. Reduced IGF-1 signaling delays age-associated proteotoxicity in mice. Cell. 2009 Dec 11;139(6):1157-69. PubMed.

Recommends

Please login to recommend the paper.

Comments

Make a Comment

To make a comment you must login or register.

Comments on Primary Papers and News

  1. The work published by Cohen et al. shows that reduced IGF-1 signaling reduces pathology in a classical mouse model of AD, the double APP/PS1 transgenic mouse. Albeit with several discrepancies, this work largely confirms a previous one by Freude et al. (Freude et al., 2009), where deletion of the IGF-1 receptor in the mouse forebrain also protected against AD pathology. From this evidence one may conclude that reduced IGF-1 signaling is a promising therapeutic strategy in AD.

    The notion that the insulin/IGF-1 signaling (IIS) pathway is involved in Alzheimer disease (AD) is gaining momentum (Craft and Watson, 2004; Gasparini and Xu, 2003). However, whether IIS is detrimental or beneficial to the disease is a matter of current debate. Pros and cons abound for each position. Probably the idea that reduced IIS has a general salutary effect stems from the observation in invertebrates that low IIS promotes longevity (Kenyon, 2001). There is also evidence favoring a similar role in mammals, including humans (Suh et al., 2008). It was next ascribed a similar protective role against cancer (LeRoith and Roberts, Jr., 2003). One of the few remaining protective actions of IIS (the other undisputed one is metabolic disturbances) was on the aging and/or diseased brain (Aleman and Torres-Aleman, 2009; Torres-Aleman and Fernandez, 1998).

    Indeed, there is evidence for a positive action of IGF-1 on the brain. In a study of growth hormone (GH)-deficient adults, GH was shown to improve cognitive functioning, and Vitiello et al. (2006) reported beneficial effects of six months of GH releasing hormone versus placebo in a group of 89 healthy older (68 years) adults. In the treatment group, IGF-1 levels increased by 35 percent, but only by 1 percent in the placebo group. In addition, increased IGF-1 levels were found of therapeutic benefit in mouse models of AD (Carro et al., 2002; Carro et al., 2006). A positive correlation between serum IGF-1 levels and health status has also been consistently reported in the elderly (Nindl and Pierce, 2010). Furthermore, brain oxidative stress, a pathological disturbance usually linked to AD (Behl et al., 1994), elicits neuronal death through FOXO3 (Lehtinen et al., 2006), a process that requires inhibition of IGF-1 signaling (Davila and Torres-Aleman, 2008), as FOXO3 is downstream of IIS. Confirming this notion is the observation that IGF-1 receptor (IGF-1R) heterozygous mice showed increased neuronal damage in response to MPTP challenge (Nadjar et al., 2008).

    However, the observations of Cohen’s and Freude´s labs questions this positive action of IGF-1 in the brain. Both groups used transgenic mice with reduced IGF-1 receptor function cross-bred with classical mouse AD models. The use of transgenic models has been crucial in furthering our understanding of human diseases. However, transgenic models alone are not sufficient to establish pathogenic mechanisms of interest to human diseases. In most cases, these models provide partial information that requires additional experimental approaches, more so when the information provided is contradictory. Thus, while Freude et al. used mice without IGF-1 receptors in forebrain neurons and found reduced Aβ plaque load, Cohen et al. used a mouse with reduced overall IGF-1R content and report increased Aβ plaque load that was equally neuroprotective as it shows lower neurotoxicity. Indeed, they report marked reduction of neuronal death, which is not usually observed in APP/PS1 mice. Notwithstanding the lack of data showing reduced brain IGF-1 signaling in these mice, and assuming that this is the case, the mouse models used express the genetic modifications already from developmental stages. Therefore, it is possible that undetermined compensatory responses develop in these mice to provide protection against AD. Alternatively, elimination of the IGF-1R may not be equivalent to inhibition of IGF-1 signaling only, as IGF-1 may play additional, not yet defined roles.

