Gnasxl, an alternative transcript of the Gnas locus, encodes an NH2-terminal variant (XLαs) of the trimeric G-protein subunit Gαs. Both proteins stimulate cAMP signalling. Gnasxl transcription is epigenetically regulated by genomic imprinting. Mice deficient for Gnasxl show a postnatal phenotype of undernutrition and growth retardation, while adults are healthy, but remain lean and hypermetabolic (1, 2). Increased expression of lipolytic genes in adipose tissues, as well as urinary catecholamine levels, indicate an increased sympathetic nervous system activity (2), which also causes higher blood pressure and heart rate in Gnasxl knock-outs (KO) (3). Gnasxl (XLαs) is expressed in regions of the postnatal and adult brain that regulate energy homeostasis and sympathetic outflow, including hypothalamic nuclei (Arc, PVH, DMH, LH), the suprachiasmatic nucleus and the preoptic area (4). In the medulla the protein is found in neurons of the raphe pallidus, NTS and ventrolateral medulla.To investigate changes in gene expression, we undertook a RNAseq analysis (Illumina) of total RNA from adult hypothalami. RNAseq data were verified by qRT-PCR and immunohistochemistry (IHC). Developmental histological analyses were performed at postnatal stages P1, P5, P10 and P15.We identified Glial Fibrillary Acidic Protein (GFAP) as a transcript that is downregulated in Gnasxl KO hypothalami (RNA seq: 2.1 fold down, p<0.0001), qRT-PCR: 2.2 fold down, p<0.01, t-test, n=6 WT, 6 KO). IHC of adult hypothalamus sections revealed a 43% reduction of GFAP-positive cells (201+16 vs 353+22 average cell number/section + S.E.M, KO vs WT, p<0.0001 t-test, n=23 sections from 4 WT and 4 KO). This reduction in GFAP-positive cells included astrocytes as well as subpopulations of ependymal tanycytes. Since Gnasxl is not expressed in glial cells, we investigated whether these changes might develop as a consequence of the early postnatal failure-to-thrive phenotype of Gnasxl KO mice. While glial cell counts of P1 and P5 samples showed no difference, a reduction became apparent on P10 (100+4 vs 125+5 average cell number/section + S.E.M, KO vs WT, p<0.001 t-test, n=21 sections from 2 WT and 2 KO) and P15 (86+5 vs 109+5, KO vs WT, p=0.001 t-test, n=37 sections from 2 WT and 2 KO). Additionally to changes in GFAP expression, we found a 2-fold increase in expression of the signalling form of the Leptin receptor (Lepr-b) by qRT-PCR (p<0.05 t-test, n=6 WT, 6 KO).Our findings of reduced GFAP expression and glial cell numbers hint at an abnormal postnatal hypothalamic development in lean Gnasxl KO mice. The data contrast with descriptions in obese mice of gliosis and elevated GFAP levels. Further work is aimed at clarifying whether lack of XLαs leads to developmental defects in the organisation of hypothalamic neural/glial circuits.