Manganese acts upon Insulin/IGF receptors to phosphorylate AKT and increase glucose uptake in Huntington’s Disease cells

Abstract

Perturbations in insulin/IGF signaling and manganese (Mn²⁺) uptake and signaling have been separately reported in Huntington’s disease (HD) models. Insulin/IGF supplementation ameliorates HD phenotypes via upregulation of AKT, a known Mn²⁺-responsive kinase. Limited evidence both in vivo and in purified biochemical systems suggest Mn²⁺ enhances insulin/IGF receptor (IR/IGFR), an upstream tyrosine kinase of AKT. Conversely, Mn²⁺ deficiency impairs insulin release and associated glucose tolerance in vivo. Here, we test the hypothesis that Mn²⁺-dependent AKT signaling is predominantly mediated by direct Mn²⁺ activation of the insulin/IGF receptors, and HD-related impairments in insulin/IGF signaling are due to HD genotype-associated deficits in Mn²⁺ bioavailability. We examined the combined effects of IGF-1 and/or Mn²⁺ treatments on AKT signaling in multiple HD cellular models. Mn²⁺ treatment potentiates p-IGFR/IR-dependent AKT phosphorylation under physiological (1nM) or saturating (10nM) concentrations of IGF-1 directly at the level of intracellular activation of IGFR/IR. Using a multi-pharmacological approach, we find that >70–80% of Mn²⁺-associated AKT signaling across rodent and human neuronal cell models is specifically dependent on IR/IGFR, versus other signaling pathways upstream of AKT activation. Mn²⁺-induced p-IGFR and p-AKT were diminished in HD cell models, and, consistent with our hypothesis, were rescued by co-treatment of Mn²⁺ and IGF-1. Lastly, Mn²⁺-induced IGF signaling can modulate HD-relevant biological processes, as the reduced glucose uptake in HD STHdh cells was partially reversed by Mn²⁺ supplementation. Our data demonstrate that Mn²⁺ supplementation increases peak IGFR/IR-induced p-AKT likely via direct effects on IGFR/IR, consistent with its role as a cofactor, and suggests reduced Mn²⁺ bioavailability contributes to impaired IGF signaling and glucose uptake in HD models.

Introduction

The essentiality of manganese (Mn²⁺) is derived from its binding to and activation of several biologically indispensable enzymes, including Mn²⁺ superoxide dismutase, glutamine synthetase, pyruvate decarboxylase, protein phosphatase 2A (PP2A), and arginase (1). In addition, Mn²⁺ is a required cofactor for a variety of kinases, and can often compete with magnesium (Mg²⁺) when at sufficiently high concentrations to activate others, including ATM and mTOR (2, 3). As the vast majority of kinases are either Mn²⁺- or Mg²⁺-dependent, Mn²⁺ can act as a potent cell signaling modifier. Mn²⁺ can activate ERK, AKT, mTOR, ATM, and JNK in vitro and in vivo (2, 4–13). As these kinases regulate transcription factors (CREB, p53, NF-kB, FOXO), Mn²⁺ can also modulate cell function at the transcriptional level (7, 14–16). Consequently, the roles of Mn²⁺ homeostasis and associated signaling in both the essentiality and toxicity of Mn²⁺ are an important area of investigation. However, it remains uncertain which Mn²⁺-dependent enzymes are most sensitive to changes in Mn²⁺ homeostasis and the relationships between Mn²⁺-biology and these signaling cascades.

In contrast, at high concentrations, Mn²⁺ can be neurotoxic, and this has been associated with risk for idiopathic parkinsonism and the Mn²⁺-induced parkinsonian-like disease known as manganism (17–20). High environmental exposure to Mn²⁺ has been associated with specific occupational settings (welding, mining), exposure to industrial ferroalloy emissions, well water consumption in some regions, or parenteral nutrition (21–25). Of particular interest, Mn²⁺-induced p-AKT has been observed in a variety of models and in Mn²⁺-exposed patient populations (4, 10, 26–29). However, it is still unclear what the role of this response is or by which upstream signaling mechanism it occurs, though Mn²⁺-induced p-AKT is not blocked by the antioxidant Trolox (30). Thus, the elucidation of the primary signaling mechanism behind Mn²⁺-responsive AKT will be informative in the context of both basal Mn²⁺ homeostasis and Mn²⁺ neurotoxicity.

Insulin and IGF-1 are highly homologous growth factors which are necessary for a variety of peripheral processes, as well as essential for synaptic maintenance and activity, neurogenesis and neurite outgrowth, and neuronal mitochondrial function (31, 32). Insulin and IGF-1 bind to highly similar cell surface receptors which initiate an autophosphorylation cascade, independent of other kinases, which activates the insulin receptor (IR) and the IGF-1 receptor (IGFR). This causes subsequent activation of phosphatidylinositol-3-kinase (PI3K), insulin receptor substrates (IRSs), and other mediators activating the pro-growth AKT, mTOR, and ERK/MAPK pathways which have widespread roles in multiple biological processes. Dysregulation of these potent neurotrophic growth factors has been associated with neurodegenerative diseases, including HD, PD, and Alzheimer’s disease (AD) (20, 33–49). However, while the vast majority of kinases in the human body are Mg²⁺ and/or Mn²⁺-dependent, few studies have mechanistically elucidated how these metals maintain kinase signaling cascades in living biological systems or contribute to kinase-dependent pathology of neurodegenerative diseases.

Evidence supporting a role for insulin/IGF-1 synergistic cross talk with Mn²⁺ has been slowly amassing, but is incompletely understood. Mn²⁺ deficiency in rodent models reduces insulin production and causes glucose intolerance, while Mn²⁺ supplementation can protect against diet-induced diabetes rescue glucose intolerance, and increase insulin and IGF-1 ligand levels in rodents (50–59). Furthermore, Mn²⁺ administration stimulates insulin-linked glucose transport and related phosphodiesterase activity in adipocytes, though insulin/IGF receptor activity was not investigated (60). Two prior studies have examined how supra-physiological Mn²⁺ (1–10mM) activates insulin receptor activity using non-living, permeabilized rat adipocytes or purified biochemical systems and have shown that Mn²⁺ directly increases net autophosphorylation of IR/IGFR by both enhancing kinase activity and inhibiting receptor dephosphorylation (61, 62). Further, the peak/maximum in vitro kinase activity of IR/IGFR is higher in the presence of Mn²⁺, than with Mg²⁺. Finally, a previous study reported that JB1, an IGFR1 antagonist, can block Mn²⁺-induced IGFR-AKT-mTOR phosphorylation in the preoptic area of prepubertal rats (10). Together, these data provide a strong premise to examine this signaling pathway at a cellular level and suggest Mn²⁺ may be a critical mediator of insulin/IGF-1 homeostasis and downstream signaling, including AKT (27–29, 52, 56–58, 63). However, the understanding of Mn²⁺-IGF synergy lacks mechanistic insight of the direct site-of-action of Mn²⁺ on IGFR/IR-AKT signaling and requires confirmation of this mechanism in living systems under biologically-relevant, sub-cytotoxic concentrations of Mn²⁺ (1–500μM). Furthermore, elucidation of an initial site-of-action for Mn²⁺ may bridge the mechanistic gap between non-living and in vivo systems—directly connecting the role of Mn²⁺-induced receptor kinase activity to the changes seen in Mn²⁺-responsive metabolism in vivo, such as glucose tolerance.

