“Manganese and the Insulin-IGF signaling network in Huntington’s disease and other neurodegenerative disorders”

Abstract

Huntington’s disease (HD) is an autosomal dominant neurodegenerative disease resulting in motor impairment and death in patients. Recently, several studies have demonstrated insulin or insulin-like growth factor (IGF) treatment in models of HD results in potent amelioration of HD phenotypes via modulation of the PI3K/AKT/mTOR pathways. Administration of IGF and insulin can rescue microtubule transport, metabolic function, and autophagy defects, resulting in clearance of Huntingtin (HTT) aggregates, restoration of mitochondrial function, amelioration of motor abnormalities, and enhanced survival. Manganese (Mn) is an essential metal to all biological systems but, in excess, can be toxic. Interestingly, several studies have revealed the insulin-mimetic effects of Mn—demonstrating Mn can activate several of the same metabolic kinases and increase peripheral and neuronal insulin and IGF1 levels in rodent models. Separate studies have shown mouse and human striatal neuroprogenitor cell (NPC) models exhibit a deficit in cellular Mn uptake, indicative of a Mn-deficiency. Furthermore, evidence from the literature reveals a striking overlap between cellular consequences of Mn-deficiency (i.e. impaired function of Mn-dependent enzymes) and known HD endophenotypes including excitotoxicity, increased reactive oxygen species (ROS) accumulation, and decreased mitochondrial function. Here we review published evidence supporting a hypothesis that (1) the potent effect of IGF or insulin treatment on HD models, (2) the insulin-mimetic effects of Mn, and (3) the newly discovered Mn-dependent perturbations in HD may all be functionally related. Together, this review will present the intriguing possibility that intricate regulatory cross-talk exists between Mn biology and/or toxicology and the insulin/IGF signaling pathways which may be deeply connected to HD pathology and perhaps, other neurodegenerative diseases (NDDs) and other neuropathological conditions.

1.1. Introduction

Between 1–3 out of 100,000 individuals are diagnosed with Huntington’s disease (HD) in the U.S. However, given the autosomal dominant etiology and near 100% penetrance of HD, generations of families are devastated by this disease. HD is caused by an expanded tri-nucleotide CAG repeat in the HTT gene. If these repeats surpass 35–40 repeats, there is a near 100% chance that the patient will present with Huntington’s disease at some point in their lifetime (usually in middle-late adulthood). While the HTTgene was discovered in 1993, there is still no cure for HD though several drugs have been used to treat symptoms (i.e. tetrabenazine for chorea). Furthermore, researchers still do not fully understand 1) the exact function(s) of wild-type HTT is in the human brain or 2) how mutant HTT (HTT >35 CAG repeats) causes neurotoxicity and HD. Two of the posited causes for HD are 1) mitochondrial dysfunction 2) autophagic dysfunction and aggregate accumulation. Recently, a series of studies have shown that insulin/insulin-like growth factor (IGF) treatment in HD models can ameliorate both of these pathogenic mechanisms.

Manganese (Mn) has only been recently implicated in HD, and studies have suggested that a Mn deficiency may underlie some of HD pathology^1–5. Interestingly, Mn can modulate insulin/IGF homeostasis, shown to be essential for mitochondrial function, and able to stimulate neuroprotective pathways associated with the activation of autophagy, namely insulin/IGF signaling (IIS). This review explores the functional intersection of these two modifiers of HD, (a) Mn biology and (b) insulin/IGF signaling (IIS)—both have been shown to regulate autophagy and mitochondrial health/function. Here we will review a role for Mn and IGF joint dysregulation in HD pathology and briefly explore some the implications of this co-regulation in the context of other neurodegenerative diseases and conditions.

While Huntington’s disease will be discussed in detail, other neurodegenerative diseases (NDDs) will also be referenced when studies provide mechanistic understanding of the roles of Mn and IGF/insulin given the shared cellular pathologies between NDDs and HD (i.e. aggregate accumulation, reactive oxygen species, mitochondrial dysfunction). It is plausible that the mechanisms of these NDD pathologies might be quite similar to HD.

