a Expression levels of Slc25a1 protein in the indicated organs derived from a 4-month-old male mouse. b Top panels: tissue microarrays derived from human normal liver (NL) and NASH patients. Bottom panels: Slc25a1 and DAPI staining in the livers of mice fed with the control (CD) or high-fat diet (HFD). c, d Quantification of Slc25a1 staining in human NASH livers (c) or murine livers (d) as shown in (b). Quantification was performed with the ImageJ program on nine NASH livers relative to three normal livers on multiple fields. e Body weight measurements in the prevention study. Mice were fed the HFD for 3 weeks and CTPI-2 treatment was started (indicated by arrow). f Body weight measurements in the reversion study. Mice fed the CD or HFD for 3 months were then randomized to receive CTPI-2 or vehicle for an additional 3 months (n = 3–4). Measurements were taken at regular intervals. The arrow indicates the time at which CTPI-2 treatment was started, after 12 weeks of HFD-feeding. In both cases CTPI-2 was administered at 50 mg/kg on alternate days via the intraperitoneal route. g Representative images of mice fed the HFD in the reversion study in the indicated treatment conditions. h Body weight measurements of mice fed control diet and treated with CTPI-2 for 5 months (n = 7). i MRI images of the visceral fat (circled in white) of the indicated mice. Insets numbers (in white) indicate the percentage of visceral fat calculated as threshold fat pixels versus threshold whole-body pixels in the abdominal region from every slice of the MRI image dataset (see “Materials and methods”). j Representative images of the hemi-lateral visceral fat (WAT) and the liver of mice fed the HFD, in the indicated treatment conditions. k Measurement (in grams) of the indicated organs in CD, HFD+Veh., or HFD+CTPI-2 (n = 3). *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.