashed line in Figure 4) becomes the limitingstep. Therefore, no plateau value could be measured in this case. Nevertheless, from the intercept of the linear plot of k0bs-l versus [OH]-' (inset of Figure
+
/ m o l dm-3
8
Figure 4. [OH-] dependence of the observed rate constant of 0 2 ' formation (A,monitored at 350 nm in the presence of 104 mol dm-3 TNM; 0 , determined by conductance decrease) of N20/02 (4:1, v/v) saturated alanine anhydride solutions (lo-' mol dm3)irradiated by an electron pulse of - 5 Gy. Inset: plot of kOh-l versus [OH-]-'. Dashed line: the limit set by the rate of oxygen addition. Dotted line: an extrapolation according to eq 16.
TABLE II: pK. Values of the Cyclic Dipeptide Peroxyl Radicals
PK, direct measmta 10.8i 0.4
calcd from 02.eliminatn kinetics 6 10.8 10.7 f 0.2 7 11.1 11.2 f 0.3 a From pulse conductivity measurements (see Figure 5).
peroxyl radical
estimd from Taft q
4) one obtains kll = 3.7 X lo6 s-I for the 02'- elimination from the peroxyl radical anion 7a. From the slope of this plot [(k-s + kll)/kSklI = 4.9 X s mol dm"] and making the same assumption for k8 = 1olodm3 mol-' s-I as above, one obtains k-8 = 1.4 X 1O7 s-I . Applying these values in eq 16, one coudl calculate the values of kob at the higher OH- concentrations (dotted line in Figure 4). Here in thealanineanhydridesystem the yield of02'-formation monitored by the optical method agrees with that obtained by the conductance detection method; Le., at pH > 9, G(NF-) = G(H+) = 5.7 X mol J-I. The product 9 from reaction 11 may have a pK, value considerably above 11, and thus no secondary deprotonationwas obsereved. It is noted that by product analysis only 15, the hydrated form of 9, is found (see Table 111). 1.3. pKa Values of the Peptide Peroxyl Radicals. With the value of k8 assumed to be 1Olodm3 mol-] s-1 and k-8 = 4.6 X 106 S-I obtained from the kinetic plot in Figure 3, the pK, value of radical 6 can be calculated. Hence, pKa(6) = pK, - pKb(6) = 14 log(k_8/ks) = 10.7 (Table 11). Similarly, for the alanine anhydride system, pKa(7) = 11.2 is also calculated from the corresponding values of ks and k-8. The pKavalue of radical 6 has also been obtainedindependently by monitoring the conductance change in the glycine anhydride solution caused by the deprotonation of radicals 6 (reaction 8). The released proton is immediately neutralized by the OH- present and hence leads to a decrease in conductance. In the inset of figure 5 a conductance "jump" within 1 ps after the pulse is shown (the slower conductancedecrease which follows this jump represents the formation of H+/02'- discussed above, cf. Figure 2 (bottom). The yieldof thisabrupt conductancechangeincreases with higher pH as more of the radicals 6 become deprotonated. In Figure 5 the corresponding G(H+) calculated from such a conductance jump is plotted against pH. From the inflection point pK, = 10.8 is obtained, which agrees well with the value obtained above. The pK, values for the radicals 6 and 7 obtained in this work compare favorably with values predicted by the Taft equation for
+
The Journal of Physical Chemistry, Vol. 97, No. 15, 1993 3187
Peptide Peroxyl Radicals
TABLE IIk Products and Their C Values lo the y Radio1 sis of Cyclic Dipeptides (1O-jmol lo N20/02 (41 Saturated Aqueous Solution at pH -6 and a Dose Rate of 0.14 Cy s-1
vi)
-J
5 4.0
G/
E
?
e.
+ -
E.