    Nevertheless, the fact that either global or brain-specific reduction of IGF-1R levels similarly protects against AD pathology, together with the reported ambivalent actions of IIS signaling on APP metabolism (Adlerz et al., 2007; Shineman et al., 2009), and the lack of protection against AD after increased IGF-1 levels (Lanz et al., 2007), opens the question of what is the role of IIS in AD. While we will probably agree that IIS has a role in AD, defining it may prove more complex. This may reflect the complexity of this ancient hormonal system. In all probability the key aspect is a balanced IIS function.

    See also:

    Torres-Aleman I, Fernandez AM (1998) The role of Insulin.like growth factors in human neurodegeneration. pp 131-149. Springer Publishing Co.

    References:

    . IGF-1-induced processing of the amyloid precursor protein family is mediated by different signaling pathways. J Biol Chem. 2007 Apr 6;282(14):10203-9. PubMed.

    . Hydrogen peroxide mediates amyloid beta protein toxicity. Cell. 1994 Jun 17;77(6):817-27. PubMed.

    . Therapeutic actions of insulin-like growth factor I on APP/PS2 mice with severe brain amyloidosis. Neurobiol Aging. 2006 Sep;27(9):1250-7. PubMed.

    . Serum insulin-like growth factor I regulates brain amyloid-beta levels. Nat Med. 2002 Dec;8(12):1390-7. PubMed.

    . Insulin and neurodegenerative disease: shared and specific mechanisms. Lancet Neurol. 2004 Mar;3(3):169-78. PubMed.

    . Neuronal death by oxidative stress involves activation of FOXO3 through a two-arm pathway that activates stress kinases and attenuates insulin-like growth factor I signaling. Mol Biol Cell. 2008 May;19(5):2014-25. PubMed.

    . Neuronal IGF-1 resistance reduces Abeta accumulation and protects against premature death in a model of Alzheimer's disease. FASEB J. 2009 Oct;23(10):3315-24. PubMed.

    . Potential roles of insulin and IGF-1 in Alzheimer's disease. Trends Neurosci. 2003 Aug;26(8):404-6. PubMed.

    . A conserved regulatory system for aging. Cell. 2001 Apr 20;105(2):165-8. PubMed.

    . Peripheral elevation of IGF-1 fails to alter Abeta clearance in multiple in vivo models. Biochem Pharmacol. 2008 Mar 1;75(5):1093-103. PubMed.

    . A conserved MST-FOXO signaling pathway mediates oxidative-stress responses and extends life span. Cell. 2006 Jun 2;125(5):987-1001. PubMed.

    . The insulin-like growth factor system and cancer. Cancer Lett. 2003 Jun 10;195(2):127-37. PubMed.

    . IGF-1 signaling reduces neuro-inflammatory response and sensitivity of neurons to MPTP. Neurobiol Aging. 2009 Dec;30(12):2021-30. PubMed.

    . Constitutively active Akt inhibits trafficking of amyloid precursor protein and amyloid precursor protein metabolites through feedback inhibition of phosphoinositide 3-kinase. Biochemistry. 2009 May 5;48(17):3787-94. PubMed.

    . Functionally significant insulin-like growth factor I receptor mutations in centenarians. Proc Natl Acad Sci U S A. 2008 Mar 4;105(9):3438-42. PubMed.

    . Growth hormone releasing hormone improves the cognition of healthy older adults. Neurobiol Aging. 2006 Feb;27(2):318-23. PubMed.

  2. This is an exciting piece of research from Ehud Cohen and coworkers from Andrew Dillin’s group at The Salk Institute (La Jolla, CA), concerning the role of insulin-like growth factor (IGF)-1 receptor signaling in the pathogenesis of Alzheimer disease (AD). Recent data suggest that IGF-1 receptor signaling might be involved in the development of AD, since postmortem studies on brains of AD patients showed decreased insulin receptor and IGF-1 receptor expression (1). However, it is unclear whether these changes in IGF-1 receptor signaling are cause, consequence, or even counter-regulation to neurodegeneration.

    The work of Ehud Cohen and coworkers revealed new insights into this highly debated field. In 2006, the same research group showed that reduced daf-2 (ortholog to insulin/IGF-1 receptors in mammals) signaling in C. elegans protects the worms from Aβ toxicity via a heat shock factor 1 (HSF-1)-dependent mechanism, which regulates Aβ disaggregation, and a DAF-16 (ortholog to FOXO in mammals)-dependent mechanism, which facilitates the formation of larger, less toxic Aβ aggregates (2). Subsequently, this hypothesis is now tested in a mouse model of Alzheimer disease.