HD is an autosomal dominant, neurodegenerative disease caused by an expanded CAG repeat within exon 1 of the Huntingtin (HTT) gene. Through a yet unknown pathogenesis, this expanded trinucleotide causes specific cell death in the medium spiny neurons of the striatum, leading to a variety of symptoms—most notably chorea. This disease has variable progression, but is ultimately fatal. There is no cure for HD, but drugs can target symptoms with variable efficacy. Recent research has shown that IGF-1 treatment in HD models results in robust amelioration of a wide-array of phenotypes via AKT signaling-mediated mechanisms. The results of this treatment include: an increase in autophagic function and mutant HTT (mutHTT) aggregate clearance; restoration of mitochondrial function; regularization of energy metabolites; HTT serine 421 phosphorylation; and medium spiny neuron health. Perhaps most importantly, IGF treatment can rescue motor abnormalities and early mortality in HD mouse models (32, 64–68). Furthermore, although HD is primarily a neurological disease, HD patients develop type 2 diabetes mellitus at a higher incidence than healthy controls and diabetic phenotypes can be rescued via IGF-1 treatment in HD models (32, 69–71). These data warrant further investigation into 1) the factors which contribute to dysregulated IGF-AKT signaling in HD and 2) the mechanism through which IGF-1 effectively treats some HD phenotypes.

Here, we assessed whether the synergistic effects of Mn²⁺+IGF occur via the ability of Mn²⁺ to directly act on IR/IGFR to elicit downstream AKT phosphorylation and determined if this mechanism is disrupted in a disease model system (HD) associated with a Mn²⁺ deficiency. While few human studies have examined Mn²⁺ levels in post-mortem HD brains, existing studies have found either no change in striatal Mn²⁺ or reductions only in the cortex (72–74). Prior studies in our lab have reported reduced Mn²⁺ uptake in in vitro HD cell models and the striatum of the YAC128 HD mouse model—indicative of a defect in Mn²⁺ accumulation (75–77). We have shown that Mn²⁺ treatment rescues deficient arginase activity in HD mice (78) and that in HD cell lines, this reduced Mn²⁺ uptake manifests as reduced Mn²⁺-induced activation of ATM-p53 and AKT cell signaling (14, 79). In contrast, changes in cholesterol metabolism appear unaffected by Mn²⁺ (80). Recently, we have established that Mn²⁺ induced p-AKT is dependent on PI3K, a downstream mediator of IR/IGFR, in STHdh cells (28). As AKT signaling mediates the restorative effects of IGF-1 administration in HD mouse models, we hypothesize here that Mn²⁺ promotes AKT signaling through an upstream IGFR-1-dependent mechanism, and that reduced Mn²⁺ uptake should manifest as impaired Mn²⁺→ IR/IGFR→ PI3K→ AKT signaling and contribute to well-established IGF-related deficits in HD. This study aimed to 1) define the synergistic co-regulation between Mn²⁺ and IGF on AKT signaling in various cellular models, 2) determine the initial, mechanistic target of Mn²⁺, which allows for downstream AKT activation, 3) elucidate the effects of Mn²⁺+IGF co-treatment on impaired IGFR/IR-AKT signaling in HD cells, and 4) investigate the effects of Mn²⁺ on glucose uptake, a downstream process of AKT signaling which is perturbed in HD patients and mouse models. These findings provide proof-of-principle evidence that Mn²⁺ supplementation could improve efficacy of IGF-centric therapies in HD.

RESULTS

Mn²⁺ and IGF exhibit synergistic regulation of p-AKT expression which is diminished in HD cells

Previous studies suggest that Mn²⁺ acts as an insulin mimetic, and also increases the kinase activity at the insulin receptor itself; however, these findings were primarily established using recombinant enzymes or permeabilized cellular models (61, 62). Furthermore, Mn²⁺ has been shown to activate insulin/IGF-responsive kinases and signaling pathways, including AKT. Given these observations, we hypothesized that Mn²⁺ potentiates IGF-induced p-AKT expression in neuronal cells. We examined p-AKT (Ser⁴⁷³) expression after 1hr serum deprivation in HBSS, followed by 3hr treatment with Mn²⁺/IGF in HBSS. For these experiments, we treated STHdh cells with 1nM IGF, which has been reported to be near physiological concentration (81). Normal human brain Mn²⁺ concentrations are estimated to be ~20–55μM (82). Mn²⁺ begins to induce in vitro cytotoxicity after 24hr, 100–200μM Mn²⁺ exposures, depending on cell type (14, 79). Across short exposures (3hrs), we did not observe any decrease in cell viability following 500μM Mn²⁺ in any cell type (Supplemental Figure 7A, B). Thus, for our experiments, we utilize sub-cytotoxic 50–500μM Mn²⁺ exposures based on cell type and exposure duration. Mn²⁺/IGF co-exposure induced a nearly 30-fold increase in p-AKT, ~15 times higher than 500μM Mn²⁺ alone, and ~3 times higher than IGF alone (Fig 1A). This confirms a synergistic regulation of AKT signaling by Mn²⁺ and IGF in living cells. Furthermore, treatment with Mn²⁺ alone resulted in an insignificant increase in p-AKT, suggesting that Mn²⁺-induced p-AKT is dependent on the presence of an upstream ligand, such as IGF or insulin. To examine this further, we treated WT Q7/Q7 STHdh cells with 50 or 500μM Mn²⁺ in normal serum-containing media or in serum-free HBSS following 1hr serum deprivation. Consistent with previous results, 3hr treatment with Mn²⁺ (50 or 500μM) did not cause a significant increase in p-AKT in serum-free HBSS (though 500μM trended towards increase p-AKT in HBSS), but did significantly increase p-AKT in serum-containing media at both concentrations. This suggested that a potential interaction with a serum component, such as insulin or IGF, is essential for Mn²⁺-induced p-AKT (Fig 1B).

Figure 1: Mn²⁺ can potentiate IGF-1 induced p-AKT and Mn²⁺-induced p-AKT is reduced in HD cells.

