| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Murdoch Children’s Research Institute (V.C.R., G.A.W.), Centre for Hormone Research, Royal Children’s Hospital, Parkville 3052, Victoria, Australia; Department of Paediatrics (V.C.R., G.A.W.), University of Melbourne, Parkville 3052, Victoria, Australia; Liggins Institute and National Research Centre for Growth and Development (P.D.G.), University of Auckland, Auckland 1020, New Zealand; and Department of Neurology (E.L.F.), University of Michigan, Ann Arbor, Michigan 48109
Correspondence: Address all correspondence and requests for reprints to: Vincenzo C. Russo, Ph.D., Centre for Hormone Research, Murdoch Children’s Research Institute, Royal Children’s Hospital, Flemington Road, Parkville, Victoria 3052, Australia. E-mail: vince.russo{at}mcri.edu.au
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
The main aim of this review is therefore to provide an up-to-date and comprehensive analysis of the pleiotropic functions of the IGF system in the nervous tissue. To maintain focus on the IGF system, the role and functions of the insulin system in brain and neuronal cells, also important but already amply reviewed elsewhere, will only be mentioned and discussed in brief in this review.
|
II. Historical Perspective |
|---|
|
TopAbstractI. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
Shortly thereafter, two mammalian somatomedins were identified by protein sequence and cDNA data (11, 14, 15, 16, 17, 18) and their structural homology with proinsulin led to their current designation of IGF-I and IGF-II.
In the middle 1990s, Sara and co-workers (19, 20) identified a brain-specific variant of IGF-I, des(1-3) or "truncated" IGF-I, which lacks the first three amino acids and is more potent than intact IGF-I in various cell culture systems (19, 20, 21), probably due to its lower affinity for IGF binding proteins (IGFBPs) (22). This finding suggested brain synthesis of IGF-I or its truncated variant. After the isolation of IGF-I mRNA from postnatal rat brain (23), a number of in situ hybridization studies (24, 25) demonstrated that IGF-I and IGF-I receptor mRNA are synthesized in the rat brain in specific regions, namely olfactory bulb, hippocampus, and cerebellum (26). In addition IGF-carrier proteins, later named IGF binding proteins, were also found expressed in similar brain regions (27, 28, 29, 30, 31, 32, 33, 34, 35, 36). The finding of IGF-I mRNA colocalization with IGF-I receptors and the presence of IGFBPs suggested a paracrine or autocrine role for IGF-I, potentially modulated by IGFBP, in the developing brain (26, 36, 37, 38).
The presence of IGF-I (39) and IGF-I mRNAs (40) in multiple tissues has necessitated the revision of the original somatomedin hypothesis to include both autocrine and paracrine actions of IGF in addition to its classical endocrine aspects.
The somatomedin hypothesis has been recently challenged by several groups employing the liver-specific IGF-I deficient (LID) mice, which have reduced circulating IGF-I and elevated GH (41, 42), and the acid-labile subunit knockout ALSKO mice (43), which have reduced circulating IGF-I and IGFBP-3 but normal GH level. The LID mice, contrary to expectations (altered IGF-I-mediated GH-stimulated growth), showed normal postnatal growth and development (43, 44), whereas the ALSKO mice, despite 65% reduction in circulting IGF-I, demonstrated only 10% reduction in body weight. Interestingly, when the LID mice were crossed with the ALSKO mice (LID + ALSKO) (43), to further affect the GH/IGF-I axis, the LID-ALSKO mice (very low IGF-I and very high GH levels) showed postnatal growth retardation and osteopenia, suggesting that circulating IGF-I is important for postnatal growth and development (44). Although investigations are ongoing in this area, it is accepted at this stage that IGF-I exerts its pleiotropic functions in an endocrine, autocrine, and paracrine fashion.
|
III. Overview of the IGF System |
|---|
|
TopAbstractI. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
- and ß-chain of the human proinsulin. This sequence homology is 43% for IGF-I and 41% for IGF-II. No sequence homology exists between the C domains of IGFs and the C peptide region of human proinsulin. The C domain of the IGFs is not removed during prohormone processing; thus the mature IGF peptides are single chain polypeptides (50, 51). The gene encoding IGF-I is highly conserved (50, 52, 53, 54, 55, 56) such that 57 of 70 residues of the mature protein are identical among mammals, birds, and amphibians (50, 52, 53, 54, 55, 56). For a more extensive review of structure, functions, and regulation of the IGF-I gene, see Refs.57, 58, 59, 60, 61 . Expression of the Igf-1 gene is affected at many levels including gene transcription, splicing, translation, and secretion. IGF-I expression is also influenced by hormonal (GH) (1, 62, 63, 64, 65, 66), nutritional (67, 68, 69), tissue-specific and developmental factors (36, 70, 71, 72, 73).
B. The IGF receptors and their functions
The biological actions of the IGFs are mediated by the type I IGF receptor (IGF-IR), a glycoprotein on the cell surface that transmits IGF binding to a highly integrated intracellular signaling system (74, 75, 76, 77). Binding of the IGFs (IGF-I and II) to the IGF-IR (74, 78, 79, 80, 81, 82, 83, 84) promotes intrinsic tyrosine kinase activity that phosphorylates the insulin receptor substrates (IRS-1 to IRS-4), which then leads to the activation of two main downstream signaling cascades, the MAPK and the phosphatidylinositol 3-kinase (PI3K) cascades (85, 86, 87, 88, 89, 90). However, IGF-II binds the IGF-IR with lower affinity (74, 78, 79, 80, 81, 82, 83, 84). Expression of the IGF-IR gene (77, 91) has been detected in many tissues and is constitutively expressed in most cells (92, 93, 94); its promoter is regulated in vitro and in vivo by transcription factors such as SP1 (95, 96) and the transcription factor p53 (97, 98).