    Cohen and coworkers present data from an AD mouse model expressing the Swedish mutation of APP and the presenilin-1 ΔE9 variant (AD mice) that were additionally heterozygous for the IGF-1 receptor (IGF-1R). These mice presented partial memory restoration and improved motor skills compared to their AD littermates. Interestingly, reduced IGF-1R signaling protected these animals from neuronal loss. Furthermore, these mice provided higher synaptic density, decreased inflammation, and smaller but more condensed Aβ aggregates compared to AD mice.

    These data corroborate the findings in C. elegans and reveal a novel and promising mechanism of Aβ oligomer detoxification via enhanced aggregation in mammals. In line with the current work, there have been two more reports this year suggesting a beneficial effect of reduced insulin receptor or IGF-1R signaling on different aspects of AD pathology (3,4). However, most of the beneficial effects of partial IGF-1 resistance described in Cohen's paper occur in later stages of disease, suggesting a complex interaction between disease progression and transcriptional changes triggered by mild IGF-1 resistance. Interesting candidates for mediating these effects might be the Foxo transcription factors, which are regulated by IGF-1R signaling, as well as "stress" kinases. The current paper might also prompt a reinterpretation of previously published studies suggesting a beneficial effect of IGF-1 treatment on AD pathology (5). In summary, the work of Cohen and colleagues reveals a novel IGF-1-dependent mechanism, conserved from C. elegans to mammals, to reduce Aβ toxicity by facilitating the formation of larger, less toxic Aβ aggregates.

    View all comments by Katharina Schilbach
  3. Maintaining the Correct Balance of IGF-1R Signaling in the Brain With Age May Protect Against AD
    This very interesting study led by Andrew Dillin shows that crossing long-lived heterozygous igf-1r+/- mice (10) with AD APPswe/PS1ΔE9 mice (11) delays age-related Aβ proteotoxicity and protects against several AD-like symptoms. The major finding of the work shows reducing total levels of IGF-1R signaling by 50 percent in the entire mouse is associated with the emergence of more dense, tightly packed Aβ plaques in the brain, which most likely sequester potentially synaptotoxic, oligomeric Aβ. This raises the exciting possibility that diminishing signaling through IGF-1R may enable the formation of more “inert” Aβ plaques and diminish Aβ oligomer synaptic toxicity in patients with AD.

    The insulin/insulin-like growth factor-1 (IGF-1) receptor signaling (IIS) pathway has long been a subject of fascination in aging research and in understanding the regulation of lifespan. DAF-2 is the one and only insulin/IGF-1 receptor in Caenorhabditis elegans, and inhibition of its expression generates long-lived, stress-resistant worms. So too, does dampening down the activity of several key components of DAF-2’s downstream signaling pathways, particularly the pathway that is analogous to mammalian PI3-K/Akt signaling, thus enabling activation of DAF-16 (FOXO transcription factors), causing upregulation of stress-resistance genes (for review, [16]). Links between AD pathology and downregulation of the IIS pathway were highlighted when earlier studies from Dillin’s group showed that knockdown of Daf-2 signaling reduces Aβ42 aggregation-induced toxicity (2). Mechanistically, this was associated with increasing the de-aggregation of Aβ42 or increasing packing of Aβ42 into less toxic Aβ plaques through activation of HSF-1 and FOXO family (DAF-16) transcription factors.

    The evolutionary conservation of the IIS pathway emphasized the possibility that it had similar function in lifespan regulation in mammals. However, in mammals the situation is more complex because distinct insulin and IGF-1 receptors exist, with specialized and overlapping roles; the insulin receptor (IR) is generally more associated with metabolic control, and the IGF-1 receptor (IGF-1R) with growth. In addition, hybrid IGF-1R/IRs can form. In 2003, Holzenberger and colleagues showed for the first time that IGF-1R is a key regulator of mammalian lifespan. Deletion of both copies of igf1r was lethal; however, deleting one copy of the gene (igf1r +/- mice) created animals that live an average 26 percent longer than mice with two copies of igf1r. Interestingly, this effect only attained significance in female mice, which live an average of 33 percent longer than wild-type littermate control mice (10).

