|Home | About | Journals | Submit | Contact Us | Français|
Type 1 and type 2 diabetes mellitus together are predicted to affect over 300 million people worldwide by the year 2020. A relative or absolute paucity of functional β-cells is a central feature of both types of disease, and identifying the pathways that mediate the embryonic origin of new β-cells and mechanisms that underlie the proliferation of existing β-cells are major efforts in the fields of developmental and islet biology. A poor secretory response of existing β-cells to nutrients and hormones and the defects in hormone processing also contribute to the hyperglycemia observed in type 2 diabetes and has prompted studies aimed at enhancing β-cell function. The factors that contribute to a greater susceptibility in aging individuals to develop diabetes is currently unclear and may be linked to a poor turnover of β-cells and/or enhanced susceptibility of β-cells to apoptosis. This review is an update on the recent work in the areas of islet/β-cell regeneration and hormone processing that are relevant to the pathophysiology of the endocrine pancreas in type 1, type 2 and obesity-associated diabetes.
The pancreas is a complex organ composed of an exocrine component, that is essential for nutrient digestion, and an endocrine component that is critical for the regulation of glucose homeostasis. The endocrine portion, which represents only ~2% of the pancreas is made up of groups of endocrine cells called the islets of Langerhans. Each islet is composed of at least five types of cells, including the insulin-producing β-cells (65-80%) (1), the glucagon releasing α-cells (15-20%) (2), the somatostatin producing δ-cells (3-10%) (3), the pancreatic polypeptide containing PP cells (1%) (4), and the ghrelin containing ε-cells (<1%) (5).
Dysfunction of the pancreas can impact either the endocrine or the endocrine components and lead to the development of two important diseases worldwide today, namely, diabetes mellitus or pancreatic cancer respectively. Diabetes mellitus is classified into type 1 and type 2 diabetes (T1D and T2D) as well as monogenic forms, termed maturity-onset diabetes of the young (MODY). T1D is prevalent in children and adolescents and is characterised by absolute deficiency of insulin secondary to autoimmune destruction of the β-cells by autologous cytotoxic T-cells. In contrast, T2D is usually diagnosed in adults and its prevalence increases with age although alarming recent data indicate that symptoms that characterize T2D also occur in adolescents (6). T2D is characterized by insulin resistance in the canonical insulin sensitive tissues including adipose, skeletal muscle and liver. MODY was originally diagnosed in individuals under the age of 25 years with an autosomal dominant inheritance (6). The underlying defects in MODY are mutations in the genes coding for transcription factors that are essential for β-cell development/differentiation and glucose-sensing proteins. Hepatic nuclear factors (HNFs) comprise a family of transcription factors that play an important role in the modulation of β-cell differentiation and/or function. Mutations in the HNFs have been associated with MODY — for example, mutations in HNF4α cause MODY1 (7), mutations in HNF1α cause MODY3 (8) and mutations in HNF1β cause MODY5 (9). Mutations in the gene coding for insulin promoting factor/PDX1 or NeuroD have been liked to MODY 4 and 6 respectively (10;11). A mutation in the glucokinase gene, which is essential for glucose sensing, is the underlying defect in MODY2, the most common of all forms of MODY (12;13). More recently, new genes associated with diabetes have been identified and include CEL (carboxyl ester lipase) (14) and the insulin gene itself (15).
Almost 6% of the world’s population suffers from diabetes with a total prevalence reaching 171 million individuals including 20.8 million in the United States alone (census reports in 2005 www.cdc.gov/diabetes) (16). Wild et al. have predicted that the total number of individuals with diabetes will rise to a staggering 366 million in less than 30 years if preventive action is not undertaken immediately (16;17). Pancreatic cancer, which largely afflicts the exocrine pancreas, is the fourth most common cause of cancer deaths in the United States, and affects ~28,000 individuals annually (18).
Both diabetes and cancer lead to pancreatic islet dysfunction that includes altered islet growth/apoptosis and hormonal secretion defects. A thorough understanding of the mechanisms that underlie these disorders is critical to plan therapeutic approaches to treat each pathological state. Despite significant progress in understanding the pathophysiology of diabetes and pancreatic cancer, the key regulators and signaling proteins involved in the two processes that allow an effective therapeutic approach are still elusive.
This review will focus on recent work related to key regulators of islet/β-cell regeneration with an emphasis on growth factor signaling. We will provide a short overview on insulin and glucagon processing and secretion, and also review the current state of knowledge on the impact of aging on islet biology and discuss the potential for pancreatic cell progenitors in islet cell regeneration.
Pancreatic β-cells are remarkably dynamic cells that are able to adapt and modulate their mass in response to a variety of physiological (pregnancy, puberty) (19;20) or pathophysiological (obesity (21)/insulin resistance (22)) states. Current literature suggests that various mechanisms exist for regulating and maintaining pancreatic β-cell mass (23;24). Early studies utilizing [3H] thymidine incorporation experiments indicated that adult pancreatic endocrine cells belong to a class of tissues that could be maintained by the self-duplication of mature differentiated cells (25;26). Recent immunohistochemical observations proposed that adult pancreatic stem or progenitor cells reside in the epithelium of pancreatic ducts (27;28), within the islets (29;30) or in the bone marrow (31), and contribute to the formation of new insulin producing β-cells. Other studies suggest that either acinar cells (32), non-insulin producing islet cells (33) or splenocytes (34) can transdifferentiate into insulin producing β-cells. β-cell expansion could also be promoted by EMT (epithelial-to-mesenchymal transition) or an EMT-like process (35). More recently, sophisticated direct lineage tracing performed on transgenic mice using the Cre/lox system demonstrated that adult pancreatic β-cells are formed by self-duplication of existing β-cells in the islets (36;37), and that all β-cells contribute equally to islet growth and maintenance with no specialized progenitors (38;39). In humans, the data on β-cell origin are scant and the limited data suggest that a balance between both apoptosis and β-cell replication determines the overall adult β-cell mass (40).
During the compensatory adaptive expansion in response to insulin resistance and hyperglycemia β-cell mass can increase either by cell size (hypertrophy), cell number (hyperplasia) and/or neogenesis together leading to enhanced insulin synthesis and secretion. However, a failure of the pancreatic islets, to compensate due to one or more pathological causes leads to a series of events characterized by β-cell hypoplasia, apoptosis and a relative increase in α-cells (41-43). Abnormal growth of islet cells can lead to the development of islet cell tumors that are generally named depending on the hormones they secrete. The most common islet cell tumor of the pancreas is the insulinoma (β-cell origin) of which up to 90% are benign. Diabetes mellitus is a risk factor for pancreatic cancer, although a small percentage of patients appear to have an inherited familial form of pancreatic cancer (44). Moreover, a recent study shows that patients with diabetes who reported smoking and were diagnosed five or more years before the study was conducted were associated with pancreatic cancer (45). However, despite several epidemiological and physiological studies, it is still unclear precisely where one disease stops and the other starts.