2.0
(3
1 10
11
-PH-
Figure 5. pH dependenceof G(H+)obtainedfrom the abrupt conductance
jump (within 1 ps after the pulse, cf. inset) following a 0.4-ps electron pulse of about 6 Gy in NzO/O2 (4:1, v/v) saturated glycine anhydride solutions (( 1-2) X 10-3 mol dm-3). Inset: conductivity decrease of a N20/02 (4:1, v/v) saturated glycine anhydride solution (2 X lo-' mol m-3) at pH 10.8 irradiated with a 0.4-ps electron pulse of 6.5 Gy. amides (pKa for RlCONHR2 = 22.0 - 3 . 1 E ~ * ) . For ~ ~ the prediction of these two pKa values we have made the assumption that substituent effects operate along all chains in a molecule. Hence, radical 6 approximates to an open-chain amide with a correctionof 0.2 pKunit added to the predicted pKvalue for ring closure around the nitrogen.23 Thus, the two substituents on the nitrogen in radical 6 are -COCH3 (a* = 1.81)23 and -CH(02)CONH2 (u* = u* for -CH2CONH2 u* for -CH202* +u* f 0 r - C H ~ = 0 . 3 1 ~ 1S524+O= ~+ 1.86. Hencqthepredicted pKa(6) = 22.0-3.1(1.81 1.86) +0.2 = 10.8. Inradical7the two substituents are -COCH2CH3 (assumed to have the same u* value as -COCH3) and -C(CH3)(02')CONH~(u* = u* for -CH(02')CONH2 u* for -CH2CH3 = 1.86 - 0.1 = 1.76). Hence, the predicted pKa(7) = 22.0 - 3.1(1.81 + 1.76) 0.2 = 11.1. A radical in an a position to an OH or N H group makes the proton at the heteroatom more acidic by about 4 pK units (cf. ref 5) compared to the parent compound. A similar but not as drastic shift results from the replacement of a H atom in an a position to the OH or N H group by a heteroatom. This situation is found in the peroxyl radical 6 where the dioxygen bound to C(3) exerts a strong electron-withdrawingeffect (the Taft u* constantof -CH200' is higher than those of -CHzF or -CH2CN). We therefore would expect that radical 4 (pKa(4) = 9.8)" is more acidic than the peroxyl radical 6 [pKa(6) = 10.81. Our data are in agreement with this expectation, but not with the low value of pKa(6) = 7.5 reported earlier.25 1.4. Bimolecular Decay Rate Constants of the Peroxyl Radicals. The overall rate constant for the bimolecular decay of the glycine anhydride peroxyl radical 6, 2k = 8.6 X lo8 dm3 mol-' s-I,has been determined by monitoring the absorption decay of 6 at 300 nm at pH 5 by pulse radiolysis of a N20/02-saturated solution of glycine anhydride. The corresponding value for the alanine anhydride peroxyl radical 7, 2k = 1.6 X lo8 dm3 mol-l s-I, obtained at pH 4.3, is significantly lower. This is expected, since radical 7 is a tertiary peroxyl radical.8 The main peroxyl radical in the sarcosine anhydride system, radical 12, lacks a dissociable N H group in the CY position. It cannot undergo the HO2*/02'- elimination reaction as in the case of radicals 6 and 7. It can only decay bimolecularly by self-termination or by cross-termination with the other peroxyl radical 15 in the system. The overall rate constant monitored at 300 nm at pH 5.8 has the value 2k = 4.0 X IOs dm3 mol-' s-I. 2. Steady-Statey Radiolysis. 2.1. Glycine Anhydride. The products in the y radiolysis of 1 in N20/0~-saturatedsolution, 2,5-dioxo-2,3,4,5-tetrahydropyrazine(8), 3-hydroxy-2,S-dioxopiperazine (14), N-glyoxylylglycinamide (16), and 2,3,5-trioxopiperazine (17), have been determined as trimethylsilylated
+
+
+
+
substrate product (lo-' mol J-1) 0.4 glycine 2,5-dioxo-2,3,4,5-tetrahydroanhydride (1) pyrazine (8)" 3-hydroxy-2,5-dioxopiperazine(14)' 3.5 N-glyoxylylglycinamide (16)' 0.1 2,3,5-trioxopiperazine(17)' 1.6 hydrogen peroxide 1.7 organic hydroperoxide (e.g. IS) 0.6 02.2.4 a1anin e 3-hydroxy-2,5-dioxo-3, 4.7 anhydride (2) 6-dimethylpiperazxine(15)" formaldehyde 0.2