COMMENTARY ON HOT TOPICS
Chronic obesity associated type 2 diabetes mellitus (T2DM) is emerging as a global healthcare crisis (World Health Organization Global Infobase: data on overweight and obesity mean body mass index, healthy diets and physical inactivity; http://www.who.int/mediacenter/). If the present trends continue nearly 350 million people worldwide will develop T2DM by the year 2030[1]. Insulin resistance in the skeletal muscle and liver accompanied by a “burnt-out” pancreas, as judged by its functional decline and reduced numbers of β-cells, is a hallmark of diabetes[2-4]. The underlying etiology of insulin resistance largely reflects a combination of chronic tissue inflammation and pancreatic β-cell failure[3,5]. Current therapies for the treatment of T2DM commonly include insulin sensitizers (e.g., metformin and pioglitazone), with or without insulin secretagogues (e.g., glyburide), or injectable insulin[6]. Decades of clinical experience have proven that current drug regimens to manage T2DM rarely achieve the physiological glycemic control elicited by a fully functional pancreas[6]. As a consequence, the cardiovascular, renal and neurological complications of poorly controlled blood sugar continue to be a major challenge in the management of T2DM[6].
A long sought after but unrealized goal of treatment of T2DM is to replenish the loss of β-cell mass in the diabetic pancreas to restore normal glycemic control. Therefore, in recent years much effort has been focused on the physiological and pathological mechanisms that regulate the biogenesis, survival and apoptosis of the pancreatic β-cells. The pancreatic mass and β-cell numbers increase rapidly during embryogenesis and in the neonatal period, but the rate of replication of β-cells slows down dramatically in adult rodents and humans[7-9]. This is in contrast to a number of physiological (e.g., gestation-induced diabetes) and pathological (e.g., obesity-associated T2DM or pancreatic injury) states of insulin resistance during which pancreatic β-cells elicit enhanced rates of proliferation[8]. These observations have prompted a number of investigators to search for and identify a plethora of cell-intrinsic (e.g., cyclin D1/D2; CDK-4, nuclear factors of activated T cells, Menin and p53) and systemic [glucagon-like peptide-1, glucose-dependent insulin-tropic polypeptide, leptin, adiponectin, interleukin (IL)-1, IL-6, tumor necrosis factor-1, fibroblast growth factor-21 and serotonin] factors that are presumed to control β-cell proliferation[8]. However, based on a rigorous functional assessment of the varied factors purported to control β-cell proliferation it is apparent that most of these factors lack sufficient potency and/or β-cell specificity for them to be warranted as serious candidates for drug development.
A number of recent observations in murine models of insulin resistance and obesity (e.g., leptin-deficient ob/ob mice) appear to offer a way out of the current predicament. It has been reported that in a number of murine models of severe insulin resistance there is an apparent increase in pancreatic mass. The increased pancreatic mass seen under these conditions have re-ignited a search for systemic factor that might control proliferation of β-cells. The strongest evidence in favor of a putative factor(s) produced by the liver and released in circulation originally came from Ronald Kahn’s laboratory[10,11]. Using liver-specific insulin receptor knockout (LIRKO) mice, these authors reported the induction of a humeral factor involved in crosstalk between liver and pancreatic islets to trigger β-cells to multiply. Subsequently, El Ouaamari et al[12] applied the techniques of parabiosis, tissue transplantation, and in vitro islet culture in LIRKO mice to provide additional evidence for a non-neural and non-cell-autonomous factor(s) capable of inducing β-cell proliferation.
