Beta-Cell growth with aging
For many years it was speculated that β cells, much like neurons, were postmitotic and that their turnover in the mammal was minimal or zero. Over the last two decades, studies in mice have suggested a more dynamic picture, wherein β-cell mass can change in response to physiologic states such as growth, pregnancy, and obesity . Following weaning in rodents, there is considerable increase in β-cell mass that reflects the increase in body mass and an adaptation to the needs for increased insulin release. β-Cell mass is reflective of the changes in β-cell formation, individual β-cell size, and β-cell death.Whereas replication can be fairly reliably estimated using thymidine analog incorporation strategies, the rate of death is much more difficult to measure, primarily because dying β cells are cleared from the islet rapidly and therefore are difficult to detect.Thus, in studies of mature rodents, the turnover of β cells is estimated in large part from rates of replication. Studies in rats have shown that β-cell mass increases in a near-linear fashion with body weight [63,73]. The replication rate of β cells declines with age (to ≤2% per day in 20 month-old animals), but does not approach zero, and β-cell volume increases with age. These results suggest a low, but clearly measurable, turnover of β cells in adult rats and a β-cell lifespan in the order of 1–3 months. Using a continuous BrdU labeling strategy in mice, a much lower rate of replication has been estimated in adult mice, leading to the conclusion that β-cell turnover is near zero . Human β-cell mass accrual and replication rates are significantly more difficult to estimate, largely because cross-sectional data from a genetically diverse population must be extrapolated using static markers of replication (e.g. Ki67 or BrdU). Nevertheless, data from human autopsy samples suggest that there is accrual of β-cell mass with increasing body mass in children, with progressively decreasing potential for replication with age [66,74].
Compensatory beta-cell growth: adaptation to demand
The concurrent growth of β-cell mass with growth of the organism is but one example of the capacity of β cells to adapt to the increasing metabolic demands of peripheral tissue. Physiologic states of tissue resistance to insulin action, such as obesity and pregnancy, pose similar challenges to the β cell. It was recognized as early as the 1930s from autopsy studies that the average size of the islets of Langerhans increases as humans become overweight . Similar findings are seen in a variety of mouse, rat, pig, and other animal models of obesity, where both β-cell size and number are reportedly increased. The increase in β-cell mass in response to obesity reflects an adaptation to the increased insulin demands imposed by the resistance to insulin action in liver, muscle, and fat. Pregnancy imposes a challenge on the β cell similar to obesity, as pregnancy causes a state of tissue resistance to insulin. In either obesity or pregnancy, inherent defects that prevent increases in β-cell mass and insulin release may be the underlying causes for the development of diabetes. For example, in mouse models haploinsufficient for the gene encoding Pdx1, there is impaired compensation for insulin resistance in terms of both β-cell mass and function, with ensuing glucose intolerance and diabetes . Similarly, humans with heterozygous mutations of Pdx1 (a disorder known as maturity-onset diabetes of the young 4, or MODY4) develop diabetes with age, typically in adolescence or early adulthood . In these individuals, it is thought that β-cell compensation for linear growth and/or age-related insulin resistance is impaired. Pdx1 is crucial not only in the regulation of genes encoding β-cell proteins that are important in insulin secretion, such as the glucose transporter Glut2, glucokinase, and insulin, but also in the regulation of genes that are downstream of the growth-promoting insulin receptor/insulin-like growth factor 1 (IGF-1) receptor signaling cascade .
Origins of new beta cells in the adult: neogenesis, transdifferentiation, and replication
Considering the relatively small mass of β cells with respect to overall body mass, there has been a vigorous attempt over the last decade to define better the potential sources of new β cells in the growing mammal and to harness such sources for the creation of new β cells for those who are deficient. As discussed in the foregoing section, new β-cell formation was largely estimated by rates of β-cell replication, but excluded potential contribution from neogenesis. Therefore, if neogenesis were a major contributor to new β-cell formation, then rates of β-cell turnover were substantially underestimated. Speculation that a precursor β cell, or a true MPC, exists in the pancreas arose from early observations in rat models that new insulin-positive cells emanated from cells within proliferating . The question of the origin of new β cells in models of pancreas regeneration has been addressed using lineage tracing analysis in mice to show that β cells arise almost exclusively by replication of pre-existing insulin-positive cells rather than via neogenesis , a finding confirmed in subsequent studies in mice . However, these findings do not exclude the possibility that a rare, insulin-positive cell type with high proliferative capacity (i.e., a cell type that would not be defined as a mature β cell) has the ability to serve as a MPC, or that under certain conditions other cell types within the pancreas (i.e., facultative stem cells) have the capacity to differentiate to β cells. Thus, investigators continue to posit the existence of these alternative cell types in the pancreas whose differentiation into mature β cells may recapitulate a pathway of transcription factor expression similar to that seen in development .