1.2.1. IIS signaling and its role in the brain

Insulin and insulin growth factor (IGF) are homologous growth hormones that classically regulate cellular metabolism. Their role in peripheral tissues has been well elucidated. However, only more recently has their role in brain health and development been studied. In the brain, IIS is necessary for synaptic maintenance and activity, neurogenesis, neurite outgrowth, neuronal survival, mitochondrial function and maintenance as well as upper-level processes including memory and feeding behavior and thus dysregulation in neurotrophic support has long been proposed as a mechanism of neurodegenerative diseases^6–31. Insulin and IGF are mainly produced in the pancreas and liver, respectively, and transported to the brain from the periphery through the blood brain barrier. Alternatively, IGF and insulin can enter the brain through CSF in the choroid plexus. IGF is also produced locally in all brain regions. Upon binding with their respective ligands, IGF receptors (IGFR) and insulin receptors (IR) undergo autophosphorylation at three tyrosine residues required for activation. Subsequently, the IR kinase domain phosphorylates IR substrates (IRSs) which act as secondary messengers, impinging upon a variety of cell signaling pathways including PI3K/AKT, mTOR, and MAPK/ERK to exert their biological effects (e.g. energy metabolism, cell stress responses)³². However, individual receptors can heterodimerize forming hybrid IGF/insulin receptors which can bind either insulin or IGF and activate both the PI3K/AKT and MAPK/ERK pathways. S6, a downstream target of mTOR acts as negative feedback, phosphorylating and inactivating IRSs. Six IGF binding proteins exist (IGFBPs) and act to regulate IGF-R binding and modulate signaling. IGFBPs show a selective expression pattern, being in distinct portions of the brain where they presumably act on specific IIS signaling within anatomical subsets of neurons. These proteins have a higher affinity for IGF than do IGF receptors, allowing tight control of IGF bioavailability. The regulation of neuronal IGFBPs is still quite unknown but evidence suggests specific mechanisms for each protein including control by epigenetic markers and neuronal activity of specific cell types^33,34.

Most kinases in humans are either magnesium (Mg) or Mn dependent. Though most are Mg dependent, several are preferentially activated by Mn including ATM and mTOR^35,36. While little research has been done to explore the role of Mn as a signaling molecule its inherent role in kinase activation suggest Mn is essential for cell signaling. Several other proteins are also activated by Mn including Arg, MRE11, Mn-SOD, glutamine synthetase, pyruvate decarboxylase, protein phosphatase 1, and many integrin-related proteins^37–46. Interestingly Mn has been shown to activate several of the same pathways as IGF/insulin including AKT, mTOR, and ERK/MAPK, and even the insulin/IGF receptor itself— all of which have been found to be neuroprotective in HD^46–56.

1.2.2. Mn and insulin/IGF homeostasis.

Mn toxicity has been linked to neuronal cell death and neurodegenerative conditions for several decades—namely Parkinson’s disease (PD) and manganism. Though recent studies have yielded greater understanding of toxic effect of Mn on neuronal function, very little is known about basic, neuronal Mn homeostasis. While brain imaging studies have revealed where Mn accumulates within the brain, there is disagreement on what sub-compartment(s) Mn primarily accumulates within a neural cell. The field is in some contention as some studies suggest mitochondria while others suggest within the nucleus^57–59. Surprisingly few studies have examined whether Mn primarily accumulates in neurons vs glial cells. Lastly, there is poor understanding of how Mn is transported within a cell, primarily due to the high promiscuity of Mn transporters for other metal ions^37,60. Muddying this understanding, at present there is only one transporter which seems specific for Mn, SCL30A10, an efflux transporter. Interestingly, mutations in this transporter lead to Mn accumulation in vitro and in vivo and have been linked to increased brain Mn and PD in patients^60–63. The answers to these basic questions could offer invaluable understanding of Mn biology in the context of both diseased and healthy brains.

Evidence of a role for Mn-dependent regulation of IIS has been steadily amassing since the 1980’s. Baly and colleagues showed Mn-deficiency caused glucose intolerance and reduced insulin production in rats, mimicking diabetic-like phenotypes^64–67. In addition, rats fed a Mn-deficient exhibited reduced pancreatic insulin output following a glucose stimulus. Furthermore, they and others found Mn to be an insulin-mimetic, promoting insulin excretion and activating insulin-related metabolic kinases^64–69. Around this same time, another study showed that Mn-deficient rats exhibited decreased circulating IGF-1 and insulin and increased IGFBP3—potentially suggesting Mn might regulate circulating IGF-1 levels via modulating IGFBP3 activity⁷⁰. Later, Lee and colleagues reported that Mn supplementation could protect against diet-induced diabetes in mice via increased insulin excretion, amelioration of glucose intolerance, and increased expression of Mn superoxide dismutase (MnSOD), a Mn-dependent anti-oxidant enzyme in mitochondria⁷¹. These results were consistent with reports that diabetic patients were responsive to oral Mn treatment as well as reports of reduced blood Mn in diabetic patients^72–74. Concurrently, other groups established that Mn deficiency was associated with reductions of IGF1 in serum and Mn supplementation could increase IGR-R and IGF1 expression in the hypothalamus of rats^{53, 54, 56, 70, 75–78} However, the mechanisms bv which Mn increases IGF1 and insulin levels remain unknown. Together, these findings suggest a functional link between Mn and the regulation of IGFl/insulin levels in both peripheral tissues and brain. While such studies clearly link Mn to diabetes and hypothalamic/pubertal development, the role of this potent regulatory mechanism has never been studied in the context of a neuronal disease or manganese toxicity.

