Metabolism and mutagenicity of halogenated olefins--a comparison of structure and activity.

Chlorinated ethylenes are metabolized in mammals, as a first step, to epoxides. The fate of these electrophilic intermediates may be reaction with nucleophiles (alkylation), hydrolysis, or intramolecular rearrangement. The latter reaction has been studied in the whole series of chlorinated epoxiethanes. The rearrangement products found were: acyl chlorides (tetrachloro-, trichloro-, and 1,1-dichloroethylenes), or chlorinated aldehydes (1,2-dichloroethylenes, cis- and trans-, vinyl chloride). The metabolities found in vivo are identical with, or further derivatives of these rearrangment products, with one important exception: trichloroethylene. With this compound, in vivo rearrangement yields chloral exclusively. The mechanism of the different rearrangement has been identified as a Lewis acid catalysis. All chlorinated ethylenes have been investigated in a tissue-mediated mutagenicity testing system. The prominent molecular feature of those with mutagenic effects (trichloro-, 1,1-dichloro-, and monochloroethylene) is unsymmetric chlorine substitution which renders the epoxides unstable, whereas symmetric substitution confers relative stability and nonmutagenic property.

Since vinyl chloride has been demonstrated carcinogenic in humans (1) and experimental animals (2), and further reports suggested carcinogenic properties of the chemical congeners vinylidene chloride (3,4) and trichloroethylene (5), we became interested in the question whether or not other members of the group of chlorinated ethylenes would share this type of biological activity, and what would be the molecular features or prerequisites, respectively, which are responsible for the carcinogenic effect. We started with a theoretical consideration of the chemical activity of the molecules which focused on the influence of the chlorine substitutions. In a second step, the metabolic bioactivation and degradation was studied in vivo and compared with the theoretical expectations from the chemical reactivity. After this, the mutagenic activity was investigated in vitro in a modified Ames testing system. Finally, the results of this mutagenicity test were compared with the in vivo carcinogenicity findings as reported by others, and a tentative rule of chemical reactivity, biotransformation and mutagenicity/carcinogenicity was worked out. *Institut fur Toxikologie und Pharmakologie, Universitat Wurzburg, D-8700 Wurzburg, West Germany.

Chemical Reactivity
The chemical reactivity of chlorinated aliphatic compounds is decisively determined by the chlorine substitution(s). However, the influence is completely different in alkanes, alkenes, and alkynes. In general, chlorine substitution exerts a stabilizing effect on account of steric protection by the bulky substituent. This interferes with the electronwithdrawal effect of the chlorine substituent. In the case of chlorinated alkanes, the result is a destabilization of C-C and C-Cl bonds. C-C and C-Cl fissions under formation of free radicals are consequent reactions in the metabolic conversions of C1 or C, compounds.
A quite different influence is encountered with chlorinated ethylenes. In this class of molecules, the electron-withdrawal effect dominates over the mesomeric donator effect of the involved carbon atom, thus decreasing the electron density in the double bond which in turn results in a chemical stabilization against electrophilic attacks (6). The stabilizing effect increases with the number of chlorine substitutions. It is optimal in tetrachloroethylene, as has convincingly been demonstrated by the reactivity of chlorinated ethylenes with ozone (7).

December 1977 61
In alkynes, chlorination results in a destabilization, as has been shown with mono-and dichloroacetylene, which are considerably less stable than acetylene (8).

Metabolic Activation and Deactivation of Chlorinated Ethylenes
The first step of metabolic transformation of chlorinated ethylenes in mammals is epoxidation. Up to now there is only indirect evidence for this reaction, (a) by the demonstration of a specific binding spectrum with P-450 in the case of trichloroethylene (9), and (b) by the type of metabolites identified in vitro as well as in vivo.
In general, chlorinated epoxides may undergo a variety of reactions (see Fig. I): reaction with nucleophilic cellular macromolecules under alkylation; conjugation with low molecular nucleophiles (mainly glutathione) both enzymatically and nonenzymatically; hydrolysis to diols, again with and without the catalytic action of enzymes (epoxide hydrase); and intramolecular rearrangement. The latter reaction represents a deactivation mechanism and is of considerable importance for the potential of acute toxicity as well as of carcinogenicity and mutagenicity of the different members of the series of chlorinated ethylenes. (1)- (5)]. The expected metabolites would be trichloroacetic acid (with tetrachloroethylene), dichloroacetic acid (triand 1,2-cis-and transdichloroethylene) or monochloroacetic acid (vinylidene and vinyl chloride), and derivatives of monochloroacetic acid (or chloroacetaldehydc) after conjugation with glutathione. The predicted metabolites have been identified in systematic experiments with the isolated perfused rat liver preparation (10), with one important exception: trichloroethylene. With this compound, essentially no dichloroacetic acid was detected; this has recently been confirmed by Leibman in a microsomal metabolizing system (11). The different behavior of trichloroethylene oxide in vitro (thermal rearrangement to dichloroacetyl chloride) and in vivo (formation of chloral) is outlined in Figure 2.  We have studied the chemical behavior of these epoxides which have been synthesized. The exception was 1,1-dichloroethylene oxide, which cannot be obtained by conventional methods, apparently because it is extremely unstable on account of its high polarity (10). They rearrange either to acyl chlorides (tetrachloro-, trichloro-, and 1,1dichloroethylene) or to chlorinated aldehydes This difference might be of crucial significance for the evaluation of the carcinogenic potential of trichloroethylene. We therefore studied the rearrangement mechanisms in a more detailed investigation. Theoretically, there are three paths of intramolecular rearrangement. A simple hydride migration which would lead to dichloroacetyl chloride has low probability of occurring (12). The formation of a-ketocarbonium ions after C-O heterolysis which is the most probable path, could occur in two versions, resulting either in a single [Eq.  7)]; the latter (disubstituted), which would yield chloral, is less stable and thus has the lower probability. The rearrangement to chloral [Eq. (7)], which is the only path detected in vivo can be forced in vitro by the catalytic action of strong Lewis acids like FeCl3 or AlCl3 (10). We have speculated therefore, that the only formation of chloral (and further oxidation or reduction products such as trichloroacetic acid and trichloroethanol, and its glucuronide) in vivo could be due to catalytic action of the iron of P450 in the trivalent form at the site of the formation of the epoxide (6); experiments with purified P450 to confirm or reject this assumption are in progress.

Mutagenicity of Chlorinated Ethylenes
The six chlorinated ethylenes have been tested for their mutagenic potential in a modified Ames tissue-mediated microbial system (13). Salmonella typhimurium as tester strains, which had previously been successfully used with vinyl and vinylidene chloride (14,15), were found unsuitable because of the high primary toxicity exerted by some of the December 1977 compounds. After extended preliminary studies, E. coli K 12 was found an acceptable tester strain.
The results are given in Figure 3. There was no direct mutagenic activity of the compounds. Only after metabolic activation by added phenobarbitalinduced mouse microsomes some of the compounds yielded positive results. The data cannot be compared on a quantitative basis because the concentrations of the test compounds varied from 0.6mM (tetrachloroethylene) to 10.6mM (vinyl chloride). However, vinyl chloride obviously is the most active compound. Vinylidene chloride and trichloroethylene exert a small but definite mutagenic effect, whereas tetrachloroethylene and 1,2-cisand trans-dichloroethylenes are inactive.