Various IGF-I receptor subtypes that present distinct structures or binding properties have also been described. Two of these subtypes, namely hybrid and atypical IGF-I receptors, have been particularly investigated in a variety of cell types (99, 100, 101, 102). The atypical IGF receptors are characterized by their ability to bind insulin as well as IGFs with relatively high affinity (103). Hybrid insulin/IGF-I receptors have been reported in cells expressing both IGF-I and insulin receptors (104, 105, 106, 107, 108); however, the physiological significance of hybrid and atypical IGF receptors is unclear.
The IGF-II ligand has greatest affinity for a distinct receptor, the type-II or IGF-II receptor (109, 110, 111, 112, 113, 114). This single chain polypeptide with a short cytoplasmic domain lacking tyrosine kinase activity is identical to the cation-independent mannose-6 phosphate (M6P) receptor (107, 115, 116). The IGF-II/M6P receptor binds two general classes of ligands: 1) non-M6P-containing ligands, the best characterized of which is IGF-II; and 2) M6P-containing ligands, including lysosomal enzymes. The multifunctional role of the receptor is evidenced by its function in the mediation of lysosomal enzyme trafficking, endocytosis, and lysosomal degradation of extracellular ligands, regulation of apoptotic/mitogenic effects, and possible intracellular signal transduction (115, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128).
More recently, high-affinity binding of IGF-II to an insulin receptor isoform (IR-A) has also been reported (129, 130, 131), thus suggesting that IGF-II might also signal via the insulin receptor. However, this insulin receptor isoform (IR-A), lacking the alternative spliced exon 11, is preferentially expressed in fetal and cancer cells (130).
C. The IGFBPs
A family of six high-affinity IGF-binding proteins (IGFBP-1 through IGFBP-6) coordinate and regulate the biological activity of IGF in several ways: 1) transport IGF in plasma and control its diffusion and efflux from the vascular space; 2) increase the half-life and regulate clearance of the IGFs; 3) provide specific binding sites for the IGFs in the extracellular and pericellular space; and 4) modulate, inhibit, or facilitate interaction of IGFs with their receptors (111, 132, 133, 134, 135). IGFBP biological activity is regulated by posttranslational modifications such as glycosylation and phosphorylation (111) and/or differential localization of the IGFBPs in the pericellular and extracellular space (111, 134, 135, 136, 137, 138, 139). It is therefore hypothesized that IGFBPs, in addition to stabilizing and regulating levels of diffusible IGFs, might regulate IGF-I cellular responses by facilitating receptor targeting of IGF-I or modulating IGF-I bioavailability in the pericellular space (111, 135).
The effects of the IGFBPs are further regulated by the presence of specific IGFBP proteases, which cleave the binding proteins, generating fragments with reduced or no binding affinity for the IGFs (111, 135, 136, 140). Some IGFBPs, including IGFBP-2 and -3, can induce direct cellular effects independent of the IGFs (135, 141, 142, 143, 144). IGFBP-3, similar to IGFBP-5, contains sequences with the potential for nuclear localization (145, 146) and detection of IGFBP-3 in the nuclei of dividing cells, as reported by several investigators (146, 147), strongly suggesting a role for IGFBP-3 in gene regulation. More recently, perinuclear or nuclear localization has also been reported for IGFBP-2 (148); however, the role of IGFBP-2 in this cellular compartment is yet to be determined.
|
IV. Expression and Pleiotropic Actions of the IGF System in the Brain |
|---|
|
TopAbstractI. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
IGF-I plays a key role in the development of the nervous system, with demonstrated effects on many stages of brain development including cell proliferation, cell differentiation, and cell survival (36, 37, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159). Although recent reports have demonstrated that postnatal circulating IGF-I might exert neurogenic/survival activity (160, 161, 162, 163, 164, 165, 166) (167, 168, 169, 170, 171), systemic IGF-I is not readily transported through the blood-brain-barrier, and therefore local production of IGF-I is considered the primary source of the ligand (autocrine and paracrine action) for brain cells.
During embryogenesis, IGF-I mRNA expression is detectable in many brain regions (23, 25, 31, 172, 173, 174, 175, 176), with its expression being particularly high in neuronal rich regions such as the spinal cord, midbrain, cerebral cortex, hippocampus, and olfactory bulb (23, 177, 178) (Fig. 1
). More precisely IGF-I is found expressed in neuronal cells with large soma and complex dendritic formations (32, 155), including sensory and projecting neurons such as the Purkinje cells (25). In most neurons, IGF-I transcription decreases significantly postnatally, a decrease that correlates with the degree of cell maturation and reaches very low levels in the adult (23, 175). However, exceptions are the mitral and tufted cells of the olfactory bulb, which undergo constant cell renewal/turnover (179); in these cells IGF-I expression persists at a high level throughout life (26, 179).
|
B. Type I and II IGF receptor expression in the nervous system
The IGF receptors are widely expressed throughout the CNS, with high levels of expression found in specific regions and located to specific cell types (25, 26, 36, 38, 119, 153, 186, 187, 188). Given that IGF receptors are expressed from early stages of embryogenesis and throughout life, and that their ligands also show similar a "temporal-spatial" pattern of expressions, it is evident that the local brain IGF circuits are crucial modulators of the processes activated during brain development. Type I receptors are expressed throughout the rat CNS (177, 187, 189, 190, 191) with high levels of expression detected in the developing cerebellum, midbrain, olfactory bulb (Fig. 1
), and in the ventral floorplate of the hindbrain (23, 24, 187). The level of IGF-IR decreases to adult levels soon after birth (192, 193) but remains relatively high in the choroid plexus, meninges, and vascular sheaths (25, 187). It is thus not surprising that knockout of the IGF-I receptor gene (194, 195, 196) produced, in addition to in utero growth retardation, a strong brain phenotype, namely a small brain.
The type II IGF receptor, which has low affinity for IGF-I, but a high affinity for IGF-II, and also binds mannose-6-phosphate, is selectively expressed in all major brain regions (119, 183, 188, 197, 198). It is highly expressed in the pyramidal cell layers of the hippocampus, the granule layer of the dentate gyrus, olfactory bulb, the choroid plexus, and in the cerebral vasculature, ependymal cells, retina, pituitary, brainstem, and spinal cord (119, 183, 188, 197). Some of these regions undergo tissue remodelling, suggesting that the IGF-II/M6P receptor, in addition to its role in transporting lysosomal enzymes, might also participate in control of neuronal growth, differentiation, and repair, processes regulated by IGF-II/M6P receptor ligands including LIF, TGFß, and retinoic acid (198).
