Insulin receptor in the brain: Mechanisms of activation and the role in the CNS pathology and treatment

Summary

While the insulin receptor (IR) was found in the CNS decades ago, the brain was long considered to be an insulin‐insensitive organ. This view is currently revisited, given emerging evidence of critical roles of IR‐mediated signaling in development, neuroprotection, metabolism, and plasticity in the brain. These diverse cellular and physiological IR activities are distinct from metabolic IR functions in peripheral tissues, thus highlighting region specificity of IR properties. This particularly concerns the fact that two IR isoforms, A and B, are predominantly expressed in either the brain or peripheral tissues, respectively, and neurons express exclusively IR‐A. Intriguingly, in comparison with IR‐B, IR‐A displays high binding affinity and is also activated by low concentrations of insulin‐like growth factor‐2 (IGF‐2), a regulator of neuronal plasticity, whose dysregulation is associated with neuropathologic processes. Deficiencies in IR activation, insulin availability, and downstream IR‐related mechanisms may result in aberrant IR‐mediated functions and, subsequently, a broad range of brain disorders, including neurodevelopmental syndromes, neoplasms, neurodegenerative conditions, and depression. Here, we discuss findings on the brain‐specific features of IR‐mediated signaling with focus on mechanisms of primary receptor activation and their roles in the neuropathology. We aimed to uncover the remaining gaps in current knowledge on IR physiology and highlight new therapies targeting IR, such as IR sensitizers.

1. INTRODUCTION

Since the first demonstration of glucose‐lowering effect of insulin in 19161 followed by the identification of insulin receptors (IRs),2 a principal role of IRs in the regulation of glucose metabolism in peripheral tissues was established. Over the past decades, IR function was thought to be restricted to the periphery, while the brain had traditionally been considered to be an insulin‐insensitive organ. This was largely based on the fact that whole‐brain glucose uptake was not affected by circulating insulin levels. The view on the role of IR in CNS functions has been revisited since the discovery of insulin3 and IR4 in the brain in 1978.

Numerous subsequent studies have revealed IR involvement in a broad spectrum of functions in the CNS, with the majority of those appearing to be unrelated to metabolic functions of insulin. By now, the IR has been established as an important regulator of synapse formation and remodeling,5, 6 neuronal survival/neuroprotective processes,7, 8 and synaptic plasticity.5, 6, 9 The IR‐mediated signaling was found to be implicated in the modulation of long‐term potentiation (LTP) and depression (LTD),9 learning and memory,9, 10 and energy balance, appetite, and feeding behavior.10, 11, 12, 13 Fundamental roles of the IR‐mediated signaling in brain development are reflected by numerous associations reported between its deficiency and neurodevelopmental disorders, such as autism14 and schizophrenia,15 as well as cancer, including cerebellar neoplasms16, 17 and other syndromes.18, 19 Altered neuroprotective functions of IR are ascribed to the contribution of IR‐related signaling to the pathophysiology of neurodegenerative diseases, such as Alzheimer's disease, Huntington's disease, and Parkinson's disorder20, 21, 22, 23, 24, 25, 26; major depression27; and neurotoxicity.28, 29

At the same time, as the implication of IR in the regulation of energy and glucose homeostasis in the brain has been revisited, these processes are now not considered to be entirely insulin‐independent,30 and compromised glucose metabolism in the brain resulting from deficient IR function is regarded as another critical factor in the risk for neuropsychiatric disorders. In particular, glucose uptake in the hippocampus was shown to be dependent on IR‐stimulated translocation of the glucose transporter GLUT4 to the plasma membrane.31, 32 This recently discovered mechanism is likely to play a role with other glucose transporters, such as GLUT1, GLUT3, and GLUT5, which are expressed in astrocytes, neurons, and microglia, respectively.31, 32, 33, 34

Dysfunctions in the IR‐mediated processes can be due to abnormalities in IR activation, lowered insulin availability, and compromised IR‐triggered downstream mechanisms and, as discussed above, result in a broad range of brain disorders, such as neurodevelopmental, affective, and neurodegenerative disorders, and cancer. Here, we sum up findings on the brain‐specific features of IR‐mediated signaling with focus on mechanisms of primary receptor activation and their roles in the neuropathology, specifically referring to Alzheimer's disease as an example. We aim to uncover the remaining gaps in current knowledge on IR physiology and discuss new therapies targeting IR, such as IR sensitizers.

