Hypothalamic AMPKα2 regulates liver energy metabolism in rainbow trout through vagal innervation.

Hypothalamic AMPK plays a major role in the regulation of whole body metabolism and energy balance. Present evidence has demonstrated that this canonical mechanism is evolutionarily conserved. Thus, recent data demonstrated that inhibition of AMPKα2 in fish hypothalamus led to decreased food intake and liver capacity to use and synthesize glucose, lipids, and amino acids. We hypothesize that a signal of abundance of nutrients from the hypothalamus controls hepatic metabolism. The vagus nerve is the most important link between the brain and the liver. We therefore examined in the present study whether surgical transection of the vagus nerve in rainbow trout is sufficient to alter the effect in liver of central inhibition of AMPKα2. Thus, we vagotomized (VGX) or not (Sham) rainbow trout and then intracerebroventricularly administered adenoviral vectors tagged with green fluorescent protein alone or linked to a dominant negative isoform of AMPKα2. The inhibition of AMPKα2 led to reduced food intake in parallel with changes in the mRNA abundance of hypothalamic neuropeptides [neuropeptide Y (npy), agouti-related protein 1 (agrp1), and cocaine- and amphetamine-related transcript (cartpt)] involved in food intake regulation. Central inhibition of AMPKα2 resulted in the liver having decreased capacity to use and synthesize glucose, lipids, and amino acids. Notably, these effects mostly disappeared in VGX fish. These results support the idea that autonomic nervous system actions mediate the actions of hypothalamic AMPKα2 on liver metabolism. Importantly, this evidence indicates that the well-established role of hypothalamic AMPK in energy balance is a canonical evolutionarily preserved mechanism that is also present in the fish lineage.

Na+/K+-ATPase is involved in the regulation of food intake in rainbow trout but apparently not through brain glucosensing mechanisms.

To assess the hypothesis that Na+/K+-ATPase (NKA) is involved in the central regulation of food intake in fish, we observed in a first experiment with rainbow trout (Oncorhynchus mykiss) that intracerebroventricular (ICV) treatment with ouabain decreased food intake. We hypothesized that this effect relates to modulation of glucosensing mechanisms in brain areas (hypothalamus, hindbrain, and telencephalon) involved in food intake control. Therefore, we evaluated in a second experiment, the effect of ICV administration of ouabain, in the absence or in the presence of glucose, on NKA activity, mRNA abundance of different NKA subunits, parameters related to glucosensing, transcription factors, and appetite-related neuropeptides in brain areas involved in the control of food intake. NKA activity and mRNA abundance of nkaα1a and nkaα1c in brain were inhibited by ouabain treatment and partially by glucose. The anorectic effect of ouabain is opposed to the orexigenic effect reported in mammals. The difference might relate to the activity of glucosensing as well as downstream mechanisms involved in food intake regulation. Ouabain inhibited glucosensing mechanisms, which were activated by glucose in hypothalamus and telencephalon. Transcription factors and neuropeptides displayed responses comparable to those elicited by glucose when ouabain was administered alone, but not when glucose and ouabain were administered simultaneously. Ouabain might therefore affect other processes, besides glucosensing mechanisms, generating changes in membrane potential and/or intracellular pathways finally modulating transcription factors and neuropeptide mRNA abundance leading to modified food intake.

Differential Role of Hypothalamic AMPKα Isoforms in Fish: an Evolutive Perspective.

