A New Activator of Hepatocyte CaMKII in Fasting and Type 2 Diabetes

Positive regulation of hepatic glucose production (HGP) and stimulation of glycogenolysis during fasting are critical processes to prevent hypoglycemia. A key hormone that carries out this function is glucagon, which is secreted by pancreatic α-cells in response to fasting and acts on its receptor on hepatocytes to promote HGP and glycogenolysis (1). However, as often happens, processes that are critical in physiology can promote disease processes in nonphysiological settings. As such, glucagon-induced HGP is a major contributor to excessive HGP and hyperglycemia in obesity and type 2 diabetes (T2D) (2). In this issue of Diabetes, Wang et al. (3) provide evidence that another circulating factor that is increased in fasting and obesity/T2D—prostaglandin F_2α (PGF_2α)—also promotes HGP in both fasting and obese, insulin-resistant mice.

Previous work has elucidated a number of pathways through which activation of the hepatocyte glucagon receptor (GcgR), which is a G-protein–coupled receptor (GPCR), promotes HGP and glycogenolysis in both fasting and obesity. Specifically, adenylyl cyclase/cyclic AMP-mediated protein kinase A (PKA) activation by G proteins of the GcgR controls the transcriptional regulation of HGP and glycogenolysis through at least three distinct but complementary processes (Fig. 1): phosphorylation of CREB, which results in recruitment of its coactivators, CBP and p300; deactivation of a kinase called SIK2; and activation of inositol 3-phosphate receptors (IP3Rs), leading to release of endoplasmic reticulum calcium into the cytosol and subsequent activation of the phosphatase calcineurin and the kinase Ca²⁺/calmodulin-activated protein kinase IIγ (CaMKIIγ) (4,5). Deactivation of SIK2 and activation of calcineurin promote nuclear translocation of the transcription factor CRTC2, which is a CREB coactivator that induces the expression of a key HGP-promoting transcription factor, PGC-1α (6). Activation of CaMKIIγ activates the kinase P38α, which phosphorylates the transcription factor FoxO1 on specific serine and threonine, leading to FoxO1 nuclear translocation (5). Nuclear FoxO1, in concert with PGC-1α, induces key genes involved in HGP and glycogenolysis, notably G6pc and Pck1 (7). During feeding, insulin-mediated activation of the kinase AKT promotes nuclear exclusion of CRTC2 and FoxO1 and thereby appropriately dampens HGP and glycogenolysis (7,8). Moreover, in the context that glucagon action on hepatocytes is elevated in obesity and T2D, both the CRTC2 and FoxO1 pathways are activated in obesity/T2D and contribute substantially to pathological hyperglycemia in this setting (4,5).

Figure 1.

Pathways through which glucagon, acting through GcgR, and PGF_2α, acting through FP, increase the expression of genes involved in HGP, notably G6pc and Pck1. CBP, CREB-binding protein; P, phosphate; PLC, phospholipase C.

The findings of Wang et al. (3) suggest that another GPCR pathway in hepatocytes also contributes to protective HGP during fasting and pathologically excessive HGP in obesity/T2D. Previous work had shown that PGF_2α, which is formed from arachidonic acid via cyclooxygenase, is elevated in hepatocytes of both fasting and obese, insulin-resistant mice and that a PGF_2α metabolite in the urine is increased in T2D patients (9,10). Interestingly, aberrant production of another cyclooxygenase product, PGI2, in fasting and insulin resistance was shown to enhance HGP by activating the cAMP/PKA/CREB pathway and to block insulin-mediated Akt phosphorylation by increasing the Akt inhibitor Tribbles homolog 3 (Trb3) (9). Using cultured hepatocytes and fasting and obese mice, Wang et al. (3) now show that interaction of PGF_2α with its receptor (F-prostanoid [FP] receptor) on hepatocytes activates the same CaMKIIγ/P38/FoxO1 pathway elucidated previously for glucagon/GcgR (Fig. 1). Based on inhibitor experiments in cultured hepatocytes, the authors suggest that the proximal step of increasing cytosolic calcium is mediated by activation of phospholipase C. Glucose response to insulin was also improved in both fasting lean mice and obese mice that lacked the FP receptor in hepatocytes. Although the mechanism of insulin sensitization in mice lacking the hepatocyte FP receptor was not addressed, it may involve enhanced insulin-mediated Akt activation in hepatocytes through suppression of a previously described CaMKII/P38/Trb3 pathway (11).

The key question to emerge from this study is how the PGF_2α/FP receptor pathway is integrated with the glucagon/GcgR pathway in both physiology (fasting) and pathophysiology (T2D). The pathways may directly interact with each other, e.g., activation of GcgR may affect FP receptor signaling and/or vice versa; they may be separate but complementary; or they may be redundant. As a first step, this question could be addressed by assaying GcgR and FP receptor activity and their common downstream pathway in hepatocytes incubated with each hormone alone versus both together. If the two pathways turn out to have complementary roles, it would be interesting to determine if they affect different stages of fasting, e.g., glycogenolysis in the earliest stage followed by HGP in later stages. In this context, inhibition of CaMKIIγ/p38 or deletion of hepatic FoxO1 regulates both glycogenolysis and gluconeogenesis (5,12), whereas the study by Wang et al. (3) reported that FP receptor signaling regulates gluconeogenic gene expression without affecting hepatic glycogen levels. Given that the mRNA of the glycogenolytic enzyme glucose-6-phosphatase was lower in FP-deficient versus control mice, it is possible that one or more other downstream mediators of the PGF_2α/FP pathway overrides G6Pc–mediated glycogenolysis during fasting. Finally, there are other pathways that boost HGP in fasting, e.g., white adipose tissue–derived asprosin, which also activates PKA in hepatocytes (13). Therefore, elucidating the specific roles and integration of these multiple activators of HGP in fasting and T2D represents an important future goal in this area of research.

What might be the translational implications of the study by Wang et al. (3)? Although treatment of T2D subjects with high-dose aspirin, which inhibits cyclooxygenase and prostaglandin synthesis, was shown to improve fasting and postprandial hyperglycemia, the beneficial effect was attributed to inhibition of IKKβ activity as opposed to an effect on cyclooxygenase (14). In another arena, there have been several phase 2 clinical trials testing glucagon receptor antagonists in T2D, and although these drugs very effectively improve hyperglycemia, their use has been stymied by elevations of plasma LDL cholesterol (15). However, studies in diabetic mice have suggested that targeting mediators downstream of the GcgR may capture the benefits of GcgR blockade while avoiding the LDL-raising effect. For example, genetic silencing or drug-mediated inhibition of CaMKII or the P38 effector MAPKAPK2 (MK2) lowers blood glucose and improves insulin sensitivity without increasing plasma cholesterol in obese mice (11,16). Whether or not FP receptor blockade would be as efficacious but safer than GcgR blockade remains to be seen. Finally, the leading cause of death associated with T2D is cardiovascular disease. In this context, studies in mice have shown that silencing of the FP receptor decreases atherosclerosis and protects against lipopolysaccharide-induced tachycardia (17,18), and silencing of CaMKII or MK2 suppresses atherosclerosis and heart failure (19–21). Thus, leveraging the knowledge of these pathways may suggest an integrated therapeutic approach to cardiometabolic disease.