Glucagon signaling is a major counterregulatory hormone to insulin. It promotes glycogenolysis and stimulates gluconeogenesis in the liver, resulting in higher glucose output to the blood stream. In many T2DM patients, in addition to insulin signaling deficiency, glucagon secretion has been found to be abnormally elevated due to dysregulation, exasperating T2DM symptoms and progression [6, 7]. Thus, attenuating glucagon signaling may provide a sound strategy to manage T2DM. Indeed, in preclinical studies, inhibition of the glucagon receptor with antagonistic small molecules, monoclonal antibodies, or antisense oligonucleotides leads to reduced plasma glucose, improved insulin sensitivity, and improved glucose tolerance [8–11]. Furthermore, genetic knockout of the mouse GCGR by homologous recombination resulted in significant reduction of blood glucose, improvement in glucose tolerance and insulin sensitivity, and resistance to diet-induced obesity [2, 3]. Whereas these data provided strong rationale for targeting GCGR as a treatment of T2DM, some side effects have also been observed in mice with severe disruption of glucagon signaling, including increased plasma cholesterol concentration and enlarged pancreas due to α-cell hyperplasia (2, 3, 8, 10]. The current studies were designed to profile mouse liver mRNA, protein levels, and plasma metabolite levels to identify changes that underlie the beneficial, as well as unwanted effects, caused by GCGR knockout.
Our data showed that GCGR knockout leads to a marked down-regulation of genes and proteins associated with liver gluconeogenesis and a modest up-regulation of those involved in glycolysis (Table 2). This is not unexpected as glucagon, in countering insulin, normally promotes expression of gluconeogenesis pathway genes and inhibits those of glycolytic pathway. The severe inhibition of liver gluconeogenesis and the modest increase of liver glycolysis likely contributed most to the improved glucose control in Gcgr-/- mice that has been published previously [2, 3]. The mRNA of enzymes involved in glycogenesis were modestly increased as expected from loss of GCGR signaling, but paradoxically, the genes and proteins involved in glycogenolysis were also up-regulated (Table 2).
An important observation in the current study is the wide-spread reduction of transcripts and proteins of enzymes involved in amino acid catabolism in the Gcgr-/- liver (Table 3). These include enzymes responsible for converting amino acid carbon skeletons into substrates for gluconeogenesis and ketone body production, e.g. Agxt, Gpt, Aass, and Haoo. Protein synthesis-related genes, on the other hand, were up-regulated only modestly at the transcriptional level in the Gcgr-/- liver (Additional file 4). As a result, blood concentration of amino acids and their derivatives increased significantly (Table 6). Whereas the scope and degree of the transcriptional dysregulation of amino acid metabolism in the current study were surprising, this could be attributed to the net effect of up-regulating insulin action while down-regulating glucagon action in the knockout mice. It is a well known fact that glucagon promotes, while insulin inhibits, amino acid catabolism to provide substrates for gluconeogenesis, ketogenesis, or direct energy source by oxidation. Such gross regulation of these enzymes at the transcriptional level, by either insulin or glucagon, has not been published previously. In fact, transcriptional regulation of amino acid catabolism has not been completely illustrated as these pathways are chiefly regulated by substrate availability and allosteric mechanisms . The dramatic transcriptional alterations of these pathways may reflect an adaptive response to innate loss of glucagon signaling in the Gcgr-/- liver. Further studies are needed to understand the complex regulatory mechanism leading to these changes.
Another important observation in this study relates to altered lipid metabolism in the Gcgr-/- liver. There is a marked increase of transcripts and proteins for the key enzymes involved in the biosynthetic pathways of fatty acids and cholesterol (Table 4 and 5). As a classical function of insulin, in opposition to glucagon, is to promote fatty acid synthesis from excessive glucose , it is consistent that fatty acid biosynthesis is up-regulated in the Gcgr-/- liver where insulin signaling dominates. Multiple lines of evidence have indicated that the transcription factor sterol regulatory element-binding protein (Srebp)-1c mediates insulin-induced stimulatory effect on fatty acid biosynthesis genes [13–15]. Indeed, our data showed that the transcripts for Srebp-1 and its chaperone Srebp cleavage-activating protein (Scap) were both up-regulated 1.3 fold in the Gcgr-/- liver (Table 4).
As alluded to above, transcripts of key genes involved in the de novo cholesterol biosynthesis such as HMG CoA reductase were increased for ~ 2 fold (Table 5) in the Gcgr-/- liver. The expression of these genes has been shown to be controlled by the transcription factor Srebp2 in a sterol-dependent manner . In contrast to Srebp1, expression of Srebp2 has not been shown to be regulated by insulin or glucagon [17, 18]. This observation is confirmed in the current study as SREBP2 transcript was unchanged in the knockout mice (data not shown). Srepb2 is synthesized first as inactive precursor bound to the ER membrane via two transmembrane domains. Upon cholesterol depletion, it is escorted by the cholesterol-sensing and escorting protein Scap to the Golgi apparatus, where the N-terminal domain is released from the membrane via proteolysis. The N-terminal domain, designated as nSrebp2, then translocates to nucleus and activates the transcription of its target genes . As noted above, SCAP expression is up-regulated in the Gcgr-/- liver (Table 4). Thus, even though the level of full-length Srebp2 is not changed in these hepatocytes, more nSrepb2 could be generated and translocated to the nucleus, where it could stimulate the transcription of key cholesterol synthesis genes. Further studies are needed to fully understand the mechanism underlying the gross transcriptional upregulation of cholesterol synthesis pathway in the Gcgr-/- liver.
