Document MJYRqb5dnQoQE7g4Dr7Jmd187

Bt 18ft NOTICE THIS ____ _ BY COPYRIGHT I A** FREE RADIC Free Radical Mechanism for Solvent Toxicity ZOLTON T. WIRTSCHAFTER, MD AND MARSHALL W. CRONYN, PhD PORTLAND, ORE It is appare ture on solvei theory of act would cover and nonhalogi quite apparen reported that 1 tion of a biok or their halogc as the product pounds are su attack in an ac Free Radicals in Solvent Toxicity A previous study (Wirtschafter and Cronyn, 1963) reported the relative toxicity of molar equivalent doses of several halogenated and nonhalogenated hydrocar bons. Previous investigators have studied the relative toxicity within homologous series, but there have been no quantitative com parisons reported for different types of solvents, von Oettingen et al (1949) made a thorough study of the relative activity of the halogenated methanes; Plaa et al (1958) ex tended their comparisons from halogenated methanes to ethanes and halogenated ethylenes; and Gerarde (1956, 1959) in vestigated the relation between structure and toxicity in a large number of alkylbenzenes. The results of our studies and those previ ously reported are consistent with the hypothesis that the primary step in the bio logical attack on all such compounds is a free radical process. The toxicity of these solvents is best explained by their reaction with the free radicals which are normal intermediates in many of the most important biological processes. A free radical bond-breaking process is one in which a pair of electrons bonding two atoms is broken in such a way as to divide the pair evenly between the two atoms, rather than leaving both electrons with one atom or the other, ie: Submitted for publication Oct 11, 1963. Research Laboratory of the Veterans Administra tion Hospital, the University of Oregon Medical School, and the Department of Chemistry, Reed College. A : a---or RO-+A c B-->R-0 :A-f-Bor for a double bond A: :B-- A: B- RO-4-A: :B---- >RO :A :B- Thus, the radical initiated polymerization oi ethylene proceeds as follows: HH HH H H HHHH RO-C::C- RO.-Cid+iJc- RO:C:C:C:C--> etc Mkit It II HH u and when a molecule such as CC14, which is susceptible to radical attack, is added, it may be attacked by the radical end of the growing chain (Walling, 1957); a ci Cl :C Ul+R1-----R> :C1+C1 :C- Cl Cl trichloromethyl radical In these examples, RO may be a ready source of free radicals such as from the breaking of the very labile peroxide bond: R-O-.O-R--*2R-0heat or light Early studies by Michaelis (1938, 1951) on biological oxidation-reduction reactions very strongly implicated free radicals as in termediates; more recently, the action of ionizing, ultraviolet, and visible radiations on biological systems has also produced evidence of the free radical nature of the reactions produced by these energy sources (Weiss, 1946). Finally, Commoner et al (1954) have succeeded in measuring by electron spin resonance technique the actual concentrations of free radicals in a variety of biological tis sues, and Temberg and Commoner (1963) have applied these techniques to the differ ential diagnosis of jaundice. Vol 9, Aug, 1964 Bioch Sc Benzene, r dichlorobenzer whether metab Williams, 195 Parke, 1956) aqueous solutic treated with a of free radic; and Weiss ( (1953 21-~), , (1956), Baerr (1941) pointe their radiation tained when th jected to biolog Within 48 he had been expire and 11/2% as C period, 35% oi form of urinar types: d>0H f h--<3-h d OH r HO-^p-OH 3 CTLC-N-1 CO __ i y-S-CH<-C-C d O HH ooc-c=c!-c!