3513659558

104

3.5.7 Statistical analysis

Data are expressed as mean ± SEM. Statistical differences between groups were further evaluated by Student T-test or two-way ANOVA with Bonferroni post test using GraphPad Prism 5.0 Software. Differences were considered significant at P<0.05.

3.6 Results

3.6.1 Leptinemia and endochondral ossification

Since our previous results demonstrated lower body weight and a global reduction of visceral adipose tissue (Kevorkova et al2013) - the primary source of leptin (Park et Ahima, 2014), - we measured basal unfasted leptinemia in mice of both genotypes. As shown in Fig. 3.1 A, plasma leptin was constantly lower in Cd36-nuli mice compared to corresponding WT mice aged up to 6 months. Adiponectin levels were unaffected, therefore causing a significant decrease in the leptin-to-adiponectin ratio (Fig. 3.IB, C). Leptin levels were positively correlated with total weight in WT mice; this correlation was very weak in younger Cd36-nuli mice, but stronger at 6 months ofage (Fig. 3.ID, E).

Lnterestingly, we measured a 2-fold decrease in Lep gene expression in bonę marrow derived MSCs from the nuli mice compared with their WT counterparts (Fig. 3.2A). Leptin has been postulated to be a modulator of endochondral ossification (Kurne et al., 2002), which led us to further investigate skeletal development in nuli mice. It was found that the expression of the chondrogenic transcription factor Sox9 was increased in Cd36-nuli MSCs (Fig 3.2B); however, the expression of the chondrocyte marker Col-IIa by nuli cells was similar to cells from WT mice (Fig. 3.2C). As