2.4. Serum ketone body determination by gas chromatography–mass spectrometry (GC‐MS)

Hundred microliter of each serum sample was dissolved in 500 µl of acetonitrile and centrifuged at 13,000 rpm for 30 min. Precipitated debris were discarded and supernatant was dissolved in 1.5 ml of water. The obtained solution was treated using NaOH 30% w/v to adjust pH to 14. The metabolites were oximated by adding 0.5 ml of hydroxylamine hydrochloride (2.5 mg/ml in water) at 60°C for 60 min. A small volume of 2.5 M H₂SO₄ was added to acidify the solution. An internal standard mix, constituted by dimethylmalonic acid (10 µM), pentadecanoic acid (10 µM), and tropic acid (10 µM), was added and then the metabolites were extracted by ethyl acetate. The organic phase was dehydrated by adding 1 g of Na₂SO₄ and evaporated at 40°C under a nitrogen flow. The sample was derivatized by adding N, O‐bis(trimethylsilyl) trifluoroacetamide (50 µl) at 60°C for 30 min and analyzed by GC‐MS analysis on GC column HP‐5MS; 30 m × 0.250 mm × 0.25 µm using GC‐MS Agilent 7890A (Agilent Technologies). The experimental set up was as follows: chromatographic gradient, 70°–280°C, 10°C/min; helium flow was 1 ml/min; solvent delay 6 min; scan range from 50 to 650 amu. The NIST and Wiley mass spectra library (2008), and the MSD Productivity Chemstation software (Agilent Technologies) was used to perform metabolite identification. Finally, the compound abundance was assessed comparing the areas of the chromatographic peaks versus areas of internal standard (Rossi et al., 2018).