Regulation of Hepatic Gluconeogenesis by the CREB:TORC2 Pathway

点击翻译按钮获取中文摘要

Abstract Under fasting conditions, increases in circulating glucagon stimulate hepatic glucose production via induction ofthe cAMP pathway. Conversely, increases in gut-derived glucagon-like peptide 1 (GLP1) during feeding enhance glucose clearance by promoting insulin release. The transcription factor CREB is thought to mediate long term effects of both peptide hormones, following its phosphorylation by PKA and association with CBP/P300. The transcriptional response to cAMP follows burst-attenuation kinetics; CREB activity peaks after 1 hour of stimulation, returning to baseline after 4-6 hours. In addition to their effects on CREB phosphorylation, glucagon and GLP1 also increase CREB activity by stimulating its association with the cAMP Regulated Transcriptional Coactivators (CRTCs/TORCs), latent cytoplasmic CREB cofactors that translocate to the nucleus following their dephosphorylation in response to cAMP. CRTC1 is expressed only in brain, while CRTC2 and CRTC3 are co-expressed in most tissues. The extent to which CRTC2 and CRTC3 function on overlapping or distinct subsets of CREB target genes is unclear, however. In the previous grant period, we showed that the CREB/CRTC2 pathway contributes importantly to fasting glucose production; acute depletion of CRTC2 in liver substantially lowers blood glucose concentrations and gluconeogenic gene expression, while over-expression of wild-type and to a greater extent phosphorylation-defective CRTC2 increases gluconeogenesis. By contrast with effects of acute hepatic CRTC2 knockdown, mice with a whole-body knockout of CRTC2 show only modest reductions in fasting glucose levels; and they develop an insulin secretion defect as they age. These results point to the involvement of additional CREB coactivators that compensate for loss of CRTC2 in liver, and they suggest that CRTC2 expression in pancreatic islets also modulates circulating glucose concentrations through its effects on insulin secretion. Supporting the latter, MafA, a beta cell transcription factor that is required for insulin secretion, is strongly upregulated by CREB and CRTC2. Proposed studies during the upcoming grant period focus on the hypothesis that members ofthe CRTC family exert overlapping effects on CREB activity. The importance of a newly identified CREB interacting protein in potentiating CREB activity and compensating for loss of CRTC2 in CRTC2 mutant mice will be tested. Finally the role of a potent CREB inhibitor, which is upregulated in pancreatic islets under hyperglycemic conditions, in promoting resistance to Gs-coupled receptor signaling, will be evaluated. Three aims are proposed; they extend the previous work by addressing the mechanisms by which the CREB pathway promotes gluconeogenesis in liver and facilitates insulin secretion from pancreatic islets. In Aim 1, we will use mice with floxed alleles of CRTC2 and CRTC3 to evaluate the relative roles of these coactivators in modulating hepatic gluconeogenesis and insulin secretion. We will generate mice with tissue ' specific knockouts of CRTC2 and CRTC3 in liver or pancreatic islets. Do CRTC2 and CRTC3 exert overlapping effects on gluconeogenic gene expression in liver? Do they promote insulin secretion by upregulating the leucine zipper factor MafA? In Aim 2, we will test the role of BRD2-a bromodomain protein identified in a proteomic screen for CREB associated proteins- in stimulating expression of gluconeogenic genes. We will characterize domains in BRD2 and CREB that mediate this interaction; and the role of CREB acetylation in modulating the BRD2:CREB association will also be tested. We will evaluate whether inhibition of BRD2, through administration of a selective bromodomain inhibitor, improves glucose levels in the setting of insulin resistance. In Aim 3, we will examine the mechanism by which CREB target gene expression in pancreatic islets is down-regulated in insulin resistance. In particular, we will investigate the role of Protein Kinase Inhibitor beta (PKIB) in interfering with GLP1 and other hormones, following its upregulation in response to hyperglycemia: PKIB knockout mice will be used to determine whether depletion of this inhibitor improves pancreatic islet function in the setting of insulin resistance. Taken together, the proposed studies will provide new insight into mechanisms by which glucagon and GLP1 promote glucose balance through their effects on the CREB pathway in liver and pancreatic beta cells. Progress Report Specific Aim 1: We will examine the role ofthe histone acetyl-transferases PSOO and CBP (P300/CBP) and the NAD+ dependent deacetylase SIRTI in modulating T0RC2 activity through acetylation and deacetylation during