    More recently, the Holzenberger group used conditional mutagenesis to specifically delete IGF-1R in the brain, showing that, as in the nervous system of worms and flies, the brain IGF-1R actually controls mammalian lifespan, and also growth, through a neuroendocrine mechanism (12). This also ties in with other studies showing mice with no IRS-1 (17), lower levels of insulin receptor substrate 2 (IRS-2) (21), or TOR (9)—key signaling proteins directly downstream of IGF-1R/IR—live longer.

    In this most recent Cell paper by the Dillin group, their findings in worms are now extended to mice. As reviewed in Alzforum, crossing long-lived heterozygous Igf1-r+/- mice with AD APPswe/PS1ΔE9 to create Igf1r+/-/AD mice delays age-related Aβ proteotoxicity and prevents several AD-like symptoms. It is not clear if this shows a female gender bias, as was shown for female igf1R+/- mice with respect to longevity. A recent publication from Markus Schubert’s group (4) supports these findings showing that both male and female Tg2576 AD mice with neuronal specific deletion of Igf1r (nIGF-1R), mostly in the hippocampus (but not with neuronal-specific deletion of IR), were protected from premature death and had decreased Aβ production with age.

    Could the same situation extend to humans? Does having less IIS signaling and less IGF-1R increase longevity? Are increased levels of IGF-1R and activity of the PI3K-Akt (the major IIS effector pathway) linked to the neurodegenerative process of AD? Should we be considering treatments that attempt to reduce and/or normalize the IIS signaling pathway in AD? Importantly, the answer to all the above questions would appear to be yes.

    Firstly, when considering human aging, functionally relevant IGF-1R mutations, with downregulated IGF-1R activity, have been discovered in female centenarians (20). In addition, low IGF-1, PI3K, and IRS-1 correlate with prolonged lifespan (1) (23). When considering age-related neurodegeneration in AD, signaling through the IGF-1R and IR is patently disturbed (5,6,8,15,18). These defects describe both increased (15) and decreased levels of IGF-1R (18), with unchanged (15) or decreased levels of IR (18). Recently, we performed a very detailed analysis of the levels and localization of IGF-1R, IR, and IRS1/2 proteins in the postmortem temporal cortex of individuals who had AD (15). Results clearly show that overall levels of IGF-1R are significantly increased in people with AD compared to people of the same age without the disease. Significantly increased IGF-1R levels (by about 50 percent) are found in activated GFAP immunopositive astrocytes, and in degenerating synapses and neurites within and surrounding Aβ plaques in AD. However, in contrast, total IGF-1R levels are actually decreased and their subcellular localization is altered in neurons in AD, particularly those with neurofibrillary tangles (NFTs).

    Unlike IGF-1R, IR levels are the same in AD and control groups with expression only evident in neurons; however, IRs show an altered localization with internalization of neuronal IR in the disease (15). Consistent with other findings (14,18), we also reported decreased levels of both IRS-1/2 in AD neurons, and increased levels of major inactivating IRS-1 phosphorylation motifs at Ser312/616. In previous work, we discovered an excessive hyperactivation of the Akt signaling pathway in the same AD neurons, with an eventual loss of Akt signaling neurons (8). Together, the findings lead us to speculate that an excessive and inappropriate hyperactivation of Akt in AD neurons, possibly induced by Aβ oligomers, as has been shown using in vitro neuronal systems (25), induces feedback inhibition of both IGF-1R and IR through inactivation of the key adaptor IRS proteins.

    With consideration of the above findings in AD brain, an exploration of the cellular and subcellular localization of IGF-1R in the newly created igf1R+/-/AD mice, and the receptor’s relationship to Aβ pathology, astroglial activation, and synaptic loss compared to AD mice with the full complement of IGF-1R, would be most informative. Are increased levels of IGF-1R associated with activated astrocytes in AD mice, as found in the AD brain, thus explaining the 50 percent reduction in astrocytosis in the igf-1R+/-/AD mouse? Are increased levels of IGF-1R associated with Aβ plaques in the AD mice but not present in the morphologically distinct denser plaques in igf1R+/-/AD mice? What is the status of IGF-R downstream signaling in neurons in igf-1R+/-/AD and AD mice?