2. ISOFORMS OF INSULIN RECEPTOR IN CNS AND THEIR LOCALIZATION

The insulin receptor is a heterotetrameric protein composed of two α‐subunits and two β‐subunits linked by disulfide bonds.35 The extracellular α‐subunit includes a ligand binding site. The cytoplasmic portion of the transmembrane β‐subunit conveys tyrosine kinase activity. There are two structurally and functionally different isoforms of the insulin receptor: a long isoform of IR, isoform B (IR‐B), which predominates in adult peripheral tissues such as muscle, liver, kidney, and fat—all well‐known targets for the metabolic effects of insulin36; and a short isoform of IR, isoform A (IR‐A). The latter arises from the alternative splicing of exon 11, where 12 amino acids from the carboxyl terminus of the α‐subunit are spliced off.36

Both isoforms display similar binding affinity for insulin in terms of half‐maximal binding inhibition in ligand‐competition assays (IC₅₀)37, 38, 39, 40 (Table 1), as well as a similar potency during in vitro activation.41 The IC₅₀ values are quite similar to insulin concentrations required to induce 50% of the maximal receptor autophosphorylation (EC₅₀), the characteristic triggering mechanism of IR activation.37 In contrast to IR‐B, IR‐A shows no negative cooperativity,36, 37 suggesting distinct mechanisms of functional receptor regulation via insulin binding.

Table 1.

Binding properties of IR isoforms A and B

Properties	Cell lines	IR‐A	IR‐B	Method	References
Insulin
IC50, nmol/L	CHO	0.9	1.6	RCBA	38
	R mouse	0.9	1.0	RCBA	37
	R¯mouse	0.4	0.5	RCBA	39
	CHO	0.3	0.5	RCBA	40
EC50, nmol/L	R¯mouse	0.8	1.1	IR pY	37
	CHO	0.6	0.7	IR pY	37
	NIH 3T3	0.7	0.8	IR pY	37
	HEK‐293	2.7	2.6	BRET	41
IGF‐2
IC50 nmol/L	R¯mouse	2.5	>20.0	RCBA	37
	CHO	2.2	10.0	RCBA	40
EC50 nmol/L	R¯mouse	3.0	24.0	IR pY	37
	CHO	3.8	30.0	IR pY	37
	NIH 3T3	3.6	22.0	IR pY	37

IC₅₀ and EC₅₀ values for binding of insulin and IGF‐2 to IR‐A and IR‐B were measured by ligand‐competition and IR tyrosine phosphorylation assays, or the BRET method in cells transfected with IR isoform cDNA.

IC₅₀, half‐maximal binding inhibition in ligand‐competition assay; EC₅₀, half‐maximal effective concentration required to achieve 50% of maximal receptor activation; R¯mouse cells, cells with targeted disruption of IGFR gene transfected with IR‐A or IR‐B; RCBA, radioligand competitive binding assay; IR pY, tyrosine phosphorylation assay; BRET, Bioluminescence Resonance Energy Transfer assay.

Remarkably, in contrast to IR‐B, the IR‐A isoform binds IGF‐2 at physiologically relevant concentrations, which also results in receptor activation.37, 38, 39, 40, 42, 43 As compared to IR‐B, IR‐A displays 5‐10 times higher binding affinity for IGF‐2.37, 40 Moreover, IR‐A can be activated by 6‐8 times lower concentrations of IGF‐2 than those which are required for its insulin‐induced autophosphorylation, that is, receptor activation37 (Table 1). In addition, the rate of internalization of IR‐A bound to insulin is 1.7‐fold higher than that of IR‐A bound to IGF‐2, suggesting different intracellular consequences of binding of these two ligands.44

Notably, recent studies established an essential role of IGF‐2 in the regulation of morphologic and functional plasticity in the brain. While these effects are mediated not only via IR, but also via IGF‐1R and IGF‐2R,6, 35, 45 distinct physiological functions of the IR in the brain versus peripheral tissues may be due to different properties of its isoforms in relation to IGF‐2. IGF‐2 is widely expressed throughout the brain and is abundant in the hippocampus, where it regulates neurogenesis.46, 47, 48 Various challenges, such as acute hypoxia, exposure to toxicity, chronic stress, and cerebral ischemia, were shown to stimulate IGF‐2 expression, which is considered to be neuroprotective.49, 50, 51, 52, 53 Our own work revealed that treatment with insulin receptor sensitizers upregulated hippocampal IGF‐2 during stress and interfered with susceptibility to a stress‐induced syndrome.53, 54 Binding of IGF‐2 to IR‐A was also shown to result in preferential activation of numerous signaling pathways regulating metabolic and mitogenic responses, which were not shown for insulin,38, 55, 56, 57 such as IGF‐2‐induced neural stem cell proliferation.58

As for brain distribution of IR, both of its isoforms are present in the brain (Table 2) and abundantly expressed in the olfactory bulb, hypothalamus, hippocampus, cerebral cortex, and cerebellum.4, 59 On the cellular level, it was found that astrocytes express both IR‐A and IR‐B, while neurons express exclusively the IR‐A.60, 61 Another important feature of IR expression is the high density of IR in neurons as compared to glia.59 Thus, neuron‐specific pattern of IR‐A that is also sensitive to IGF‐2 may underlie functions of IR that are characteristic for the brain.

