Potentially mutagenic impurities: analysis of structural classes and carcinogenic potencies of chemical intermediates in pharmaceutical syntheses supports alternative methods to the default TTC for calculating safe levels of impurities.

Potentially mutagenic impurities in new pharmaceuticals are controlled to levels with negligible risk, the TTC (threshold of toxicological concern, 1.5 µg/day for a lifetime). The TTC was based on the more potent rodent carcinogens, excluding the highly potent "cohort of concern" (COC; for mutagenic carcinogens these are N-nitroso, Aflatoxin-like, and azoxy structures). We compared molecules with DEREK "structural alerts" for mutagenicity used in drug syntheses with the mutagenic carcinogens in the Gold Carcinogenicity Potency Database. Data from 108 diverse synthetic routes from 13 companies confirm that many "alerting" or mutagenic chemicals are in structural classes with lower carcinogenic potency than those used to derive the TTC. Acceptable daily intakes can be established that are higher than the default TTC for many structural classes (e.g., mono-functional alkyl halides and certain aromatic amines). Examples of ADIs for lifetime and shorter-term exposure are given for chemicals of various potencies. The percentage of chemicals with DEREK alerts that proved mutagenic in the Ames test ranged from 36% to 83%, depending on structural class, demonstrating that such SAR analysis to "flag" potential mutagens is conservative. We also note that aromatic azoxy compounds need not be classed as COC, which was based on alkyl azoxy chemicals.

The lack of binding of methyl-n-amyl ketone (MAK) to rat liver DNA as demonstrated by direct binding measurements, and 32P-postlabeling techniques.

It has been reported that 14C-labeled methyl-n-amyl ketone (MAK, 2-heptanone) is able to bind spontaneously, in vitro, to isolated rat liver DNA to the extent of 400 pmol/mg DNA; and that 14C-MAK, when given by gavage to female Fischer 344 rats, resulted in HPLC chromatograms of isolated, hydrolyzed liver DNA in which some radiolabel was not associated with the four normal DNA bases dA, dT, dC, and dG. The present studies were undertaken to re-examine the hypothesis that MAK is able to bind to rat liver DNA. In the in vitro study, liver nuclear DNA was incubated with [2-14C]-labeled MAK (25 mCi/mmol) in the absence, or in the presence of rat liver microsomes, precipitated, washed free of unbound MAK, and counted by scintillation spectrometry. No binding to DNA by MAK was detectable. In the in vivo study, groups of five female F344 rats were exposed by inhalation to 0, 80, 400, or 1000 ppm MAK for 6 h/day for 10 days. DNA was purified from the liver nuclei of the 0 and 1000 ppm dosed animals, and 32P-postlabeling techniques were used to assay for adducts. No DNA adducts were detected using these techniques. It was concluded that MAK lacks the ability to bind to rat liver DNA in vitro and in vivo.