Quantitative image analysis in adipose tissue using an automated image analysis system: Differential effects of peroxisome proliferator-activated receptor-α and -γ agonist on white and brown adipose tissue morphology in AKR obese and db/db diabetic mice

Morphometric analysis of adipocytes is widely used to demonstrate the effects of antiobesity drugs or anti-diabetic drugs on adipose tissues. However, adipocyte morphometry has been quantitatively performed by manual object extraction using conventional image analysis systems. The authors have developed an automated quantitative image analysis method for adipose tissues using an innovative object-based quantitative image analysis system (eCognition). Using this system, it has been shown quantitatively that morphological features of adipose tissues of mice treated with peroxisome proliferator-activated receptor (PPAR) agonists differ dramatically depending on the type of PPAR agonist. Marked alteration of morphological characteristics of brown adipose tissue (BAT) treated with GI259578A, a PPAR-α agonist, was observed in AKR/J (AKR) obese mice. Furthermore, there was a 22.8% decrease in the mean size of adipocytes in white adipose tissue (WAT) compared with vehicle. In diabetic db/db mice, the PPAR-γ agonist GW347845X decreased the mean size of adipocytes in WAT by 15.4% compared with vehicle. In contrast to changes in WAT, GW347845X increased the mean size of adipocytes in BAT greatly by 96.1% compared with vehicle. These findings suggest that GI259578A may activate fatty acid oxidation in BAT and that GW347845X may cause adipocyte differentiation in WAT and enhancement of lipid storage in BAT.