Because all pancreatic epithelial cell types arise from a common Pdx1-positive precursor, it has been proposed that mature pancreatic cells of either exocrine or endocrine origin may have the capacity to directly differentiate into β cells without the need for de-differentiation into a precursor form (a process known as “transdifferentiation”). In this respect, although lineage tracing analyses have all but ruled out the possibility that mature acinar cells transdifferentiate under normal conditions to β cells in mice , the ectopic expression of the key developmental transcription factors Pdx1, Neurog3, and MafA in acinar cells enables a program that allows their conversion to insulin-expressing cells . Similarly, under specific experimental conditions in mice, mature α cells have the capacity to transdifferentiate into β cells . Taken together, these studies reinforce the theme that different mature cell types of the pancreas that arise from a common origin have the capacity to exhibit phenotypic characteristics of one another, and leave open the possibility that under specific conditions such cell types may transdifferentiate to offset loss of β-cellmass.
Whether any of the mechanisms discussed earlier—neogenesis or transdifferentiation—play a role in human β-cell replenishment remains uncertain. Interestingly, studies in vitro suggest the potential existence of precursor cell types in the human pancreas , but it is unknown whether and to what extent such cells give rise to β cells normally in humans. To date, the best available data indicate that replication of preexisting β cells is the likely mechanism for accrual of β-cell mass during human growth [66,74,82].
Regulators of beta-cell growth: growth factors and cell cycle regulators
A host of circulating factors appears crucial in the stimulation of early postnatal β-cell growth. As the growth of β cells closely parallels the growth of the organism during this early phase, it is relevant to note that nutrients, particularly glucose, remain among the most important factors contributing to β-cell replication during this period.Thus, intravenous glucose infusions for even short time periods (96 hours), which only kinase B). In recent years, a host of other growth factors has also been shown to positively influence β-cell replication and/or function (see Table 4.3), and include factors released not only from the islet, but also from a variety of organs, such as bone (osteocalcin), the anterior pituitary (growth hormone, prolactin), gut (glucagon-like peptide 1), fat (leptin, adiponectin), and brain (serotonin). Whereas these metabolites and growth factors can directly or indirectly impact β-cell replication, it should be noted that their effects are much greater in younger mice and humans and much less so as aging occurs.
Although the effects of the aforementioned growth factors result in enhanced β-cell replication and insulin release, the pathways leading to activation of cellular replication machinery differ depending upon the factor . Nevertheless, all factors ultimately impinge upon the components of the cell cycle. Transit through the cell cycle requires the β cell to exit the resting state (G0) and traverse G1, S, G2, and M states . For the most part, replication of β cells is largely driven by factors that control the G1/S transition of the cell cycle. Genetic manipulation studies in mice have emphasized the importance of not only activators of the G1/S transition, but also inhibitors, such that the balance between the two appears to regulate the overall drive for β-cell replication. Cyclins and cyclin-dependent kinases (Cdks proteins) are major activators of β-cell replication. Cyclins and Cdks negatively regulate the major pocket protein known as pRb, which functions as a mildly increased serum glucose concentrations, result in fivefold increases in β-cell replication in young mice . Although an effect of glucose to directly stimulate β-cell replication has been proposed, it is possible that its effect may be caused by its stimulation of insulin release from β cells, such that insulin in an autocrine manner serves as the mitogen. Insulin and insulin-like growth factor 1 are classic growth factors that signal through related transmembrane receptors with associated receptor tyrosine kinases. Mice lacking the insulin receptor in β cells display impaired insulin release associated with reduced β-cell mass, whereas mice lacking the IGF-1 receptor in β cells display impaired insulin release without associated loss of β-cell mass. Interestingly, loss of both the insulin receptor and IGF-1 receptor in β cells results in severe reductions in β-cell mass and frank diabetes . These data suggest that the insulin and IGF-1 signaling pathways function in distinct, but complementary ways, notwithstanding that both receptors share similar downstream signaling molecules (insulin receptor substrate proteins, phosphatidyl inositol-3 kinase, and protein “molecular brake” on the G1/S transition. Cyclins and Cdks appear crucial in the accrual of early postnatal β-cell mass, but interestingly not in the generation of β-cell mass in the embryo.
