S9 Incubations and subsequent LC-UV/MS analysisChemicals and reagents

All solvents used were of LC–MS grade quality, if not stated otherwise. Magnesium chloride hexahydrate, glucose 6-phosphate (G6P) disodium salt hydrate, glutathione, 37% formaldehyde solution, ammonium acetate, potassium chloride, dimethylformamide suitable for HPLC, dibasic sodium phosphate, acetonitrile, formic acid, 2,4-dinitrophenylhydrazine (DNPH) phosphoric acid solution, formaldehyde-DNPH solution, nicotinamide adenine dinucleotide phosphate (NADP) disodium salt, nutrient broth No.1 medium and sodium dihydrogen phosphate dihydrate were obtained from Merck. NO-HCTZ was synthesised as described in section S5. S9 fraction from phenobarbital/5,6-benzoflavone-induced hamster liver was purchased from Trinova Biochem (Giessen, Germany). The water used in this part of the study was prepared with a Milli-Q system (Q-POD). As a suitable reference standard was not commercially available for S‑(hydroxymethyl)glutathione, a standard serving for its qualitative identification in the incubation samples was prepared by adding an excess of formaldehyde to a glutathione standard solution applying a modified approach described in literature (Hopkinson et al. 2010).

S9 incubationsPreparation of the S9 mix

The S9 mix was prepared immediately before use with final concentrations of 0.1 M phosphate buffer (pH 7.4), 8 mM MgCl2, 33 mM KCl, 4 mM NADP, 5 mM G6P and 10% (v/v) S9 fraction from phenobarbital/5,6-benzoflavone-induced hamster liver.

For the incubations with supplemented glutathione, appropriate aliquots of a glutathione stock solution prepared in phosphate buffer were administered by replacing the equivalent volume of phosphate buffer in the S9 mix.

The incubations samples for the − S9 conditions in the absence of the S9 mix were prepared solely in 0.1 M phosphate, with any required glutathione supplemented by replacing the respective volume of phosphate buffer with the glutathione stock solution.

Incubation procedure

500 µl of phenobarbital/5,6-benzoflavone-induced hamster S9 mix (+ S9 conditions) or 0.1 M phosphate buffer (− S9 conditions) containing either 0, 3.3, 6.7 or 10 mg glutathione were added to a centrifuge tube and mixed with 100 µl nutrient broth medium. After addition of 50 µl NO-HCTZ (100 mg/ml in dimethylformamide) and brief mixing, the samples were incubated at 37 °C and shaken at 500 rpm. After selected time points (0, 2, 10, 30, 60 min), the individual incubations were terminated by addition of 650 µl ice-cold acetonitrile and the tubes were shaken on a vortex shaker for a few seconds. The samples were subsequently centrifuged for 5 min at 15,000 g at room temperature, followed by LC-UV or LC–MS analysis of the obtained supernatant as described below.

Furthermore, appropriate chemical control samples for the + S9 and − S9 conditions were prepared. For that purpose, 500 µl of S9 mix or 0.1 M phosphate buffer supplemented with or without glutathione were mixed with 100 µl nutrient broth medium prior to addition of 50 µl dimethylformamide instead of NO-HCTZ. The subsequent incubations and treatments were carried out as described above, with the exception that only the incubation time points 0 min and 60 min were prepared for the individual chemical control samples.

LC-UV and LC–MS analysisLC-UV analysis of formaldehyde after derivatisation with 2,4-dinitrophenylhydrazine

50 µl of the supernatant obtained after termination of the incubation and subsequent centrifugation were mixed with 945 µl acetonitrile and 5 µl of 2,4-dinitrophenylhydrazine phosphoric acid (0.2 M) for derivatisation. After brief shaking, the sample was diluted with 500 µl of water prior to LC-UV analysis according to the method details given in the Supplementary Material (Supplementary Table 1). Identification and assignment of the peak corresponding to formaldehyde-2,4-dinitrophenylhydrazine was performed based on comparison to a commercially obtained reference standard; an exemplary LC-UV chromatogram is shown in Supplementary Fig. 1.