fasting. We will identify residues in T0RC2 (CRTC2) that undergo acetylation, and we will test the importance of P300/CBP and SIRT1 in this process by hepatic over-expression or depletion of each protein. The importance of acetylation in augmenting T0RC2 activity through protein stabilization will also be addressed. We found that CBP/P300 enhanced TORC2/CRTC2 activity in part by promoting CRTC2 acetylation at Lys628, a residue that is well conserved amongst CRTC family members (1). Conversely, SirTI was found to inhibit CRTC2 activity in part through deacetylation of CRTC2 at Lys628. Lys628 appears to regulate CRTC2 stability. Following its nuclear entry in response to cAMP signaling, CRTC2 undergoes mono-ubiquitination at Lys628 via an association with C O P I , the adaptor component of a Cul4A E3 ligase. When it exits the nucleus, mono-ubiquitinated CRTC2 undergoes poly-ubiquitination and proteasome-mediated degradation. In addition to their effects on CRTC2, CBP and SirTI also appear to regulate CREB target genes through acetylation /deacetylation of CREB itself at Lys136 (2). To explore the underlying mechanism, we have developed an acetyl-lys136 specific CREB antiserum. CREB acetylation is constitutively upregulated in SirTI-/- mouse embryo fibroblasts and it is absent in CBP-/-, P300-/- mutant cells. In Aim 2 ofthe proposed studies, we will address the potential role of the bromodomain coactivator BRD2 in associating with and potentiating the activity of Lysl 36-acetylated CREB. Specific Aim 2: We will investigate the role of Salt Inducible Kinases (SIKs) in modulating hepatic T0RC2 (CRTC2) activity by phosphorylating P300/CBP and inhibiting their association with T0RC2. We will identify SIK2 phosphorylation sites in P300/CBP, and we will test the importance of SIK2 in catalyzing P300/CBP phosphorylation by over-expression or RNAi mediated depletion of SIK2 in liver. The role of P300/CBP phosphorylation in modulating hepatic T0RC2 activity will also be determined using phosphorylation-defective PSOO mtitant proteins. The potential role of 14-3-3 proteins in binding to phosphorylated PSO0/CBP and thereby disrupting the TORC2:P300/CBP interaction will also be examined. SIK2 was found to phosphorylate CBP and PSOO at Ser89 (in PSOO) (1). In turn Ser89 phosphorylation reduced CBP/PSOO activity over CREB target genes in hepatocytes exposed to glucagon. Conversely, Ser89Ala mutant PSOO was more active than wild-type PSOO in supporting CREB dependent transcription. Although Ser89 forms part of a consensus motif for 14-S-S binding, we were unable to detect any association of either CBP or PSOO with 14-3-S proteins in hepatocytes or other cells. In addition to their effects on CBP/PSOO, we found that SIKs also regulate the activities of class Ila HDACs in liver (3, 4). Class Ila HDACs are sequestered in the nucleus under basal conditions through phosphorylation dependent interactions with 14-S-3 proteins; and they move to the nucleus in response to cAMP agonist, when SIKS are inhibited by PKA-dependent phosphorylation. Although first identified in skeletal muscle, cAMP stimulates the translocation of Class Ila HDACs in most cell types, including liver. Indeed, increases in nuclear class Ila HDACs during fasting appears to promote hepatic gluconeogenesis, in part through the de-acetylation of F O X 0 1 . In Aim 1 of the proposed studies, we will investigate whether the class Ila HDAC pathway compensates for loss of CRTC2 and CRTCS expression in liver. Specific Aim 3: We will investigate the role of T0RC2 (CRTC2) in triggering the gluconeogenic program during fasting through its association with a histone methyl transferase (HMT) complex. We will monitor histone methylation over gluconeogenic genes, and we will evaluate the importance of T0RC2 in mediating this process through RNAi mediated depletion or over-expression of mutant TORC2 proteins that are defective in the HMT interaction. The potential role of HMTs in modulating gluconeogenic gene expression by methylating T0RC2 will also be investigated. We found that CRTC2/TORC2 associates with WDR5, a core component of histone methyl transferase (HMT) complexes (5). RNAi-mediated knockdown of WDR5 reduces CRTC2 activity over gluconeogenic genes in cells exposed to glucagon; but knockdown of other core components such as Ash2l and RbbP4 had no effect, on gluconeogenic gene expression, despite substantial reductions in HSK4 tri-methylation in Ash2l or RbBp4 depleted cells. Rather we found that WDR5 regulates gluconeogenic gene expression by modulating the activity of histone acetyl transferase complexes, which contain the paralogs GCN5 (KAT2A) and PCAF (KAT2B). Indeed, KAT2A/B associate directly with the trans-activation domain (TAD) of CRTC2 (5); and mutations'in