Early studies from our laboratory have demonstrated that some of these regions expressing both the IGF-I and IGF-II receptors are also rich in insulin receptors (199). Insulin receptor density is elevated in the choroid plexus, olfactory bulb, limbic system, and hypothalamus but is much lower in other brain regions (cortex, cerebellum, pituitary gland, thalamus), all regions concerned with olfaction, appetite, and autonomic functions (199). More recently, ablation of the insulin receptor in nestin-positive neurons (NIRKO mouse) (200) suggests a role for insulin in control of appetite suppression and reproduction (200).
C. The IGF ligands are pleiotropic factors
Whether the IGFs are produced locally (23, 25, 31, 172, 173, 174, 175, 176, 201) or reach brain cells systemically (160, 161, 162, 163, 164, 165, 166), these molecules exert potent neurotrophic, neurogenic, and neuroprotective/antiapoptotic activities. A large number of in vitro studies have demonstrated that IGF-I promotes mitogenesis and differentiation in glial cells (202, 203, 204, 205, 206, 207), oligodendrocytes (208, 209, 210), neuronal cells (36, 37, 211, 212), adult stem cells (213), and brain explants (21, 214, 215, 216, 217), and regulates axon myelination (218, 219, 220). IGF-I enhances growth cone motility and promotes neurite outgrowth (221, 222).
Studies using the IGF-I –/– mice models (152, 162, 220) have clearly demonstrated that most of the IGF functions determined in vitro also apply to the in vivo situation, affecting a wide range of brain cells. The IGF-I null mice have reduced brain size and altered brain structures (151, 155) and show alteration of myelination processes (168, 223). This appears to be the result of reduced oligodendrocyte proliferation and maturation, which is also associated with reduction of axonal growth (168, 196, 223). However, some early studies from Bondy and co-workers (220) have reported instead that in IGF-I null mice, myelination is reduced but proportionate to the neuronal composition. Furthermore, ablation of the Igf-1 gene has revealed deficit in the numbers of specific neurons, oligodendrocytes in the olfactory bulb, dentate gyrus and striatum (194, 196, 220) and in cochlear ganglion neurons (169). These anatomical differences are likely to be the consequences of alteration in proliferation, survival, or differentiation caused by the absence of IGF-I during development. A study from Vicario-Abejon et al. (224) suggests that reduction of stem cell proliferation/differentiation in the IGF-I null mice might be the cause of reduced plasticity/maturation normally occuring in some brain regions (i.e., olfactory bulb) postnatally (224). In addition to gross structural brain abnormalities, the IGF null mice also show alteration of important brain metabolic functions such as reduced glucose uptake, the major source of energy of neuronal cells (225). These in vivo models thus show that the IGF system affects several steps involved in development and organization (number of cells, their connection, and the extracellular matrix) of CNS architecture.
However, actions of IGF are not limited to fetal life (25, 149, 150, 226), but extend into postnatal (21, 217) and adult life (185, 227, 228, 229), with effects on proliferation, neuronal differentiation, and maturation maintained in some regions of the adult brain. Furthermore, alterations in levels of IGF-I have been reported in the brains of aging rats (230, 231), alterations that also correlate with deterioration of cognitive functions observed in elderly patients with low circulating IGF-I (232, 233, 234).
D. The IGFs are neurotrophic factors
In support of studies by the Gluckman group (235) and others on the use of IGFs as therapeutic agents (see also Section VII), many investigators have examined the cellular and molecular mechanisms of IGF action in nervous system diseases. In this regard, the IGFs have been shown to be neurotrophic factors, i.e., they promote the survival and differentiation of neuronal cells, including sensory (171, 221, 236), sympathetic (211, 237), and motor neurons (MNs) (236, 237, 238). In fact, the IGFs are the only known growth factors that support both sensory and motor nerve regeneration in adult animals (160, 238, 239, 240, 241, 242, 243). IGF-I is involved in brain plasticity processes (21, 217, 224, 244), and it specifically modulates synaptic efficacy by regulating synapse formation, neurotransmitter release, and neuronal exicitability (245, 246, 247, 248).
IGF-I also provides constant trophic support to neuronal cells in the brain and in this way maintains appropriate neuronal function (245). Alteration of this trophic input may lead to brain disease as seen in neurodegenerative disorders such as Alzheimer’s disease (AD) (162, 249), Ataxia telangiectasia (250), Huntington’s disease (251), and Parkinson’s disease (252), all variably responsive to IGF-I treatment (see also Section VII). A recent study by Lichtenwalner et al. (253) showed that neurogenesis declines in brains of aged mice, but it is efficiently restored after IGF-I administration via intracerebroventricular infusion.
E. IGF-I prevents neuronal apoptosis
The ability of IGF-I (and IGF-II) to promote neuronal survival is associated with the ability of these factors to prevent apoptosis, and IGF-I appears to be a potent agent for rescuing neurons from apoptosis. For example, IGF-I prevents N-methyl-D-aspartate (NMDA)- and nitric oxide-induced apoptosis in hippocampal and cortical neurons (254, 255, 256, 257, 258).
IGF-I has been shown to enhance the survival of rat embryo cerebellar granule cells and stimulate their terminal differentiation into cerebellar granule neurons (259, 260). Additionally, IGF-I promotes the survival of rat hypothalamic and hippocampal neurons in vitro (261). IGF-I and -II rescue chick MNs from injury-induced and developmentally regulated death (262, 263). A role for IGF-I in the regulation of Schwann cell survival has also been reported (264), as has the ability of IGF-I to prevent apoptosis in sympathetic neurons exposed to high glucose (265). Rat hippocampal neurons are also rescued by IGF-I from the induction of apoptosis by amyloid-derived peptides (162, 266, 267, 268). A similar antiapoptotic effect of IGF-I is observed in rat cerebellar neurons subjected to serum or potassium withdrawal or high KCl levels (269, 270), okadaic acid treatment (271), and in MNs during normal development or (262, 272) after axotomy or spinal transection (273, 274).
