In Vitro Evaluation of DNA Damage Effect Markers toward Five Nitrogen Mustards Based on Liquid Chromatography–Tandem Mass Spectrometry

Endogenous DNA lesions frequently occur due to internal effects such as oxidative stress, inflammation, endogenous alkylation, and epigenetic modifications. However, exposure to chemical toxicants from the environment, diet, or drugs can also induce significant endogenous DNA damage. The quantification of endogenous DNA damage effect markers might reflect the actual DNA damage level of chemical toxicants. Herein, we report a liquid chromatography–triple quadrupole tandem mass spectrometry (LC–QqQ MS/MS) method for simultaneous determination of eight representative endogenous DNA damage biomarkers, including five endogenous DNA damage effect markers (oxidative damage, 8-oxo-dG; lipid peroxidation, εdA and N2-Et-dG; inflammation, 5-Cl-dC; and endogenous alkylation, O6-Me-dG), and three epigenetic modifications (5-m-dC, 5-hm-dC, and N6-Me-dA). The method validation was performed, and the linear range was 0.05 pg to 2 ng (on-column), the limit of detection was 0.02 pg (on-column), and the precision, accuracy, matrix effect, and recovery were all between 85 and 115%. We then applied this method to evaluate endogenous DNA damage to human embryonic lung fibroblast cells exposed to five nitrogen mustards [NMs, i.e., HN1, HN2, HN3, chlorambucil (CB), and cyclophosphamide (CTX)], where curcumin exposure was used as a control due to its inability to induce the formation of endogenous DNA adducts. The amounts of eight DNA adducts in the low-, middle-, and high-concentration exposure groups of five NMs were almost all significantly different from those in the blank group (P < 0.05). We obtained a positive correlation between the contents of eight DNA damage biomarkers and the inhibition dose of five NMs, except for N2-Et-dG and 5-Cl-dC. Via further principal component analysis and partial least squares discriminant analysis, we clustered all NMs into three units with different cytotoxicity levels, that is, HN2 and HN1 (highly toxic), HN3 and CB (moderately toxic), and CTX (less toxic). Moreover, for the same concentration of HN1/2/3 exposure groups, as the cytotoxicity increased according to the order of HN3 < HN1 < HN2, the contents of 8-oxo-dG, 5-m-dC, 5-hm-dC, and N6-Me-dA increased, whereas the content of O6-Me-dG decreased. Therefore, the contents of these DNA damage effect markers were somewhat related to the cytotoxicity and concentration of NMs. We hope that this method will provide an alternative evaluation approach for the toxicological effects of NMs and the safety of the medication.