1.3.1. HD pathobiology

HD is an autosomal dominant neurodegenerative disease which results in hyperkinetic movements, behavioral changes in cognition and mood, and ultimately death. An expanded trinucleo-tide (CAG) repeat in the Huntingtin gene (HTT) resulting in a mutant HTT protein (mHTT) causes HD. Higher CAG repeats are correlated with increased disease severity and younger age of onset though both are highly variable even between patients with similar repeat size^79,80. Usually, the disease manifests in adulthood (though juvenile cases do occur), and gives rise to a combination of motor, cognitive, and psychiatric symptoms which ultimately result in death. A hallmark symptom of HD is chorea, uncontrolled hyperkinetic movements, which has been associated with mHTT-dependent cell death within the striatum. Degeneration in other brain regions (cortex, hypothalamus) usual follows, contributing to the variability in symptoms. As HTT is ubiquitously expressed, the basis for the selective neurotoxicity of mHTT for striatal medium spiny neurons (MSNs) and a handful of other neuronal sub-populations remains a mystery^80–84.

1.3.2. Mn dysregulation in HD.

Mn dysregulation has only recently been implicated in HD. In normal brains, Mn accumulation is enriched in the basal ganglia—the part of the brain which most severely degenerates in HD—suggesting Mn serves an important role in this brain region^{59,85, 86}. Recently a set of studies revealed a Mn transport deficit, indicative of a brain-specific Mn deficiency, in an HD immortalized striatal neuroprogenitor cell line (STHdhQ111/Q111), in HD hiPSC-derived striatal NPCs cells, and also in the striata of YAC128Q mouse model of HD^{1, 4}. The mechanism of this Mn-transport deficit has been difficult to resolve as so little is known about Mn sub-cellular transport. Analysis of Mn homeostasis is complicated by the high promiscuity of proposed Mn transporters for other essential metals^{37, 60, 87, 88}.

However, Mn is known to activate several of the signaling pathways dysregulated in HD including ATM/p53 and AKT/mTOR^{1, 53, 54, 88–90}. STHdh Q111/Q111 and hiPSC-derived striatal neuroprogenitor HD cell models exhibit decreased net Mn up-take leading to diminished ATM activation, a Mn-responsive kinase upstream of p53 and other cellular stress response proteins¹. Similar to ATM/p53, Mn robustly activates AKT and mTOR, both of which are neuroprotective in HD^91–98. AKT activation can increase HTTSer421 phosphorylation, shown to facilitate axonal transport, restoring mitochondrial and autophagic function in HD models^{92, 95–97, 99–102}. In contrast, Guilarte and colleagues reported decreased HTTSer421 phosphorylation by Mn in YAC128 mouse cortical and hippocampal primary cultures, though striatal levels were not assessed^{4, 5}. Lastly, reinstatement of aberrant mTOR activity in HD models restores autophagic function, enhances aggregate clearance, and increases MSN health, though some reports have shown mTOR inhibition to be neuroprotective in HD^91,103,104.

1.3.3. IIS dysregulation in HD.

Recently, several groups observed impaired IIS in HD. Paradoxically, reduced IGF1 expression has been detected in patient caudate tissue and skin-fibroblasts as well as other non-human HD models, while increased IGF1 has been found peripherally in HD and this has been correlated with cognitive decline^96,105,106. Previous studies have shown mutant HTT disrupts intracellular transport and secretion of insulin while others have shown Mn can act as a potent insulin-mimetic in vivo⁶⁹. Additionally, several groups reported robust neuroprotective effects of IGF1 treatment in HD cell and mouse models via increased 1) AKT/ERK signaling 2) IRS2/VPS34 (Class III PI3K) signaling and 3) increased HTT Ser421 phosphor-ylation. Upregulation of these pathways increased autophagic function, aggregate clearance and ameliorated mitochondrial dysfunction^{96, 97, 100, 101, 107–111}. Administration of IGF and insulin can also rescue microtubule transport, amelioration of motor abnormalities, MSN health, and enhanced survival in cell and rodent models. IGF1 is also neuroprotective in models of other NDDs^112–119.