Table 2.

Distribution of IR isoforms A and B in adult mammalian tissues

Tissue/cell type	Species	Method	IR‐A	IR‐B	References
Whole brain	Human, nonhuman primate	PCR	++	+	36
Whole brain	Rat	EP	++	−	59, 60
Synaptosomes
Hip	Rat	EP	++	−	58, 59
Hyp	Rat	EP	++	−	58, 59
OT	Rat	EP, C	++	−	58, 59
BC	Rat	EP	++	−	58
Synaptic membranes BC, Th, Hip, Hyp, OT, Cer, Str, OB	Rat	EP	++	−	4, 60, 61
Neuronal precursors	Human	RT‐PCR	++	−	60
Astrocytes	Human	RT‐PCR	+	++	60
Astrocytes	Human	RT‐PCR	++	+	59
Liver	Human	PCR	+	++	36, 148
Muscle	Human	PCR	+	++	148
Adipose tissue	Human, rat	PCR, EP	+	++	58, 148

Symbols indicate the absence (−), presence at level <50% (+), or presence at level >50% (++) of the specific isoform in the tissue/cell.

PCR, polymerase chain reaction; EP, electrophoresis of photoaffinity‐labeled proteins; C, chromatography of photoaffinity‐labeled proteins; Hip, hippocampus; Hyp, hypothalamus; OT, olfactory tubercle; BC, brain cortex; Th, thalamus; Cer, cerebellum; Str, striatum; OB, olfactory bulb.

3. MECHANISMS OF ACTIVATION OF THE BRAIN INSULIN RECEPTOR

The IR was shown to possess intrinsic tyrosine kinase activity. Binding of insulin to the extracellular α‐subunits of the receptor results in a dose‐dependent autophosphorylation of three tyrosine residues in the activation loop of the kinase domain within the cytoplasmic portion of the β‐subunits, upon which the receptor tyrosine kinase becomes fully active.62 The phosphorylation of tyrosine kinase is considered to be the principal mechanism of the IR activation that triggers subsequent phosphorylation mediating cellular effects of insulin.

The analysis of crystal structures of the unphosphorylated and phosphorylated forms of IR suggests an autoinhibitory mechanism whereby the activation loop restricts access to protein substrates and ATP for binding to the active site.63 Upon autophosphorylation, the activation loop undergoes a major conformational change resulting in unrestricted access of ATP and protein substrates to the kinase active site. As binding of insulin or IGF‐2 is required for IR activation, their availability in the brain is essential for IR functions. However, growing evidence suggests that effective autophosphorylation is also critically dependent on other molecular interactions within primary signaling mechanisms of IR, which we will discuss below.

4. G PROTEIN SIGNALING IN THE ACTIVATION OF THE INSULIN RECEPTOR

The IR autophosphorylation in neurons was shown to be completely abrogated by pertussis toxin (PTX), a classic inhibitor of cAMP‐inhibiting G protein Gi,64 suggesting a signaling pathway involving Gi in the activation of IR in the brain. Similar inhibitory effects of PTX on the IR autophosphorylation and activation of IR kinase were demonstrated for FAO cells65 and adipose tissue.66 Gαi2 was found to be the specific Gi protein isoform that is recruited by the IR and affects its autophosphorylation.66

Next, NADPH oxidase (Nox) was found to be the effector that is coupled to the IR via a heterotrimeric G protein.67 The coupling of the Gαi2 isoform to the NADPH oxidase homolog Nox4 was found to be involved in production of the hydrogen peroxide (H₂O₂) in fat cells.68 In line with these findings, it was shown that changes in this molecular machinery affect IR functions. Mice expressing constitutively active Gαi2 revealed enhanced glucose tolerance and IR‐related activities of GLUT4,69 as well as markedly increased tyrosine phosphorylation of the IR in fat and muscle tissues following insulin challenge.70 In contrast, mice deficient in Gαi2 expression had reduced insulin sensitivity in peripheral tissues.71 While no information is currently available regarding IR/Gαi2 interactions in the brain, with the exception of unaltered brain Gαi2 levels in Alzheimer's disease45 and aging,72 the inhibitory G protein signaling is considered, in analogy to peripheral tissues, as a therapeutic target for pharmacologic correction of dysfunctional activation of IR.73

