Upon careful analysis and comparison between
Upon careful analysis and comparison between our study  and those from Gutierrez et al. and Chmelar et al. [4,7], we noticed dietary differences among these studies, in addition to the fact that we used Kit-dependent MC-deficient Kit mice and Gutierrez et al. and Chmelar et al. used Kit-independent Cpa3 and Mcpt5-Cre R-DTA mice. In our study, we fed mice a high-cholesterol Western diet (Cat# D12108C, Research Diets, Inc., New Brunswick, NJ), whereas all mice in Gutierrez's and Chmelar's studies consumed a HFD with 60% kcal% fat (Cat# D12942, Research Diets, Inc.) that contained no cholesterol (Table 1). The dietary cholesterol content differences might have yielded the conflicting obesity and diabetes phenotypes among these different studies [1,4,7]. Several lines of unexplained observations support this hypothesis. First, if reduced obesity and diabetes in Kit mice from our prior study did not result from the loss of MCs but from alterations in other immune Loxapine Succinate [2,3], we cannot explain the restored obesity and diabetes phenotypes in Kit-recipient mice after adoptive transfer of donor MCs from WT mice . Although we detected only partial phenotype recovery in WT MC-reconstituted Kit mice, this can be explained by the partial recovery of adipose tissue MCs after donor MC engraftment . Criticism of donor MC reconstitution reports that it may show potential differences in both numbers and anatomical distribution in recipient mice. Therefore, donor MCs may never act the same as native endogenous MCs [4,8]. Second, we showed that MC inhibition with either DSCG or ketotifen reduced obesity and insulin resistance in Western diet-fed WT mice . Yet, using the same doses of DSCG, Gutierrez et al. did not show any inhibitory activity of DSCG in HFD-induced body weight gain from WT mice (Fig. 2R, ref. ). These two experiments used the same mice, same mast cell inhibitor, and same dose, but yielded different conclusions.
All these prior observations point to a possibility that dietary cholesterol may have made the metabolic phenotype discrepancies among these different studies [1,4,7]. In this study, we induced obesity and insulin resistance in 6-week-old male WT mice with a cholesterol-free HFD, a high-cholesterol (1.25%) Western diet, or the same HFD but supplemented with 1.25% cholesterol (HFD + Cho, Cat# D08063002, Research Diets, Inc.) (Table 1), while mice received the same doses of DSCG and ketotifen as previously used by us and Gutierrez et al. [1,4].
Materials and methods
Discussion Published studies have consistently demonstrated increased MCs in adipose tissue from obese humans and mice [1,27,28]. Such accumulation of MCs in adipose tissue may not just serve as a signature of tissue inflammation. MCs are known to interact with macrophages [1,29], regulatory T cells (Tregs) [30,31], and group 2 innate lymphoid cells (ILC2) , suggesting other potential MC effects through cross-interactions with these immune cells when considering obesity and diabetes [, , ]. We report here that mice that consumed a cholesterol-free HFD demonstrated much weaker activity in local (adipose tissue) and systemic MC activation than those that consumed a high-cholesterol Western diet or a high cholesterol-supplemented HFD + Cho diet (Fig. 5A–G). We demonstrated that the plasma from mice that consumed a Western diet or the HFD + Cho diet was much more potent in MC activation than the plasma from HFD-fed mice (Fig. 6A). Such differences in dietary cholesterol intake may explain why MC-deficiency or MC inhibition by DSCG did not affect HFD-induced obesity and insulin resistance in some studies [4,7]. Low degree of MC activation may cause minimal contribution of MCs to HFD-induced obesity and diabetes. Diet-induced obesity and diabetes in HFD-fed mice may be MC-independent. In contrast, high cholesterol diet-fed mice showed much high levels of MC activation. The high activity of plasma from these mice in MC activation may affect the pathobiology of other immune cells, such as macrophages, Tregs, and ILC2 [1,, , , , , , ]. Therefore, MC function in these mice may become significant.