A) p-AKT expression in STHdh Q7/Q7 following a 1hr serum deprivation, then 3hr exposure in HBSS with 1nM IGF-1 and/or 500μM Mn²⁺. Vehicle=dotted line. One-way ANOVA; treatment= F(2,6)=20.12; p=0.0022. B) Quantification of p-AKT after 3hr, 50/500μM Mn²⁺ exposure in serum free HBSS (following 1hr serum deprivation) or media containing 10% FBS. Two-way ANOVA; treatment= F(2,6)=40.84; p=0.0003; media/HBSS= F(1,3)= 671.6; p=0.0001; treatment-media interaction= F(2,6)= 24,37; p=0.0013. C) p-AKT expression in STHdh WT and HD cells following 1hr serum deprivation then 3hr Mn²⁺ (0/200/500μM) + IGF (1nM) exposure in HBSS. Two-way ANOVA; treatment= F(2, 4)= 10.29; p=0.0265. D) p-AKT expression after 24hrs treatment with 10nM IGF-1, 50μM Mn²⁺, or both in STHdh Q7/Q7 and Q111/Q111. Two-way ANOVA; treatment= F(2,10)=40.84; p=0.0064; genotype= F(1,5)= 671.6; p=0.0163; treatment-genotype interaction= F(2,10)= 1.587; p=0.2515. E,F) p-AKT (Ser⁴⁷³) and p-S6 (Ser ^235/236) in uninduced PC12 cells following treatment with 100μM Mn²⁺, 10nM IGF-1, or both. For these PC12 experiments, all uninduced (i.e. were only expressing WT rat HTT) samples from the 23Q, 74Q, and 140Q (3 biological replicates each) were used. Representative blot Supp Fig 1C. One-way ANOVA; p-AKT treatment= F(2,10)=36.71; p=<0.0001; One-way ANOVA; p-S6 treatment= F(2,10)=20.57; p=<0.0003; Error bars= SEM. Dotted line= vehicle (=1). N=3 for Panel A-C; N=4 for panel D; N=5 for panel E; N=6 for panels F and G. *= significant by Tukey’s (A, F, G), Dunnet (B), and Sidak multiple comparison (D-E). *P<.05, **P<.01, ***P<.001.

The STHdh Q111/Q111 HD cell model exhibits both a basal Mn²⁺ uptake deficit as well as a reduced net Mn²⁺ accumulation after an exogenous exposure, making it an ideal model to study Mn²⁺-induced IGF signaling and the consequences of perturbations to this system (14, 79). Thus, we hypothesized that this cell model would also exhibit reduced Mn²⁺/IGF-induced p-AKT expression. This would be consistent with other studies demonstrating defects in AKT signaling in HD (65, 66, 83–85). Indeed, the Q111 HD cell model exhibited reduced Mn²⁺/IGF-induced p-AKT expression following a 3hr Mn²⁺ exposure in serum-free media (Fig 1C). However, treatment with IGF+500μM Mn²⁺ in the Q111 HD cells restored p-AKT expression to levels seen with IGF+200μM Mn²⁺ in the Q7 WT model. Total AKT levels were unchanged by Mn²⁺ after 3hrs in media or HBSS (Supp Fig 1A). This confirms reduced Mn²⁺-induced p-AKT in this HD cell model, and demonstrates that the Mn²⁺-uptake defect can be attenuated via higher doses of Mn²⁺ treatment, compensating for the uptake deficit.

We reasoned that if Mn²⁺ acts as an insulin/IGF “mimetic” by increasing ligand concentration or ligand-receptor occupancy, Mn²⁺ should be unable to further activate p-AKT in the presence of saturating concentrations of insulin/IGF. We determined that the saturating concentration of IGF-1 in serum-containing media for p-AKT after 24hrs is approximately 10nM (Supp Fig 1B). Co-treatment with 10 nM IGF-1 and 50 μM Mn²⁺ for 24hrs in normal (serum-containing) media resulted in supra-additive p-AKT responses (>2fold compared to IGF or Mn²⁺ alone) in STHdh and p-AKT and p-S6 in uninduced, differentiated PC12 cells mirroring the effects seen in the 3hr exposures above (Fig 1D–F, Supp Fig 1C). P-p53, another Mn²⁺-responsive pathway in these cells, was indistinguishable between Mn²⁺ and Mn²⁺+IGF exposed conditions, demonstrating that this is not a broad effect across all Mn²⁺-responsive pathways (data not shown). As expected, Mn²⁺-induced p-AKT was blunted in HD cells following 24hr exposure. Furthermore, Mn²⁺/IGF co-treatment significantly increased p-AKT activation compared to the effects of Mn²⁺ or IGF alone in the Q111 HD cells (Fig 1D), similar to Fig 1C. Total AKT levels (Pan-AKT) were not significantly different in any of the conditions and p-AKT (Thr³⁰⁸) showed a highly similar trend to Ser⁴⁷³; thus, going forward, we only quantified p-AKT (Ser⁴⁷³) (Supp Fig. 1D). Together, these data suggest that Mn²⁺ synergistically increases the maximal activity of the AKT pathway, even under saturating concentrations of ligand, but expression of mutHTT dampens this effect.

Phosphorylation of AKT is specific to Mn²⁺ and not shared by other cation exposures at sub-cytotoxic concentrations

We hypothesized that Mn²⁺-induced p-AKT is a consequence of a Mn²⁺-responsive kinase upstream of AKT rather than a broad effect of heavy metal exposure, such as reactive oxygen species (ROS) accumulation. In other words, if Mn²⁺ is acting as a cofactor for an upstream kinase, then its activity on downstream proteins should be unique to Mn²⁺ vs other metal cations not capable of serving as kinase cofactors. Thus, we determined 1) whether other metal cations are capable of increasing p-AKT similarly to Mn²⁺ and 2) if HD genotype cells exhibit reduced p-AKT in response to other metal cations. We tested a battery of cations (Fe³⁺, Cu²⁺, Mg²⁺, Zn²⁺, Cd²⁺, Ni²⁺, and Co²⁺) and examined p-AKT expression after 24hrs. For these experiments, we used concentrations which are near the toxic threshold in these cells after 24hr exposures as shown in our previous work (79) and found that Mn²⁺ was the only cation which elicited a significant p-AKT response, and thus, the only metal which exhibited an HD phenotype. Cu trended towards an increase in p-AKT but this was not significant (Fig 2A–C). Although other cations were unable to significantly increase p-AKT under these sub-cytotoxic concentrations, we hypothesize that higher concentrations of some cations may also increase p-AKT via a cytotoxic response. As Mn²⁺ and Mg²⁺ often act as cofactors with the same enzymes, we supplemented the media for these cells with an additional 50μM Mg²⁺ or Mn²⁺ (though DMEM contains high physiological concentrations of Mg²⁺) increasing the available combined pool of Mg²⁺/Mn²⁺, and did not observe Mg²⁺-induced p-AKT (Fig 2C). Consistent with our results in living cells, these studies observed that the combination of Mn²⁺+Mg²⁺ evoked higher insulin receptor activity (61, 62). This supports the hypothesis where the role of Mn²⁺ to increase p-AKT expression is metal ion-specific, and furthermore, that Mn²⁺ is able to do this under saturating concentrations of Mg²⁺ (a known, competing cofactor for many kinases).

Figure 2: Assessing the specificity and dynamics of the Mn²⁺-AKT interaction.