Various nutrients and hormones have been implicated in regulating β-mass including glucose, insulin, IGFs (insulin like growth factors), GH (growth hormone), GLP-1 (glucagon-like-peptide-1), HGF (hepatocyte growth factor/scatter factor), PTHrP (parathyroid hormone-related protein) and lactogens (reviewed in (46;47)). In addition to its role as a nutrient, glucose has been shown to increase β-cell mass in several models (48-52). GLP-1 controls blood glucose levels by stimulating insulin secretion, insulin biosynthesis, β-cell proliferation, islet neogenesis and inhibiting gastric emptying and glucagon release (53). The effects of GLP-1-induced proliferation of β-cells are non-additive with those of glucose and the growth effects have been proposed to be mediated through different intermediates including PI3K/Akt, PKC, and Jun. HGF and PTHrP are two potent β-cell mitogens and recently both have been shown to act via PKCζ (54). Among lactogens, prolactin (PRL) is a hormonal regulator of pregnancy and β-cell replication induced through the PRL-Receptor is dependent on the JAK2/STAT5 (Janus Kinase/Signal Transduction and Activators of Transcription) signaling pathway. Moreover, PRL has been recently reported to repress menin levels and stimulate β-cell proliferation (55). Menin is a gene underlying the endocrine tumor syndrome termed multiple endocrine neoplasia type 1 (MEN1), which affects most notably the pancreas. In addition, intracellular Ca2+, a second messenger, has been reported to regulate NFAT (nuclear factor of activated T cells) and to integrate mitogenic signals from glucose, GLP-1, GH and insulin to induce β-cell proliferation (47;56). Insulin and IGF-I that mediate two ubiquitous signaling pathways will be discussed in more detail below. Leptin and adiponectin are two major adipokines, that have been suggested to regulate compensatory β-cell growth in response to obesity (57). Leptin seems to have a dual role in β-cell growth since this adipokine has been demonstrated to lower PTEN activity and promote p27 nuclear exclusion which in turn promotes β-cell proliferation (58) on the other hand leptin also mediates negative effects on insulin secretion by increasing PDE3B (Phosphodiesterase 3B) activity and decreasing cAMP availability (59). Indeed, a mutation in the leptin receptor in the diabetic db/db mouse causes hyperinsulinemia and islet hyperplasia but it was unclear whether the islet dysfunction is due to a lack of direct leptin action in β-cells or secondary to the effects of peripheral insulin resistance. Morioka et al. recently reported that mice lacking the leptin receptor specifically in the pancreas and fed normal chow manifested improved glucose tolerance, enhanced first-phase insulin response to glucose, and islet hyperplasia implying a direct role for leptin action on the β-cells (60). Surprisingly, challenging these mice with a high-fat diet, led to significantly greater glucose intolerance and poor compensatory islet growth. Together, these data provide genetic evidence for a critical role for leptin signaling in islet growth and function (60). Interactions between proteins in the leptin and insulin signaling pathays in the overall growth response during physiological () and pathological (obesity) states are worth exploring in the context of β-cell failure and development of T2D.
Despite significant progress in identifying β-cell growth inducers and their signaling pathways, the cell cycle molecules responsible for tightly modulating β-cell proliferation during embryonic growth and underlying the compensatory growth response to insulin resistance have not been completely defined. Since β-cell replication is now accepted as essential for maintenance of adult β-cell mass (35;36;61-63), several investigators have focused on examining the proteins that are involved in the regulation of the G1/S transition in the β-cell cycle. Cell replication is generally modulated by the interaction of a diverse set of proteins that comprise the cyclins, the CDKs (cyclin-dependent-kinase) and the CDKIs (CDK inhibitor) and requires a balance between the active complexes formed by cyclin-CDK and the CDKIs (reviewed in (47;61). For example, cyclin D partners with CDK4 or CDK6 in early G1, and cyclin E partners with CDK2 in late G1, and both complexes phosphorylate the pRb (retinoblastoma protein) on different sites. This hyperphosphorylation of pRb leads to the release of E2F and the initiation of a transcriptional program required for entering the S phase of the cell cycle (64). There are two families of CDKIs: the INK (inhibitor of CDK4) family composed of p15Inkb, p16Inka, p18Inkc and p19Inkd, and the CIP/KIP family including p21Cip1 and p27kip1. The INK proteins are specific inhibitors of CDK4, that partners cyclin D during the early G1 phase (65). The CIP/KIP proteins are inhibitors of CDK2, the partner of cyclin E in the late G1 phase, and they stabilize and enhance CDK4 function during cell cycle progression (65;66). Although these cell cycle proteins are expressed in the islet (see reviews: (61;67)) only some of them have been directly implicated in β-cell proliferation (Figure 1).
Cyclin D2, the most abundant cyclin in the islet, is critical for postnatal pancreatic β-cell growth (62;68). Indeed, global knockout mice for cyclin D2 display decreased β cell mass and glucose intolerance leading to diabetes by 12 months of age (62) and the additional deletion of cyclin D1 in cyclin D2 deficient mice further decreases the β-cell mass (68). Interestingly, transgenic mice overexpressing c-myc in β-cells, which has been shown to drive oncogenic cell proliferation through cyclin D (69), manifest β-cell proliferation and apoptosis, down regulation of insulin gene expression, and diabetes in early neonatal life (70). Mice with global knockouts for CDK4 develop β-cell hypoplasia and diabetes (71;72) likely due to a lack of compensation secondary to the absence of CDK6 in β-cells. Conversely, mice transgenic for CDK4 manifest islet hyperplasia but do not develop hypoglycemia (71). Although mice deficient for E2F1 exhibit a defect in postnatal islet growth, reduced islet number and β-cell mass and glucose intolerance, the mutants do not develop diabetes (73). Nevertheless the additional deletion of E2F2 with E2F1 in these mice leads to insulin deficiency and pancreas dysplasia together leading to diabetes (74). Thus, the cyclin D/CDK4 complex and E2F family appear to have a crucial role in the regulation of β-cell mass particularly in the β-cell compensatory response that is necessary to delay the onset of diabetes.
pRb and p53 are central regulators of the cell cycle underlying G1/S arrest and are also tumor suppressors that have been reported to be frequently mutated in different forms of human cancer. Surprisingly, β-cell specific deletion of pRb in mice does not lead to a severe phenotype, and the mutants only manifest limited effects on β-cell replication, mass and function (75). Mice deficient for p53 are developmentally normal and despite a susceptibility to develop spontaneous tumors (76) the mice do not develop insulinomas (77;78), while mice lacking both pRb and p53 develop insulinomas and other islet tumors (77;78). To our knowledge, no studies have been reported on islet and glucose metabolism in p53 global knockout mice.
