Identification of mutagens in drinking water and classification of mutations on DNA

The objective of this work was to develop a method to recover and identify mutagenic electrophiles present in drinking water and then to develop a method for rapidly determining the types of mutation such mutagens create in DNA. I. 5,5$\sp\prime$-dithiobis-(2-nitrobenzoic acid) was esterified to a carbohydrate polymer coating on controlled pore glass and reduced to the free thiol, creating an immobilized nucleophile, 5-mercapto-2-nitrobenzoyl (MNB) glass. MNB glass was used to trap the electrophilic mutagens present in drinking water concentrates (DWC). After washing to remove non-covalently bound material, the nucleophile-electrophile adducts were released by treatment with strong base. HPLC was used to separate and purify these products. II. A series of polycyclic aromatic hydrocarbon diol epoxides (PAHDEs) and small alkylating agents was tested for mutagenicity in Salmonella typhimurium strains TA100 and TA104. The relative levels of mutagenicity observed in the two strains correlated with the known amounts of adduct formed at dG and dA by the compounds, suggesting that the relative response of the two strains might be used to predict unknown ratios of chemical attack of a mutagen on dG vs dA in DNA. The analysis with TA100 and TA104 was then extended to DWC and the DWC mutagen MX, whose extent of reaction with dG vs dA is unknown. The results indicated MX is a dG favoring mutagen while the mutagens in DWC other than MX are more dA favoring compared to MX. The F$\sp\prime$ episomes of six LacZ$\sp-$ Escherichia coli (E. coli) strains, CC101-106, each of which is reverted by one of the six possible base pair substitutions, were transferred to E. coli strain EE122, which carries $\Delta$uvrB and rfa, to create six intermediate strains, CL101-106. Plasmid pKM101 was transferred to CL101-106 to create six new strains, CL101P-106P. The CL101P-106P strains were treated in parallel with a set of PAHDEs to generate mutational spectra or quantitative measurements of the relative amounts of the six kinds of point mutation that are produced by the compounds. The results were close to the published mutational spectra for these compounds, which were determined by various DNA sequencing studies, except CL103P (reverted by GC $\to$ CG transversions) responded very poorly. Excluding the underresponsiveness of CL103P, PAHDEs mainly created GC $\to$ TA and AT $\to$ TA transversions (70-80%) while MX mostly generated GC $\to$ TA and AT $\to$ CG transversions (90%).