In mammals, hypothalamic AMP-activated protein kinase (AMPK) α1 and α2 isoforms mainly relate to regulation of thermogenesis/liver metabolism and food intake, respectively. Since both isoforms are present in fish, which do not thermoregulate, we assessed their role(s) in hypothalamus regarding control of food intake and energy homeostasis. Since many fish species are carnivorous and mostly mammals are omnivorous, assessing if the role of hypothalamic AMPK is different is also an open question. Using the rainbow trout as a fish model, we first observed that food deprivation for 5 days did not significantly increase phosphorylation status of AMPKα in hypothalamus. Then, we administered adenoviral vectors that express dominant negative (DN) AMPKα1 or AMPKα2 isoforms. The inhibition of AMPKα2 (but not AMPKα1) led to decreased food intake. The central inhibition of AMPKα2 resulted in liver with decreased capacity of use and synthesis of glucose, lipids, and amino acids suggesting that a signal of nutrient abundance flows from hypothalamus to the liver, thus suggesting a role for central AMPKα2 in the regulation of peripheral metabolism in fishes. The central inhibition of AMPKα1 induced comparable changes in liver metabolism though at a lower extent. From an evolutionary point of view, it is of interest that the function of central AMPKα2 remained similar throughout the vertebrate lineage. In contrast, the function of central AMPKα1 in fish relates to modulation of liver metabolism whereas in mammals modulates not only liver metabolism but also brown adipose tissue and thermogenesis.

Changes in the levels and phosphorylation status of Akt, AMPK, CREB and FoxO1 in hypothalamus of rainbow trout under conditions of enhanced glucosensing activity.

There is no available information about mechanisms linking glucosensing activation in fish and changes in the expression of brain neuropeptides controlling food intake. Therefore, we assessed in rainbow trout hypothalamus the effects of raised levels of glucose on the levels and phosphorylation status of two transcription factors, FoxO1 and CREB, possibly involved in linking these processes. We also aimed to assess the changes in the levels and phosphorylation status of two proteins possibly involved in the modulation of these transcription factors: Akt and AMPK. Therefore, in pooled preparations of hypothalamus incubated for 3 and 6 h in the presence of 2, 4 or 8 mmol l-1 d-glucose, we evaluated the response of parameters related to glucosensing mechanisms, neuropeptide expression and levels and phosphorylation status of the proteins of interest. The activation of hypothalamic glucosensing systems and the concomitant enhanced anorectic potential occurred in parallel with activation of Akt and inhibition of AMPK. The changes in these proteins relate to neuropeptide expression through changes in the level and phosphorylation status of transcription factors under their control, such as CREB and FoxO1, which displayed inhibitory (CREB) or activatory (FoxO1) responses to increased glucose.

In vitro evidence supports the presence of glucokinase-independent glucosensing mechanisms in hypothalamus and hindbrain of rainbow trout.

We previously obtained evidence in rainbow trout for the presence and response to changes in circulating levels of glucose (induced by intraperitoneal hypoglycaemic and hyperglycaemic treatments) of glucosensing mechanisms based on liver X receptor (LXR), mitochondrial production of reactive oxygen species (ROS) leading to increased expression of uncoupling protein 2 (UCP2), and sweet taste receptor in the hypothalamus, and on sodium/glucose co-transporter 1 (SGLT-1) in hindbrain. However, these effects of glucose might be indirect. Therefore, we evaluated the response of parameters related to these glucosensing mechanisms in a first experiment using pooled sections of hypothalamus and hindbrain incubated for 6 h at 15°C in modified Hanks' medium containing 2, 4 or 8 mmol l(-1) d-glucose. The responses observed in some cases were consistent with glucosensing capacity. In a second experiment, pooled sections of hypothalamus and hindbrain were incubated for 6 h at 15°C in modified Hanks' medium with 8 mmol l(-1) d-glucose alone (control) or containing 1 mmol l(-1) phloridzin (SGLT-1 antagonist), 20 µmol l(-1) genipin (UCP2 inhibitor), 1 µmol l(-1) trolox (ROS scavenger), 100 µmol l(-1) bezafibrate (T1R3 inhibitor) and 50 µmol l(-1) geranyl-geranyl pyrophosphate (LXR inhibitor). The response observed in the presence of these specific inhibitors/antagonists further supports the proposal that critical components of the different glucosensing mechanisms are functioning in rainbow trout hypothalamus and hindbrain.