Among the proteins that comprise the CDKI families, an increased expression of p15Inkb in TGFβ (tumor necrosis factor) transgenic mice was observed to correlate with pancreas hypoplasia, hypoinsulinemia, and diabetes in mice on both the NOD and B6 genetic backgrounds (79). Surprisingly, p21Cip1, a well-known target of p53, was recently reported not to be essential for maintaining β-cell mass or β-cell function in vivo, similar to the observations on pRb (80). While global knockout mice for p27kip1 exhibit no phenotype on postnatal β-cell expansion (81), β-cell specific overexpression of p27kip1 leads to islet hypoplasia and diabetes (81). Moreover, an increase in nuclear localization of p27kip1 is observed in mice with pancreas-specific deletion of FoxM1 - a member of the Forkhead Box family (82). These mice show a decrease in β-cell mass and develop diabetes by 9 weeks of age. Furthermore, double knockout for p27kip1 and p18Inkc display β-cell hyperplasia but fail to develop insulinomas (83;84).
The genetic studies described above support an important role for cell cycle inhibitors in β-cell growth. The relative levels of expression and activity of one or more of the cell cycle proteins leads to either diabetes in the case of Cylin D2, CDK4, E2F or p27kip1, or to the development of cancer as discussed in the context of a combination of p53 with pRb, or p27kip1 with p18Inkc. Further studies are necessary to directly delineate the role of p53 in both and β- and α-cell pathophysiology.
Since insulin resistance is an early predictor of type-2 diabetes and IGF-I has been largely linked to cancer, the insulin/IGF-I system has been a popular area of investigation. Insulin and IGF-I constitute two primary members of the growth factor family, their receptors are expressed ubiquitously and mediate the growth and metabolic effects of the hormones in virtually all tissues in mammals. Insulin and IGF-I classically bind to their own receptors but can also cross-react and activate common downstream proteins. Receptor activation transmits signals by phosphorylating insulin receptor substrates (IRS) including the four IRS proteins, Shc, Gab-1, FAK, Cbl, or potentially other substrates leading to a cascade activation mainly through PI3K (phosphatidylinositol-3 kinase)/Akt to regulate multiple cellular processes such as glucose transport and utilization, protein synthesis, cell growth, proliferation and antiapoptosis (reviewed in (85;86)) (Figure 1). Mouse and human β-cells express both the insulin and IGF-I receptors and most components in their signaling pathways (87-89). Interestingly, human α, β and δ cells have been reported to exhibit distinct expression patterns of proteins in the insulin signaling cascade(67;90). Recent studies on the role of the insulin receptor (IR) signaling in β-cells have provided cumulative evidence for a role for autocrine action of insulin on its own receptor. Two early studies that provided direct genetic evidence for a role for insulin/IGF-I signaling in the regulation of β-cell biology include the β-cell-specific knockout of the insulin receptor (βIRKO) (91) and the global knockout of insulin receptor substrate-2 (IRS-2) (92). While the βIRKO mouse manifested a phenotype most resembling human type 2 diabetes, mice with IRS-2KO failed to maintain their β-cell mass and developed diabetes. Following these two studies, multiple laboratories have reported the creation and characterization of transgenics and knockouts complemented by in vitro and ex vivo approaches to indicate the significance of proteins in the insulin/IGF-I cascade for the regulation of β-cells (91;93-100). Several recent reviews provide an excellent resource for interested readers (85;101-104).
While the potential of insulin and IGF-I as β-cell growth factors has been a topic of investigation for several years we will focus largely on in vivo experiments that provide direct genetic evidence using homologous recombination to create mouse models to study phenotypes linked to β-cell biology. Global deletion of the insulin gene in mice leads to intra-uterine growth retardation and death due to ketoacidosis and hepatic steatosis (105). Interestingly, examination of pancreas during the postnatal period revealed islet hyperplasia due to an increase in islet cell proliferation and a reduction in apoptosis (106). Whereas global IGF-I knockout mice display growth retardation (107) (108) mice with a pancreas-specific knockout of the IGF-I gene manifest islet hyperplasia and are resistant to streptozotocin-induced diabetes (109). Transgenic mice expressing IGF-II, the other member of the IGF family which can activate both insulin and IGF-1 receptors in addition to its own IGF-II-mannose phosphate receptor, exhibit hyperplasic islets but surprisingly developed diabetes (110;111). While the mechanisms underlying these defects are still unclear, one interpretation of these data is that insulin and IGF-I, contrary to the effects of IGF-II, are negative regulators of islet growth. Interestingly, mice expressing IGF-I in β-cells have been shown to regenerate pancreatic islets and counteract cytotoxicity and insulitis after treatment with multiple low doses of streptozotocin, suggesting that IGF-I gene transfer to the pancreas might be a suitable therapy for type 1 diabetes (112;113).
Contrary to traditional thought and despite a role in β-cell proliferation, we and others have used genetic approaches to directly demonstrate that the insulin/IGF1 signaling pathway is not critical for early development of β-cells (91;114-116). Indeed, although mice with a global knockout of insulin receptors or IGF-I receptors die immediately after birth, or manifest a variable survival rate in the case of the IGF-1RKO, they are born with mature β-cells (117). Consistent with the observations that double or single knockout mice show unperturbed endocrine α- and β-cell development compared to control mice (117), β-cell-specific knockout mice for the insulin receptor (βIRKO) (91) or IGF-I receptor (βIGFRKO) (118;119) are both born with a normal complement of islet cells. However, although both βIRKO and βIGFRKO mice exhibit impaired glucose tolerance and reduced GSIS (glucose-stimulated insulin secretion), only βIRKO mice show an age-dependent decrease in β-cell mass and an increased susceptibility to develop overt diabetes (91;97) suggesting a dominant role for insulin signaling in the regulation of adult β-cell mass (99). Mice lacking functional receptors for both insulin and IGF-1 only in β-cells were also born with a normal number of islet cells, but 3 weeks after birth, they developed diabetes, in contrast to mild phenotypes observed in single mutants (99). Therefore, IR and IGF-IR are not critical for development of β-cells but a loss of action of these hormones in β-cells leads to diabetes. It has been observed for several decades that patients with type 2 diabetes as well as mouse models of diabetes and obesity exhibit a remarkable ability to compensate for the increase in insulin demand in response to insulin resistance in liver, muscle and adipose tissue (43;120). One such model - the liver-specific IR knockout (LIRKO) mouse develops severe insulin resistance and glucose intolerance but mice do not become overtly diabetic due, in part, to a ~30-fold increase in β-cell mass (121). Furthermore, the enhanced β-cell proliferation shows a striking and persistent positive correlation with high circulating insulin levels in contrast to blood glucose levels that decrease in aging LIRKO mice (122). It is possible that insulin-resistant hepatocytes in these mice transmit signals to the endocrine pancreas via circulating growth factors that promote β-cell compensation. Consistent with a role for insulin signaling in β-cell proliferation, compound LIRKO/βIRKO mice fail to develop compensatory islet hyperplasia in response to insulin resistance (122). Besides, insulin itself is an obvious candidate for β-cell proliferation considering that insulin levels are elevated due to insulin resistance. Indeed, a recent in vitro study shows that insulin, at levels that have been measured in vivo, can directly stimulate β-cell proliferation (123).