A) Representative blot for p-AKT expression in STHdh Q7/Q7 and Q111/Q111 cells following 24hr exposures with Mn²⁺, Fe³⁺, Cu²⁺, Mg²⁺, Zn²⁺, Cd²⁺, Ni²⁺, or Co²⁺. B) Quantification of p-AKT expression, blot shown in panel A. C) Quantification of only Mn²⁺ and Mg²⁺-induced p-AKT after 24hrs, blot shown in panel A. Two-way ANOVA; treatment= F(8, 24)=2.509; p=<0.0388. N=4; Error bars= SEM; Normalized to respective vehicle. D) Representative blot for p-AKT and p-S6 expression in STHdh Q7/Q7 and Q111/Q111 cells following 7-day exposure with Mn²⁺ (1/5/10μM) or IGF (10nM). E) Quantification of p-AKT expression, blot shown in panel D. N=4. Error=SEM. Two-way ANOVA; treatment= F(3,9)=.5284; p=0.6738; genotype= F(1,3)= 33.39; p=0.0103; treatment-genotype interaction= F(3,9)= 4.912; p=0.0273. F) Representative blot for p-AKT and p-p53 expression following 24hr exposure with 50μM Mn²⁺, 10μM KB-R7943, or both. G) Quantification of p-AKT expression, blot shown in panel H (p-p53 quantification not shown). Two-way ANOVA; treatment= F(3,12)=24.91; p=<0.0001; genotype= F(1,4)= 2.840 p=0.1672; treatment-genotype interaction= F(3,12)= 4.761; p=0.0207. N=4; Error bars= SEM. *= significant by Dunnet (B, C, E), and Sidak multiple comparison tests (G,H). *P<.05, **P<.01, ***P<.001.

Reduced Mn²⁺-induced p-AKT in HD cells persists under prolonged physiologically-relevant Mn²⁺ exposures

While we observed reduced Mn²⁺-induced p-AKT in HD cells after high dose, acute Mn²⁺ exposure, we wanted to assess whether this phenotype would persist under lower dose, subacute, week-long exposures. Thus, we treated STHdh cells for 1 week with 1, 5, or 10 μM Mn²⁺. In Q7/Q7 cells, Mn²⁺-induced p-AKT was observed after a 7-day exposure, with 5 and 10 μM eliciting the highest effect (Fig 2D, E). Phosphorylation of AKT was almost completely unresponsive in Q111/Q111 HD cells to low-dose Mn²⁺ exposure, confirming that the HD genotype perturbation persisted under these subacute treatments. Additionally, this suggests Mn²⁺-induced p-AKT occurs with changes in Mn²⁺ homeostasis well below the toxic threshold.

Normalization of net Mn²⁺ uptake ameliorates Mn²⁺-induced p-AKT defect in HD cells

The Q111/Q111 HD STHdh cells exhibit reduced Mn²⁺-induced p-AKT across several treatment paradigms (Fig 1C–D, 2D–E). If reduced Mn²⁺-induced p-AKT in HD cells is dependent on intracellular Mn²⁺ levels, normalization of intracellular Mn²⁺ uptake in the HD cells to match WT levels would be predicted to ameliorate Mn²⁺-induced p-AKT differences between the genotypes. KB-R7943 is a drug which inhibits the sodium-calcium uniporter (NCX1/3) and normalizes Mn²⁺ uptake in these HD cells by an unknown mechanism. This drug was previously used to normalize Mn²⁺ uptake and Mn²⁺-induced p-p53, concurrently(86). We confirmed this result and observed significant increases in Mn uptake in HD cells when treated with Mn and KB-R7943 (Supplemental Fig 2A). When HD cells were exposed to KB-R7943 and 50 μM Mn²⁺, Mn²⁺-induced p-AKT was restored to levels observed in WT cells treated with Mn²⁺ alone (Fig 2F, G). This suggests reduced Mn²⁺-induced cell signaling levels are driven by the decreased intracellular Mn²⁺ in this model. Although the mechanism by which KB-R7943 restores Mn²⁺ uptake is unknown, these data also demonstrate that the drug can increase the bioavailable pool of Mn²⁺, as opposed to merely sequestering Mn²⁺ in metabolically inaccessible regions of the cell.

Expression of mutHTT is sufficient to reduce Mn²⁺-induced p-AKT in differentiated PC12 cells

Next, we sought to confirm if other HD model cells lines exhibit reduced Mn²⁺-induced p-AKT. First, we utilized differentiated PC12 cells which express WT HTT but are capable of additional ponasterone A-induced HTT expression (23, 74, 145 CAG or CAA). These cells were differentiated into a neuronal phenotype by treatment with nerve growth factor (NGF) over the course of a week and expressed tyrosine hydroxylase, indicating a catecholaminergic population (data not shown). After 7 days of differentiation with NGF and mutHTT induction with ponasterone A, 145 CAG-expressing PC12 cells exhibited reduced Mn²⁺-induced p-AKT compared to uninduced counterparts (Fig 3A,B). Induction of 74CAG mutHTT resulted in a modest, but insignificant, reduction in Mn²⁺-induced p-AKT, and, as expected, induction of 23 CAG HTT had no effect. (data not shown). Pan AKT and pan S6 were unaffected in all conditions, similar to STHdh cells (Fig 3A, Supp Fig 1C). Mn²⁺ induced p-S6 was unaffected by mutHTT expression (data not shown). Together, these data demonstrate impairments in Mn²⁺ homeostasis and signaling are present in a variety of HD cell lines and occur within days of mutHTT expression.

Figure 3: Assessing the effects of WT and mutant HTT in Mn²⁺-induced pAKT.

A) Representative western blot of PC12 cells in differentiated and 145Q HTT-induced PC12 cells following 24hr exposure with 100μM Mn²⁺ and/or IGF (10nM) B) Quantification of p-AKT expression. 145Q induced HTT cells (red) are compared to uninduced counterparts (blue). Image of WT HTT (bottom arrow) and 145Q HTT (top arrow) expression using mAb 2166 inset within graph- blue= non-induced, red= induced. N=3; Error bars= SEM, *= significance genotype difference by student’s t-test. *P<.05, **P<.01, ***P<.001.

Mn²⁺ potentiates IGF-induced p-IR/IGFR expression which is blunted in HD cells

We established a specific, synergistic effect on p-AKT signaling by Mn²⁺/IGF co-treatment which was diminished in several HD cell lines. We sought to elucidate the mechanistic target of Mn²⁺ which allows for the synergistic effect on p-AKT. This target is likely responsible for driving the reduced Mn²⁺-induced p-AKT in HD models. As Mn²⁺ is a known cofactor for a variety of kinases, including IR/IGFR(61, 62), we hypothesized that Mn²⁺ may be directly interacting with IR/IGFR to induce p-AKT. To our knowledge, phospho-specific antibodies do not exist which distinguish between p-IR or p-IGFR, as they are phosphorylated at highly similar residues (87, 88). Thus, to examine this, we co-treated Q7 WT and Q111 HD STHdh cells with Mn²⁺/IGF for 3hrs following a 1hr serum deprivation and assessed p-IR/IGFR levels. We found that Mn²⁺ potentiated IGF-induced IR/IGFR phosphorylation. Further, Mn²⁺-induced p-IR/IGFR levels were blunted in HD cells but co-treatment with IGF+500μM in HD cells elicited similar p-IR/IGFR expression as IGF+200μM Mn²⁺ in WT cells (Fig 4A,B). These observations mirror the effects of Mn²⁺ on p-AKT expression in Fig 1C. Altogether, these data support a role for Mn²⁺ directly increasing IR/IGFR activity, thereby activating downstream signaling including AKT.