1.3.4. Autophagy deficits in HD, potential links to Mn and IIS.

The inability to clear toxic mHTT aggregates may be a principle mechanism of HD-related cell death though there is contention about which form(s) and fragment(s) are truly toxic and which are a compensatory/protective reaction to cellular toxicity^120–123. Autophagy, a process by which cells degrade complex organelles and proteins to base nutrients, is also the primary process in clearing mHTT aggregates^{121, 123–129}. HTT acts as a scaffold for autophagy and this activity is altered or impaired by mHTT, potentially exacerbating pathogenesis^{126, 130–133}. In HD, autophagic impairment causes failure of cargo-recognition and lysosomal function resulting in the accumulation of cellular waste and protein aggregates¹³⁴. This may trigger a feed-forward pathogenic loop with ever increasing mHTT levels further impairing clearance¹²⁶.

IGF treatment incurs robust amelioration of autophagy defects in HD models. Rothman and colleagues observed that IGF1 upregulates autophagy via an IRS2/VPS34-dependent mechanism in HD cells, resulting in a marked increase in aggregate clearance. This is an AKT/mTOR-independent process, though both AKT and mTOR are activated by IGF1¹⁰⁷. Additionally, other groups have shown that upregulation of mTOR in HD models increases autophagy and aggregate clearance, rescuing HD-related phenotypes even though mTOR canonically acts as a negative regulator of autophagy by inhibiting ULK1^91,135. Interestingly, published studies indicate Mn both increases and decreases autophagy in neuronal systems in a biphasic, time-dependent manner^{55, 136}. Given this regulation of autophagy by Mn and Mn-responsive pathways, it seems plausible that correcting Mn homeostasis in HD models may ameliorate aspects of autophagic dysfunction. To date however, there has been only a handful of studies exploring the role of Mn in autophagy—and the majority have been done in the context of Mn toxicity, instead of Mn essentiality^{55, 137}. Given clear ties of Mn biology to pathways upstream of autophagy, future studies should interrogate the role of Mn in autophagy during normal neuronal function, in addition to disease states. In particular, we need to establish whether Mn plays a role in basal autophagy or only in the context of Mn toxicity.

1.3.5. Mitochondrial pathology in HD, possible links to Mn and IIS

Mitochondrial dysfunction is another mechanism by which mHTT may cause selective neurodegeneration in HD. Mitochondrial dysfunction may contribute to neurodegenerative diseases (NDD) for several reasons; 1) High mitochondrial respiration is needed to accommodate high ATP usage in neurons, 2) mitochondria, out of all organelles, produce the highest amount of intracellular reactive oxygen species (ROS), 3) mitochondria are a critical regulator of cell death, a common feature of most NDDs, 4) mitophagy (mitochondrial selective autophagy) is often defective in NDD, and 5) perturbations in various metabolic processes, indicative of mitochondrial dysfunction, are often associated with NDD^138–140. In HD, specifically, overt metabolic effects such as rapid weight changes and defects in glucose homeostasis have been observed in HD patients and models^141–151. Also, WT HTT has been shown essential for mitochondrial health¹⁵². To this end, several basic studies and clinical trials have investigated metabolic targets as potential therapeutics for HD including creatine and Coenzyme Q10, but have found little success^153–159.

Several landmark studies demonstrate IGF1 restores mitochondrial health in HD models^{97, 100, 101}. Given the IIS-mimetic effects of Mn, correcting Mn homeostasis may ameliorate some facets of mitochondrial dysfunction in HD. This hypothesis is consistent with established roles for Mn in mitochondria: 1) Mn accumulates in mitochondria more so than other organelles supporting a functional need in this organelle; 2) Mn has anti-oxidant functions via the Mn-dependent, mitochondrial enzyme, MnSOD; and 3) Mn is essential for the function at least two gluconeogenesis enyzmes^{37, 57,58,60,88}. Rego and colleagues have reported a series of studies providing a mechanistic understanding of how IGF is capable of such robust amelioration of HD symptoms^{100, 101, 139, 153, 160–169}. They found HD models exhibit reduced ATP/ADP ratio, decreased O2 consumption, increased mitochondrial ROS and fragmentation, aberrant lactate/pyruvate levels and decreased mitochondrial membrane potential— all of which indicates mitochondrial dysfunction. Each of these was shown to be ameliorated by IGF treatment via upregulation of PI3K/AKT signaling in cellular and mouse models of HD.