IRSs and Akt, important downstream signaling molecules in the IR/IGF-IR signaling pathway have been reported to play a dominant role in β-cell growth. Indeed, global knockout of IRS-1 in mice leads to postnatal growth retardation, hyperplastic and dysfunctional islets. However, the mice do not develop overt diabetes due to β-cell compensation (94;95;124;125). In contrast, IRS-2 global knockouts develop only mild growth retardation, and depending on their genetic background, develop either mild glucose intolerance (92;126) or β-cell hypoplasia and overt diabetes (92). β-cell specific deletion of IRS-2 (βIRS2KO) also leads to mild diabetes (127-129). To avoid potential extra-pancreatic effects of the RIP-Cre promoter used to create the βIRS2KO mice, knockouts have also been generated using a cre recombinase driven by the pancreatic homeobox domain (PDX-1) promoter (130). βIRS2KO mice exhibit reduced β- and α-cell mass and impaired glucose homeostasis (130). Thus, IRS-2 appears to be a positive regulator of β-cell mass and β-cell compensation while IRS-1 predominantly regulates insulin secretion. A similar scenario is observed in the context of the Akt isoforms. Thus, global knockouts for Akt2 develop overt diabetes largely due to insulin resistance in peripheral tissues and β-cell failure, despite islet hyperplasia and hyperinsulinemia (131;132). Transgenic mice expressing a kinase dead mutant of Akt1 showed increased susceptibility to develop glucose intolerance and diabetes following fat feeding (133). Islet hyperplasia, β-cell hypertrophy, and hyperinsulinemia are observed in β-cell specific transgenic mice expressing constitutively active Akt1. However, these mice show improved glucose tolerance and a resistance to experimental diabetes (134;135). These in vivo data are supported by a recent in vitro study in other cell types (myoblasts and fibroblasts) that Akt1 is necessary for proliferation while Akt2 promotes cell cycle exit, and support a dual role for the Akts in the regulation of β-cell mass (136) . Moreover, mutation in the human Akt2 gene has been described in a family with severe insulin resistance and diabetes, indicating a role for Akt signaling in maintaining tissue insulin sensitivity in humans (137).
The role of phosphatidyl inositol 3-kinase (PI3K) — a critical node between the insulin receptor substrates and Akt — has also been examined in the context of islet biology although its role is not fully understood (86). PI3K appears to play a role in the differentiation of acinar cells following pancreas resection (138) and pancreatic duct cell differentiation into insulin producing cells in part, by regulating PDX-1 (139). These results are based on experiments aimed at inhibiting PI3K chemically, using Wortmannin or siRNA against the p85α subunit of PI3K (139). Multiple roles of the PI3K/Akt pathway have been described in cell cycle progression that are notably linked to tumor-promoting effect of this pathway (reviewed in (140)) but also in physiological β-cell proliferation. Considering the above reports on IRS and Akt isoforms it is likely that Akt1 mediates its proliferative effect following activation of IRS-2 whereas Akt2 is associated with anti-proliferative effects following IRS-1 stimulation. A recent review provides a discussion of the role of Akt in the regulation of β-cell proliferation (Reviewed in (141)).
One of the downstream molecules of Akt known to be involved in proliferation is GSK3. Akt mediates phosphorylation and inactivation of GSK3 upon stimulation (142) (Figure 1). While no such direct evidence has been demonstrated in β-cells, GSK3 phosphorylation was observed in mice over-expressing a constitutively active form of Akt1, which correlated with an increase in cyclin D1 levels in islet lysates (143). This data suggests a role for GSK3 inhibition in β-cell cycle progression since GSK3 has been shown in other cell types to phosphorylate and promote degradation of cyclin D1 (144;145). More recently, Huang et al. have shown that the GSK3β/β-catenin/TCF pathway in cooperation with CREB can downregulate the cyclin D2 expression by PTEN, suggesting a convergence of the PI3K/PTEN and the Wnt pathways in the modulation of cyclin D2 expression (146). Recent experiments have also linked the cyclin/CDK4 complex to Akt in β-cell proliferation showing that Akt1 upregulates cyclin D1, cyclin D2 and p21Cip1 levels (but not p27Kip1) and CDK4 activity (143). Moreover, this study shows that CDK4 is indispensable for β-cell proliferation induced by Akt1 in these transgenic mice. While only Akt1 appears to be required for proliferation, Akt2 has been shown in myoblast to promote cell cycle exit through specific p21Cip1 binding (136). Indeed, Heron-Milhavet et al. have demonstrated in vitro that silencing Akt1 resulted in decreased cyclin A levels and inhibition of S phase entry whereas no effects were observed with Akt2 knockdown except for reduced p21Cip1 levels. In contrast, overexpression of Akt2 reduced cyclin A expression and delayed cell cycle progression with increased nuclear expression of p21Cip1. Thus, the cyclin D/CDK4 complex is a likely target or mediator of PI3K/Akt pathways in the islet (Figure 1).
The FoxO proteins are other important downstream targets of Akt that are known to regulate the β-cell cycle (Reviewed in (147;148)). Phosphorylation by Akt regulates the subcellular localization of FoxO proteins, translocating them from the nucleus to the cytoplasm where they are inactive and ubiquinated. Akt therefore negatively regulates the transcriptional activity of FoxO proteins. Transcriptional repression of cyclin D has been shown to be required for FoxO mediated inhibition of cell cycle progression and transformation (149). Moreover, FoxO proteins have been reported to upregulate the transcription of p27Kip1 (150). Nevertheless, p27Kip1 does not seem to be a main factor regulated by Akt for β-cell progression (143). Thus, one downstream mechanism that has been proposed for β-cell proliferation by FoxO is the regulation of Pdx1 expression (151). Indeed, the phosphorylation of FoxO1 by Akt is believed to prevent the transcription factor Foxa2 from driving the expression of Pdx1 which is required for β-cell differentiation (152;153).