Figure 4: Mn²⁺ increases phosphorylation of IR/IGFR and decreases total protein and mRNA expression.

A,B) Representative western blot and quantification of p-IGFR expression in STHdh cells following 1hr serum deprivation and 3hr Mn²⁺ (200/500μM) + IGF(1nM) exposures. Note representative blot is the same samples/run as Figure 1C. Two-way ANOVA; treatment= F(2,6)=5.461; p=<0.0446; genotype= F(1,3)= 30.98; p=0.0114; treatment-genotype interaction= F(2,6)= 3.275; p=0.1093. N=4. C,D) Representative western blot and quantification of pan IGFR protein expression in STHdh cells following 24hr exposure with Mn²⁺ (50μM) and/or IGF (10nM). Dotted line= Vehicle (=1). Two-way ANOVA for IGFR; treatment= F(3,9)=16.71; p=<0.0005. N=4. E,F) Representative western blot and quantification of pan IR protein expression in STHdh cells following 24hr exposure with Mn²⁺ (50μM) and/or IGF (10nM). Dotted line= Vehicle (=1). Two-way ANOVA for IR; treatment= F(3,6)=20.59; p=0.0015. N=3. G-I) mRNA expression of IGFR (G), IR (H), and IRS2 (I) after 24hr Mn²⁺ (50μM) and/or IGF (10nM) exposure. Two-way ANOVA treatment for IGFR: F(3,6)=85.01; p=<0.0001. For IR: F(3,6)=7.204; p=<0.0205. For IRS2: F(1,3)=213.8; p=<0.0007.Error bars= SEM; N=4 for panels A-D, G-I; N=3 for E,F. P<.05, **P<.01, ***P<.001. *= significant by Tukey’s (B) or Dunnet’s (D, F, G, H, I) multiple comparison tests.

Mn²⁺ reduces total IGFR protein expression and IGFR mRNA expression

Our data suggest that Mn²⁺ can increase maximal IGF-induced p-AKT activity. Thus, we wanted to assess whether this leads to negative feedback on total IGFR expression. We hypothesized that Mn²⁺ (in conjunction with IGF) exposure reduces total IGFR expression after 24hrs to reduce overall activity of the AKT pathway. Indeed, 50μM Mn²⁺ treatment, similarly to IGF treatment, decreased total IGFR expression by ~50%—perhaps acting through a negative feedback mechanism to prevent overactivation of the pathway (Fig 4C,D). Unlike IGFR, Mn²⁺ did not reduce total IR expression in WT or HD cells, but IGF treatment alone decreased total IR expression specifically in HD cells (Fig 4E,F). Mn²⁺ exposure for only 3hrs in media or HBSS left total IGFR unchanged (Supp Fig 1A).

To determine whether the Mn²⁺-induced reduction in total IR/IGFR expression is mediated, at least in part, by a transcriptional mechanism, we assessed the effects of Mn²⁺ on IR/IGFR mRNA expression. We found that 24hr Mn²⁺ (or Mn²⁺+IGF) treatment reduced IGFR mRNA expression and Mn²⁺+IGF co-treatment modestly reduced expression of IR mRNA expression in both genotypes, suggesting that Mn²⁺ may have a more specific effect on IGFR signaling than that of IR (Fig 4G,H). Mn²⁺ specifically reduced IR mRNA expression in HD cells only. The HD genotype did not affect basal IGFR or IR mRNA expression. Additionally, we assessed mRNA expression of IRS2, a downstream partner in insulin/IGF signaling. IRS2 mRNA was greatly reduced in HD cells and Mn²⁺ exposure reduced IRS2 mRNA in Q7 cells only (Fig 4I). Together these data demonstrate that Mn²⁺ also modulates total IR/IGFR protein expression via transcriptional downregulation.

Mn²⁺-induced p-AKT is completely abrogated by pharmacological IR/IGFR inhibition

After observing that Mn²⁺ and IGF exert synergistic effects on p-AKT and p-IR/IGFR, we hypothesized that IGF and Mn²⁺ may be cooperatively activating on the same upstream kinase to increase AKT phosphorylation. Previously, we found that Mn²⁺-induced p-AKT is dependent on PI3K in these cells, as LY294002, a PI3K inhibitor, can completely abrogate Mn²⁺-induced p-AKT (28). This narrows down the possible targets to upstream activators of PI3K including IR/IGFR. Given these data and the prior evidence of a role for IR/IGFR signaling, we hypothesized some or all of Mn²⁺-induced p-AKT is IR/IGFR-dependent. IR/IGFR are essential to neuronal differentiation, development, and homeostasis which complicates the generation and use of knockout cell lines, and less-than-complete efficiency of siRNAs would allow for remaining receptors to compensate (31, 32). Thus, we turned to pharmacological modulation of IR/IGFR. While small molecules allow dosing to achieve near-complete inhibition, they are prone to off-target effects. Thus, to increase the rigor of our findings we employed a battery of four ATP-competitive IR/IGFR inhibitors (BMS-536924, BMS-754807, Linsitinib (OSI-906), and NVP-AEW541) so that multiple inhibitors could be used across each cell line. Additionally, these inhibitors have different known off-target proteins (Fig 5A), so any shared effects by the inhibitors are likely via IR/IGFR only (89–92).

Figure 5: IR/IGFR inhibitors block Mn²⁺-induced p-AKT.

A) List of IR/IGFR inhibitor names, IC50s for IR/IGFR, off-targets, and concentrations used here. B) Representative western blot of pAKT expression STHdh Q7/Q7 and Q111/Q111 after 24hr treatment with Mn²⁺ (50μM), IGF (10nM), and/or BMS-536924 (100nM).C) Quantification of Mn²⁺/IGF induced p-AKT expression with BMS536924 (100nM). Two-way ANOVA for BMS5 treatment; F(1.610, 4.831)=20.03; p=.0039. N=4. D) Representative western blot of p-AKT expression in STHdh Q7/Q7 and Q111/Q111 after 24hr treatment with Mn²⁺ (50μM), Linsitinib (100nm/1μM), NVP-AEW541-AEW541 (100nM/1μM) and Mn²⁺+IGF(10nM). E, F) Quantification of Mn²⁺/IGF-induced p-AKT expression with Linsitinib (E) or NVP-AEW541 (F) Two-way ANOVA for LLinsitinib treatment; F(4,8)=8.414; p=.0182; Two-way ANOVA for NVP treatment; F(4,8)=10.68; p=.0027. N=3–4. G, H) Quantification of p-AKT (G) and p-S6 (H) expression in uninduced PC12 cells treated with 100μM Mn²⁺ or Mn²⁺ (100μM) + BMS536924 (100nM) after 24hr exposures. Representative blot Supp Fig 1C. *= significance by student’s t-test. For these PC12 experiments, uninduced samples from the 23Q, 74Q, and 140Q (total N=5) were used. Note representative blot Supplemental Fig 1C. Error bars= SEM. Dotted line= Vehicle (=1). *P<.05, **P<.01, ***P<.001.