1.4.1. IIS signaling and Mn in other NDDs

Abnormal levels of IGF/insulin and decreased IIS signaling (namely, reduced AKT signaling) have been observed in all neuro-degenerative diseases including PD, Alzheimer’s disease (AD), amyotrophic lateral sclerosis (ALS), multiple sclerosis (MS), spinocerebellar ataxias (SCA), and other NDD-like conditions such as ataxia telangiectasia (AT). In the case for many models of these diseases, IGF or insulin have been successfully used to ameliorate pathologies in vitro and in vivo; and they have been used or targeted in clinical trials^{114,170–172}. Unfortunately, these clinical trials have reported little success. One possible reason for this is control of IGF-1 bioavailability by IGFBPs. This could be overcome by using a modified IGF-1-like peptide which is unable to bind IGFBPs¹⁷³. Furthermore, although many studies have shown perturbation in metal ion homeostasis in these diseases, few have explored a more specific role for Mn dysregulation. Recent studies elucidating Mn or IGF/Insulin dysregulation in NDDs will be reviewed next, emphasizing developments in recent years.

1.4.2. PD and IIS/Mn

PD is a neurodegenerative disorder resulting in bradykinesia and motor rigidity affecting an estimated 10 million people worldwide. Symptoms of the disease mostly occur in late adulthood as a threshold of dopaminergic neurons in the substantia nigra degenerate. Unlike HD, there is no clear genetic predisposition for most cases of PD, though mutations in some genes are correlated to increased risk for PD. Given this and the late-onset of the disease, many studies have focused on environmental modifiers of the disease¹⁷⁴. PD has long been associated with perturbations in metal ion homeostasis—particularly iron (Fe) and Mn. Mn toxicity causes parkinsonian-like symptoms and a disease-state known as manganism, but most agree that its pathology is different from that seen in PD. This is mainly because neurodegeneration in PD occurs primarily in the dopaminergic neurons of the substantia nigra while Mn toxicity manifests within the globus pallidus. Furthermore, at least some patients with Mn induced parkinsonism do not produce Lewy bodies and can be unresponsive to levodopa treatment^175–177. While these two diseases may be distinct, several lines of evidence support a role for Mn dysregulation in PD. Chronic exposure to Mn is associated with increased risk for PD. Also, Mn toxicity has been linked to reduced tyrosine hydroxylase and dopamine levels and DAT cell surface expression but reports regarding impaired neuro-transmission and viability in dopaminergic neurons have been inconsistent^{37, 177–181}. Mn toxicity has also been associated with increased alpha synuclein build-up, but it is unclear if this response is neuroprotective or enhances neurodegeneration^182–184.

IGF has been studied in the context of PD as well. Previous studies have revealed neuroprotective effects of IGF in PD models and associated with increased dopaminergic survival in the substantia nigra^{112, 182, 185–187}. However, the majority of recent studies mainly focus on plasma IGF-1 levels as a biomarker for PD progression. Several groups published studies suggesting IGF-1 levels were increased in the sera of PD patients compared to control^{188, 189}. Furthermore, studies revealed that increased plasma IGF-1 levels correlate with cognitive decline and motor symptoms^{188, 190}. While these studies have great utility as a clinical tool and seem to be quite sensitive, they have added minimal mechanistic insight as to if or why IGF-1/insulin and related signaling may be dysregulated or pathogenic consequences. Thus, continued basic and mechanistic experiment to understanding of IGF’s role in PD are needed to resolve inconsistencies and provide detail.

AKT has received considerable attention in the PD field via its neuroprotective roles in the brain. Aside from reduced p-AKT levels found in post-mortem PD brains, several studies have linked increased AKT and IIS signaling to both reduced dopaminergic cell death, reduced alpha synuclein toxicity and complex interactions with PD-related proteins including PARKIN, PINK1, and DJ1^{112,185, 191–195}.

1.4.3. AD and IIS/Mn

AD results primarily from the degeneration of hippocampal neurons which leads to severe cognitive defects in late-adulthood. Disease is defined by two hallmark pathological features, neurofibrillary tangles (hyperphosphorylated tau) and amyloid beta plaques, two aggregates which incur neurotoxic stress. Heavy metals have also been associated with AD and its aggregate pathology, though few studies have examined Mn levels or dysregulation^196,197. However, two recent studies investigated plasma Mn levels in AD and reported opposing results. Dehua and colleagues reported elevated Mn levels which were correlated with increased amyloid beta expression and reduced cognition while Bush et al reported reduced Mn levels in sera but no difference in patient erythrocytes^{198, 199}.