=c- k It Wirtschafter--Cr, TSCHAFTER, MD :RONYN, PhD >A-+B>R-0 :A+B A:B>R0 :A :Bymerization of HHHH IIII ):C:C:C:C--- etc iiliU CCU, which is added, it may )f the growing Cl ci romethyl radical ^ be a ready as from the oxide bond: (1938, 1951) :ion reactions radicals as in die action of : radiations on luced evidence the reactions irees (Weiss, 1 (1954) have electron spin toncentrations biological tisloner (1963) to the differ- ol 9, Aug, 1964 FREE RADICALS IN SOLVENT TOXICITY 187 It is apparent from a review of the litera The phenolic products indicate quite clear ture on solvent toxicity that no satisfactory ly that* the benzene ring is being attacked by theory of action has been suggested which H-O or H-O-O radicals in the same man would cover equally well both halogenated ner that has been observed when benzene or and nonhalogenated toxicity. However, it is benzene derivatives are treated with Fenton's, quite apparent from all of the results so far reagent (Fe++ and H202): reported that the products obtained in the ac tion of a biological system on hydrocarbons or their halogenated derivatives are the same as the products obtained when the same com pounds are subjected directly to free radical attack in an aqueous medium. Biochemical Breakdown of Solvent Molecules Benzene, nitrobenzene, and the three dichlorobenzenes yield phenolic products whether metabolized in the rabbit ( Parke and Williams, 195321-22; Azouz et al, 1955; Parke, 1956), subjected to radiation in aqueous solution (Stein and Weiss, 1949), or treated with a metal ion plus peroxide source of free radicals (Konecny, 1954). Stein and Weiss (1949), Parke and Williams (195321'22), Azouz et al (1955), Parke (1956), Baemstein (1945), and Bernhard (1941) pointed out the similarity between their radiation products and the phenols ob tained when the same compounds were sub jected to biological metabolism. Within 48 hours, 45% of the total benzene had been expired, 43% as unchanged benzene and lVi% as CO*. At the end of the 48-hour period, 35% of the benzene appeared in the CfH.+H,0,+Fe**--*GH,,0H+H,0 or Cu* ion and HjOi (Konecny, 1954). Similarly, when benzene is irradiated in aqueous solution in the presence of oxygen, phenols are produced by the attack of radicals generated by the action of the radiation on the water (Bacq and Alexander, 1955). HiO-f- O' --OH+H-O-OArH+OH --*Ar- +H^) or ArH-P-O-O-H--Ar- Ar-POH --ArOH (Stein.and Weiss, 1949) The finding by Gerarde (1956) that alkylbenzenes are much less myelotoxic than benzene provides another indication that the metabolism of the hydrocarbons is initiated by a free radical attack. The radical attack on alkylated benzene generally proceeds at the methylene next to the aromatic ring rather than on the ring itself. The resulting radicals have a much greater stability than those which would result from the removal of a ring hydrogen. ArCHs-p-OH--> Ar-CHr+H.O Harper and Calcutt (1961) have found that the hydroxylation of polynuclear hydro carbons is promoted by a microsomal fraction of liver homogenate and is inhibited by ribo flavin. In Gerarde's work on the alkylben form of urinary metabolites of the following zenes (1956), the primary products were types: found to be hydroxy derivatives with the <>OH phenol, 23% H-O-OOH (fihydroquinone, 4.8* >-OH resorcinol, 22% major substitution occurring on the C-H adjacent to the aromatic ring. In barbiturates, also, the reaction products indicate side chain hydroxylation followed by oxidation to HO-C3-OH 3-hydroxy-l,4-<iihydroquinone. OJ* carboxyl adds (Brodie, 1956; Maynert and Losin, 1955). Butler (1949) has found that O CfiUC-N-H <* j>-S-CHr-C-COOH L-phenylmercapturic acid, 0.5% d HH HOOC-C=c!-C=C-CC)OH (ra*j-/rmi-muconic acid, 1-3% bb the metabolism of trichloroethylene, aa \/ C ::C\ ah in dogs, gives chloral, which is further metab olized to trichloroethanol and trichloroacetic acid. IVirtschafter--Cronyn 188 ARCHIVES OF ENVIRONMENTAL HEALTH aa \/ C::C r\ ah < ?a Cl-C-C. I a CCUCOOH C1-C-CH.OH I a The conversion of trichloroethylene to chloral finds a parallel in the free radical mediated radiation conversion of tetrachloro- ethylene into trichloroacetyl in the presence of oxygen (Walling, 1957). a C( .Cl catalytic amount I JD ci-c-q a a +o. -- i Cl a The fate of C14CU absorbed by inhalation in monkeys has been studied by McCollister et al (1951). They have found that 40% of the C14Cl4 is exhaled in 1,800 hours as C14C14 and 11% as C14C>2. The remaining C14 was found