5. THE ROLE OF H₂O₂ SIGNALING IN THE ACTIVATION OF INSULIN RECEPTORS

Forty years ago, insulin‐stimulated H2O2 generation in peripheral tissues74 has been established as a triggering event of the IR activation75, 76 and the inhibition of protein tyrosine phosphatases (PTPs), which are involved in the dephosphorylation and inactivation of IR (Figure 1).77 The role of insulin‐induced H₂O₂ in the autophosphorylation of central IR was initially demonstrated for cerebellar granular neurons78 and later confirmed in organotypic cultures of hypothalamus.79 Both mitochondria78 and NADPH oxidase79 were identified as insulin‐sensitive sources of H₂O₂.

Figure 1.

Regulatory factors of insulin receptor activation in neurons. Insulin binding to insulin receptor (IR) evokes a brief (15 s) release of mitochondrial H₂O₂ that, upon exceeding the activation threshold, triggers the autophosphorylation (pYpYpY) of the IR. Factors that regulate H₂O₂ release may consequently alter the IR activation: They include inhibitory G protein (Gi) signaling, mitochondrial dysfunction, low activation of succinate dehydrogenase in response to insulin, and elevated capacity of the antioxidant system to metabolize the insulin‐induced H₂O₂ signal and eventually affect IR functions during physiological and pathological conditions, as well as the use of IR‐targeting therapies (see the text)

The autophosphorylation of IR was shown to be highly sensitive to H₂O₂ signaling.64, 80 The H₂O₂ scavenger N‐acetylcysteine (NAC) inhibits autophosphorylation of IRs in neurons in a dose‐dependent manner with a markedly sigmoidal (Hill slope = 8) dose‐response curve, suggesting that minor decreases in the magnitude of H₂O₂ levels can completely abrogate the insulin‐stimulated activation of IR. Thus, H₂O₂ release can control the IR activation in an “all‐or‐nothing”, and thus rate‐limiting, fashion. Consequently, H₂O₂ scavengers regulate IR function.

Among them, peroxiredoxins (Prx) and glutathione peroxidase (Gpx1) may be physiologically relevant and effective scavengers of H₂O₂ 81 that alter IR function, as supported by studies in vivo. It was shown that overexpression of Gpx1 results in insulin resistance associated with a profound decrease in IR autophosphorylation, hyperinsulinemia, hyperglycemia, and obesity in mice.82 On the contrary, a genetic deficit of Gpx1 diminishes their vulnerability to diet‐induced insulin resistance, and the administration of N‐acetylcysteine, the Gpx1 substrate, has an opposing effect.83 An essential role of H₂O₂ scavengers in the regulation of IR functions was further suggested by results from studies on human postmortem brain, which demonstrated elevated expression of glutathione peroxidase in the brain of patients with Alzheimer's disease.84 Other studies in Alzheimer's disease showed that two isoforms of Prx, peroxiredoxin 1 (Prx1) and peroxiredoxin 2 (Prx2), were elevated in the hippocampus and cortex.85 Moreover, a marked increase in peroxiredoxin 6 (Prx6), which is expressed in astrocytes, was found in the gray and white matter in the brain of these patients.86 It can be speculated that the capacity to metabolize the insulin‐induced H₂O₂ signal is pathologically elevated in Alzheimer's disease. Accordingly, therapeutic targeting of insulin‐induced H₂O₂ signaling may represent a novel treatment strategy for improvement of IR activation in neurodegenerative disorders.

6. MITOCHONDRIAL CONTROL OF INSULIN RECEPTOR ACTIVATION IN NEURONS

H₂O₂ signaling in mitochondria was shown to be an integral part of IR activation in neurons.64, 78, 80 In cerebellar cultures, insulin‐induced stimulation evokes a brief and transient, 15‐second increase in rates of H₂O₂ production preceding IR autophosphorylation; both processes are abolished by Prt and thus dependent on G protein signaling. Carbonyl cyanide‐4‐(trifluoromethoxy)‐phenylhydrazone (FCCP), the protonophore and uncoupler of oxidative phosphorylation in mitochondria, was shown to inhibit both insulin‐induced H₂O₂ production and IR autophosphorylation. The activation of IRs was reported to be sensitive to the changes in the mitochondrial inner membrane potential (ΔΨ), where a limited depolarization by FCCP leads to a reduction in the IR autophosphorylation.

