Biomarkers of aging: correlation of DNA I-compound levels with median lifespan of calorically restricted and ad libitum fed rats and mice

Abstract

I-compounds are species-, tissue-, genotype-, gender-, and diet-dependent bulky DNA modifications whose levels increase with animal age. While a few of these DNA modifications represent oxidation products, the majority of I-compounds appear to be derived from as yet unidentified endogenous DNA-reactive intermediates other than reactive oxygen species. Circadian rhythms of certain I-compounds in rodent liver imply that levels of these DNA modifications are precisely regulated. Caloric restriction (CR), the currently most effective method available to retard aging and carcinogenesis, has been previously shown to elicit significant elevations of I-compound levels in tissue DNA from Brown-Norway (BN) and F-344 rats as compared to age-matched ad libitum fed (AL) animals. The present investigation has extended this work by examining liver and kidney DNA I-compound levels in three genotypes of rats (F-344, BN, and F-344 × BN) and two genotypes of mice (C57BL/6N and B6D2F1) under identical experimental conditions in order to determine whether correlations exist between I-compound levels, measured in middle-aged animals, and median lifespan. Levels of a number of liver and kidney I-compounds were found to display genotype- and diet-dependent, statistically significant positive linear correlations with median lifespan in both species. In particular, the longer-lived hybrid F-344 × BN rats and B6D2F1 mice tended to exhibit higher I-compound levels than the parent strains. CR enhanced I-compound levels substantially in both rats and mice. Thus, I-compounds, measured at middle age, reflected the functional capability (‘health’) of the organism at old age, suggesting their predictive value as biomarkers of aging. The positive linear correlations between levels of certain I-compounds (designated as type I) and lifespan suggest that these modifications may be functionally important and thus not represent endogenous DNA lesions (type II), whose levels would be expected to correlate inversely with lifespan.