Neuroblastoma cells are rescued from hyperosmotic shock-induced apoptosis by IGF-I (275) or metabolic stress (265, 276) including exposure to low glucose (277). IGF-I inhibits mature oligodendrocyte apoptosis during primary demyelination (278) and prevents apoptosis in neurons after nerve growth factor withdrawal (279) or serum withdrawal in brain explants (280) and neuronal cells (281).
F. IGFBP action in the nervous system
A growing body of evidence suggests an important role for IGFBP in the nervous system. The mRNA expression profiles and location of the most abundant IGFBP-2, -4, and -5 in the normal developing and adult CNS are well defined (31, 32, 33, 35, 282, 283, 284). IGFBP-3 (35, 284, 285) and IGFBP-6 (286) are also discretely expressed in the CNS but at lower levels, and therefore data on their mRNA expression distribution are limited. IGFBP-1 is not expressed in the CNS, but its expression is induced under certain experimental conditions (287).
1. IGFBP-1.
Although IGFBP-1 mRNA is normally not detectable in brain and the effects of Igfbp-1 gene deletion on CNS have not been reported, the overexpression of IGFBP-1 mRNA in brain has instead generated an interesting brain phenotype, namely small brain (288, 289, 290, 291, 292).
IGFBP-1 is known to inhibit somatic linear growth, weight gain, tissue growth, and glucose metabolism (293). Increased expression of IGFBP-1 has been documented in a variety of situations, such as fetal nutritional deprivation and hypoxia, and has been considered to be a marker of metabolic disturbances that cause fetal growth retardation. Of interest, IGFBP-1 transgenic mice demonstrate both intrauterine and postnatal growth retardation phenotypes (288, 289, 290, 291, 292) including alteration of the somatotrophic axis (294).
In a study by Doublier et al. (295), a line of transgenic mice with liver-specific expression of IGFBP-1 also showed impaired brain development and hydrocephalus.
Furthermore, a study by Zhou and co-workers (287) described disrupted cerebellar morphogenesis and reduction in cerebellar size in mice overexpressing the hepatocyte nuclear factor-3ß (HNF-3ß). The transgenic cerebella displayed levels of IGFBP-1 elevated to 22 times greater than those measured for wild-type cerebella, an elevation consistent with the reduction in transgenic cerebellar size.
Together, these in vivo observations suggest that increased expression of IGFBP-1, as observed in a variety of clinical situations, may itself contribute to the growth retardation and impaired fetal brain development. However, it is suggested that the underlying mechanisms in these events involve IGFBP-1-sequestration of IGF-I.
2. IGFBP-2.
IGFBP-2 is expressed early in embryogenesis (296) and by embryonic d 10 is highly expressed in neuroectoderm structures including the neural tube and neuroepithelium (296). During development, the most prominent sites of IGFBP-2 expression in the CNS comprise cells with nonneuronal phenotypes including the epithelium of the choroid plexus, the floor plate, and the infundibulum (30). Later in development, IGFBP-2 mRNA is detectable throughout the brain (33, 297), particularly in brain regions undergoing continuous remodelling as is the olfactory bulb (Fig. 1), the cerebellum, and the hippocampus (33, 35, 36, 297). IGFBP-2 expression correlates and complements that of IGF-II, and both IGFBP-2 and IGF-II protein are highly abundant in the cerebrospinal fluid (CSF) and choroid plexus (36). It is also known that IGFBP-2 binds IGF-II with a moderate preferential affinity over IGF-I (135).
IGFBP-2 associates to cell surface proteoglycans in rat brain tissue (134) and neuronal cells (136), and IGF-I/IGFBP-2/proteoglycan complexes have been identified in rat brain tissue (298), but the role of these cell membrane complexes is not completely understood. However, it has been suggested that differential localization of the IGFBPs in the pericellular and extracellular space, involving components of the extracellular matrix (111, 134, 135, 136, 137, 138, 139), might regulate levels of diffusible/free IGFs. We have recently shown that IGF-I complexes with IGFBP-2 to promote neurogenesis in adult stem cells (213) and further demonstrated that neurogenesis was inhibited by IGFBP-2 antibody blockade (213), thus suggesting a key role for IGFBP-2 in this process. Additional evidence in support of an IGF-facilitating role for IGFBP-2 in the brain comes from colocalization of injected IGF-I with IGFBP-2 (299) and subsequent neuroprotection after hypoxic-ischemic injury, an effect not seen with an IGF variant that does not bind IGFBP-2 (300). It is therefore possible that brain IGFBP-2 regulates IGF-I cellular responses by facilitating receptor targeting of IGF-I or modulating IGF-I bioavailability in the pericellular space in vivo (111, 135). Cell-associated IGFBP-2 may therefore act as a "linker" molecule allowing pericellular storage of IGF.
This IGF-I storage might be affected by the presence of IGFBP-2 proteases (136, 281, 301, 302) that generate IGFBP-2 fragments which have markedly reduced binding affinity for the IGFs (135, 136). An IGFBP-2 fragment was identified on the cell surface of neuronal cells (136). The presence of this proteolytic fragment of IGFBP-2, capable of binding IGF-I while simultaneously being bound to the cell surface, might point to a process whereby proteolysis of membrane-bound IGFBP-2 provides a mechanism for creating peri-receptor low affinity IGF-I binding sites (136). Although we have recently demonstrated that IGFBP-2 proteolysis occurs during neuronal differentiation (281), the precise physiological significance of IGFBP-2 proteolysis in the nervous system remains to be determined.
