Insulin and IGF-1 signaling in mammals
Although the core of the insulin/IGF-1 signaling pathway is conserved between invertebrates and mammals (
Figure 1), the mammalian IIS network has greatly increased in complexity [
18]. In mammals, three different insulin/IGF-1 receptor ligands are present: insulin, IGF-1 and IGF-2. Three different mammalian insulin/IGF tyrosine kinase receptors have been identified: the insulin receptor (IR), IGF-1 receptor (IGF-1R), and the orphan IR related receptor (IRR). In addition, a structurally and functionally distinct mannose-6-phosphate IGF-2 receptor exists, which is thought to have evolved primarily as a scavenger receptor for IGF-2 [
19]. After ligand binding, the activated IGF-1 or insulin receptor phosphorylates several intracellular substrates, including IR substrates (IRS) and the Src-homology-2-domain containing transforming protein (Shc). The phosphorylated substrates provide specific docking sites for intracellular effectors, including the p85 regulatory subunit of PI-3K and Growth-factor-receptor-bound protein-2 (Grb2), thus leading to the activation of two major signaling pathways, the PI-3K-PKB/AKT pathway and the Ras-MAPK pathway (
Figure 1). The PI-3K- PKB/AKT pathway has been shown to regulate most of the metabolic effects of insulin/IGF-1 signaling, whereas the Ras-MAPK pathway had been shown to regulate most of the mitogenic effects of insulin/IGF-1 signaling [
18].
Further adding to the complexity, for most of the critical components of the mammalian insulin/IGF-1 signaling cascade different forms encoded by different genes and/or different isoforms encoded by a single gene have been identified [
18]. Two isoforms exist of the insulin receptor (encoded by one gene), IR-A (lacking exon 11) and IR-B (including exon 11), that show pronounced functional differences [
20]. IR-A was found to exhibit high affinity for IGF-2, and has been predominantly implicated in mitogenic IGF signaling, whereas IR-B has been predominantly implicated in metabolic insulin signaling. Moreover, hybrid IR-IGF-1R complexes might be formed that show distinct affinities for insulin, IGF-1 and IGF-2. Likewise, three different isoforms exist of PKB/AKT (PKB α, β and γ or AKT1–3) [
18]. Four distinct IRS proteins have been identified in various mammals (IRS1–4), and two additional IRS proteins have been detected in humans (IRS5–6). IRS5 and IRS6 were shown to be involved in insulin signaling, but not in PI-3K activation [
21]. For class 1A of PI-3K, eight different isoforms (derived from three genes [
22]) have been identified of the regulatory subunit that can associate with three different forms of the PI-3K catalytic subunit (p110α, β and δ) [
23]. Moreover, four different members of the mammalian FoxO family of transcription factors have been identified (FOXO1A, FOXO3A, FOXO4 and FOXO6) [
24]. The existence of different forms and isoforms that differ in their tissue distribution, subcellular localization and interactions with downstream targets and components from other signaling pathways has greatly enhanced the possibilities for tissue specificity, diversification and fine-tuning of IIS transduction under various physiological states [
18]. Concordantly, the complexity of the IIS network has increased enormously. In addition, the mammalian IIS network is tightly linked to growth hormone (GH), that appeared relatively late in evolution within vertebrates. GH produced by the anterior pituitary regulates the biosynthesis and release of IGF-1 by the liver and peripheral tissues. However, GH has other effects on IIS in addition to the effects mediated by IGF-1. Binding of GH to the GH receptor (GHR) results in activation of the associated Janus kinase (JAK2) [
25]. In turn, JAK2 activates several intracellular mediators leading to different signaling pathways. The main GH signaling pathways comprise besides STAT (signal transducers and activators of transcription) signaling, the PI-3K- PKB/AKT and the MAPK pathways. Moreover, additional pathways, independent of JAK2, have been described [
25].
Go to:
Insulin and IGF-1 signaling and longevity in mammals
Various mouse mutants with reduced GH/IGF-1/insulin have been shown to be long-lived (
Table 1), but the increased complexity of the mammalian IIS network has made it difficult to disentangle the roles of GH, IGF-1 and insulin.