In addition, the activation of IR was found to depend on the inhibition of mitochondrial complex II at flavine site (II_F). Malonate, an inhibitor of II_F, was shown to dose‐dependently suppress insulin‐stimulated receptor autophosphorylation that was abolished completely with the use of high concentrations. As the dose‐response curve was highly sigmoidal (Hill slope = 3.4) with the application of malonate, it can be speculated that minor decreases in succinate dehydrogenase activity can entirely prevent insulin‐induced IR activation.

Several lines of evidence indicate that the flavin site of mitochondrial complex II, succinate dehydrogenase, is implicated in the generation of the insulin‐induced H₂O₂ signal. First, it was shown that the H₂O₂ arises only from the flavin site (site II_F) of the complex II, and not from the ubiquinone‐binding site (site II_Q), as malonate, but not II_Q site inhibitors, reduces H₂O₂ production. At the same time, it is known that mitochondrial complex II generates H₂O₂ at high rates in the presence of micromolar succinate concentrations.87 Second, insulin stimulates both succinate oxidation and succinate‐supported H₂O₂ production in mitochondria. Given this, it is remarkable that a rapid stimulatory effect of insulin on the succinate oxidation by succinate dehydrogenase was demonstrated, using isolated rat hepatocytes.88 In addition, a dramatic increase in rates of succinate‐supported H₂O₂ generation was observed in mitochondria which were isolated from pretreated with insulin tissues.89 Notably, the effects of insulin are maximal at the physiological range of succinate concentrations, that is, 1‐9 μmol/L.90

Third, succinate triggers IR autophosphorylation in neurons in the presence of subthreshold insulin concentrations.78 The application of succinate, and other respiratory substrates, stimulates insulin‐dependent H₂O₂ production in the mitochondrial respiratory chain leading to an enhancement of IR autophosphorylation in cerebellar neurons.78, 91 Succinate elevates the insulin‐stimulated nonbasal autophosphorylation of IR, and it is considered an important endogenous sensitizer of the neuronal IR. Together, the H₂O₂ signaling pathway is considered as an important regulatory mechanism of neuronal IR activation.

The physiological significance of the mitochondrial control mechanism over IR activation remains to be elucidated. It can be hypothesized that this mechanism may serve as a factor that adjusts the IR activation to the level of synaptic activity. The finding that IR is highly expressed at the postsynaptic density of dendritic spines60 argues in favor of this hypothesis. In contrast, dendritic spines have a low content of mitochondria, present in less than ten percent of spines.92 Of note, increases in synaptic activity may result in insulin release into the synaptic cleft93 and the accumulation of mitochondria in dendritic spines.92 Thus, impairment of mitochondrial functions/transport to the site of synaptic activity is likely to impair IR activation, leading to subsequent pathological changes in the brain. In support of this view, numerous reports suggesting an association between synaptic dysfunction, insulin resistance, and mitochondrial dysfunctions have been accumulated at least for Alzheimer's disease.94, 95 Dysfunctional mitochondrial signaling during IR activation may therefore be regarded as a therapeutic target in disorders linked to deficient IR function.

7. REGULATION OF INSULIN LEVELS IN THE CNS

The availability of insulin is critical for the activation of IR in the brain, where it preferentially transported by the receptor‐mediated mechanism from blood across the blood‐brain barrier (BBB).96 The rate of insulin transport into the brain was shown to be influenced by various factors: Proinflammatory factors increase insulin transport,97 and exposure to high‐fat diet and obesity are associated with decreased insulin transport into the mammalian brain,98 which is restored by caloric restriction.99 Recent findings suggest a prominent role of astrocytes in the regulation of insulin transport across the BBB, as the ablation of astrocytic IR reduces this process.100

Correspondingly, it was demonstrated that insulin can be produced in neurons and astrocytes in substantial amounts.101, 102, 103, 104 In neurons, insulin was shown to be released in an activity‐dependent manner93 and be triggered by the transcription factor NeuroD1, a critical target gene of Wnt/β‐catenin signaling pathway.104 Insulin synthesis was found in GABAergic neurons of the rat cerebral cortex,105 as well as in olfactory bulbs, hypothalamus, and hippocampus.101, 104, 106 Insulin and C‐peptide production was found to be abundantly synthesized in adult neural progenitor cells of the hippocampus and olfactory bulb.104 The insulin and C‐peptide immunoreactivities were shown to colocalize and mainly detected at the cell soma and proximal dendrites, at concentrations that exceed blood levels.101, 106 Central insulin synthesis is apparently independent from its peripheral concentrations: It was reported that insulin levels in the hypothalamic parenchyma were not related to pancreatic insulin during fasting.107

The insulin production in the CNS decreases during aging and sporadic Alzheimer's disease,108 where the insulin expression and insulin protein levels in the frontal cortex, hippocampus, and hypothalamus become strongly reduced.109 Interestingly, in patients with Alzheimer's disease, APOE ε4 gene variation was associated with lowered insulin levels in cerebrospinal fluid (CSF).110 In vitro studies showed that the application of β‐amyloid peptide (1‐42) can inhibit insulin expression and reduce insulin levels in cultured astrocytes.103 Together, insulin supplementation can be of therapeutic value for patients with altered IR functions that can be particularly warranted as a treatment strategy in Alzheimer's disease.