Using these various inhibitors across STHdh and PC12 cells, we found 1) Mn²⁺-induced p-AKT was inhibited in STHdh cells with 100nM BMS-536924 (Fig 5B, C), 0.1–1μM Linsitinib (Fig 5D,E), or 0.1–1μM NVP-AEW541 (Fig 5D,F) and 2) Mn²⁺-induced p-AKT and p-S6 was blocked with 100nM BMS-536924 in PC12 cells (Fig 5G–H, Supp Fig 1C). In PC12 cells, >100% of Mn²⁺-induced p-AKT (below vehicle levels) and 59% of Mn²⁺-induced p-S6 was inhibited using 100nM BMS-536924 (near the estimated IC50 concentration) (Fig 5G–H, Supp Fig 6B). BMS-536924 inhibited 46% of Mn²⁺-induced p-AKT at this approximate IC50 in SThdh cells (Fig 5B, C, Supp Fig 6A). 100nM linsitinib (approximately the IC50 for IR, 3X IC50 for IGFR) inhibited 71% of Mn²⁺-induced p-AKT (Fig 5D, E, Supp Fig 6A). 100nM NVP-AEW541 (approximately 65–70% the IC50 for IR and IGFR) inhibited 54% of Mn²⁺-induced p-AKT (Fig 5D, F, Supp Fig 6A). At 1μM (10-fold higher), Linsitinib and NVP-AEW541 inhibited 81% and 65% of Mn²⁺-induced p-AKT. Additionally, Linsitinib, the most specific of the inhibitors, was able to inhibit 60% of Mn²⁺+IGF induced p-AKT in STHdh cells at 100nM (Fig 5D, E, Supp Fig 6A). We confirmed that these inhibitors blocked IGF-induced p-IR/IGFR in the STHdh cells (data not shown). The high degree of dose-dependent inhibition by multiple IGFR inhibitors suggests that the vast majority of Mn²⁺-induced IGFR/AKT signaling is IR/IGFR-dependent. IR/IGFR exhibit autophosphorylation upon binding to insulin/IGF (and necessary co-factors) and do not require the kinase activity of other upstream kinases for activation. With this in mind, this data in addition to our observed Mn²⁺-induced p-IGFR/IR (Fig 4A) also suggest Mn²⁺ is activating directly at the level of the IR/IGFR (or affecting extracellular insulin/IGF production or binding) as opposed to another kinase upstream of these tyrosine kinase receptors

Non-neuronal cell types also exhibit IR/IGFR-dependent, Mn²⁺-induced p-IGFR/p-AKT

We hypothesized Mn²⁺-induced p-AKT was not specific to NPCs and would be present in other cell types. To investigate this, we exposed isolated primary rat astrocytes to 500μM Mn²⁺, 1nM IGF, or both for 3hrs following a 1hr serum deprivation. As in neuroprogenitors, addition of Mn²⁺+IGF increased p-AKT significantly more than either Mn²⁺ or IGF alone (Fig 6A, B). Additionally, we tested whether Mn²⁺-potentiated p-IGFR/p-AKT occurred in other non-neuronal, peripheral cell types—immortalized 3T3 fibroblasts and HEK293 kidney cells. Similar to what we observed in STHdh cells, a 3hr 500μM Mn²⁺ exposure did not activate p-IGFR, p-AKT, or p-S6 in the absence of IGF-1 in these cells (Fig 6C–E, Supp Fig 2B). This confirms that Mn²⁺-induced IGFR/AKT signaling requires the presence of ligand (IGF-1 or insulin) across a variety of cell types. In NIH3T3 and HEK293 fibroblasts, Mn²⁺+IGF increased p-AKT and p-IGFR expression more than IGF or Mn²⁺ alone after 1hr serum deprivation (Fig 6C, E, Supp Fig 2B). This corroborates the finding that the synergistic effects between Mn²⁺ and IGF are not restricted to neuronal cells. Additionally, Linsitinib (1μM) completely inhibited p-IGFR/IR (>98%) in 3T3 cells and the vast majority of Mn²⁺+IGF-induced p-AKT (≥80%) could also be blocked by Linsitinib or LY294002 (PI3K inhibitor). In HEK293 cells, the vast majority of p-IGFR (≥98%) was inhibited with Linsitinib, but only 40% of Mn²⁺-induced p-AKT was blocked, suggesting Mn²⁺-induced p-AKT is less dependent on IGFR/IR in HEK293 cells than in 3T3 cells. (Fig 6F,H, Supp Fig 6B). Furthermore, by comparing the degree of p-IGFR/IR inhibition to p-AKT inhibition by Linsitinib, we can estimate the percentage of Mn²⁺-induced p-AKT which is dependent on p-IGFR/IR [(% of p-AKT inhibition) / (% of p-IGFR/IR inhibition)]. Using this calculation, we estimate that 99% of Mn²⁺-induced p-AKT in 3T3 cells and 41% in HEK293 cells is dependent on p-IGFR (Supp Fig 6B). This confirms that Mn²⁺-induced p-AKT is IR/IGFR-dependent in these non-neuronal cell types as well, particularly 3T3 cells. P-S6 also trended with p-AKT across conditions in 3T3 cells, but due to the variability in Mn²⁺+IGF treatment, this trend was not significant (Fig 6D,G). Additionally, we found that there was only a 62% correlation between p-S6:p-IGFR/IR inhibition in 3T3 cells and 19% correlation in HEK293 cells, suggesting Mn²⁺-induced p-S6 is much less dependent on IGFR/IR than p-AKT (Supp Fig 6B). These data provide strong evidence that Mn²⁺-induced p-IGFR and p-AKT are not restricted to specific neuronal lineages.

Figure 6: Mn²⁺ increases p-AKT and p-IGFR expression in the presence of IGF-1 in astrocytes and non-neuronal cell lines.

A, B) Representative western blot and quantification of p-AKT in astrocytes after 1hr serum deprivation followed by 3hr exposure with Mn²⁺ (500μM), IGF (1nM), or both. One-way ANOVA; treatment= F(1.167, 3.501)=24.37; p=<0.0104. N=4. Error= SEM. C-E) Quantifications of western blots of 3T3 and HEK293 p-AKT (C) p-S6 (D), p-IGFR (E) expression after 1hr serum deprivation followed by 3hr exposure in HBSS with Mn²⁺ (100μM), IGF (1nM), or both. F-H) Quantifications of western blots of 3T3 and HEK293 p-AKT (F), p-S6 (G), p-IGFR (H) expression after 1hr serum deprivation followed by 3hr exposure in HBSS with Mn²⁺ (100μM) + IGF (1nM) with LY294002 (7μM), and/or LLinsitinib (1μM). Representative blot in Supp. Fig 3. N=3; Error bars= SEM; Dotted line= IGF alone (=1). One-way ANOVA for p-IGFR; HEK: F(1.284, 2.568)= 38.12; p=.0129; 3T3: F(1.221, 2.442)=99.63; p=.0048. One-way ANOVA for p-AKT; HEK: F(4,8)= 9.391; p=.0041; 3T3: F(4,8)=8.544; p=.0055. One-way ANOVA for p-S6; HEK: F(1.274, 2.548)= 5.081; p=.1270; 3T3: F(1.014, 2.028)=3.123; p=.2180 *= significance by Dunnet’s multiple comparison test.*P<.05, **P<.01, ***P<.001.