AD may have the most significant ties to IGF dysregulation of all NDDs. AD has been heavily correlated to diabetic status and mechanistic understanding of metabolic dysfunction in AD has led to it being referred to as “type 3 diabetes”, a form of diabetes that specifically affects the brain²⁰⁰. In recent years, studies have focused primarily on the effects of IGF/insulin on amyloid beta accumulation and the use of IGF-1 levels as a biomarker for disease risk and progression. Two studies in 2009 reported that reduced IGF signaling protects against AB accumulation, potentially by acting on the plaques themselves, condensing them to less toxic forms^201–204. These were contrary to a flurry of studies in the early-mid 2000’s revealing IGF resistance and ameliorative effects by IGF treatment on AB accumulation and cognitive function^{115,205–212}. A few years later, insulin resistance and reduced IIS signaling was found in postmortem AD brain tissue and soon after that, lower serum IGF-1 levels were correlated to an increased risk for AD and dementia while higher levels were associated with greater brain volume^{212, 213}. Interestingly, increased IGF has been reported in CSF of patients^214,215. Thus, even though conflicting results have been reported, these studies reveal that AD is deeply tied to IGF biology.

Contrary to PD, excessive AKT signaling has been observed in AD. Several studies have reduced or inhibited IIS signaling and observed delays in symptoms and reduced AB pathology^{206,216, 217}. These results, of course, are contrary to aforementioned studies utilizing IGF treatment in AD models. Such conflicting results may be explained by an initial hyperactivation of IIS signaling which eventually desensitizes the pathway. In this way, both IIS inhibition early or IIS treatment late in disease progression may result in ameliorative effects. However, further research will have to be done across disease progression to see if this is indeed the case.

1.4.4. ALS and IIS/Mn

ALS is a neurodegenerative disease which affects more than 12,000 people in the U.S. Disease onset is more variable than other diseases and can often occur in younger people. The cause of ALS is unknown but pathology is attributed to loss of motor neurons in the brain and spinal cord resulting in loss of voluntary muscle control and, in late-stage, patients are unable to move or breathe without ventilator support. ALS has also been associated with metal ion dysregulation. Again, few studies focused on Mn levels but a few studies have reported increased Mn in CSF and plasma while the other reports no change in Mn but significant increases in copper and zinc and a reduction in selenium^218–221.

IGF dysregulation and insulin resistance has been reported in ALS^222–224. These data led to a few in vivo studies using IGF-1 treatment in ALS models. While subcutaneous injection into the periphery with IGF-1 was largely found to be ineffective, direct intrathecal injections directly into the CSF resulted in some decrease in motor atrophy^170,225. Given these results, a few clinical trials have been attempted in ALS but have found little success^{171, 226,227}. One reason may be that these treatments are given peripherally instead of intrathecally¹⁷³. More recently, IGF2 has been found to be neuroprotective in ALS models¹¹³.

1.4.5. Autophagy in other neurodegenerative diseases

Autophagy has been linked to every neurodegenerative disease—namely because most NDDs develop aggregate pathology which is often processed by autophagy. While autophagy is activated as a protective process in order to maintain healthy homeostasis of the cell, if hyperactivated can result in autophagy-mediated cell death. Thus, interactions between aggregates and autophagy play a precarious role in NDDs²²⁸. Recent studies have begun to explore the effects of metal toxicity on autophagy as well^{55, 137}. In PD, autophagy has primarily been investigated in the context of mitophagy (mitochondrial specific autophagy). PD has been linked to mitochondrial toxicity and dysfunction which incurs mitophagy in an attempt to remove unhealthy mitochondria from the neurons to reestablish cellular integrity. PARKIN and PINK1, two proteins associated with familial forms of PD, are essential members of the mitophagy process^229–234. In AD, autophagy is known to regulate both the secretion and degradation of AB which adds increased complexity to its role in disease pathology. Several studies have revealed increased autophagasome accumulation in AD models, but these results have been inconsistent across disease progression^235–238. Recently, ALS studies have revealed that two ALS associated proteins, TDP-43 and SOD1, are often dysregulated in ALS patients and models^239–241. Interestingly, mutations in these proteins (amongst several other observed ALS mutation-associated proteins) cause aberrant autophagic processing in neuronal and spinal cord neurons²⁴². Further studies are needed to elucidate mechanistic understanding of these complex relationships to determine whether dysregulated autophagy is a pathogenic mechanism or compensatory “rescue” response. Future investigation must interrogate autophagic flux rather than commonly used end-point measurements as the directionality and capacity of autophagy is necessary for forther understanding and therapeutics. The connections that have been drawn between autophagy and Mn or IGF/insulin warrant continued exploration but studies should consider potential co-regulation of Mn and IGF/insulin on autophagy processes and dysregulation.