8. INSULIN RECEPTOR–MEDIATED SIGNALING AND CNS PATHOLOGIES

Abnormalities in the IR‐mediated processes can be due to an impairment in the initial IR activation, reduced insulin availability, and compromised IR‐triggered downstream signaling.6, 30, 35 A large body of evidence has established their link to a broad range of CNS disorders, including, as mentioned above, neurodevelopmental syndromes, neurotoxicity, cancer, and neurodegenerative and neuropsychiatric disorders.9, 21 Compromised IR‐mediated signaling is commonly termed “insulin resistance”.111 In line with a commonly accepted definition, central insulin resistance can be viewed as an impaired biological response to exogenous or endogenous insulin which involves primary and/or downstream IR‐mediated signaling in the brain. Given multiple functions of IR in the brain, central insulin resistance is regarded as a multifaceted phenomenon that is not limited by aberrations in glucose metabolism45, 112 and associated with obesity13 and diabetes,10 but also was established as a feature of Alzheimer's20, 21 and Parkinson's diseases.22 For example, IR activation by insulin in Alzheimer's disease was shown to be diminished by 29%‐34%.45, 113

Up to now, lowered response of central IR during neuropsychiatric conditions was systematically studied and reported for a limited number of syndromes, with Alzheimer's disease investigated most extensively. This is because the issue with central IR‐mediated signaling during CNS disorders is a new field that started to attract the interest of researchers only recently. Yet, a number of indirect findings for altered IR‐mediated signaling and associated molecular cascades in various CNS pathologies have been accumulated. These disorders that are presumably related to dysfunctional properties of IR both during brain development and during degeneration of various origins are apparently linked to altered expression of IGF‐1/IGF‐2, their receptors (IGF‐1R and IGF‐2R), and other critical regulatory elements, as well as regulatory factors of the IR.23, 46

For instance, during Alzheimer's disease, apart from deficient IR activation, imbalanced phosphorylation of insulin receptor substrate‐1 (IRS‐1) was reported.45, 113 Studies of recent years have demonstrated that IR signaling is important for CNS development playing a role in dendritic outgrowth, neuronal survival, circuit development, synaptic plasticity, and postsynaptic neurotransmitter receptor trafficking. This may explain clinical observations suggesting that IR‐mediated processes are involved in the pathogenesis of neurodevelopmental conditions and cancer.14, 15, 18, 19, 114, 115 In particular, genetic studies accompanied by additional assays revealed a link between deficiencies in the IR‐mediated signaling and autistic‐like spectrum disorders that particularly concern insulin‐induced gene‐1 (INSIG1).116 In addition, genetic mutations of gene encoding a receptor of IGF‐1 were shown to be associated with developmental delay.117

The implication of IR and IR‐related indirect mechanisms was documented during various challenges in adulthood that include aging, neurodegeneration, stress, and exposure to toxins such as alcohol, and are related to the roles of IR in the neuroprotection.23, 24, 25, 26, 27, 28, 29, 118

9. CEREBELLUM PATHOLOGY AND INSULIN RECEPTOR–MEDIATED SIGNALING

As the cerebellum consumes about 70% of total CNS glucose, IR‐mediated signaling plays an exceptional role in its functions under normal and pathological conditions, as, for instance, in responses to insulin as compared to the forebrain cortex.119, 120, 121, 122 This may explain particular vulnerability of cerebellar functions during diabetes mellitus, many of which, such as a loss of pain sensitivity, impaired touch perception, and position sense, are decreased.123 Atrophy of the cerebellum has been often reported in patients with diabetes, and this is not associated with the duration of disease or glycemic control.124 Upregulation of IR gene expression and glucose transporter GLUT3 in the cerebellum during experimentally induced diabetes is regarded as a compensatory change of IR function in this structure,125 where inhibitory phosphorylation of IRS‐1 was found.112 Moreover, diabetes during pregnancy may result in the developmental anomalies of the cerebellum in offspring which are accompanied by abnormal levels of both IR and IGF‐1R.126

The cerebellum is also believed to be highly sensitive to IR‐mediated processes at various stages of its development that may underlie the association between a very frequent and highly malignant oncological condition in children, that is, medulloblastoma, constituting about 25% of all pediatric intracranial neoplasms, and abnormal functions of insulin receptor substrate‐1 (IRS‐1), a major signaling molecule of IR‐related cascades.16, 17, 127