Mn²⁺-induced IGF signaling is dampened in human induced pluripotent stem cells (hiPSC)-derived striatal neuroprogenitors and is blocked by IGFR inhibition

We wanted test whether the HD-dependent IGF signaling defect observed in cell line models of HD would also be seen in a non-transformed human neuronal model. We chose to utilize the hiPSC-derived Islet-1-expressing striatal neuroprogenitor cells (NPCs) which are derived via a protocol designed in our lab (14). After a 24hr exposure, Mn²⁺-induced p-AKT in Islet-1+ hiPSC-derived striatal-like NPCs from HD patients was not significantly different than control cells; also, p-IGFR and p-S6 (often used as a readout for mTOR activity) were both induced by Mn²⁺, but neither was differentially affected in HD cells (data not shown). However, in striatal NPCs, we observed that under these conditions, IGF-1 was unable to activate p-AKT, likely because their normal NPC media conditions contain saturating insulin (Supp Fig 3). Thus, this standard striatal NPC media conditions are not a biologically-relevant setting to study IGF signaling or disease-relevant IGF/AKT-centric phenotypes. However, the data do suggest that Mn²⁺ can upregulate maximal AKT phosphorylation even under insulin-saturating concentrations.

To circumvent this issue of saturating insulin in the media, we found that growth factor/insulin deprivation prior to exposure allowed detection of p-IGFR in these NPCs and decreased variability in downstream p-AKT/p-S6 signaling. Thus, we used a 3hr growth factor deprivation followed by 3hr exposure with Mn²⁺ and/or IGF, and then assessed p-AKT, p-S6, and p-IGFR expression. Cells were treated under insulin/IGF-deprived conditions (though there are likely trace amounts of insulin/IGF left behind or produced by the cells themselves) and in the presence of 1 nM IGF. IGF, as expected, stimulated p-IGFR and p-AKT expression, though there were no apparent differences in magnitude between control and HD patient cells. Surprisingly, IGF did not significantly increase p-S6 expression in control or HD cells (Supp Fig 5A, B).

A 200–500μM Mn²⁺ treatment for 3hrs increased p-AKT and 500μM treatment increased p-IGFR and p-S6 in control cells (Fig 7A–C, Supp Fig 4). p-AKT and p-IGFR responses to Mn²⁺ were greater in conditions with IGF, suggesting a synergistic relationship similar to the one observed in STHdh cells (Fig 7A,B, Supp Fig 4). Mn²⁺ was able to significantly increase p-IGFR, p-AKT, and p-S6 in the absence of added IGF (Fig 7A–C, Supp Fig 4). This may be a result of trace IGF left behind in the media or generation of IGF by the cells themselves. In HD cells, Mn²⁺ induced p-AKT was only significant after 500μM Mn²⁺ treatment, and neither p-IGFR or p-S6 were significantly impacted (Fig 7A, B, Supp Fig 4). Mn²⁺-induced p-S6 is completely diminished without the addition of IGF-1 to HD cells (Fig 7C, Supp Fig 4). We found that total levels of IGFR, AKT, and S6 did not change with Mn²⁺ or IGF exposures, and thus we only quantified p-IGFR, p-AKT, and p-S6 (Supp Fig 4C–F). These results confirm that Mn²⁺-induced IGF signaling is diminished in HD patient-derived striatal-like NPCs.

Figure 7: Mn²⁺-induced p-AKT, IGFR is reduced in HD hiPSC-derived neuroprogenitors.

A-C) Western blot quantification of hiPSC-derived neuroprogenitors from three control patients and three HD patients (CAG repeat 58, 66, 70) for p-AKT (A), p-IGFR (B), and p-S6 (C), following treatment 3hr treatment with Mn²⁺ (200/500μM) or Mn²⁺+IGF (1nM) in growth factor/insulin free media, after 3hr serum deprivation. N=8–12 for control from three separate patients, N=7–9 for HD including three separate patients; Error bars= SEM. All data normalized to respective control or HD treated IGF-1 alone. Representative blot in Supplemental Fig 5A,B. One-way ANOVA stats listed in Supp Fig 4D. *= significance from vehicle by Dunnet’s multiple comparison test. *P<.05, **P<.01, ***P<.001.

We then assessed whether BMS-536924 and/or BMS-754807 could block Mn²⁺-induced IGF signaling in these cells, to validate that IR/IGFR signaling was required for the effects on p-IGFR, p-AKT and/or p-S6 effects. Similar to other cells tested, with some trending exceptions, both inhibitors blocked a portion of Mn²⁺-induced p-IGFR, p-AKT, and p-S6 in both control (Fig 8A,B) and HD (Fig 8C,D) cells at ~IC50 concentrations (100nM for BMS-536924 and 2nM for BMS-754807). In HD cells, we detected significant inhibition on Mn-induced p-IGFR and p-S6 as drug application often decreased expression below vehicle levels. 100nM BMS-536924 inhibited 48% of Mn²⁺-induced p-AKT, but at higher concentrations (1μM) could inhibit Mn²⁺-induced p-AKT even further, achieving >71% inhibition (Fig 8A). BMS-754807 inhibited 57% of Mn²⁺-induced p-AKT at its ~IC50 concentration (100nM) (Fig 8B). We estimate ≥68% of Mn²⁺-induced p-AKT is dependent on p-IGFR/IR across both inhibitors (74–80% with BMS-536924, 68% with BMS-754807) in Mn²⁺+IGF treated control cells. These inhibitors were less effective at inhibiting Mn²⁺-induced p-S6, similar to 3T3 and HEK293 cells (Supp Fig 6C). These two inhibitors also inhibited basal p-IGFR, p-AKT, and p-S6 (Supp Fig 5C,D). Together, these results suggest that, similar to STHdh andPC12 cells, in human derived NPCs: 1) Mn²⁺-induced IGF signaling is dampened in HD cells, and 2) Mn²⁺-induced p-AKT is dependent on IR/IGFR.

Figure 8: Mn²⁺-induced p-AKT, IGFR is blocked by IR/IGFR inhibition in HD hiPSC-derived neuroprogenitors.

Western blot quantification of control and HD patient iPSC-derived neuroprogenitors following treatment 3hr treatment with Mn²⁺ (500μM) or Mn²⁺+IGF (1nM) in growth factor/insulin free media, after 3hr N2 supplement (insulin/growth factor) withdrawal. A-D) Quantification of p-IGFR, p-AKT, and p-S6 in control (Blue, A,B) or HD (Red, C,D) after treatment with BMS-536924 (100nM/1μM; A,C) or BMS-754807 (2nM; B,D). All data is represented as fold change compared to Mn²⁺ alone or Mn²⁺+IGF (both normalized=1). Samples were from the exact same patient cell lines and differentiations as Figure 7. N=4–6 per condition across three control and three HD patients. Representative blot in Supplemental Fig 5.A,B. #= significant difference to Veh (dotted line=1 for -IGF conditions) or IGF alone (dotted line=1 for +IGF conditions) by 95% CI.

