Evidence That Species-Specific Postnatal Regulation of the Human Growth Hormone Gene Is Linked to E-Box Sequences and Flexibility of an Intrachromosomal Loop
Presentation Number: SAT 427
Date of Presentation: April 1st, 2017
Hana Vakili*, Yan Jin and Peter A Cattini
University of Manitoba, Winnipeg, MB, Canada
Prenatal activation and somatotroph-specific expression of the human (h) growth hormone (GH) gene (hGH1) was defined in transgenic mice in vivo. Activation is linked to physical interaction between pituitary-specific transcription factor Pit-1 sites located 14.5 kb upstream of hGH1 in the CD79b gene locus, and those in the hGH1 proximal promoter region. This interaction linking multiple distant Pit-1 binding sites is essential for hGH1 activation, and results in a long-range intrachromosomal loop of the intervening sequences that is detected by the chromosome conformation capture (3C) assay. We also showed that this hGH loop is a dynamic structure and under postnatal regulatory control. A high-fat diet (HFD) of 60 kcal% for 3 days with accompanying hyperinsulinemia was associated with loop disruption and decreased rhythmic daily hGH1 RNA levels, as well as reduced Pit-1 binding. This disruption was also linked to an enhancer motif (E-box) element in the hGH loop sequences, capable of binding the circadian factors Bmal1/Clock and responding to insulin levels in vitro and in vivo.
Like hGH1, the mouse GH gene (mGH) is also activated in developing somatotrophs, but unlike hGH1, mGH expression is not similarly rhythmic or affected by a HFD. However, the presence and importance of the long-range loop in the mGH locus has not been reported. Thus, we pursued the possibility that generation of a long-range loop for mGH activation can be uncoupled from postnatal loop disruption/reformation that might allow for differential regulatory control of hGH1 and mGH.
Sequence analysis revealed three Pit-1 binding sites ~16 kb upstream of the mGH promoter in addition to the two conserved Pit-1 binding sites in the proximal promoter region. A physical interaction between these two clusters of Pit-1 sites, consistent with looping out of intervening mouse sequences, was detected by 3C assay. However, the mGH loop is resistant to disruption by excess caloric intake, and does not contain the E-box element previously characterized in the equivalent hGH loop sequences. This was further supported by lack of rhythmicity in mGH RNA levels as observed for hGH1. Unlike negative regulation of hGH1expression by a HFD, mGH RNA levels as well as Pit-1 association was unaffected, thus consistent with the presence of an intact mGH long-range loop.
Our observations support a role for a long-range intrachromosomal loop in the activation of both hGH1 and mGH during development. However, the hGH and mGH loops and expression respond differently to excess caloric intake. This suggests the potential for distinct postnatnal control of hGH1 and mGH related to different “regulatory sequences” associated with their “chromatin loops”. These differences can potentially contribute to divergent responses to a HFD challenge and circadian regulation, and highlights the difficulty with extrapolating data from mGH (rodents) to hGH (primates).
Nothing to Disclose: HV, YJ, PAC