br Quantification of hemoglobin adducts br Blood samples fro
2.2. Quantification of SB431542 adducts
Blood samples from the treated mice and rats were prepared ac-cording to the FIRE procedure to measure the adduct levels from gly-cidol to the N-terminal valine in Hb (Aasa et al., 2017; von Stedingk et al., 2010). This procedure enables the detachment from Hb of the formed adduct, N-(2,3-dihydroxypropyl)valine (diHOPrVal), through derivatization with the Edman reagent fluorescein isothiocyanate (FITC), giving the corresponding fluorescein isothiohydantoin (FTH). Quantification of diHOPrVal-FTH and the corresponding (13C5)-sub-stituted internal standard was performed by ultra-pressure liquid chromatography with high resolution mass spectrometry (UPLC/ HRMS) after clean-up of the blood samples. The calibration curve, prepared by adding known amounts of diHOPrVal-FTH to derivatized human blood samples, was processed in parallel. See Supplementary Data and Fig. S1 for further details.
2.3. Calculation of the internal dose
The internal doses of glycidol, expressed as AUC, in the treated mice and rats were calculated from the measured diHOPrVal adduct levels (A), using the second-order reaction rate constant (kval) for glycidol binding to the N-terminal valine in Hb. The rate constant, kval, was determined by incubation of glycidol in pooled blood from both sexes of mice and of rats, respectively, from the same strains as in the in vivo dosimetry study. The procedure was similar to the previously described experiment with glycidol and blood from mice (Aasa et al., 2017). Glycidol at five different concentrations (in duplicates) was incubated in lysed blood from respective species for 1 h (at 37 °C), followed by derivatization with FITC. Following work-up and analysis of the sam-ples, the second-order rate constant (kval) given for mouse and rat, re-spectively, was derived from the linear slopes of Hb adduct levels versus the glycidol AUC in vitro (r2 > 0.91). The unit [pmol/g Hb per μMh] is used for kval when the adduct level (A) is expressed as [pmol/g Hb] and the AUC as [μMh]. See Supplementary Data for details of the procedure.
The decrease in adduct level due to elimination of Hb (erythrocyte lifetime) during the exposure period in the short-term animal exposure was considered by adjusting A by multiplying by a factor (dependent on exposure and sampling time and erythrocyte lifetime) of 1.1 (mice) or 1.07 (rats), denoted species-specific factor in Equation (1) (cf. Granath et al., 1992). The daily adduct level increment (a) was then obtained by dividing by five (days of administration) according to Equation (1). The daily AUC's (AUCday) were then calculated using Equation (2).
For projection to the lifetime AUC (AUClifetime) in the carcinogeni-city studies, the AUC per daily administered dose (mg/kg per day) (that
Fig. 1. Internal doses (AUC) of glycidol, calculated from Hb adduct levels in blood from both sexes of mice (A) and rats (B), after treatment with glycidol by gavage for five consecutive days. Three animals per dose level and for the controls were used. For the AUC calculations the measured ad-duct levels were adjusted for the ery-throcyte lifetime of each species. The slopes of the linear regressions correspond to the mean AUC per administered dose (note different scales on the y-axis for the two species).
is AUCday) obtained from the short-term exposure studies were multi-plied by the number of days in the carcinogenicity studies, Transcription is 5 days per week and 103 weeks for the different exposure doses (Irwin et al., 1996; NTP, 1990).
2.4. Glycidol data from carcinogenicity studies
Published data from NTP (Irwin et al., 1996; NTP, 1990) on gly-cidol-induced neoplasms in dose groups of 50 female and 50 male F344 rats and B6C3F1 mice were used for evaluation of the applicability of the relative cancer risk model to glycidol. In those carcinogenicity studies the animals were administered water (control) or glycidol via gavage, 5 days per week for 103 weeks at dose levels of 37.5 and 75 mg/kg (rats) and of 25 and 50 mg/kg (mice). Different types of neoplasms were observed in rats (13 types) and mice (10 types); see Supplementary Tables S1–S2 for details.
2.5. Multiplicative cancer risk model
We assume that the observed relative frequencies of tumours in the published animal carcinogenicity studies represent the cumulative risk over two years. If Si(t) denotes the survival function for cancer in site i, and λi (t) is the corresponding hazard function among unexposed, the cumulative risk up to time t, among unexposed animals, Pi(0), can be expressed as:
where i is the cumulative hazard function