    Previous investigation of IGF-1R status in Tg2576 AD model mice showed increased levels of IGF-1R in neurons at six months prior to the emergence of extracellular Aβ plaque pathology (19), indicating a possible upregulation of IGF-1R by Aβ species that precedes overt plaque pathology, possibly as a protective mechanism. Other work has shown a highly significant upregulation of IGF-1R surrounding extracellular Aβ plaques at later ages in these mice (15). Together, this indicates that IGF-1Rs may be an important functional responder to Aβ oligomers. Of note, diverse Aβ species, including monomers (24) and soluble oligomers such as ADDLs (22, 25), bind the closely related IRs, but it is not clear whether Aβ also interacts with the IGF-1R. It is possible that there may be reciprocal competition between IGF-1R ligands and Aβ oligomers, as shown for the IR, where ADDLs block IR activation in vitro (25). In addition, ADDLs can increase Akt activation (25) and cause the inactivation of IRS-1 by phosphorylation at serine residue in vitro (14). Together this indicates that soluble Aβ oligomers may derail and/or compete for the IGF-1R/IR signaling system in AD neurons through inappropriate increased activation of the IIS signaling pathway.

    Thus, in the newly described igf-1R+/-/AD mouse, it is possible that decreasing the actual level of IGF-1R limits the amount of IGF-1R available to respond or interact with oligomeric Aβ, thereby diminishing hyperactivated IIS signaling. This, in itself, may be mechanistically important in protecting these mice from AD-like symptoms. In addition, the decreased IGF-1R levels will allow increased activation of transcription factors such as FOXO and HSF-1, which could either de-aggregate or trap the potentially pathological Aβ oligomers in denser plaques, keeping them away from IGF-1R. Of interest is the further possibility that decreasing IGF-1R and IIS signaling to a moderate level may also impact tau biology and NFT formation in AD, as increased mTOR (which is suppressed by IIS signaling) can induce cell cycle activation and increase neurodegeneration in a Drosophila tauopathy model (13). Moreover, Akt and GSK3β, other downstream players, have strong regulatory roles in NFT formation.

    Maintaining appropriately responsive IGF-1R/IR signaling in neurons would appear to be crucial to protect against potential Aβ oligomer synaptotoxicity. This is because in vitro findings show that AG1024, a tyrophostin that specifically inhibits IGF-1R and IR family tyrosine kinase activity, selectively mimics the detrimental effects of cell-derived Aβ oligomers on synaptic signaling (22). Furthermore, IGF-1R/IR tyrosine kinase activity can prevent the pathogenic binding of Aβ oligomers to neurons, thereby blocking their synaptotoxic effects(3). This stresses the importance of fine-tuning the balance and type of response elicited by IGF-1R and IIS signaling in neurons in AD.

    Finally, an association between IGF-1R polymorphisms and dementia has been reported (7), and as mentioned above, long-lived human female centenarians have IGF-1R genotypes that have low levels of IGF-1R activity (20). It is thus possible that polymorphisms in IGF-1R and components of the IGF-1R signaling pathway in humans could predispose both to increased longevity and protection from AD neurodegeneration. In addition, it will be important to understand how IGF-1R levels and IIS signaling are controlled in the human brain with age. It is possible that being able to maintain appropriate levels of IGF-1R and/or IIS signaling components in the face of increased age-related stressors, including oligomeric Aβ, could delay or protect from the onset of AD with age. Blocking induction of increased levels of IGF-1R and IIS responses and maintaining them at moderate but not heightened, potentially toxic levels as may occur in AD, could possibly be achieved by caloric restriction and/or exercise, or by targeting components of this pathway directly. It will thus be vital to determine more clearly the broader molecular mechanisms by which maintaining IGF-1R levels and IIS signaling at lower levels can protect against AD symptoms and Aβ proteotoxicity.

    View all comments by Cora O'Neill