as carbonate and nonvolatile bound C14 in urine and feces. In the case of CCU, unlike benzene and trichloroethylene, there are two processes by which the mole cule might be degraded to CO2, the free radical path, such as observed by Minder (1948), who measured the HQ generated when CQ4 was exposed to x-rays in aque ous solution and the nucleophilic decomposi tion which has been studied by W. F. von Oettingen et al '(1949). However, the rela tive toxicities of the various halogenated methanes do not seem to relate to the large difference in nucleophilic reactivity between chloroform and carbon tetrachloride. Recently, Butler (1961) has determined that CCU is converted into CHCI3 both in vivo and in vitro by various tissues and tis sue constituents and has proposed that in certain halogenated hydrocarbons such as CCI4 and CHCI3, toxicity is the result of a free radical cleavage and the subsequent chemical fate of the radical products. Butler limits his suggestion to halogenated hydro carbons and even more strictly to those in which the C-Q dipole is reduced to lower the activation energy for a homolytic rather than heterolytic cleavage of the carbon-halogen bond. We believe that the notion of a homolytic process as Butler has suggested is an excel lent pne, since it not only explains his ob served reduction, but it provides a clue to the behavior of other solvents in biological sys tems. Furthermore, Commoner et al (1954) demonstrated the presence of free radicals in biological material at a level of about 10~s mol/gm. Because Butler (1961) correlated the potential free radical activity of CCU and CHCI3 with their C-Cl dipole moment and the activation energy for C-C1 bond splitting as a unimolecular process, he did not extend his proposal to simple hydrocar bons and less highly substitute halogenated derivatives. However, we would like to suggest that it may be more profitable to analyze the rela tive propensity for free radical formation using another criterion, namely the relative rate of reaction of various hydrocarbons and their halogenated derivatives directly with free radicals already present in the medium. In other words, to consider this as at least a bimolecular (or higher) process at the bond-breaking step; ie, X -f-A :B-- X.A+Brather than A:B--A-)-BThe best data for this type of compari.-on are to be found in the transfer constants which measure the relative rate of attack of a growing free radical catalyzed polymer chain on solvent molecules compared to their rate of attack on more of the monomer units (Walling, 1957). The advantage in using this type of sys tem as a measure of relative energy for free radical formation is apparent if we examine the behavior of CHQ3. If we assume that under a unimolecular process the weakest bond will be broken, then the C-Q bond should cleave since the bond energies in CHClj are C-C1, 72 k-cal/mol and C-H, approximately 90 k-cal/mol (Cottrell, 1958). However, when CHGs actually reacts by a free radical process, such as with poly merizing 1-octene, the ratio of C-H to C-Cl bond cleavage is about 50 or 60 to 1 (Wal ling, 1957). Thus, in spite of the much greater bond strength for the C-H bond, it is preferentially cleaved by a free radical. The reason for this is that in a reaction with Vol 9, Aug, 1964 3 FREE RADICALS another radical, th( coupled with a be the activation enei i point in this proce i the isolated bondenergy of partial b favorable- interact ' tronic configuratic i and the new radio a ! X+H:C C These character would account for CH2CI2 found by experiments and ; vivo. It would alsc of CC14 which is to CO2 (McCol though the first st pears to be a red lowed through t CH2CI2, CH3CI s of CO2. That the relativ different bonds, s vary with the sou demonstrated in stants of CCU anc tion of several all An examinatioi for CCU, benzen clearly that CC14 radical reactor ai ference between hydrocarbons. The implicatioi proposal and the Table 1.