Ectopic lipid accumulation, notably in the liver, is a key feature of metabolic disease in the scope of obesity. Hepatic steatosis may be both a consequence and a cause of insulin resistance¹. For example high-fat diet (HFD)-induced steatosis was shown to lead to hepatic insulin resistance without perturbations in peripheral metabolism². On the other hand, it is well known that insulin resistance causes hepatic lipid accumulation due to impaired insulin action in adipocytes, namely in suppressing lipolysis¹. To help clarify the role of hepatic lipid alterations in the development of metabolic dysregulation in diet-induced obesity, we performed a longitudinal study where hepatic lipids were non-invasively monitored by ¹H-MRS in vivo, in HFD-fed mice whose glucose homeostasis was also characterized.11-week old C57BL/6J mice were fed either a high-fat diet (60% kcal from fat, Research Diets D12492), HFD (N=10), or a control diet (10% kcal from fat, Research Diets D12450B), CTRL (N=9) and monitored as depicted in figure 1. Hepatic lipid content (HLC) and composition were estimated as previously described³, by localized ¹H-MRS in vivo with STEAM, in a 14.1T-26cm magnet interfaced to a Direct Drive console (VnmrJ, Agilent Technologies) using a ¹H quadrature surface coil as trans-receiver. All MR acquisitions were respiration-triggered. HLC was expressed as a percentage of total ¹H signal in spectra acquired without water-suppression. Contributions of saturated- monounsaturated- and polyunsaturated fatty-acids (SFA, MUFA and PUFA) were resolved as shown in figure 2, using the fatty-acid composition indices derived from water-suppressed spectra³. Oral glucose tolerance tests (OGTT) were performed after a 6h-fasting and glycemia was monitored from tail tip samples before the glucose gavage (1.5 g/kg) and up to 2h afterwards. Fasting insulin was measured by ELISA immunoassay. Data are means ± SEM and were analyzed with 2-way ANOVA followed by Student t-test for specific comparisons.Both experimental groups gained weight throughout the study (Figure 3). Mice in HFD group developed obesity with a 72 ± 4% increase in body weight by week 18 of the dietary regimen. In comparison, the weight gain for CTRL group was four times smaller: 18 ± 3%. Increased body weight in the HFD group, relative to CTRL, was observed as little as one week after the dietary switch (P<0.01). Also within a week, HFD-mice showed higher AUC during the OGTT, relative to controls (P<0.01). Dysregulation of glucose homeostasis was established by week 4 with increased AUC during the OGTT (P<0.05), 2h-post load glycemia (P<0.05), 6h-fasting glycemia (P<0.05) and 6h-fasting insulinemia (P<0.01). These alterations preceded hepatic lipid accumulation that was only observed by week 9, when total HLC was 3.4 ± 0.5% in HFD and 1.2 ± 0.1% in CTRL groups (P<0.001). Such lipid accumulation resulted from increased contributions from both SFA (P<0.01) and PUFA (P=0.01) in the HFD group relative to CTRL, while MUFA contribution was similar between the two groups (Figure 4). Interestingly, already by week 4, despite the similar total HLC observed between the two groups, significant alterations were detected in the contributions of SFA (increased, P<0.01), MUFA (decreased, P<0.01), and PUFA (increased, P<0.01), relative to controls. By week 18, HLC ranged from 3.4 to 27.4% in the HFD group. Overnight-fasting induced in a 3-fold increase in HLC in CTRL, driven by accumulation of both SFA and PUFA (P<0.001 vs ad libitum conditions). On the other hand, no changes were observed in HLC after an overnight-fast in HFD-mice although a small decrease in MUFA contribution was noted (P<0.05 vs ad libitum conditions).In our experimental setting, glucose intolerance appeared as a very early metabolic abnormality in diet-induced obesity. Significant alterations in the contributions of SFA, MUFA and PUFA to HLC were detected within a month of HFD feeding before frank hepatic lipid accumulation took place. Namely, PUFA were the major contributors to increases in HLC in HFD-fed mice. Since those species are more readily mobilized from adipose stores when compared to SFA or MUFA⁴, our finding likely reflects inappropriate inhibition of adipose tissue lipolysis due to early perturbations in insulin action. Indeed, contributions of PUFA, MUFA and SFA to liver lipids resembled those found in fasted mice fed a control diet, consistent impaired insulin action. Moreover, aggravation of insulin resistance, denoted by increased fasting glucose and insulin levels, was observed in parallel with hepatic lipid accumulation, stressing the link between the two events. We conclude that under HFD-feeding initial defects in adipose tissue lipid storage (notably increased lipolysis) lead to hepatic steatosis that itself contributes to aggravate the loss of glucose homeostasis, namely by enhancing insulin resistance.