LC–MS analysis of S-(hydroxymethyl)glutathione and NO-HCTZ

20 µl of the supernatant obtained after termination of the incubation and subsequent centrifugation were diluted with 980 µl water followed by LC–MS analysis for S‑(hydroxymethyl)glutathione and NO-HCTZ using two separate LC–MS methods. All samples were analyzed for their levels of S‑(hydroxymethyl)glutathione, whereas only the incubation samples obtained after 0 min, 30 min and 60 min were subjected to LC–MS analysis for NO-HCTZ.

Details on the LC–MS conditions and exemplary chromatograms for the analysis of S‑(hydroxymethyl)glutathione are shown in the Supplementary Material (Supplementary Table 2, Supplementary Fig. 2, Supplementary Fig. 3). Identification and assignment of the peak corresponding to S‑(hydroxymethyl)glutathione in the incubation samples was carried out based on the respective m/z as well as comparison to a prepared qualitative reference standard.

The LC–MS method used for the detection of NO-HCTZ and an illustrative LC–MS chromatogram are shown in Supplementary Table 3 and Supplementary Fig. 4, respectively. Identification and assignment of the peak corresponding to NO-HCTZ in the samples was performed based on comparison to the available reference standard.

Ames testsMetabolic activation system

Experiments were conducted in either the presence of a metabolic activation system (S9 mix) (based on the livers of male Sprague–Dawley rats or male Syrian hamsters induced with either Aroclor 1254 or phenobarbital/5,6-benzoflavone), or in the absence of S9 mix (phosphate buffer pH 7.4). S9 fraction (purchased from Trinova Biochem (Giessen, Germany)) was diluted to 30 mg/ml protein with 0.1 mol/l phosphate buffer (pH 7.4) before incorporation into the S9 mix. The S9 fraction was thawed and complete activation systems (S9 mix) were prepared immediately before use. Final concentrations in S9 mix were: 100 mmol/l phosphate buffer (pH 7.4), 8 mmol/l magnesium chloride, 33 mmol/l potassium chloride, 4 mmol/l NADP, 5 mmol/l glucose-6-phosphate (G6P) and 10 % v/v of 30 mg/ml protein rat or hamster liver homogenate (S9 fraction).

Note that the inducing agent used (Aroclor 1254 or phenobarbital/5,6-benzoflavone) for the S9 fraction varied between experiments due to the discontinuation of S9 induced with Aroclor 1254 following the banning of Aroclor 1254 production. Within any given experiment that included both rat and hamster liver S9 mixes, the inducing agent was the same for both. All batches of rat liver S9 fraction (regardless of inducing agent) used in this work had metabolic capacity demonstrated by the supplier by key enzyme assays and the ability to activate reference agents to bacterial mutagens. All batches of hamster liver S9 fraction (regardless of inducing agent) were evaluated by the supplier for the ability to metabolise the standard reference oil HC235 to mutagenic intermediates. All batches of S9 fraction (for either species, regardless of inducing agent) were additionally tested in-house by the test facility using two promutagens, cyclophosphamide and benzo[a]pyrene, in strains TA1535 and TA98, respectively. The transition to S9 induced with phenobarbital/5,6-benzoflavone is supported by a body of research (Callander et al. 1995; Elliott et al. 1992; Paolini et al. 1991). The type of S9 used in any given experimental figure in this publication is described in the corresponding figure legend.

Bacterial strains

The bacterial strains used were Salmonella typhimurium LT2 strains TA1535, TA1537, TA98 and TA100, and Escherichia coli WP2 strain uvrA/pKM101 (referred to as WP2 uvrA/pKM101 hereafter). Further details are available in SI and Supplementary Table 8.

Ames experimental conditions

Ames tests were conducted based on recommendations of the OECD 471 test guideline (OECD 2020) and using the pre-incubation test (pre-incubations were of 60 min duration at 37 °C in an orbital incubator set at 120 rpm) based on the methods of Maron and Ames (1983), Mortelmans and Riccio (2000), Mortelmans and Zeiger (2000) and Venitt et al (1984). Dosing was performed in ≤ 100 µl of solvent (water for FA and NDEA, DMF for NO-HCTZ). Further details are available in the supplementary information (chapter S9).