In addition to the transcriptional up-regulation, fatty acid and cholesterol biosynthesis in the Gcgr-/- liver may also be boosted by the potential accumulation of their common substrate acetyl-CoA, due to increased glycolysis, reduced oxidation via the TCA cycle, and inhibited gluconeogenesis (Table 2). Furthermore, our data showed that enzymes regulating mitochondrial β-oxidation of fatty acids were down-regulated at the transcriptional level, which is in line with a recently published biochemical study of fatty acid oxidation using 1-14C-palmitate labeling in the knockout mice . Taken together, one would expect the plasma levels of free fatty acids (FFA), TG, and cholesterol to rise in Gcgr-/- mice. Indeed, several publications, as well as our data (not shown), have shown that plasma cholesterol and/or LDL-C increase significantly in the knockout mice [2, 3]. With regard to blood levels of FFA and TG, earlier reports indicated that there was little change in Gcgr-/- mice compared to the wild-type littermates under fed or fasted condition [2, 3, and 20]. However, a recent study by Longuet and colleagues demonstrated that under prolonged fasting (16 hr), plasma FFA, TG, and liver VLDL secretion all rose significantly in Gcgr-/- mice . This discrepancy merits further study.
A surprising finding in the current study is the highly elevated level of cholic acid and glycocholic acid in the blood of Gcgr-/- mice (Table 6). Bile acids (BAs) are synthesized from cholesterol in hepatocytes through a multistep enzymatic process and are secreted into the bile via the bile salt export pump ABCB11. BAs are released from the gall bladder into the intestinal lumen upon feeding to facilitate digestion of lipids. The majority of secreted BAs are reabsorbed efficiently into portal blood by specific transporters in the terminal ileum, and then taken up by hepatocytes through the action of basolateral uptake transporters, thus fulfilling an "enterohepatic cycle" .
Although the transcription of critical genes involved in liver BA synthesis, such as the rate-limiting enzyme Cyp7a1, were not up-regulated in the Gcgr-/- liver (data not shown), BA production most likely increased in these mice owing to elevated level of cholesterol, the substrate for BA synthesis. BA excretion to the canalicular bile and uptake from the portal vein is predicted to increase moderately based on up-regulation of the transporter genes associated with these processes (Table 5). Hepatic BAs can be secreted to the systemic circulation via alternative basolateral efflux transporters and disposed of in urine . In adaptive response, this alternative secretive pathway is up-regulated dramatically in mice when hepatic BAs accumulate as a result of bile duct ligation or suppression of ABCB11 expression in FXR knockout mice . Interestingly, three of these alternative efflux transporters: Abcc3, Abcc4, and Ostβ are up-regulated by 1.6, 7.0, and 1.5 fold at the transcript level, respectively, in Gcgr-/- mice (Table 5). As a result, BA excretion to the systemic circulation via these alternative transporters could increase dramatically in these mice, leading to the higher plasma level of BAs observed (Table 6). Thus, in Gcgr-/- mice, hepatic BA production probably rises due to increased cholesterol synthesis, and BA excretion to bile and urine via blood is predicted to increase in adaptive responses.
It has been reported that plasma levels of glucagon-like peptide-1 (GLP1), as well as glucagon, were significantly elevated in Gcgr-/- mice [2, 3]. The increase in circulating GLP1 was caused, at least partly, by increased production and processing of the preproglucagon mRNA from hyperplasic alpha cells in these animas . As BAs were recently found to be able to stimulate GLP1 secretion from enteroendocrine cells via the TGR5 receptor [5, 21], it is tempting to speculate that elevated BAs may partially contribute to the increase of plasma GLP1 in these animals. Further studies are warranted to investigate this possibility as well as potential effects of elevated BAs on glucose homeostasis, lipid metabolism, and thermogenesis.
In summary, genetic knockout of the glucagon receptor in mice brought about significant metabolic changes in the liver. These include up-regulation of glycolysis, severe inhibition of gluconeogenesis and amino acid degradation, marked reduction of plasma glucose, and increased levels of plasma amino acids. Meanwhile, there is evidence indicating the up-regulation of fatty acid and cholesterol biosynthesis pathways and bile acid generation in these mice. In assessing the significance of these findings to anti-GCGR therapies for T2DM, we are mindful of the limitations to the current study. First, the tissues were collected for analysis when the animals were in the fed state. As many transcripts and proteins were likely turned over rapidly in response to fasting and feeding, our observation might not reflect the full metabolic status that exists in the fasted state. While glucagon signaling may contribute to the metabolic disorder in T2DM throughout the day, its most profound effects are thought to occur in the postprandial state. Second, our findings reflect the profound metabolic changes in animals with genetic ablation of the GCGR, while pharmaceutical inhibition of GCGR is unlikely to result in a complete and constant blockade of GCGR signaling. As inhibition of the glucagon signaling pathway presents an attractive therapeutic strategy for T2DM, the challenge will be to design drugs with optimal pharmacokinetic and pharmacodynamic properties that render glucose-lowering benefits while avoiding the potential side effects of lipogenesis.