A parallel line of investigation in another model of obesity and insulin resistance also hinted at the existence of inter-organ metabolic communication between liver and pancreas. It had been suspected for some time that the livers of obese animals elicit enhanced activation of mitogen-activated protein kinase (MAPK) pathway. To investigate if MAPK signaling was involved in inter-organ relay of information between liver and pancreas, Imai et al[13] expressed a constitutively active extracellular regulated kinase (ca-ERK) in the livers of mice. Consistent with their hypothesis, ERK activation in the liver induced an obesity-associated metabolic signal(s) that was relayed via neuronal circuits to induce pancreatic β-cell proliferation. A number of additional studies have also hinted at the existence of a circulating factor(s) that might induce β-cell proliferation[10,14]. These tantalizing observations in experimental models of obesity notwithstanding, the identity of the putative liver-derived molecule(s) responsible for increased pancreatic β-cell mass and their implications for humans remain elusive. Apparently, Yi et al[15] were inspired by the tantalizing observations in LIRKO or ca-ERK mice to mount a systematic search for a putative liver-derived pancreatic β-cell growth factor produced in insulin resistance states.
Yi et al[15] began by co-opting a previously described model of insulin resistance in rat caused by infusion of S961, an insulin receptor (IR) antagonist peptide[16]. The authors delivered S961 by an osmotic pump and noted that S961 infusion caused dose-dependent peripheral insulin resistance and hyperglycemia that were accompanied by a dose-dependent surge in plasma insulin. These findings led the authors to posit that the observed surge in plasma insulin in mice most likely resulted from a compensatory increase in the numbers of pancreatic β-cells as a direct result of S961-induced insulin resistance. The authors experimentally tested this hypothesis and stained pancreatic sections of normal and S961-treated mice with specific antibodies to assess cell division [e.g., Ki67 and Proliferating Cell Nuclear Antigen (PCNA)]; insulin- and nuclear Nkx6.1-specific antibodies were used to specifically stain the β-cells in the pancreatic islets. The histological assessments of pancreas from control and insulin-resistant mice were further complimented by quantitative polymerase chain reaction (qPCR) to measure expression of cell cycle specific genes (e.g., cyclins, CDKs and E2Fs). Based on these analyses the authors concluded that S961 infusion-mediated insulin resistance in mice was accompanied with a potent (about 12-fold) and preferential induction of cell proliferation.
Since a direct exposure of pancreatic islets to S961 did not cause β-cell proliferation, Yi et al[15] reasoned that the IR antagonist acted by altering gene expression in insulin sensitive tissues, other than pancreas, and potential inter-organ relay of signals indirectly impinged on the mechanism of pancreatic β-cell proliferation. To test this hypothesis, the authors analyzed global gene expression in untreated control and S961-treated liver, adipose tissue, skeletal muscle and pancreatic β-cells. The analyses of gene expression by microarrays revealed that a number of mRNAs were specifically induced in the liver and white adipose tissue (WAT) of S961-infused mice. Yi et al[15] chose a candidate gene from microarray analysis for more detailed in silico and experimental studies. They discovered that betatrophin mRNA was specifically induced by S961 in the liver and WAT of mice but not in skeletal muscle and pancreatic β-cells. The authors expressed exogenous betatrophin in the liver and detected the recombinant betatrophin protein in the serum. Interestingly, an in silico analysis of the putative betatrophin gene/transcript revealed that these had been annotated by other names in the mouse and human genomes. It became obvious from these analyses that Gm6484 (gene) and EG624219 (protein) as annotated in the mouse genome and C19orf80 (gene) and TD26 (hepatocellular carcinoma-associated protein), annotated in the human genome described the same genetic entity[17].
Yi et al[15] noted that the expression of betatrophin mRNA was the most abundant in the liver and WAT of mice. Similarly, the expression of betatrophin was more than 250-fold higher in the human liver compared with WAT, heart, brain, pancreas or skeletal muscle. The authors extended their observations in S961-infused mice by exploring the expression of betatrophin in a number of other well-studied murine models of insulin resistance. Consistent with their hypothesis that betatrophin expression was a compensatory response to insulin resistance, Yi el al[15] observed that the livers of db/db and ob/ob mice expressed 3-4-fold higher levels of betatrophin mRNA. Similarly, a about 20-fold surge in hepatic expression of betatrophin mRNA was seen in mice during gestation, a time known to be associated with insulin resistance, accompanied with pancreatic β-cell expansion. Based on these observations the authors concluded that a state of insulin resistance was specifically involved in the induction of betatrophin gene expression. This conclusion was further bolstered by authors’ observation that regeneration response of pancreas after its exposure to diphtheria toxin did not lead to induction of betatrophin.