Akt can also downregulate p21Cip1 activity indirectly through MDM2/p53. Indeed, Akt can phosphorylate and stabilize the E3 ubiquitin ligase, MDM2, responsible for p53 degradation, and impact the expression of its main target in cell cycle regulation, p21Cip1 (154). Therefore, insulin and IGF-I can regulate β-cell mass and β-cell adaptation to insulin resistance to prevent diabetes through multiple mechanisms regulated by Akt (Figure 1). In addition, crosstalk with several other pathways can occur or interfere, for instance, with other growth factors though PI3K/Akt or the Wnt pathway via GSK3. Insulin and IGF-I can promote proliferation independently of the PI3K pathway by acting through the ERK pathway. The mechanisms regulated by the insulin/IGF-I signaling pathway have been described to promote abnormal growth and tumors in several tissues including β-cells and islets. However, it is unclear at what point the growth of islets that occurs as an important protective compensatory mechanism against insulin resistance, develops towards the formation of tumors.
Although we have focused on the growth and function of β-cells the other islet cell types also play a role in the β-cell and islet dysfunction observed in diabetes. For example, a relative increase in α-cell number is a recognized feature in adult patients with type 1 and/or type 2 diabetes (42;155). Nevertheless very few studies have focused on α- or δ-cell proliferation and the detailed mechanisms that underlie the changes in growth and secretory responses in type 1 and type 2 diabetes. Recent studies describe the development of α-cell hyperplasia in mouse models of varying etiology. For example, mice with a knockout of PC2, a prohormone convertase participating in the processing of proinsulin, proglucagon and a variety of other neuroendocrine precursors, develop marked hyperplasia of α- and δ-cells and a relative diminution of β-cells (156) (see below). In a second model, knockout of the glucagon receptor leads to hypoglycemia, hyperglucagonemia, and α-cell hyperplasia indicating that glucagon is essential for the maintenance of normal glycemia and in the postnatal regulation of islet and α and δ cell numbers (157). A role for IRS-2 in the regulation of α-cell growth is based on the observation that pancreas-specific deletion of IRS-2 leads to a reduced α-cell mass (130) and IRS-2 has been reported to be localized predominantly to α-cells compared to β-cells (94). Further studies are necessary to examine the contribution of IRS-1 and other proteins in the modulation of α-cell growth and its consequent paracrine effects on the adaptive responses of β-cells in response to insulin resistance.
While a failure to expand the β-cell mass in response to an increased demand is an important pathogen factor, the β-cells also manifest defects in insulin synthesis and secretion that contributes to the hyperglycemia in the development of overt T2D. In addition to defects in secretagogue-stimulated secretion (158;159), the regulation of insulin synthesis itself may be a site of abnormal regulation and warrant investigation of the mechanisms and factors that regulate the synthesis of hormones (160).
Insulin is exclusively synthesized in the β-cells and is first secreted in a precursor form similar to other islet hormones - glucagon and somatostatin. In contrast to humans, rodents have two insulin molecules, insulin 1 and insulin 2, and the latter is homologous to human insulin (161). Insulin mRNA within a β-cell accounts for 10-15% of the total RNA of a β-cell. Its translation is minimal at glucose concentrations below 3 mM and increases by ~50-fold when glucose levels are raised (162). Prohormone convertases (PCs) are basic amino acid specific enzymes that are essential for cleaving protein precursors into the active and mature hormones. The tissue specific co-localization of the prohormone with the convertases determines the site of conversion. Seven members of the family have been identified to date and include: PC1/3, PC2, furin/PACE, PACE4, PC4, PC5/6, and PC7/SPC7/LPC/PC8 (163). Substrates for these enzymes include proinsulin, proglucagon, prosomatostatin, proalbumin, proopiomelanocortin, proparathyriod hormone and others. The cleavage of the precursor forms by prohormone convertases creates fragments that are further modified by several carboxypeptidases (e.g. Carboxpeptidase E/H (CPE/H) or Carboxypeptidase B-like enzymes) (164;165). All pancreatic hormones are processed by PC1/3 and PC2. It is worth noting that PC1/3 is found to co-localize with insulin, while PC2 is co-localized with insulin, glucagon, somatostatin and pancreatic polypeptide (PP) (166). The dominant enzyme of proinsulin cleavage in pancreatic β-cells is PC1/3.
Proinsulin conversion occurs in secretory granules of β-cells, which provide a calcium-rich and acidic environment that is necessary for the full activity of the processing enzymes. Proinsulin is composed of two chains (A and B chains) which are connected by a connecting peptide (C-peptide)). Elevated fasting proinsulin concentrations and/or a switch of the ratios between mature insulin and proinsulin has been reported to occur in states of impaired glucose regulation (167). Indeed, the ratio between proinsulin and insulin has been used as a marker of β-cell dysfunction and has been used to predict the development of T2D (168). The circulating levels of proinsulin molecules is increased 4 to 5 fold in T2D compared to healthy individuals and this effect appears to be proportional to the level of hyperglycemia (169;170).
PC1/3, in human β-cells, cleaves almost exclusively at the carboxyl site of Arg31/ Arg32 at the B/C junction whereas PC2 prefers the Lys64/Arg65 site at the A/C junction. PC2 has a higher Vmax with des-31, 32 proinsulin intermediate compared to intact proinsulin and PC1/3 converts intact proinsulin as well as des-64, 65 proinsulin similarly efficient (171) (Figure 2). From this data it is estimated that 90% of the mature insulin is processed via des-31, 32-proinsulin intermediates (162). Interestingly, des-31, 32 intermediates have been reported to accumulate in patients with T2D. Whether the accumulation of des-31, 32 intermediates in PC2 knockout mice suggests a potential lack of active PC2 in humans with type 2 diabetes requires further investigation (172) (Figure 3). The exoprotease CPH/E removes basic residues exposed by PC1/3 and/or PC2 and accelerates the processing. Traditionally, cleavage of proinsulin results in an insulin molecule and a C-peptide fragment that are secreted in a 1:1 molar ratio. Whether this ratio is altered in response to differential regulation of insulin versus C-peptide processing making it a potential site for dysregulation during states of increased demand is not fully explored. Some studies suggest a biological role for C-peptide, which is independent of insulin action (173;174).