Thus, IGFBP-2 possesses endocrine functions elicited by modulating serum IGF half-life and transport as well as autocrine/paracrine functions that result from blocking or enhancing the availability of IGFs to bind cell surface receptors. Despite these key functions, ablation of the Igfbp-2 gene (303) generated a phenotype less dramatic than that initially predicted. Selective alterations were reported for spleen and liver size, whereas the level of other circulating IGFBPs was found to be increased in the adult animals (304). The absence of a "brain phenotype" in the Igfbp-2 –/– mouse (303) suggests functional redundancy in the IGFBP family during development of the CNS (305).
The IGFBP-2 transgenic mouse model developed by the Hoeflich group (306) has suggested that IGFBP-2 may be a negative regulator of postnatal growth, including brain growth, in rodents, potentially by reducing the bioavailability of IGF-I. These effects were seen even when hemizygous cytomegalovirus-IGFBP-2 transgenic mice were crossed with hemizygous phosphoenolpyruvate carboxykinase-bovine GH transgenic mice (with very high GH and IGF-I serum levels) (307), thus suggesting an inhibitory role for IGFBP-2 in vivo. However, in both studies (306, 307), IGFBP-2 was constitutively overexpressed in most tissues, including the brain, resulting in very high IGFBP-2 levels, and therefore it might be possible that any specific local enhancing activities of IGFBP-2 were overridden by the high systemic level of IGFBP-2. It is therefore possible that although IGFBP-2 is abundantly expressed in the brain, expression at nonphysiological levels (up to 9-fold over the normal endogenous levels) in the transgenic mice would dramatically change local brain IGF availability leading to the reported effects on brain size.
An example of physiological change in local brain levels of IGFBP-2 and how this might affect IGF action has been reported by Cardona-Gomez et al. (308) in a specialized group of glial cells of the third ventricle named tanycytes. These cells have the ability to accumulate IGF-I and thus regulate IGF-I availability (309, 310, 311). Cardona-Gomez’s data show that estradiol and progesterone regulate local levels of IGFBP-2, including "peri-IGF-I-receptor" IGFBP-2 (308), and that these changes affect accumulation of IGF-I in tanycytes. This accumulation process might involve interaction of IGF-I with cell surface-associated IGFBP-2 before IGF-I being intracellularly translocated (308).
3. IGFBP-3.
IGFBP-3 is normally expressed at a low level in the CNS, mainly in nonneuronal structures including epithelial cells (35, 37, 312, 313), and the effects of IGFBP-3 gene deletion on CNS are either unknown or have not been reported.
Nevertheless, in a study by Ajo et al. (201), IGFBP-3 was found to be up-regulated in rat cerebral cortical cells after GH stimulation. Ajo et al. show that GH promotes proliferation of neural precursors, neurogenesis, and gliogenesis and that these responses are mediated by locally produced IGF-I and its modulator IGFBP-3.
Conversely, a study by Rensink et al. (314), investigating the mechanisms of amyloid-ß (Aß) deposition in cerebral blood vessel walls and Abeta-induced toxicity in AD, proposed that IGFBP-3 might contribute to neuronal degeneration in AD. It is therefore possible that, as seen in other systems, IGFBP-3 might promote either enhancement or inhibition of IGF-I action in brain cells in vitro, depending on the experimental conditions.
The function of IGFBP-3 in vivo has been in part elucidated by the IGFBP-3 transgenic mouse model. This mouse showed modest intrauterine and postnatal growth retardation despite elevated circulating IGF-I levels (315). These mice showed a significant reduction in birth weight, reduction in litter size, and postnatal reduction of both body weight and length and organ size. No change in brain growth or phenotype has been reported in IGFBP-3 transgenic mice (316).
4. IGFBP-4.
IGFBP-4 is normally expressed at a very low level in the CNS, where its mRNA is found in a variety of brain cell types including meningeal cells, astrocytes, and fetal neuronal cells (317). During early brain development, IGFBP-4 expression is increased, and its mRNA is easily detectable in regions such as the choroid plexus, meninges, and the basal ganglia (318). Postnatally, IGFBP-4 mRNA is found in the meningeal cell layer surrounding the developing cerebellum in the hippocampal formation and olfactory bulb (Fig. 1) (35, 317). In some of these IGFBP-4-expressing brain regions, which maintain a degree of tissue remodelling/plasticity (38), local expression for IGF-I and its receptor is also seen, thus suggesting that IGFBP-4 may play a role as a local modulator of IGF action (318).
Overexpression of IGFBP-4 gene has only been recently investigated in smooth muscle cell-rich tissue, and therefore the effects to the nervous tissue are unknown.
As discussed earlier, IGFBP-4 generally acts in vitro and in vivo as a potent inhibitor of IGF-I action (147, 319, 320, 321, 322), and it is therefore intriguing to find IGFBP-4 in sites where IGF-I exerts its primary mitogenic and trophic actions. A potential explanation might come from the work of Pintar et al. (305) on IGFBP-4 null mice, which indicates, contrary to expectation, that these mice have lower weight at birth. In this case, the authors argued that the absence of IGFBP-4 diminished tissue IGF storage capacity (305). This would imply that physiological levels of IGFBP-4 are indeed required for normal growth. Whether ablation of the IGFBP-4 gene also affects the CNS of these mice was not reported.
5. IGFBP-5
IGFBP-5 gene expression is highly abundant during brain development (32). The early expression of IGFBP-5 at embryonic d 10.5 indicates a key role of this IGFBP during embryogenesis (323). This hypothesis is further supported by a recent study from Pera et al. (150) which shows that IGFBP-5, as well as three IGFs expressed in early embryos, promoted anterior development by increasing the head region in Xenopus embryos. Thus, active IGF signals, including IGFBP-5, appear to be required for anterior neural induction in Xenopus (150). Whether IGFBP-5 has similar functions in early mammalian neural development is not known.
In rodents, IGFBP-5 appears to be coexpressed with IGF-I in principal neurons of sensory relay systems, cerebellar cortex, hippocampal formation, and many other neuron-rich regions (32, 324), including the olfactory bulb (Fig. 1) (35). These data point to the presence of potential autocrine and paracrine interactions between IGFBP-5 and IGF-I in specific brain regions where IGFBP-5 may act as a modulator or determinant of IGF action (325).
