We read with great interest the article by Kremer et al., 1 on the role of interleukin-12 (IL-12) production by Kupffer cells in fatty liver, and its possible impact in the reduction of hepatic resident natural killer T (NKT) cells using an animal model of hepatosteatosis (mice fed choline-deficient diet for 0-20 weeks). As previously demonstrated, 2 in this study the authors found that hepatosteatosis in mice that were fed a chronic choline-deficient diet is characterized at the hepatic level by the induction of interferon-gamma and tumor necrosis factor-alpha, elevated content of IL-12, and by a remarkable reduction in the number and activity of NKT cells. Interestingly, all these effects are undetectable in IL-12–depleted mice fed a choline-deficient diet, even if fatty liver is equally present. Several studies 2- 4 conducted on different metabolic animal models of hepatic steatosis have reported the relationship between lipid accumulation, increased Th1 cytokine production (i.e., tumor necrosis factor, IL-12, and interferon-gamma), and hepatic NKT cell depletion. IL-12 is well known as a NKT cell inductor able to stimulate the production/release of large amounts of interferon-gamma and to activate specific transcription factors, including signal transducer and activator of transcription 4 (STAT4). 5 However, the IL-12 increase may not be the sole factor involved in death-dependent NKT cell depletion: others factors, such as dietary factors, might interfere, for example, with mechanisms that mediate the hepatic homing and apoptosis of NKT cells. 6 All these findings noticeably demonstrated that NKT cell reduction is an effect of a combination between intrahepatic fat accumulation and IL-12 increase rather than a cause of hepatosteatosis. However, the real unresolved question is the connection between intrahepatic fat accumulation and up-regulated hepatic IL-12 messenger RNA levels. Once again, as demonstrated by Kremer et al., 1 and as anticipated by other studies, 7 the activation of Kupffer cells by an endotoxin-mediated mechanism could be the link between fatty liver, inflammatory response, and a reduction of NKT cell population. Noteworthy, this phenomenon could be either an effect of fatty liver or an early signal for the development of fibrosis. 8 Therefore, it might be very interesting to investigate whether the number of NKT cells may be a predictive marker of liver fibrosis. Furthermore, research favors the hypothesis of the role of endotoxin and the toll like receptor-4 in diet-induced steatohepatitis. 9 Yet, we would emphasize how many points are still obscure on the molecular mechanisms regulating this intricate network of interactions between cells of the immune system and hepatocellular damage. Anna Alisi Ph.D.*, Nadia Panera*, Valerio Nobili M.D.*, * Liver Unit, Bambino Gesù Children's Hospital and Research Institute, Rome, Italy.