Mice homozygous null for the gene encoding CyclinD2 or Cdk4 exhibit no alterations in β-cell mass at birth, but show loss of mass accrual with age [84,85]. Similarly, loss of the gene encoding CyclinD1 does not affect embryogenesis, but heterozygosity of the CyclinD1 gene in combination with homozygous loss of CyclinD2 results in even further loss of β-cell mass with age and severe, life-threatening diabetes . The cyclin-dependent kinase inhibitors (CKIs)—including p15Ink4b, p16Ink4a, p18Ink4c, p19Ink4d, p21Cip, p27Kip1, and p57Kip2—are major negative regulators of β-cell proliferation, and their actions appear to predominate in later life, where these factors may be responsible for inhibition of β-cell proliferation in aging mammals .
Recent studies have clarified the human islet G1/S cell cycle protein expression pattern . Whereas murine and human islets differ in their expression of the G1/S cell cycle activator Cdk4 (humans express Cdk6), they have virtually all G1/S CKIs in common. This latter observation may be crucial in the understanding of why β-cell replication is so dramatically reduced in aging humans. A particularly intriguing target in this respect is p16Ink4a, whose expression in β cells is up-regulated as mice age, and may serve as a target to prevent the age-induced limitations in β-cell mass . The potential for β cells to undergo uncontrolled replication as a result of deregulation of G1/S cell cycle proteins is dramatically emphasized by mutations in the gene encoding Menin in both mice and humans. Menin is a tumor suppressor transcription factor that negatively regulates the expression of p18Ink4c and p27Kip1, and its absence or mutation results in the tumorigenic transformation of a variety of endocrine tissues (including β cells) in a syndrome known as multiple endocrine neoplasia 1 (MEN1) .
Conclusions and areas of future study
In the early twentieth century, the discovery of insulin dramatically transformed the treatment of diabetes mellitus. Indeed, it was thought that the administration of insulin might reduce stress and allow for the time necessary to regrow new β cells, a consequence that was never observed. In the ensuing decades, the incidence of type 2 diabetes rose to dramatic proportions, and as a result the quest for β-cell-based therapies for diabetes has seen broader appeal. As discussed, more recent research has led to dramatic insights into pancreas and β-cell development, and into the postnatal life cycle of the β cell. Although most of these insights derive from studies in lower animal species, their applicability to the treatment of human diabetes mellitus has risen to the forefront of discussion in recent years. Importantly, we know now that although β cells have the capacity to expand in the postnatal period, in humans the window for such expansion may be limited to the first 2–3 decades of life, and thereafter the ability to compensate for physiologic stressors (such as obesity) diminishes with age. As such, strategies for therapies for diabetes in the future may well focus on ways to enhance β-cell replication or to engineer new β cells. With respect to the latter, studies of embryonic development have enabled important strides in generating β-like cells from primitive biologic precursors (e.g. human embryonic stem and induced pluripotent cells). Yet, these engineered cells do not exhibit the full phenotypic spectrum of true β cells, such as the ability to release insulin in response to a physiologic glucose challenge. The knowledge that all cells of the pancreas arise from a common progenitor has raised awareness that plasticity of fully differentiated pancreatic cell types may be much greater than originally thought. In this respect, studies of transdifferentiation of other abundant pancreatic cell types (such as α cells or acinar cells) to β cells in vivo may hold promise for the treatment of human diabetes, but to date no clear examples of human cell transdifferentiation have emerged. As the burden of diabetes increases, the need to translate research from lower animals to humans increases, and in the coming years it is likely that the generation of better model systems that mimic the human condition will become a greater priority.