in the liver, lungs, and kidneys, with approximately 20 percent of the ³⁵S activity being excreted in 12 hours. In rodents the majority of IV-injected sulfur mustard was excreted in the urine within 72 hours (Davison et al., 1961). The urinary metabolites included thiodiglycol and its conjugate (15 percent), glutamine-bis(b-chloroethylsulfide) conjugates (45 percent), glutamine-bis(b-chloroethylsulfone) conjugates (7 percent), and bis(b-chloroethysulfone) and conjugate (8 percent), with minute amounts of cysteine conjugates. These findings are comparable to subsequent work in rodents after intraperitoneal injection (Roberts and Warwick, 1963).

The reactive cyclic intermediate, the sulfonium ion, reacts avidly with proteins and nucleic acids, producing alkylation products (see Chapter 5). The alkylation of DNA by sulfur mustards has been studied by many investigators (Ball and Roberts, 1972; Boursnell et al., 1946; Davison et al., 1957, 1961; Gross et al., 1981; Habraken and Ludlum, 1989; Kohn et al., 1965; Lawley and Brookes, 1965; Ludlum et al., 1986; Meier et al., 1984; Papirmeister and Davison, 1964; Papirmeister et al., 1969, 1970, 1984a,b; Price et al., 1968; Roberts et al., 1971; Walker, 1971; Wheeler, 1962).

The sulfur mustards can be bifunctional, in that some ion intermediates covalently bind adjacent strands of DNA (a DNA cross-link). This interstrand link has been the subject of much of the study of the genotoxic effect of these agents. DNA cross-links induced by these mustards were shown by Wheeler (1962) to be extremely lethal to cells. Several workers also studied the cell cycle-specific toxicity of this bifunctional agent (Ludlum et al., 1978; Mauro and Elkind, 1968; Roberts et al., 1968, 1986). They have shown that cells in late G ₁ phase or S phase of the cell cycle are particularly sensitive to the biologic effects of alkylation.

In addition, the repair of DNA lesions induced by sulfur mustards has been studied in many systems, including those employing cells known to be naturally deficient in certain repair enzymes (Ball and Roberts, 1970; Fox and Fox, 1973; Gilbert et al., 1975; Lawley and Brookes, 1968; Murnane and Byfield, 1981; Plant and Roberts, 1971; Reid and Walker, 1966, 1969; Roberts and Kotsaki-Kovatsi, 1986; Roberts et al., 1986; Savage and Breckon, 1981; Walker, 1966; Walker and Reid, 1971; Walker and Smith, 1969). As expected, DNA repair-deficient cells generally are much more sensitive to the DNA cross-linking, and the cells die at significantly lower doses.

Recent work has specifically shown that ring nitrogens on DNA are the primary sites of attack. Among the products identified are N-7