1.5.1. Manganese toxicity and IGF

Little investigation has been done to examine the role of IGF in manganese toxicity. Tong and colleagues found Mn toxicity caused reduced ATP and insulin/IGF receptor expression. Additionally, as mentioned before, Hiney and colleagues have been revealing a role for Mn-induced toxicity in hypothalamic development via IGF/mTOR related pathways^{56,53, 54, 56, 76–78}. It is likely that Mn toxicity in other brain regions are regulated a similar manner. Given that Mn accumulates in the brain primarily in the basal ganglia, not the hypothalamus, it seems likely that and IGF/Mn interaction may play even more crucial roles in other brain regions, particularly in aged model systems. Thus, future studies on Mn toxicity and IGF could be informative on developmental toxicity, chronic environmental exposures, and overall brain health.

1.5.2. The co-regulation of ATM, Mn, and insulin/IGF

Interestingly ATM, a Mn activated kinase, has been linked to both IGF/insulin and Mn signaling. Previous studies have shown that Mn induced p53 activity is regulated by ATM. Furthermore, this Mn-induced activity is blunted in HD due to lack of bioavailable Mn¹. Separately, low levels of the IGF-1 receptor and loss of IGF-1 sensitivity have been observed in Ataxia Telangiectasia (AT), the disease resulting from loss of function mutations in ATM, and in loss-of-function ATM models^243–246. Additionally, studies have shown patients with AT have significantly decreased IGF-1 levels^247–250. Additionally, others have shown ATM is essential for IGF and IGF-R transcription by phosphorylating and relieving transcription factors and complexes including p53 from their respective promotors, allowing for transcription^{243, 244, 251–254}. Concurrently, downregulation of IGF-R results in increased radiosensitivity and decreased ATM protein levels (mRNA was unchanged) revealing a potential circular regulation between ATM and IGF-R^{244, 255, 256}. Also, given that ATM is required for full activation of AKT, it seems likely that the connections between ATM, Mn, and IGF carry some biological relevance in the context of Mn/IGF co-regulation in NDD²⁵⁷. Mn could act as an initiating signaling molecule within this cascade where Mn activates ATM/p53 which results in increased IGF/IGF-R transcription and subsequent activation of the PI3K/AKT pathway. This hypothetical, albeit plausible, interaction could explain how a Mn deficiency in HD might contribute to decreased IIS (AKT/mTOR) and Mn-induced ATM/p53 signaling.

1.5.3. IIS signaling, Mn and cancer

Given the striking parallels and potential co-regulation between Mn and IIS and the pronounced and well-studied roles of IIS in cancer progression, one must wonder if there is role for Mn/IIS co-regulation in cancer etiology. As a pro-growth signaling pathway, IIS is often highly upregulated in cancers particularly during tumor progression²⁵⁸. However, most findings suggest Mn is not significantly carcinogenic, even to exposed workers. In fact, Mn deficiency leads to a higher risk. A plethora of studies, namely clinical examination of Mn levels in cancer patients, support the role for Mn deficiency in cancer via reduced MnSOD activity and enhanced ROS accumulation in various cancer types^259–263. Of note, Mn has been shown to be essential for the activation of ATM and MRE11, two DNA-damage repair proteins, and able to increase phosphorylation of p53, the most-well studied tumor suppressor gene which exerts control on cell cycle supporting a role for Mn deficiency in cancer. In fact, many cancers contain mutations in these same proteins. Somewhat paradoxically, HD is associated with reduced Mn bioavailability and reduced risk for cancers²⁶⁴. Accumulating data, studies, and clinical trials support a hypothesis that perturbations in IIS and metal ion homeostasis separately contribute to both NDDs and cancer in somewhat opposite fashions while a dearth of investigation exists to study their potential co-regulation in either disease.

1.6. Conclusions

The roles for IGF and Mn separately in HD are far from being fully elucidated. However, the sizeable overlap between their homeostasis and downstream effects supports a need to consider their coregulation in the context of diseased and healthy states. Neuroprotective cell signaling (i.e. AKT, mTOR, ERK/MAPK), mitochondrial health, and autophagic function have been implicated in all NDDs repeatedly by multiple groups. Past and present research has revealed an essential role for IIS in coordinating these cellular processes. However, little attention has been given to Mn role even though distinct lines of evidence substantiate its essentiality in these very same processes and even the upstream regulation of insulin/IGF. There is not enough evidence one way or another to draw a clear conclusion whether Mn may be at the heart of IIS dysregulation in NDDs, but there is certainly enough to warrant serious consideration of its role as a contributing factor.