FIRKO mice, in which the insulin receptor was specifically deleted in fat tissue, provide evidence for a link between longevity and reduced insulin signaling [
26]. In addition to being long-lived,
FIRKO mice exhibit a reduction in fat mass and lessened age related loss of insulin sensitivity [
27]. Available evidence indicates that enhanced mitochondrial capacity of white adipose tissue may contribute to the resistance of
FIRKO mouse to diet induced obesity [
28]. Data from other mutant mouse models support a link between reduced GH/IGF-1 signaling and longevity. In mammals, GH produced by the anterior pituitary regulates the biosynthesis and release of IGF-1 by the liver and peripheral tissues to control mammalian growth. Four dwarf mouse models with impeded IGF-1 production, namely
Prop1df/df [
29],
Pit1dw/dw [
30],
GHRHRlit/lit [
30] and
GHR−/− [
31] all show a long-lived phenotype. Common characteristics of these long-lived GH-deficient dwarfs and GH-resistant dwarfs include reduced circulating levels of insulin and glucose and enhanced insulin sensitivity [
32]. Results obtained with mice mutated for the IGF-1 receptor hint at a direct role for reduced IGF-1 signaling in mammalian longevity:
Igf1r+/− females, but not males, exhibit a long-lived phenotype as well as increased resistance to oxidative stress [
33]. Overexpression of
Klotho can inhibit IIS [
34] and extend lifespan, whereas
Klotho mutant mice age prematurely [
35]. Thus far, the strongest effects on life span in mouse mutants with defective GH/IGF-1 and/or insulin signaling have been observed in the GH deficient hypopituitary dwarfs and the GH resistant
GHR−/− dwarfs. Recent evidence strongly suggests that enhancement of insulin sensitivity, in conjunction with reduced insulin levels, is a key factor in the longevity phenotype of these mice as well as in wild type mice subjected to caloric restriction [
36]. However, insulin resistance has been reported for other mouse models with extended longevity, including
Klotho transgenic mice [
34],
IRS1−/− mice [
37] and mice with a brain specific deletion of IRS2, which are long-lived when fed a high fat diet [
38]. It has been suggested that a key feature shared between these insulin resistant mice and the insulin sensitive dwarfs is a reduced strength of the insulin signal in specific key insulin target tissues or organs [
36].
Go to:
IIS and human longevity: studies on genetic polymorphisms
Based on the observed associations between reduced insulin/IGF-1 signaling and longevity in organisms as diverse as worms, flies and mice and given the evolutionarily conservation of the core IIS pathway components, it could be speculated that the genes involved in insulin/IGF-1 signaling might be important for human longevity as well. However, results from human studies have been conflicting and controversial. In humans, defects in insulin signaling have been associated with insulin resistance and diabetes [
18]. Also, defects in GH/IGF-1 signaling have been associated with defects in growth and increased risk of cardiovascular disease [
39]. However, despite their obesity, patients with Laron syndrome, a human dwarf disease that is associated with IGF-1 deficiency, do not exhibit premature death and seem protected against cancer [
40]. Moreover, common polymorphisms in several of the IIS genes have been associated with longevity across diverse cohorts. Genotype combinations at
IGF-IR and
PI3KCB genes were found associated with lower free IGF-I plasma levels and were found to be enriched in Italian centenarians [
41]. In the Leiden 85-plus Study, a composite score was calculated based on the expected effects (increased or reduced signaling) of genetic variants in the
GHRHR,
GH1,
IGF-1,
INS and
IRS1 loci [
42]. In nonagenarian women of the Leiden 85-plus Study, a lower composite score was found to be associated with shorter stature and improved old age survival [
42], as well as with reduced cognitive decline [
43]. In studies on Ashkenazi Jewish centenarians and their offspring, higher serum levels of IGF-1 were associated with smaller stature in female offspring of centenarians [
44]. Sequence analysis showed overrepresentation of heterozygous mutations in the
IGF-1R gene among centenarians that were associated with high serum IGF-I levels and reduced activity of the IGF-IR as measured in transformed lymphocytes. Also in Italian centenarians, a higher plasma IGF-I/IGFBP3 molar ratio was found that was positively associated with whole body glucose disposal rate [
45]. Another study showed enrichment of a haplotype in the
INSR gene in Japanese semisupercentenarians [
46]. Variants in
AKT were found associated with longevity across three Caucasian cohorts [
47]. Variants in
FOXO3A have been associated with longevity in an ethnic Japanese population in Hawaii [
48], as well as in four different Caucasian cohorts [
47,
49,
50] and in a Chinese cohort [
51]. Variation in
FOXO1A was found associated with higher Hb1Ac levels and mortality in old age [
52]. To date, although only few findings have been systematically replicated in different cohorts and confirmed in meta-analyses, these data seem to indicate that of the single genes of the IIS pathway that have been systematically analyzed across different cohorts, variation in
FOXO3A is most consistently associated with human longevity [
1].