In addition to this spatiotemporal coexpression of IGFBP-5 and IGF-I, it is now becoming clear that IGF-I specifically regulates IGFBP-5. Using two IGF-I transgenic mice lines, Ye and D’Ercole (326) demonstrated that IGF-I up-regulates IGFBP-5 expression in vivo. This increase is specific for IGFBP-5 mRNA, because the level of expression of IGFBP-2 and IGFBP-4 mRNAs in these mice was found unchanged (326). The effects of IGFBP-5 gene ablation to the brain have not been reported.
6. IGFBP-6.
IGFBP-6 is poorly expressed in the nervous system, and information regarding its mRNA expression distribution, in both the developing and adult nervous system, is limited.
IGFBP-6’s unique property of preferential binding to the IGF-II ligand (327), coupled with the fact that this ligand is the most abundantly expressed IGF in the adult CNS, suggests that the IGFBP-6/IGF-II complex has a unique role in modulating IGF-II function in the adult brain (286).
During CNS embryogenesis, IGFBP-6 expression is tightly restricted to trigeminal ganglia and, relative to the rest of the embryo, this structure has the highest expression (286). The expression in the forebrain and cerebellum does not occur until after postnatal d 21 and then is primarily associated with GABAergic interneurons (286). The highest levels of expression in the adult animal are in the hindbrain, spinal cord, cranial ganglia, and dorsal root ganglia (286). These nuclei in the hindbrain and periphery that express IGFBP-6 are all associated with the coordination of sensorimotor function in the cerebellum, which indicates an important role for the IGFBP-6/IGF-II complex in the function and maintenance of these systems.
As introduced earlier, IGFBP-6 preferentially binds IGF-II and is regarded as a relatively specific inhibitor of IGF-II actions (327). IGFBP-6 is often expressed in nonproliferative cells, and its expression is associated with inhibition of growth of tumor cells in vitro (281, 327, 328) and in vivo (327). These findings are also supported by overexpression of IGFBP-6 in vivo, a model recently developed by the Babajko group (329) with strong expression of IGFBP-6 in glial fibrillary acidic protein-positive cells. Preliminary analysis of the IGFBP-6 transgenic mouse shows reduced cerebellum size and weight combined with altered differentiation of astrocytes (329). Abnormalities in the hypothalamus and pituitary were also reported (329). The effects on the brain of IGFBP-6 gene ablation remain unknown.
|
V. The IGF System and Neuroendocrine Cross-Talk in the Nervous System |
|---|
|
TopAbstractI. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
) may further modulate the biological activity of IGFs and cellular responses. This modulation can be seen in the neurogenic activity of IGF, which in many cases requires the presence of basic FGF or FGF-2 to function (21, 330, 331).
|
In a study by Drago and co-workers (330), it was shown that by blocking the endogenous stimulation of IGF-I with an antibody, the ability of FGF-2 to stimulate the proliferation of precursor cells isolated from the neuroepithelium of embryonic d 10 mice was abrogated. In later work by Frodin and Gammeltoft (333), it was also found that the IGFs act synergistically with FGF-2 and nerve growth factor to promote chromaffin cell proliferation. A recent study from our laboratory (281) demonstrated, in neuroblastoma cells, that FGF-2 overrides IGF-I mitogenic and survival activity via induction of neuronal differentiation and blockade of the antiapoptotic response of IGF-I.
Furthermore, Torres-Aleman (243) recently highlighted the importance of a number of serum growth factors (in contrast to brain-produced), which included IGF-I, FGF-2, and other blood-borne neurotrophic factors (neurotrophins), in neuroprotective surveillance. All these factors exerted a tonic trophic input on brain cells, providing defense mechanisms ranging from blockade of neuronal death to up-regulation of neurogenesis.
B. IGF-I and neurotrophic factors
Increased interest has recently been devoted to the interaction of IGF-I with other "classic" neurotrophic factors. A study by Corse et al. (334) has investigated the potential effects of neuroprotective factors including IGF-I, glial cell line-derived neurotrophic factor, brain-derived neurotrophic factor, neurotrophin (NT)-4 and -3 (NT-4/5, NT-3), and ciliary neurotrophic factor on mature motor neurons. Only IGF-I, glial cell line-derived neurotrophic factor, and NT-4/5 were found to be potently neuroprotective, pointing to the potential combined use of these neuroprotective factors in treatment of neurodegenerative diseases. Subsequently, the results from a number of phase III clinical trials with nerve growth factor, brain-derived neurotrophic factor, and IGF-I (335) indicate that these neurotrophic factors may find an application in degenerative disorders or injury of peripheral nerves and motor neurons (336).
C. IGF-I and GH
Although GH gene expression also occurs in the central and peripheral nervous system, with brain GH imunoreactivity not affected by hypophysectomy (337) (178), the potential cross-talk between the IGF system and GH in the nervous system has not been fully investigated. Several experimental models, in vivo and in vitro, support a key role for GH in brain development. The Snell dwarf mouse (Pit1) and the GHRH receptor-deficient little mouse exhibit microcephalic cerebra with hypomyelination and retarded neuronal growth with poor synaptogenesis. Postnatal GH administration normalizes neuronal growth in these mice (338, 339). Recent in vivo studies by the Williams group have demonstrated that GH is involved in neuroprotection during hypoxic/ischemic brain injury (340, 341, 342). In these studies, they also showed GH-like immunoreactivity on injured brain cells and demonstrated that GH administered intracerebroventricularly is capable of preventing brain cell loss (340, 341, 342). Whether the neurogenicprotective effects of GH involve IGF-I induction is not clear.
However, some recent in vitro studies by Ajo et al. (337) have demonstrated that GH promotes proliferation and differentiation of fetal cerebral cortical cells in primary culture, and these effects are mediated by IGF-I. Furthermore, a study by Edmondson and co-workers (343) has demonstrated that GH and IGF-I promote growth and survival of melanocytes (neural-crest derived cells) by a process synergistic with FGF-2.