As predicted by in silico analysis, Myc-tagged betatrophin expressed in Hepa1-6 or 293T cells behaved as a secreted protein, as judged by its intracellular location in vesicle-like structures, and its presence in the culture medium. Similarly, the hepatic expression of Myc-tagged betatrophin led to its appearance in circulation and was accompanied by a robust increase in the rate of replication of β-cells (17 to 33-fold). To examine the specificity of betatrophin action on β-cells, Yi et al[15] stained pancreatic sections with Ki67- and PCNA-specific antibodies, and used qPCR to quantify expression of cyclins, CDKs, E2F1 and E2F2 genes. These studies revealed that exogenous hepatic expression of betatrophin was highly specific for β-cell proliferation since no other exocrine or endocrine cell types in the pancreas underwent multiplication. Thus, betatrophin appears to be a bona fide pancreatic β-cell specific hormone.
Yi et al[15] experimentally assessed the authenticity and normalcy of the newly formed β-cells by a number of criteria. They observed that pancreatic islets from control and betatrophin-expressing mice elicited similar glucose induced insulin secretion. The insulin tolerance tests in control and betatrphin-producing mice showed that they were similarly insulin sensitive. The normal insulin sensitivity of the betatrphin-expressing mice was in a sharp contrast to the situation seen in S961-infused mice that elicited robust insulin resistance. Finally, Yi et al[15] discovered that mice with forced expression of betatrophin elicited lower fasting glucose levels and had improved scores on glucose tolerance tests. These observations are consistent with the authors’ conclusion that betatrophin, a hormone capable of inducing β-cell proliferation, is specifically produced by the liver under conditions of insulin resistance. Furthermore, the functional phenotype of the newly made β-cell in response to a forced expression of betatrophin appeared to be normal.
The discovery of a circulating hormone capable of inducing β-cell proliferation fulfills a cherished goal of many investigators engaged in T2DM research[14]. My commentary on the publication of Yi et al[15], would not be complete without a few words about the re-discovery of betatrophin in its current guise and it molecular characterization by three other names by three independent laboratories. Thus, the discovery of so-called RIFL (re-feeding induced fat and liver) was reported as an adipocyte-enriched insulin target gene with a role in lipid metabolism[18]. A related strategy to analyze differential gene expression by RNA-seq expression experiments in the liver and WAT of fasted and high fat-fed mice, lead Zhang[19] to discover lipasin as a nutritionally regulated inhibitor of lipoprotein lipase that was involved in regulating serum triglyceride levels. Finally, Quagliarini and co-authors characterized angiopoitin-like 8 (ANGPTL8) protein, a member of angiopoitin-like family of proteins[20] that was shown to regulate postprandial fatty acid metabolism; apparently, ANGPTL8 could not carry out this function in the absence of its association with ANGPTL3, a paralog of ANGPTL8[21]. In retrospect, it is obvious that betatrophin, RIFL, lipasin and ANGPTL8 are one and the same gene/protein. However, all three reports that preceded Yi et al[15] had described these genes as regulators of fatty acid metabolism and lipid homeostasis. In light of these observations the question of how mechanistically betatrophin exerts its pleiotrophic actions in different tissues under physiological and pathological conditions becomes even more urgent[17]. The current progress notwithstanding, a fuller understanding of the varied functions attributed to betatrophin/RIFL/Lipasin/ANGPTL8 must be clarified before full therapeutic implications of Yi et al[15] findings are realized.
