Identification of a novel fluorescent adduct formed in the reaction of malonaldehyde with adenosine

Malonaldehyde was reacted with adenosine in aqueous solution at acidic conditions and the reaction mixtures were analysed by HPLC. Four major product peaks were observed in the chromatogram recorded by the UV detector at 320 nm. Two of the peaks could be deduced to the previously characterised malonaldehyde–adenosine reaction product 9-(β-D-ribofuranosyl)-6-(5,7-diformyl-2H-3,6-dihydro-2,6-methano-1,3-oxazocin-3-yl)purine (M3A) and to the ribose analogue of the 2′-deoxy adduct 9-(2′-deoxy-β-D-ribofuranosyl)-6-(3,5-diformyl-4-methyl-1,4-dihydro-1-pyridyl)purine (M2AA-dA). The two other peaks were due to previously uncharacterised adducts. Upon isolation of these adduct peaks it was found that the peaks are interconverted to each other, one adduct peak dominating over the other one. On the basis of data recorded by UV, fluorescence and 1H NMR spectroscopy, and mass spectrometry, the structure of the major adduct could be determined as 9-(β-D-ribofuranosyl)-6-(3,5-diformyl-4-etanal-1,4-dihydro-1-pyridyl)purine (3M-A). The adduct is most likely formed by reaction of adenosine with a malonaldehyde condensation product consisting of three units of malonaldehyde. The highest yield of 3M-A was obtained in the reaction performed at pH 4.6 and 80 °C for 75 h. The minor form of the interconverting peaks could not be characterised on the basis of the spectral data. However, it is concluded that the minor peak most likely represents the hydrated form of 3M-A. Since these adducts are formed only in trace amounts at neutral pH and 37 °C and the formation requires high amounts of malonaldehyde, it is likely that the adducts are not formed in DNA in vivo and thus not contribute to the malonaldehyde genotoxicity.