Go to:
IIS and human longevity: studies on insulin sensitivity
In mammals, carbohydrates are an essential fuel source for the central nervous system and the immune system. In response to high levels of circulating glucose, the pancreas secretes insulin, which stimulates the uptake of glucose and its subsequent metabolism (glycolysis and glucose oxidation in the muscle) and the storage of excess carbohydrates (glycogenesis in the liver and lipogenesis in adipose tissue). In response to low circulating glucose levels, metabolism is shifted towards the breakdown of fat reserves (lipolysis in adipose tissue): the liberated fatty acids are used for fatty acid oxidation in the muscle and glycerol is used for the synthesis of glucose in the liver (gluconeogenesis), thus proving carbohydrates for the central nervous system. Under conditions of low circulating glucose levels, glycolysis and lipogenesis are suppressed. With age, insulin sensitivity progressively declines, which significantly contributes to the increased incidence of type 2 diabetes mellitus in older people [
53]. Remarkably, centenarians [
54] and their offspring [
55], as well as the offspring of nonagenarian siblings were found to have a reduced risk of diabetes [
56]. In a sample of 52 healthy subjects representing three different age categories of the Italian population (adults, aged subjects and centenarians), centenarians were found to exhibit preserved glucose tolerance, as well as preserved insulin sensitivity as assessed by the hyperinsulinemic euglycemic clamp technique [
57]. Whole body glucose disposal rate (per kg fat-free mass) was significantly higher in centenarians (mean age: 102 years) than in aged subjects (mean age: 78 years) and was comparable to that of adults (mean age: 44.5 years) [
57]. In a sample of 466 healthy Italian subjects, covering an age range from 28–110 years, insulin resistance (as determined by homeostasis model assessment) was shown to increase with age and reach a peak around the age of 80 years. However, beyond the ages of 85–90 years, insulin resistance declined again and a group of subjects with a lower degree of insulin resistance emerged [
58]. It is unresolved to what extent the preserved insulin sensitivity in centenarians reflects selective survival of subjects with genetically determined favorable insulin sensitivity. Recently, offspring of long-lived nonagenarian siblings were also found to have lower levels of fasting glucose and insulin, a hallmark of enhanced insulin sensitivity, as well as better glucose tolerance compared to a control group of similar age and body composition [
59]. Likewise, it is not clear which biological mechanisms contribute to the preservation of insulin sensitivity in centenarians. Interestingly, centenarians were shown to have higher serum levels of insulin sensitizing hormones, most notably adiponectin [
60].
Taken together these data suggest that, as in the GH deficient hypopituitary dwarfs and the GH resistant
GHR−/− dwarfs, low glucose, low insulin and preserved insulin sensitivity may represent a key metabolic feature of a human longevity phenotype. Although speculatively, these metabolic features might reflect a state of reduced flux through the IIS pathway and enhanced FoxO activation.
Go to:
Strength of the insulin signal: insulin versus Foxo
Class O forkhead box transcription factors (FoxOs) may act as a “master-switch” to adapt cells and organisms to food shortage and ensure metabolic stability under conditions of food shortage and thus opposes many of insulin functions [
24]. In addition to PI-3K/PKB-AKT, other kinases, including AMPK (5’ adenosine monophosphate-activated protein kinase), JNK (c-Jun NH2-terminal kinase), and IKKβ (inhibitor of nuclear factor kappa-B kinase subunit beta) can also phosphorylate FoxO, centralizing its role as a master-switch [
61]. Moreover, in addition to phosphorylation, other posttranslational modifications, including acetylation and ubiquitination are important in the regulation of FoxO activity [
24,
61].