Characterisation of Arcuate Nucleus Projections to the Paraventricular Nucleus of Thalamus

The metabolic state (fed or hungry) has been shown to influence varying responses to food-associated cues. Recent studies reveal that the paraventricular nucleus of the thalamus (PVT) may play a critical role as it receives inputs from the hypothalamic arcuate nucleus (ARC), a major regulator of appetite and energy homeostasis. The PVT also relays outputs to the nucleus accumbens, known to mediate behavioural responses to reward-associated cues. The ARC receives information about metabolic state via hormones from the periphery, but it is unknown if this information is transmitted to the PVT. This study investigated leptin, a key metabolic hormone penetrating the ARC, which signals the brain to modulate food intake. Produced in adipose tissue, leptin suppresses food intake and stimulates metabolic processes to dissipate excessive energy stores; with intrinsic levels being low in hungry states. However, it is not known how or if the leptin signal in ARC is transmitted to the PVT. To address this question, I investigated whether ARC neurons projecting to the PVT respond to leptin. Subsets of this goal included mapping the location of these neurons in the ARC; and evaluating their response to leptin. I hypothesized that these ARC neurons express leptin markers. For this purpose, male Wistar rats (n = 11) were injected with a GFP (green fluorescent protein)-expressing retrograde viral vector (AAVrg-Syn-ChR2(H134R)-GFP) in the PVT. A subset of rats (n = 7) were later maintained on food restriction and injected with leptin (80µg/kg s.c.) 90 minutes before perfusion. Using immunohistochemical staining for GFP, coronal brain sections (20/rat; 40µm thickness) through the extent of the ARC were examined and every retrogradely-labelled GFP cell quantified, mapped and correlated with the injection spread in the PVT. In leptin-injected rats, double immunohistochemistry was performed for GFP and pSTAT3 (phosphorylated signal transducer and activator of transcription 3), a downstream signalling marker for leptin. Results showed that in 9 rats, the injection site clearly involved PVT, and this correlated with the presence of GFP-labelled neurons in the ARC. Furthermore, these GFP-labelled, PVT-projecting neurons (n = 150) were mostly located in the medial (n = 74) and lateral ARC (n = 61) with some cells located in the dorsal ARC (n = 15). Interestingly, in leptin-treated rats co-localization of GFP cell bodies with pSTAT3-labelled nuclei occurred in ~one-third of GFP-labelled ARC neurons (n = 21/64), indicating that a sizable proportion of ARC neurons with projections to PVT respond to the key metabolic hormone leptin, thus confirming my hypothesis. These co-labelled neurons were located only in the medial (n = 15) and lateral ARC (n = 6), completely excluding the dorsal ARC. This distribution is consistent with that of ARC neurons expressing the feeding-related neuropeptides, NPY/AgRP (neuropeptide Y and agouti-related peptide) and POMC/CART (proopiomelanocortin and cocaine-and-amphetamine-regulated transcript). This ARC-PVT pathway is potentially important for regulating responses to food cues according to metabolic state. Future studies are needed to determine if neurons in the PVT that receive this input are involved in integrating it with reward signal information