Thus, PCs have been linked to various pathologies including metabolic disorders such as diabetes (175) and obesity (176). PC1 knockout mice exhibit hyperproinsulinemia but maintain normal glucose tolerance in response to intraperitoneal (IP) glucose injections (177). PC2 knockout mice manifest chronic fasting hypoglycemia, improved glucose tolerance, a deficiency in circulating glucagon (156) and α-cell hyperplasia (156). Interestingly, the phenotype of the PC2 knockouts is also observed in mice lacking the glucagon receptor (157), suggesting that signaling by surface membrane receptors is linked to hormones and their precursors. The importance of CPE in hormone processing is evidenced by phenotyping CPE/H knockout mice. Besides behavioral defects, CPE/H knockouts display elevated plasma glucose and 100-fold higher insulin-like immunoreactivity that is identified as proinsulin. Although there is no further characterization of the proinsulin product, it is likely that spilt insulin products accumulate in these mutants since CPE plays a central role in both cleavage pathways (178), via des-31, 32-proinsulin conversion intermediates or des-64, 65-proinsulin (Figure (Figure2,2, ,3).3). Signaling pathways and mechanisms that regulate the activity of the convertases are not fully explained and further studies are necessary to examine whether defects in insulin action in β-cells promote hyperproinsulinemia.
Nutrient-stimulated insulin secretion is an important component of the endocrine response to maintain glucose concentrations in the physiological range. While glucose is clearly the critical physiological nutrient stimulus for insulin secretion, several other secretagogues including amino acids and free fatty acids and the incretin hormones glucagon-like peptide (GLP-1) and glucose dependent insulinotropic peptide (GIP), neurotransmitters and drugs also play important roles in different physiological and pathophysiological settings (reviewed in (179) and (40)). The mechanisms and pathways by which each of these secretagogues promote insulin release are different and will be discussed only briefly in this review.
Glucose enters the β-cell using the glucose transporter protein GLUT2, which is specific to the β-cells. Following rapid phosphorylation by glucokinase, glucose is further metabolized to yield pyruvate as an end-product of glycolysis. Pyruvate in turn is shuttled into the mitochondrial matrix where it is further oxidized to carbon dioxide in the Krebs cycle. Electrons generated in the oxidation process are fueled into the ATP generating process of oxidative phosphorylation. Increasing concentrations of ATP lead to an elevation of the cytoplasmic ratio ATP to ADP. This triggers the closure of ATP-sensitive K+ channels (180) resulting in depolarization of the cell membrane and an influx of extracellular Ca2+ ions, which in turn triggers insulin exocytosis (181). Several drugs including sulphonylureas act at the level of the ATP-dependent K+ channel and have been a subject of studies for several years (reviewed in (182)).
Several mouse models created with mutations in some of the key proteins that are relevant for insulin secretion have been reported to develop diabetes (183). Thus, global GLUT2 knockout mice exhibit impaired insulin secretion (184) and the absence of the glucose sensing protein, glucokinase, leads to perinatal death due to severe hyperglycemia (185). Moreover, the global knockout mice for glucokinase die perinatally due to severe hyperglycemia and diabetes (185). The Kir6.2 subunit of ATP-sensitive K+ channels also impact insulin secretion, as demonstrated by different experimental approaches (186). Kir6.2 is thought to form the K+ ion-selective pore of ATP-sensitive K+ (KATP) channels, the disruption of which leads to unregulated insulin secretion, where resting membrane potentials as well as basal Ca2+ concentrations are high and fail to respond appropriately to a glucose challenge (187-189). However, Kir6.2 knockout animals show only mild impairment in glucose tolerance. In addition, knocking out HNF1α, which underlies the defect in MODY 3, blunts insulin secretion and is accompanied by dwarfism and abnormal expression of genes involved in islet development and metabolism (190-192).
Insulin secretion generally exhibits a biphasic characteristic that is composed of a rapid but transient first phase of insulin release that is followed by a sustained second phase (193). A reduced or absent first phase insulin secretion in response to glucose is a common distinct feature in T2D. As the disease progresses, one can also observe a reduced second phase insulin release and diminished response to non-glucose stimuli. Among qualitative defects an increased proinsulin release in patients with T2D has been reported several decades ago, however, the mechanism underlying this response is not fully understood (40).
Hyperproinsulinemia and hyperinsulinemia also occur in pathological conditions such as insulinomas, which are rare endocrine tumors, characterized by uncontrolled β-cell proliferation. Inappropriate secretion of high levels of insulin leads to symptoms of hypoglycemia, which is counter-regulated by a high catecholamine output. Little is known about the minimum threshold levels for insulin and C-peptide to conclude that their secretion levels are inappropriately high and this is further complicated by the availability of specific assays to distinguish between proinsulin and insulin in rodents. The proportion of secreted proinsulin from insulinomas is generally higher when compared to normal β-cells. In humans, normal circulating levels for fasting insulin are 12.2–222.4 pmol/l (1.7–40 mIU/l), for C-peptide 0.26–0.63 nmol/l (0.78–1.89 ng/ml), and for proinsulin 7.9±1.5 pmol/l (194). Proinsulin levels exceeding 5 pmol in the presence of blood glucose levels below 2.5 mmol/l is generally accepted as a preoperative indicator for an insulinoma ((195) and references therein).
Most studies investigating the molecular pathways that underlie alterations in proinsulin processing have focused on the processing enzymes PC1/3 or PC2. One study investigated changes in gene expression profiles in sporadic insulinomas and found that genes for PC2 and islet amyloid polypeptide were upregulated (196). It is interesting to note that pro-islet amyloid polypeptide is also a substrate for PC2 and that these two genes, PC2 and pro-islet amyloid, coexist in secretory granules of β-cells (196). Since insulin itself is not up regulated at the transcriptional level, the authors suggest that the elevated insulin secretion in insulinomas is due to perturbations in the post-translational processing of proinsulin (196). While high circulating levels of insulin observed in T2D are associated with β-cell failure that potentially occurs due to insulin resistance in β-cells, the functional effects of sustained high levels of circulating insulin (or proinsulin) on β-cells function and growth is unclear.
