Kinetics of in Vitro Guanine-N7-Alkylation in Calf Thymus DNA by (2S,3S)-1,2-Epoxybutane-3,4-diol 4-methanesulfonate and (2S,3S)-1,2:3,4-Diepoxybutane: Revision of the Mechanism of DNA Cross-Linking by the Prodrug Treosulfan

Prodrug treosulfan, originally registered for treatment of ovarian cancer, has gained a use in conditioning prior to hematopoietic stem cell transplantation. Treosulfan converts nonenzymatically to the monoepoxide intermediate (EBDM), and then to (2S,3S)-1,2:3,4-diepoxybutane (DEB). The latter alkylates DNA forming mainly (2′S,3′S)-N-7-(2′,3′,4′-trihydroxybut-1′-yl)guanine (THBG) and (2S,3S)-1,4-bis(guan-7′-yl)butane-2,3-diol cross-link (bis-N7G-BD) via the intermediate epoxide adduct (EHBG). It is believed that DNA cross-linking by DEB is a primary mechanism for the anticancer and myeloablative properties of treosulfan, but clear evidence is lacking. Recently, we have proved that EBDM alkylates DNA producing (2′S,3′S)-N-7-(2′,3′-dihydroxy-4′-methylsulfonyloxybut-1′-yl)-guanine (HMSBG) and that free HMSBG converts to EHBG. In this paper, we investigated the kinetics of HMSBG, bis-N7G-BD, and THBG in DNA in vitro to elucidate the contribution of EBDM and DEB to treosulfan-dependent DNA–DNA cross-linking. Calf thymus DNA was exposed to (A) 100 μM treosulfan, (B) 200 μM treosulfan, and (C) DEB at a concentration 100 μM, exceeding that produced by 200 μM treosulfan. Following mild acid thermal hydrolysis of DNA, ultrafiltration, and off-line HPLC purification, the guanine adducts were quantified by LC–MS/MS. Both bis-N7G-BD and THBG reached highest concentrations in the DNA in experiment B. Ratios of the maximal concentration of bis-N7G-BD and THBG to DEB (adduct Cmax/DEB Cmax) in experiments A and B were 1.7–3.0-times greater than in experiment C. EHBG converted to the bis-N7G-BD cross-link at a much higher rate constant (0.20 h–1) than EBDM and DEB initially alkylated the DNA (1.8–3.4 × 10–5 h–1), giving rise to HMSBG and EHBG, respectively. HMSBG decayed unexpectedly slowly (0.022 h–1) compared with the previously reported behavior of the free adduct (0.14 h–1), which revealed the inhibitory effect of the DNA environment on the adduct epoxidation to EHBG. A kinetic simulation based on the obtained results and the literature pharmacokinetic parameters of treosulfan, EBDM, and DEB suggested that in patients treated with the prodrug, EBDM could produce the vast majority of EHBG and bis-N7G-BD via HMSBG. In conclusion, EBDM can produce DNA–DNA lesions independently of DEB, and likely plays a greater role in DNA cross-linking after in vivo administration of treosulfan than DEB. These findings compel revision of the previously proposed mechanism of the pharmacological action of treosulfan and contribute to better understanding of the importance of EBDM for biological effects.