--Transfer to Radical Source Styrene (60 C) Vinyl acetate (60 C) Ethylene (70 C) Wirtschafter--Cron <L HEALTH i*-*' his.ob. a to the logical sys:tal (1954) ree radicals about10~8 i correlated ty of CC14 ile moment C-Cl bond ess, he did : hydrocarlaiogenated gest that it :e the rela- formation he relative arbons and ectly with e medium, s at least a 5S at the < * l 1 omnarison stants attack of l polymer :d to their imer units ie of sysy for free : examine iume that : weakest -Cl bond ergies in nd C-H, 11, 1958). reacts by ith poly: to C-Cl 1 (Walle much bond, it radical, ion with iug, 1964 I 1 FREE RADICALS IN SOLVENT TOXICITY 189 another radical, the bond-breaking process is coupled with a bond-forming process, and the activation energy required for the mid point in this process is decreased relative to the isolated bond-breaking process both by energy of partial bond formation and by any favorable- interactions between other elec tronic configurations of the radical source and the new radical being generated. aa X +H :C:C1--X :H+C :C1 a, a 'a X:CH-C:C1 H These characteristics of radical reactions would account for the very small amount of CH2CI2 found by Butler (1961) for in vitro experiments and account for none at all in vivo. It would also explain the oxidative fate of CCI4 which is converted, in part at least, to CO2 (McCollister et al, 1951) even though the first step observed by Butler ap pears to be a reductive process, and if fol lowed through the series, CCU, CHCI3, CH2CI2, CH3CI should lead to CH4 instead of C02. That the relative ease of bond breaking for different bonds, such as C-H and C-Cl, will vary with the source of the radical is clearly demonstrated in the relative transfer con stants of CCI4 and CHCI3 in the polymeriza tion of several alkenes (Walling, 1957). An examination of the transfer constants for CCI4, benzene, and cyclohexane shows clearly that CCU is by far the most effective radical reactor and that there is little dif ference between benzene and the saturated hydrocarbons. The implications of these results for our proposal and the comparison of the biologi- Table 1.--Transfer Constants for CCU and CHCU Radical Source Styrene (60 C) Vinyl acetate (60 C) Ethylene (70 C) Transfer Constants CCU CHCl. 90X10-* 1.0 0.7 0.5X10"4 0.016 0.8 lVirtschafter--Cronyn Table 2.--Transfer Constants Styrene With Typical Solvent * Solvent Cyclohexane Benzene Methylene chloride Chloroform Ethylene dichloride Tetrachloroethane Carbon tetrachloride n-Butyl mercaptan 60 C 0.0000024 0.0000018 0.000015 0.000050 0.0090 22 100 C 0.000016 0.000018 0.00066 0.0018 0.0180 Ethylene t 70 C Methyl chloride Methylene chloride Chloroform Carbon tetrachloride Ethyl chloride Ethylidene dichloride Methyl chloroform CHiCl 0.0004 CHsCls 0.07 CHCli 0.8 CC1. : 0.7 CHaCHiCl 0.012 CSiCH Cl-i 0.15 CHiCClt 0.05 Walling, reference 28, p 132. t Walling, reference 28, p 237. cal activity of various solvents is apparent. Since the relative rate of attack within a homologous series of hydrocarbons or halogenated hydrocarbons for a radical process will depend on the nature of the attacking radical, we may not expect to find a very clear parallel between any simple chemically determined relative order of activity and a biological one. Not only is there no single chemically determinable reference order of activity, but in the biological system, the relative order will vary with the many pos sible radical sources within the living sys tem. The combination of differences in activity as a function of the radical gen erating entity and the differences in tissue distribution (von Oettingen et al, 1949; Gerarde, 1956; Minder et al, 1948), due in part to differences in distribution coefficients and vapor pressure (von Oettingen et al, 1949; Gerarde, 1956), is no doubt responsi ble for the observation that there is no cor relation between the dose causing liver damage and that causing death. Thus, a com parison of the hepatotoxic effects at ED5o (the dose producing the desired effect in 50% of the cases) shows that CCU is more