The insulin secretion defects in mice with β-cell-specific deletion of insulin receptor (βIRKO), IGF1R (βIGFRKO), and IRS-2 (βIRS2KO) (91;119;128;197) or global deletion of IRS-1 (124), underscores the role of the proteins in these signaling cascades for the maintenance of appropriate insulin secretion (91;118;119). In agreement with this, the overexpression of insulin receptors in βTC6-F7 cells leads to both basal and stimulated insulin secretion (198). Both βIRKO and βIGF1RKO mice show a selective loss of first phase insulin secretion in response to glucose, but not arginine. βIRKO mice maintain second phase insulin secretion in response to glucose, which is blunted in βIGF1RKO mice. Both mice exhibit defects in glucose sensing, specifically suppression of GLUT2 transporters and the glucokinase gene, which are both crucial for glucorecognition. Furthermore, mice lacking both, insulin receptors and IGF1R simultaneously in β-cells exhibit, severe fasting hyperglycemia, glucose intolerance, altered Ca2+ oscillations, reduced oxygen consumption and blunted insulin secretion (99). Although IR and IGF1R signaling share several common signaling proteins downstream such as the IRS, PI3-K, PDK1 and Akt, the effects of insulin and IGF-I on insulin secretion overlap and are not identical. IRS-1KO mice display no differences in fasting glucose levels despite elevated basal circulating insulin levels but develop hyperglycemia in response to intraperitoneal glucose injections (124). Studies on islets isolated from the knockouts and SV-40 transformed β-cell lines derived from IRS+/- or IRS-/- mice suggest a significant role for IRS-1 in insulin secretion in response to glucose as well as arginine (95). Aspinwall et al. reported a significant function for IRS-1 in the regulation of [Ca2+]i levels (199), which was supported by the observation that IRS1-/- islets exhibit alterations in intracellular Ca2+ flux in response to glucose, which is accompanied by impaired insulin exocytosis (95). Another study with clonal β-cells demonstrated that overexpression of IRS-1 and insulin receptor elevated [Ca2+]i levels and enhanced fractional insulin secretion (200). A dramatically reduced islet expression of Ca2+-ATPAse (SERCA)-2b and-3 was identified in the mutants (201). Abnormalities in intracellular Ca2+ levels are related to some forms of T2D in both humans and rodents (202;203).
Several strategies have been adopted to determine the role of IRS-2 in β-cells. The significance of IRS-2 in the maintenance of β-cell mass, without alterations in non-β cells was reported by Withers et al. (92). However, Kubota et al. using mice of a different strain reported only a mild phenotype in the IRS-2KO suggesting a role for the genetic background in modifying the phenotypes in mice (126). Indeed, comparison of standard parameters of glucose homeostasis in inbred strains or in mice with mutations in the insulin receptor and IRS-1 on the 129Sv, C57Bl/6 or DBA/2 background confirms the impact of the genetic background on the phenotypes (204;205). Interestingly, insulin secretion in response to glucose stimulation, measured ex vivo, was significantly increased in IRS-2KO mice (126). Mice with a β-cell and hypothalamus-specific IRS-2 knockout (βHT-IRS2) also exhibited glucose intolerance and impaired glucose-induced insulin secretion due to reduced β-cell mass and number (128). Insulin secretion normalized to the same number of cells per islet, showed a significantly increased effect at high glucose concentrations in the mutants (128). Contrary to previous reports mentioned above (126) IRS-2 knockout generated by Cre recombinase driven by the PDX1 promoter with a consequent deletion of IRS-2 in all islet cells, and the mutants displayed attenuated insulin secretion in response to glucose, most likely due to impairment of intracellular calcium flux (130) and reduced expression of GLUT2 (130). One explanation for this difference in phenotypes is the significant role of IRS-2 in non-β-cells in the pancreas that may be linked to paracrine effects impacting β-cell function. In tissues such as liver, muscle and adipose PI3K determines sensitivity to insulin, while in β-cells the downstream component of the IR/IGF1R pathway plays a role in adjusting the insulin secretory response to glucose stimulation (206).
Non β-cells isolated from normal humans also express components of the insulin/IGF-I signaling pathway (88;100). Most studies to date have utilized whole islets derived from patients with T2D and reveal a significant reduction in gene expression for protein in the insulin/IGF pathway including the insulin receptor, IRS-2 and Akt2 (100). Islets from T2D patients also manifest the expected reduction in expression for insulin, GLUT1, GLUT2 and glucokinase (207;208). However, the molecular mechanisms underlying these changes in β and non-β-cells and potential crosstalk between signaling proteins among the different islet cell types is difficult to examine in human islets due to a limited source from patients with T2D.
The glucagon-secreting α-cells are important in the maintenance of overall glucose homeostasis. The precursor of the mature hormone - proglucagon - is encoded by a single gene in mammals and is expressed in pancreatic α-cells, intestinal L-cells and some neurons in the arcuate and paraventricular nuclei of the hypothalamus (209). Differences in posttranslational processing of the 160 amino acid precursor peptide within these cells lead to several biologically active proglucagon cleavage products (209). Cleavage of proglucagon in the α-cell due to the action of PC2 results in GRPP (glicentin-related pancreatic peptide), glucagon and the major proglucagon fragment (MPGF) which contains GLP1 and GLP2 in one single molecule (209). The excision of GLP1 requires PC1, which is not expressed in pancreatic α-cells (210). Glucagon, a 29 amino acid peptide, is a major product of the pancreatic α-cells, whereas glicentin, oxyntomodulin, GLP1 and GLP2 are produced in the L-cells of the intestine (209) (Figure 4).
The mechanisms underlying glucagon secretion are complex and include endocrine and paracrine effects as well as effects secondary to insulin, zinc, GABA, glutamate, somatostatin, ghrelin, GLP1 and glucagon (reviewed in (211)). Although hypoglycemia is a strong signal for glucagon secretion, some studies suggest that glucose is a direct stimulator of α-cells (212;213). The close proximity between the α- and β-cells and the direction of blood flow from the β-cell to the α-cells (214), has led to the proposal that intra-islet insulin impacts α-cell secretory function. Several secretory products including insulin and zinc, which are co localized in secretory granules of β-cells, have been suggested to regulate glucagon release (reviewed in (211).) Other mechanism that have been suggested to modulate glucagon release in response to decreasing plasma glucose levels include glucose sensing directly at the level of the α-cell membrane and sensing that is regulated via the autonomous nervous system (211).
Glucagon is one of the major counter regulatory hormones that opposes insulin action and increases hepatic glucose output. While glucagon is critical for preventing pathological hypoglycemia in healthy individuals, there is also evidence, that hyperglucagonemia is a feature of both T1D and T2D and several reviews address this topic (211). Glucagon receptor knockout mice display lower blood glucose accompanied by hyperglucagonemia and α- cell hyperplasia (157). PC2 knockout mice, which lack mature glucagon due to processing defects, display a similar reduction in blood glucose and manifest α and δ-cell hyperplasia (156).