Furthermore, there is some evidence that the GH-related hormone prolactin (PRL) also possesses neurogenic properties. In female mice during pregnancy, PRL stimulates production of neuronal progenitors in the forebrain subventricular zone (344, 345). Whether the neurogenic properties of PRL are a direct effect of its action or mediated by other factors, including IGF-I, remains to be determined.
D. IGF-I and erythropoietin
Erythropoietin (EPO) is traditionally known as a hematopoietic cytokine produced by the fetal liver and adult kidney in response to hypoxia (346, 347, 348, 349, 350). However, the expression of EPO and EPO-receptors in the CNS and the up-regulation of EPO by hypoxic-ischemic insult (346, 347, 348, 349, 350) suggest that this cytokine is an important mediator of the brain’s response to injury. In fact, in vivo EPO administration protects hippocampal CA1 neurons and retinal neurons from ischemic damage and prevents brain injury after a number of "insults" (346, 347, 348, 349, 350) by mechanisms involving both inhibition of apoptosis and neurotrophic actions (346, 347, 348, 349, 350, 351, 352). In the brain, EPO expression, regulated by the hypoxia-inducible factor (HIF)-1, is mainly found in astrocytes.
Recent work by Wang et al. (353) demonstrated that IGF-I protection of primary neuronal cells exposed to low oxygen concentration correlates with activation of HIF-1 expression (351, 352). In the same studies using an in vivo model of hypoxia/ischemic brain injury, they also show that IGF-I transcriptional activation correlates with that of HIF-1, suggesting that HIF-1 might mediate some of the IGF-I responses. In another study by Chavez and LaManna (354), it was shown that IGF-I induces HIF-1 transcriptional activity in rat cerebral cortex and neuronal cells (PC12) and that this induction is abolished by a selective IGF-I receptor antagonist (JB-1) (354). Furthermore, an early study by Feldser and co-workers (355), in hepatoma cells, not only showed that insulin, IGF-I, and IGF-II induced HIF-1 expression but also demonstrated that HIF-1 is required for expression of IGF-II, IGFBP-2, and IGFBP-3 (355).
These findings suggest the presence of a potential complex synergistic cross-talk between the IGF and the EPO system involving both activation of common intracellular signaling pathways and regulation of gene expression. The presence of synergistic cross-talk between the IGF and the EPO system in neuronal cells has recently been demonstrated by Digicaylioglu and co-workers (356), showing that EPO can exert a more immediate neuroprotective action when administered in concert with IGF-I. The neuroprotective mechanism, after coadministration of EPO and IGF-I, involved synergistic activation of the PI3K–Akt pathway (356). These findings further support the concept that the coadministration of synergistic neuroprotective agents rather than a single agent might provide improved therapeutic outcome. Thus, treatment with appropiate combinations of EPO and IGF-I could be a future therapeutic strategy for a variety of acute neurological events (356).
E. IGF-I and sex steroids
Among the numerous endocrine systems regulating brain physiology, sex steroids play an important role. However, it appears that part of the effects of these hormones on the brain are mediated by trophic factors including IGF-I (357, 358, 359, 360, 361). Studies from the Garcia-Segura group (357, 358, 359, 360, 361) have demonstrated that estradiol and IGF-I increase survival and differentiation of developing fetal rat hypothalamic neurons. They have demonstrated that estrogen-induced activation of the estrogen receptor requires the presence of IGF-I and that both estradiol and IGF-I use the estrogen receptor to mediate their trophic effects on hypothalamic cells. In vivo sex steroids affect IGF-I levels in the endocrine hypothalamus, arcuate nucleus, and median eminence (357, 358, 359, 360, 361).
Furthermore, increased clinical and basic evidence suggests that gonadal steroids affect the onset and progression of several neurodegenerative diseases and schizophrenia and the recovery from traumatic neurological injury such as stroke. In the brain, similarly to the IGF system, both estrogen synthesis and estrogen receptor expression are up-regulated at sites of injury (362, 363). Once again, it is suggested that the neuroprotective effects of estrogen may be exerted independently of the classical nuclear estrogen receptors involving modulation of the IGF-I signaling (362, 363). This is supported by the fact that estrogen receptors and IGF-I receptor interact in the activation of PI3K and MAPK signaling cascades and possibly in the promotion of neuroprotection (362, 363). It is therefore possible that the decrease in estrogen and IGF-I levels with aging may thus result in an increased risk for neuronal pathological alterations. Also, in this case these findings suggest the presence of a potential complex synergistic cross-talk between the IGF and sex steroids involving activation of common intracellular signaling pathways, however the precise mechanism remains unclear.
|
VI. Altered Expression of the IGF System in Response to CNS Injury |
|---|
|
TopAbstractI. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
, A–C). IGF-I mRNA induction is seen within infarcted regions by 3 d after hypoxia (369) (Fig. 3A). In addition, IGF-II (365), IGFBP-2 (365, 370) (Fig. 3C), IGFBP-3 (369) (Fig. 3A), and IGFBP-5 (368) (Fig. 3B) genes are differentially induced in specific regions after hypoxic/ischemic injury in the same model, suggesting that they may modulate the actions of IGF-I in a spatiotemporal-specific manner. The induced expression of IGF-I and IGFBP-2, -3, -4, and -5 in reactive microglia, oligodendrocytes, astrocytes, and surviving neurons (313, 370) of the periinfarcted area, including areas distant from the region of cell loss, suggests a role for the local IGFBPs in transporting IGF-I from its sites of production to the sites of action (313) (Fig. 3, A–C). Furthermore, in the periinfarcted regions, IGFBP-2 was highly expressed by reactive astrocytes that were juxtaposed to surviving neurons (Fig. 3C) (313), therefore suggesting a specific role for IGFBP-2 in augmenting IGF-I action (313) (Fig. 4).
|
|
). These findings might point to a role for cell-associated IGFBP-2 (134) to modulate local IGF-I bioavailability (Fig. 4).