10. EXOGENOUS SUPPLEMENT WITH INSULIN AS POTENTIAL THERAPY FOR CNS DISORDERS

In the face of well‐established IR dysfunction and evidence for reduced brain insulin concentrations in Alzheimer's disease, supplementation of insulin was extensively tested as potential therapy. Intranasal recombinant human insulin was successfully applied in several studies in patients with neurodegenerative and other disorders, being considered as a safe alternative to insulin injections.128, 129, 130 Initial clinical trials have demonstrated the efficacy of this therapy in improving cognitive performance of patients with Alzheimer's disease. Several clinical studies have demonstrated beneficial effects of intranasal insulin during mild cognitive impairment (MCI) and early stages of Alzheimer's disease. In patients with these conditions, a single intranasal administration of regular human insulin had an immediate positive dose‐dependent effect on verbal memory only in patients that are negative for APOE ε4‐, but not in carriers of APOE ε4+. The dose‐response curve of these effects was bell‐shaped within the dose range of 10‐60 IU and a peak at about 20 IU.131 The use of intranasal treatment at the dose of 20 IU twice a day for 21 days resulted in memory improvement in patients with MCI and early symptoms of Alzheimer's disease.132 In a similar study, daily intranasal treatment of subjects with these two conditions using 20 or 40 IU of intranasal insulin dosing for 4 months significantly improved memory and general cognition, where women were responsive to a lower dose of insulin as compared to men.133, 134, 135

Of interest, intranasal insulin administration had procognitive effects in nondemented subjects as well, while effective dose of insulin needed was much higher than in patients with dementia, constituting 160 IU/d, suggesting critical importance of IR‐mediated signaling during the above‐described pathologies.128, 129, 130 Therapeutic efficacy of intranasal insulin application was also reported in patients with depression, where fast effects of improving mood, self‐confidence, and general well‐being and decreasing anger, anxiety, and depressed mood were found.130 The use of intranasal insulin was shown to normalize eating behavior during obesity, balancing food intake and fat mass.129, 130

Altogether, the application of exogenous supplementary insulin, in particular via the intranasal route, was shown to be useful at least with some CNS conditions, where the disrupted IR primary activation was demonstrated. Extensive studies with this approach, however, revealed noteworthy limitations of this therapy. The primary one is a narrow range of therapeutic window as an increase in the dose of insulin above certain levels can worsen symptoms of dementia.131, 133

11. INSULIN RECEPTOR SENSITIZERS AS NEW PHARMACOTHERAPY OF NEUROPSYCHIATRIC DISORDERS

As an alternative to the use of intranasal insulin administration, a new class of compounds called “insulin receptor sensitizers” were generated and investigated for their preclinical and clinical efficacy during the last years. This term stems from the fact that these drugs can potentiate the binding of insulin to its receptor or its immediate molecular consequences via various mechanisms.78, 91, 136, 137, 138, 139 For example, recent clinical and translational studies have revealed antidepressant‐like effects, increased mitochondrial biogenesis in neurons, and decreased neuronal damage and antiinflammatory properties of the thiazolidinediones rosiglitazone and pioglitazone.137, 140, 141 Rosiglitazone and pioglitazone were reported to be effective in the treatment of major depressive disorder that was refractory to standard antidepressant treatment and accompanied by insulin resistance.138, 139

Currently, metformin and thiazolidinediones are considered as promising candidates for the treatment of Alzheimer's disease and other forms of dementia. Recent meta‐analysis showed potency of these two insulin sensitizers to reduce the combined relative risk for the incidence of dementia in patients with diabetes by 22% during combination therapy, and by 21% and 25% during monotherapy with metformin or thiazolidinediones, respectively. Moreover, a clinical pilot study with metformin used at the dose 2000 mg/d for 12 months showed improvement of MCI symptoms; however, only 10% of patients tolerated such a high dose of the compound.142 At the same time, the use of rosiglitazone in patients with Alzheimer's disease resulted in mixed outcomes that were seemingly dependent on the comorbidity with diabetes.143 Another thiazolidinedione, pioglitazone, was shown to induce therapeutic effects in patients suffering from both Alzheimer's disease and diabetes,144 but not Alzheimer's disease alone.145 Therefore, the issue of whether oral antidiabetic medications may be useful in the treatment of Alzheimer's disease patients without comorbid diabetes needs further studies.