toxic than CHCI3. The ED30 values in millimols per kilogram were determined by Plaa et al 190 ARCHIVES OF ENVIRONMENTAL HEALTH (1958) to be 1.4 for CHC13 and 0.45 for CCU, whereas for the lethal dose the re verse is true: LDjo, 5.9 for CHC13 versus 200 mM/kg for CC14. This is also the order of toxicity observed by von Oettingen et al (1949) for the lethal effect when both com pounds were administered at the same con centration. Where the structures being compared con stitute a homologous series or a group of compounds of very similar structure, but without a wide range of solubility character istics, the order of reactivity correlates more often than not with the order to be expected for a free radical process in which there is a radical attack upon the hydrocarbon or its derivative. For example, Plaa et al (1958), using a sensitive test for hepatotoxicity, found the following ED50 values (mM/ kg): CClt (carbon tetrachloride)-0.45; CHCI3 (chloroform)-1.40; CH2C1CHC12 (l,l,2-trichloroethane)-2.1; CHCI2CHCI2 (jyw-tetrachloroethane) -7.2; CHCI=CCl2 (trichloroethylene)-11; CC12=C02 (tetrachloroethylene)-27; CH3CCI3 (1,1,1-trichloroethane) -84. If the unsaturated compounds are considered separately, since they would be expected to react by radical additions, the others in this series correspond quite well to the expected reactivity toward a free radical source. Between the two unsaturated com pounds, the order is also what would be ex pected on this basis (Walling, 1957). Gerarde (1956) rated alkylbenzenes ac cording to their lethal effect for rats. There is a similarity between the transfer constant pattern and the lethal ratio (Table 3). In the unbranched side chain series, the ef fect of solubility is apparent since the tox- Table 3.--Transfer Constants and Lethal Ratios for Alkylbenzenes Benxeba Tolnene Ethylbenzene Isopropylbenzene Transfer Constant Styrene (60 C) 0.24X10-* 1.25X10"* 6.70X10"* 8.20X10"* Lethal Ratio 0/10 3/10 7/10 /10 icity rises to a maximum and then drops off to 0/10 for w-decylbenzene. It is not surprising that the new anesthetic, Cl halothane, H-C-CF,, has caused liver damI Br age since it has one hydrogen which should be readily removed under free radical at tack. It should also be expected that com pounds other than hydrocarbons and their halogenated derivatives may produce similar toxic effects if they are highly susceptible to attack by free radicals and are not metabo lized in some other way. For example, mercaptans have one of the highest transfer constants (Walling, 1957), being about 10* times as effective as carbon tetrachloride in styrene polymerization, so that in addition to any possible disruption of physiological processes via -- SH exchange, complexing with metal ions, etc, radical formation may give rise to further toxic effects. It should also be expected that compounds which are presumed to be useful for their antiradia tion effects may be expected to show disrup tion of normal processes as a consequence of their particular reactions with radicals. Aminoethylthiourea was studied by Doh erty and Burnett (1955) and found to have considerable protective effect, showing a sur vival rate of 70% at day 28 in mice sub jected to a lethal dose of x-irradiation. Cheymol et al (1959) have tested amino ethylthiourea as to its effect on dehydro genase activity of the liver of the mouse subjected to x-irradiation. They determined that y-radiations (900 r) not only do not in hibit succinodehydrogenase activity of the liver of the mouse, but on the contrary, they appear to increase it. They further demon strated that aminoethylthiourea exercises the same effect on the enzymatic system as do the 7-radiations and that it does not com pensate for the ill-effects caused by the radiation. Summary We believe that our own results and all of the evidence previously reported on the Vol 9, Aug, 1964 FREE R relative 1 for the halogen: tirely cc mechani A tacked i I initial f It is for a he index o based o quency | attacked gated pc tors wh of a pa-, lower tl J can be 1 *1 any seriof react radical t coefficiei sure is A. Zolton tory, VA Ore 9720 1. Azoi R. T.: B 2. Bacc of Radiol 1 19S5- 'I 3. Baei 4. Bert Acta 24 -.: 5. Bro< 6. Bull 1949. 