By knocking out Gsα, a G protein essential for hormone stimulated cAMP generation, specifically in the liver (LGsKO mice), Chen et al. described mice with increased glucose tolerance, increased glucose stimulated insulin secretion, and enhanced insulin sensitivity in muscle and liver as well as very high serum glucagon and GLP1 levels and α-cell hyperplasia (215). The mechanisms and pathways that modulate α-cell growth and function warrant further investigation to develop therapeutic approaches to prevent hyperglycemia in both T1D and T2D.
Aging is associated with an increasing incidence in type 2 diabetes and a decline in glucose tolerance is reported to begin in the third or fourth decade of life (216). Glucose stimulated insulin release in human islets is inversely related with the age of the donor and independent of the donor’s BMI, sex or cause of death (217). The decline in β-cell function appears to occur as a result of impaired compensation of β-cells in response to increasing glucose levels which normally includes an adaptation by increasing the mass of β-cells either by modifying the number or size of cells, (replication and neogenesis) and/or by increased insulin secretion. β-cell function related genes have also been demonstrated to be differentially expressed at different ages for example PDX-1, insulin-2, GLUT2 and PC1/3 were all down regulated with age (217). Other observations in the elderly include an increased proinsulin to insulin ratio in response to glucose (218), and a decrease in the β-cells sensitivity to incretin hormones (219). Moreover, differential expression of components in the insulin/IGF-I pathway and cell cycle proteins during aging have also been observed in rodent pancreatic islet cells (67). Aging effects on β-cell proliferation appears to be linked with alterations in the expression of cyclin D2 in β-cells while non-β-cells express higher levels of p53, both proteins being key regulators of cell proliferation (67). Other investigators have suggested an important role for p53 in regulating senescence (220). In regard to cell cycle proteins, the CDKI p16INK4a, which is a stress dependent inducer of (premature) senescence, has been recently reported to constrain islet proliferation and regeneration in an age dependent manner (221). The proliferative kinase, CDK4, is effectively inhibited by the tumor suppressor p16INK4a, and this inhibition is specifically relevant in aging individuals as shown by overexpression and deletion studies, whereby effects were dramatically observed only with increasing age (221).
A thorough characterization of the role of each of the cell cycle proteins during aging and their potential link with proteins in the growth factor signaling pathway and metabolic pathways will allow a better perspective on potential candidates that can be harnessed for developing therapeutic approaches to enhance β-cell mass and/or function to prevent the development of T2D.
Understanding the mechanisms, that are critical to maintain β-cell function and for the compensatory β-cell adaptive response to hyperinsulinemia, is of critical interest in the field of islet biology. A related area of interest is the regeneration of β-cells to replace dying and/or senescent cells in diabetes and aging. Several reviews have thoroughly discussed strategies for increasing the availability of β-cells with the potential aim of treating patients with diabetes (23;222-224). Some of these transplantation approaches include islet encapsulation, using islets from species other than humans, immortalized human β-cell lines, insulin-producing cells derived from embryonic stem cells (ESCs) and pancreatic progenitors and transdifferentiation. The observation that adult rodent and human β-cells are formed by self-duplication rather than stem cell differentiation suggest that replication is a major source of increasing functional β-cells (35;36;39;63). Although lineage trace analysis are not possible in human, β-cell replication has been recently reported to be a major mechanism subserving the postnatal expansion of β-cell mass in humans (63). However, it is possible that other mechanisms also contribute to maintaining the β-cell mass. While replicating β-cells have an increased vulnerability to apoptosis in an proapoptotic environment of type 1 or 2 diabetes (224), a recent study using lineage trace analysis in rodents reported enhanced proliferation of the surviving β-cells in the diabetic environment (225). Recent work from Xu et al. described that multipotent progenitor cells exist in the pancreas of adult mice and can be activated cell autonomously to increase the functional β-cell mass by differentiation and proliferation rather than by self-duplication of pre-existing β-cells only (226). Thus, potential use of ESCs remains one of the most important tools for β-cell replacement therapy and developing optimal conditions for isolation, growth and differentiation of ESCs for mammals is a critical area of research. While, some progress has been made in this area, a fundamental question relates to the mechanism(s) that underlie pluripotency and differentiation of ESCs. The use of mouse ESC has greatly progressed whereas human ESC technology has only slowly become a reality. Notably, Kroon et al. have recently shown that pancreatic endoderm derived from human ESCs efficiently generates glucose-responsive endocrine cells after implantation into mice (227). Theses results are consistent with the generation of mature and functional islets. Moreover, they demonstrated that this implantation of hESC-derived from pancreatic endoderm protect mice against streptozotocin-induced hyperglycemia. Thus, these data are promising for the use of hESCs as a renewable source of islets for diabetes cell-replacement therapies. A word of caution is necessary in this context since some cell lines resulted in the development of teratomas.
We have discussed earlier above the potential role of insulin and IGF-I signaling pathways in the regulation of β-cell replication and β-cell regeneration. It is possible that this pathway is also significant for maintenance of pluripotency and early differentiation of stem cells based on the observation that both insulin and IGF-I are present in the growth media required for culture and maintenance of ESCs. Moreover, both transcripts are already present at the one-cell stage in human and mouse embryos (228;229). Although IR and IGF-IR do not seem to be essential for β-cell development single global knockouts for either of the receptors display general growth retardation and the compound IR/IGF-IR results in even smaller embryos (117). Moreover, IGF and insulin have been shown, in vitro, to accelerate preimplantation and to increase blastocyst formation (230). Besides, insulin and IGF-I can differentiate ESCs into different cell types in vitro including adipocytes (231), cardiomyocytes (232) or osteoblasts (233). Together, these data suggest a significant role for insulin and IGF-I for growth and differentiation of ESCs and warrant further investigation.
Since the discovery of insulin several decades ago a large body of work has defined the pleiotropic effects of the hormone and its cross-signaling effects mediated by the IGF-1 receptor to promote metabolism and regulate growth in virtually every cell in the mammal. However, the differential signaling pathways that distinguish insulin versus IGF-I effects in common target tissues are still elusive. Further questions remain especially in the context of endocrine lineage determination of embryonic β-cell growth and differentiation, the answers to which are likely to move efforts closer towards a cure for both forms of the disease.
We thank Lindsay Huse for excellent assistance with the preparation of the manuscript; R.N.K. acknowledges support from NIH RO1 DK 67536, RO1 DK 68721, American Diabetes Association Research Grant 7-04-RA-55, and a Grant from the Harvard Stem Cell Institute; Charlotte Hinault is supported by a NIH Training Grant (1RL9EB008539-01) (SysCODE).