Other insulin receptor sensitizers were recently proposed. Among them, commercially available analogues of hormone GLP‐1 were tested for the treatment of patients with Alzheimer's disorder. Among them, liraglutide was shown to restore normal hippocampal response to insulin in the IR/IRS‐1/Akt pathway and improve cognition in the APP/PS1 mouse model of AD.113 The 6‐month treatment of patients with this drug ameliorated brain glucose uptake but not cognition scores.146

It may be suggested that insulin receptor sensitizers with direct action on the IR are likely to be more effective therapeutically. As mitochondrial H₂O₂ signaling directly affects IR, it can be targeted during dysfunctional activation of IR. Given that the magnitude of the insulin‐induced H2O2 signal depends on succinate concentrations in the cell89, 91 of mitochondrial complex II, dicholine succinate (DS), a bioavailable succinate source, was tested as potential therapeutic insulin receptor sensitizer. DS was found to dose‐dependently stimulate insulin‐dependent H₂O₂ production of the mitochondrial respiratory chain in cerebellar neurons leading to an enhancement of the IR via insulin‐stimulated autophosphorylation of the IR kinase at tyrosine residues in neurons.31, 64, 78, 91 A P‐NMR in vivo study demonstrated that DS preserved whole‐brain ATP decline in a rat model of global ischemia.147 Based on this and that DS demonstrated positive effects in animal models of brain disorders, including Alzheimer's disease, depression, chronic cerebral hypoperfusion, aging, and exposure to β‐amyloid peptide toxicity,10, 20, 21, 22, 36, 37, 45, 111, 112, 148 it can be regarded as a drug candidate that normalizes IR functions.53, 54, 91, 147, 149, 150 Its further clinical use can be promising in providing effective stimulation of IR‐mediated signaling while overcoming reported limitations with the use of insulin or insulin receptor sensitizers.

12. CONCLUSIONS

The ultimate goal of this review is to highlight the physiological and clinical importance of mechanisms of the primary activation of IR in the context of the CNS disorder pathophysiology and treatment response. We explicitly point out the remaining gaps in knowledge that await resolution in an area of research currently going through a renewed and dynamic development. Nevertheless, presently available data are already sufficient to draw wide‐ranging conclusions regarding the importance of IR activation mechanisms in the pathogenesis and treatment of CNS disorders and the therapeutic potential of supplementary insulin and insulin receptor sensitizers.

13. PERSPECTIVES

Evidently, some fundamental questions of IR physiology have not yet been addressed experimentally, but can be investigated using quite basic approaches. First, while the role of H₂O₂ in the mitochondrial signaling was identified for IR‐A, it was not studied in IR‐B. In the context of CNS functions, this could be of particular relevance for understanding of the role of IR‐mediated signaling in astrocytes, which express both IR subtypes. Second, it was not studied whether or not IGF‐2 induces similar cellular effects in response to those of insulin on IR‐A; for instance, H₂O₂‐related mechanisms are involved in these cellular effects. As IGF‐2 is present in neuronal precursors and was suggested to govern their differentiation to neurons, it would be of significance to study the role of IR‐A and H₂O₂ in these processes. Several particular findings regarding IR would be of interest to address. For example, the clarification of the physiological role of IR‐A in olfactory bulbs that express maximal density of this IR subtype would help to elucidate the functions of central IR in general. Finally, whether IR‐A and IR‐B are differentially involved during various medical conditions apart from AD, such as obesity, metabolic syndrome, depression, and other known neurodegenerative and neurodevelopmental disorders, remains unexplored, as well as the role of primary IR autophosphorylation during various diseases. This information is required for the elaboration of a conceptual framework for further insight into the mechanisms through which insulin facilitates critical brain functions, including development, metabolism, cognition, and motivated behaviors. Eventually, this knowledge is likely to be useful in the development of new therapies targeting IR. Given the limited ability of succinate to penetrate the blood‐brain barrier, a strategy is proposed for the development of succinate donors and systems delivering succinate to the CNS.

ACKNOWLEDGMENTS

The work reported here was supported by the European Community (EC: the European Union's Seventh Framework Programme (FP7/2007–2013) under Grant No. 602805 (Aggressotype), the Horizon 2020 Research and Innovation Programme under Grant No. 728018 (Eat2beNICE), the “5‐100” Russian Academic Excellence Project (to KPL and TS), and the German Research Foundation (DFG: SFB TRR58‐A05 to KPL).

CONFLICT OF INTEREST

The authors declare no conflict of interest

Pomytkin I, Costa‐Nunes JP, Kasatkin V, et al. Insulin receptor in the brain: Mechanisms of activation and the role in the CNS pathology and treatment. CNS Neurosci Ther. 2018;24:763–774. 10.1111/cns.12866