7. Bull 1961. Wirischc. iL.HEALTH FREE RADICALS IN SOLVENT TOXICITY en. drops off v anesthetic. liver dam- rhich should radical at- d that coms and their iuce similar isceptible to tot metabor example, lest transfer about 104 tchloride in in addition lysiological :omplexing lation may It should which are antiradiaow disrupequence of i oy Dohad to have dng a surmice subrradiation. id amino- dehydrohe mouse etermined do not iny of the rary. they r demonrcises the im as do not com1 by the l relative toxicities and metabolic products 8. Cheymol, J., et al: C R Soc Biol (Paris) for the entire range of hydrocarbon and 153:1965, 1959. I halogenated hydrocarbon solvents are en 9. Commoner, B.; Townsend, J.; and Pake, G. E.: Nature (London) 174:689, 1954. tirely consistent with the hypothesis that the 10. Cottrell, T. L.: Strength of Chemical Bonds, mechanism by which these molecules are at London: Butterworth Scientific Publications, Inc, j tacked in the biological system involves an 1958. initial free radical process. 11. Doherty, D. G., and Burnett, W. T., Jr.: Proc Soc Exp Biol Med 89:312, 1955. It is clear that the thermodynamic value 12. Gerarde, H. W.: AMA Arch Industr Health for a homolytic cleavage is not as useful an 13:468,1956. , index of reactivity as the transfer constant 13. Gerarde, H. W.: AMA Arch Industr Health based on observations of the relative fre 19:403, 1959. 14. Harper, K. H., and Calcutt, G.: Nature quency with which a particular molecule is (London) 192:165, 1961. )4 attacked by a growing free radical propa 15. Konecny, J. O.: J Amer Chem Soc 76:4993, gated polymer. There are several other fac 1954. I tors which influence the toxic effectiveness 16. Maynert, E. W., and Losin, L.: J Pharmacol I of a particular compound and consequently lower the probability that direct correlation Exp Ther 115:275, 1955. 17. McCollister, D. D., et al: J Pharmacol Exp Ther 102:112, 1951. l can be made. These factors are: (1) For 18. Michaelis, L., and Schubert, M. P.: Chem any series of solvents, different relative rates Rev 22:437, 1938. i of reaction are to be expected for different 19. Michaelis, L., in The Enzymes, J. B. Sumner - radical sources; (2) water/lipid distribution and K. Myrback, ed.. New York: Academic Press, Inc., 1951, vol 2, pt 1. chap 44. 1 coefficients are significant; (3) vapor pres 20. Minder, W.; Knuchel, H.; and Gurtner, P.: sure is important. Experientia 4:219, 1948. 21. Parke, D. V., and Williams, R. T.: Biochem Zolton T. Wirtschafter, MD, Research Labora J 54:231, 1953. tory, VA Hospital, Sam Jackson Park Rd, Portland, f Ore 97207. i REFERENCES i1 1. Azouz, W. M.; Parke, D. V.; and Williams, 22. Parke, D. V., and Williams, R. T.: Biochem J 55:337, 1953. 23. Parke, D. V.: Biochem J 62:339, 1956. 24. Plaa, G. L.; Evans, E. A.; and Hine, C. H.: J Pharmacol Exp Ther 123:224, 1958. ! R. T.: Biochem J 59:410, 1955. 25. Stein, G., and Weiss, ,J.; J Chem Soc, pt 4, 2. Bacq, Z. M., and Alexander, P.: Fundamentals i of Radiobiology, New York: Academic Press, Inc., j 1955. suppl 1-2:3245, 1949. 26. Temberg, J. L.. and Commoner, B.: JAMA 183:339, 1963. j 3. Baemstein, H. D.: J Biol Chem 161:685, 1945. 27. von Oettingen. W. F., et al: National 4. Bernhard, K., and Gressly, E.: Helv Chim Institutes of Health Bulletin No. 191, United States Acta 24:83, 1941. Public Health Service, 1949, p 1. 1 5. Brodie, B. B.: J Phartn Pharmacol 8:1, 1956. 28. Walling, C.: Free Radicals in Solution, New 6. Butler, T. C.: J Pharmacol Exp Ther 97:84, York: John Wiley & Sons, 1957. 1949. 29. Weiss, J.: Nature (London) 157:584, 1946. 7. Butler, T. C.: J Pharmacol Exp Ther 134:311, 30. Wirtschafter, Z. T., and Cronyn, M. W.: 1961. Relative Hepatotoxicitv, this issue, pp 180-185. * 41J and all 1 on the Aug, 196* 4 Wirtschafter--Cronyn 13 ff tr