40 mm millipede coil

The magnetic resonance imaging experiments were conducted on the same mouse at week 14 and week 25 using a 9.4 T horizontal bore magnet (Varian Yarnton UK) equipped with a 40 mm millipede coil, as previously described [49 (link)]. Localized ¹H-MRS from visceral and subcutaneous fat depots were acquired from 2 × 1.5 × 1.5 mm³ voxels positioned in the upper gonadal abdominal fat (as representative of visceral fat) and in the inguinal abdominal fat (as representative of the subcutaneous fat) (Fig. 2a).

Animals were anesthetized using isoflurane (4% for sleep induction and ~2% for sleep maintenance) in a 3:7 mixture of oxygen and air, before being positioned prone in the MR-compatible animal holder. Respiration was monitored during scanning (SA-instruments, Stony Brook, NY, USA). Core body temperature was maintained at 37 °C during scanning using a warm air system (SA-instruments, Stony Brook, NY, USA).
The MRI experiments (n = 6–7 per group) were conducted using a 9.4 T horizontal bore magnet (Varian, Yarnton, UK) equipped with a 40-mm millipede coil, as previously detailed^{45 (link)}. All images were collected on the matrix size 256 × 96 and a field-of-view of 51.2 × 51.2 mm². Fiji software (http://fiji.sc) was used to compute the volume of total fat (TF), visceral fat (VAT), and total subcutaneous fat (SAT). VAT was calculated as the difference between the TF signal and SAT signal in the abdominal region. MRI experiments were performed on the same mouse (F1) at P80 (MID) and P150 (END).