IGFBP-2 mRNA is induced throughout the injured hemisphere (313) (Fig. 3C) whereas IGFBP-3 is only moderately induced in reactive microglia and glial cells and is substantially decreased in neuronal cells of the region of injury (372) (Fig. 3A). However, in the early stages of the injury response, IGFBP-3 expression increases rapidly in vascular endothelial cells throughout the affected hemisphere with maximal expression levels at 24 h, which then become undetectable 48 h later (372). This early transient induction of IGFBP-3 in brain vascular endothelial cells is a likely mechanism by which brain vascular endothelial cells potentially regulate up-take of circulating IGFs after the hypoxic-ischemic insult (372). However, the precise mechanism involved remains to be determined. Nevertheless, these data might point to a potential role for IGFBP-3 as carrier/transporter of vascular IGF-I into the brain tissue in the early phases of the injury response, when local IGF-I availability and expression might not be sufficient to trigger the neuroprotective mechanisms (Fig. 4).
After hypoxic-ischemic insult, IGFBP-4 mRNA is also moderately up-regulated (365, 368). This is also true for IGFBP-5 (Fig. 3B), which is up-regulated after severe hypoxic-ischemic injury in the infant rat brain (368), and its mRNA is up-regulated in cerebellar granule cells during apoptosis (325). In both cases, the increase of IGFBP-5 expression is believed to be required to maximize the availability of IGF and thus potentiate the IGF-triggered survival signaling, as has previously been described for cerebellar granule cells (325).
After peripheral nerve injury (lumbar motoneurons), IGFBP-6 mRNA and its protein expression are strongly up-regulated in the spinal motoneurons. This increased expression of IGFBP-6 in the damaged nerve is spatiotemporally associated with a local increased expression of the IGFs (373). It is therefore likely that IGFBP-6 might play a specific role in controlling IGF availability to lesioned motoneurons and thus regulates axonal regeneration (373). These findings all suggest that the IGFs in the brain, after injury, may act as endogenous neuroprotective agents and limit the degree of neuronal and glial loss (Fig. 4).
|
VII. Altered Expression of the IGF System in Malignancies of the Nervous System |
|---|
|
TopAbstractI. IntroductionII. Historical PerspectiveIII. Overview of the...IV. Expression and Pleiotropic...V. The IGF System...VI. Altered Expression of...VII. Altered Expression of...VIII. IGF-I Therapy in...IX. Perspectives for Future...References |
|---|
Within the CNS, the IGF family is critical for normal development: aberrant expression and/or activation may be associated with CNS malignancy (381, 382). Components of the IGF system, in particular IGF-I, IGF-II, and the IGF-IR, are overexpressed to varying degrees in numerous CNS tumors (381, 383), including low-grade gliomas, glioblastomas, medulloblastomas, astrocytomas, ependymomas, and meningiomas.
Expression of the IGF family in CNS malignancy has been most thoroughly investigated in glioblastoma. IGF-IR expression is higher in glioblastomas than in normal brain, and glioblastoma cell lines exhibit a dose-dependent IGF-I stimulation of both receptor autophosphorylation and thymidine incorporation, indicating a functional receptor (384). Both ligands for the IGF-IR, IGF-I and IGF-II, have increased mRNA transcript formation in glioblastoma compared with normal brain, and the major IGF-II transcript of 6.0 kb found in glioblastoma is similar to that found in fetal brain (385). The IGF ligands are also secreted into the CSF of glioblastoma patients (386). Most of the IGF binding proteins, including IGFBP-1, -2, -3, -5, and -6, are expressed in glioblastomas, with the most abundant being IGFBP-3 (387). In addition, glioblastomas express membrane-bound forms of the binding proteins, and IGFBP-3 is secreted into culture media (387). Much of the in vitro work on the role of IGF-IR in transformation has been performed in rat glioblastoma cell lines. In this work, researchers have found through mutational analyses that distinct regions of IGF-IR are critical for the biological actions of the IGF ligands, including mitogenesis, transformation, and differentiation, through the induction of specific downstream targets (388, 389, 390, 391, 392).
IGF ligands acting through the IGF-IR are known to affect several aspects of glioblastoma tumor formation in addition to growth and transformation. IGF-I stimulates the migration (393) and invasion (394) of glioblastoma cell lines in culture, prerequisites for tumor cell metastasis. IGF-I acts as a potent survival factor in glioblastoma cell lines, even overcoming proapoptotic stimuli (394, 395). Glioblastoma cell treatment with IGF-I triple helix-forming DNA (396, 397, 398), antisense IGF-IR (399, 400, 401, 402), kinase-defective IGF-IR (403), or additional IGF-IR mutant dominant-negative constructs (404) all induce growth suppression and/or apoptosis, resulting in decreased tumor growth. These studies have led the way for future work on IGF-IR inhibition and ligand interference in glioblastoma patient therapy (382).
As previously indicated, expression of IGF system components is also altered in other CNS tumors. In infiltrating astrocytomas, IGF-IR expression increases early in tumor development, appearing in stage II tumors and continuing into stage IV (204). IGF-IR protein levels are increased in the majority of medulloblastoma cases, with concomitant increases in phosphorylated forms of the downstream signaling molecules Erk-1, Erk-2, and Akt/PKB (405, 406). When IGF-II or activated Akt/PKB is retrovirally transferred into neural progenitors along with sonic hedgehog-patched, medulloblastomas form (407), suggesting a role for this pathway in tumor formation. In microarray analyses of ependymomas, high IGF-II expression was detected (408). High IGF-II expression occurs in meningiomas, and the IGF-II/IGFBP-2 ratio, indicating free IGF-II levels, correlates with tumor anaplastic histopathology (381). In fact, meningiomas express the highest IGF-II mRNA levels of all intracranial tumors (409). Although IGF ligand levels are increased in the majority of CNS tumors, ligand increases are not detected in the CSF, indicating that IGFs act locally in an autocrine and paracrine fashion to control tumor cell proliferatio