It is still unclear how Mn is exerting its effects on IGF/insulin levels and signaling. Is Mn acting at the levels of transcription, translation, or post-translationally? The intriguing possibility that Mn might regulate IGF and IGF-R transcription through ATM/p53 is one that merits further study as it may have implications in not only NDDs but cancer and diabetes as well. Furthermore, given the widespread transcriptional targets of p53, Mn could be widely essential for the transcription of various proteins. Mn could also be exerting its control post-transcriptionally - potentially at the blood brain barrier or via interactions with IGF binding proteins. Clegg and colleagues reported that Mn deficiency resulted in increased IGF-BP3 which they suspected might reduce IGF bioavailability⁷⁰. However, little investigation has been done to follow up on these findings or explore Mn’s role on other IGF-BPs which could offer a clear mechanism of Mn’s regulation of IGF.

We discussed here many examples of overlap between HD etiology, IGF/insulin biology, and Mn homeostasis. While these connections have been more fully elucidated in HD, the inherent overlap between NDD pathology suggests similar roles for Mn and IGF/insulin in other NDDs. However, as reviewed here, there is preliminary evidence that these NDDs often exhibit different trends in Mn and/or insulin/IGF homeostasis— for example PD is associated with increased Mn while HD is associated with Mn-deficiency. However, these observations lead to the following additional questions——1) are we exploring IGF/insulin and Mn dysregulation at the “right” times during disease progression 2) are we inspecting the levels of Mn or IGF/insulin in the correct tissues and 3) is this dysregulation truly a contributor to disease pathology or simply a downstream effect of a higher mechanism? If IGF/insulin and/or Mn are truly dysregulated in NDDs, one would imagine that there are defined stages of disease progression when specific defects can be observed. Mn or insulin/IGF could be affected early on in the disease prior to symptoms, during early symptom manifestation, or during late-stage progression once significant brain atrophy has occurred (or across the entirety of disease progression). Furthermore, it is likely that this dysregulation may differ in not only magnitude, but directionality, between each stage of the disease as molecular signaling attempts to compensate or desensitize. While serum and plasma levels offer a potential biomarker of brain Mn dysregulation, further studies must examine how these levels correlate to what is seen in actual brain tissue. Studies have found that changes in IGF by age, sex, diet, BMI, and secondary disease status can cause immense variability between patients, particularly in peripheral samples¹¹⁸. Several heavy metals are reported to accumulate in the brain with age and can differ by similar confounds suggesting peripheral Mn may also be an inappropriate measurement for brain Mn. Furthermore, regulation of IGF/insulin and Mn across the blood brain barrier has been somewhat elucidated, but strict regulation of these molecules is needed to establish brain integrity suggesting that they might be very different from what is seen in serum/plasma or even CSF. Confirming consistencies between serum, plasma, blood, CSF and the brain should be done in rodent models across disease progression to validate IGF/insulin and Mn biomarkers—substantiating their use in clinical studies. For other NDDs, a higher mechanistic understanding of IGF/insulin and Mn biology should be explored at the molecular and cellular levels, similar to what has been done in the HD field. Lastly, given the extended time it takes prior to NDD manifestation, one must ask whether observed defects in IGF/insulin or Mn are either a cause of the disease or instead a consequence of the neurodegeneration. This is a difficult question to answer given the inherent difficulty in working with aged models—namely mouse models which often do not fully recapitulate the pathology observed in humans.

Currently, available methods and technology make it quite difficult to truly investigate these questions in a high-through put manner. Highly sensitive biomarkers for Mn and IGF/insulin levels in the brain are likely required to observe changes across disease progression which are currently unavailable. The high variability and contradictory data of IGF/insulin levels in serum/plasma compared to brain suggest these are not always appropriate measurements for brain levels. While existing techniques can quantify levels of Mn in tissues or cells (ICP-MS, graphite furnace, cellular fura-2 Mn extraction assay (CFMEA)) as well as techniques that allow a cellular/sub-cellar resolution of Mn localization (XANESX-ray absorption near edge structure), high costs and complexities related to maintaining in vivo patterns has limited understanding of Mn brain homeostasis². Thus, creative approaches will be necessary to answer the outstanding questions.