Mn²⁺ promotes IGF receptor phosphorylation via intracellular interactions

Since the vast majority of Mn²⁺-induced p-AKT was dependent on IR/IGFR (Fig 5, 8) which autophosphorylate and do not require the presence of upstream kinases, our data suggest that Mn²⁺ is either 1) directly impinging on IGFR/IR activity intracellularly (kinase domain) or 2) affecting insulin/IGFR production or receptor binding, extracellularly. To test this, we exposed hiPSC-derived NPCs to Mn²⁺ for 3hrs, followed by a thorough set of media washes. After washing off the exogenous Mn²⁺, we added Mn²⁺-free HBSS with or without 1nM IGF-1 to the cells. We reasoned that if Mn²⁺ activates IGFR/AKT extracellularly, pre-exposure with Mn²⁺ should not potentiate IGF-induced IGFR/AKT/S6 phosphorylation, as extracellular Mn²⁺ would be negligible in the presence of IGF. However, if Mn²⁺ acts intracellularly to promote IGFR activity, pre-exposure with Mn²⁺ will lead to a lasting increase of intracellular Mn²⁺ and should still cause additive increases to IGF-1-induced signaling. Consistent with an intracellular action of Mn²⁺, we found that pre-exposure with Mn²⁺ was able to increase p-IGFR and p-AKT expression (Fig 9A–C). Similar to WT cells in Fig 7C, Mn²⁺ only increased p-S6 in conditions when IGF was absent (Fig 9D), suggesting the synergy between Mn²⁺+IGF co-treatment on p-S6 signaling may be quite different from that of p-AKT and p-IGFR in these cells. Together, these data support our hypothesis that Mn²⁺ acts on IGFR via intracellular interactions.

Figure 9: Mn²⁺-induced p-IGFR/AKT is due to intracellular, not extracellular Mn²⁺.

A) Representative western blot of control hiPSC-derived neuroprogenitors for p-IGFR, p-AKT, and p-S6 following 3hr Mn²⁺ exposure with 200/500μM Mn²⁺ in HBSS, 3X HBSS washes, then exposure with 1nM IGF (without Mn²⁺) or HBSS (no IGF or Mn²⁺) for another 3hrs. B-D) Quantifications of p-IGFR (B) One-way ANOVA for treatment F(2,6)= 116.5; p<.0001, p-AKT (C) One-way ANOVA for treatment F(5,15)= 107.2; p<.0001, and p-S6 (D) One-way ANOVA for treatment F(5,15)= 14.12; p<.0001. N=6 (two differentiations of three separate control patients); Error bars= SEM. *= significance by Tukey multiple comparison test. *P<.05, **P<.01, ***P<.001.

Acute Mn²⁺ exposure increases glucose uptake in HD cells only

To determine whether Mn²⁺-associated IGF/AKT signaling is associated with the decreased IR/IGFR-related functional processes in HD cells, we examined glucose uptake. IGF/AKT signaling has been associated with energy and glucose homeostasis, both of which are perturbed in HD and even precede onset of primary symptoms in HD mice (93–99). Accordingly, we observed that HD STHdh cells exhibit ~60% reduction in glucose uptake compared WT, consistent with published reports in HD patients (100). We then measured glucose uptake in STHdh cells after a 24hr Mn²⁺ exposure. Consistent with an insulin-mimetic role of Mn²⁺, 100 μM Mn²⁺ was able to modestly, but significantly, increase glucose uptake in HD cells, but not WT cells—although glucose levels in HD cells did not reach WT levels (Fig 10A). We did not observe any effect at 50 μM (data not shown). This increase in glucose uptake was highly consistent with significant increases in p-AKT in HD cells-occurring only at 75 or 100uM (p-p53 was also used as a positive control for Mn-induced signaling on the representative blot, but not quantified) (Fig 10B, C). 100uM Mn treatment in HD cells increased Mn²⁺ uptake to levels equal in magnitude to WT cells treated with 50uM Mn²⁺, although there were still significant differences in Mn uptake between genotypes at 50uM, 75uM, and 100uM (Supplemental Figure 7H). This HD-specific effect provides evidence that homeostatic Mn²⁺ levels play a role in maintaining glucose homeostasis, and suggests that reductions in bioavailable Mn²⁺ (such is the case is with Q111/Q111 cells) contribute to perturbations in glucose uptake, a known phenotype in HD patients.

Figure 10: Mn²⁺ increases glucose uptake in Q111 cells.

A) Quantification of glucose uptake in STHdh Q7/Q7 and Q111/Q111 following 24hr exposure with 50/100μM Mn²⁺. “no 2DG” is a negative control in which no glucose was added to cells. N=4; Error bars= SEM. Two-way ANOVA; treatment= F(2,6)=1.443 p=<0.3079; genotype= F(1,3)= 5.585 p=0.0991; treatment-genotype interaction= F(2,6)= 5.939 p=.0370. *= significance by Tukey’s multiple comparison test. B) Quantification of p-AKT (p-p53 is not quantified) across 12.5–100uM Mn exposure. Two-way ANOVA; treatment= F(5,10)=29.75; p=<0.0001; genotype= F(1,2)=35.45; p=0.0271. *= significant by Sidak’s multiple comparison. *P<.05, **P<.01, ***P<.001. C) Representative western blot of p-AKT and p-p53 after 24hr Mn exposure with 12.5–100uM in WT and HD STHdh cells

Discussion

Several laboratories have observed Mn²⁺-induced insulin/IGF-related signaling (primarily AKT) in cell and rodent models, but the direct mechanism by which Mn²⁺ stimulates these pathways was unclear (10, 27–29). Here, we sought to define the ambiguous relationship between Mn²⁺ and insulin/IGF signaling, and how this interaction mediates Mn²⁺-induced signaling—namely AKT. We also wanted to investigate how perturbations in this relationship may manifest in HD—as models of HD exhibit reduced Mn²⁺ uptake and defective insulin/IGF signaling.

Here, we observed that IGF and Mn²⁺ work synergistically to regulate insulin/IGF activity. In the presence of IGF, Mn²⁺ was able to increase peak AKT signaling more than near-physiological (1nM) or saturating (10nM) concentrations of IGF alone (Fig 1C, D) (101). Since Mn increased p-AKT more than saturating concentrations of IGF alone, this suggest Mn can increase maximal IGF-dependent activation of p-AKT in living cells. Furthermore, Mn²⁺ was able to increase phosphorylation of IR/IGFR, in the presence of physiological levels of IGF (1nM) (Fig 4A, 6, 7B). Additionally, after 24hr exposure, Mn²⁺ decreased total IGFR protein expression and IGFR mRNA expression (Fig 4C–E). We hypothesize that sustained exposure with Mn²⁺ causes a negative feedback loop to reduce total IGFR protein by decreasing IGFR mRNA, thus dampening total IGF signaling activity, though we have no explicitly tested this mechanistically. These findings are consistent with a previous study showing that Mn²⁺ inhibits insulin receptor dephosphorylation, in addition to increasing kinase activity (61). This sustained activity would likely require a negative feedback loop to prevent continuous stimulation of the pathway, facilitating restoration of homeostasis. However, presently, it is unclear why Mn-induced decreases in IGFR mRNA and protein are similar between WT and HD cells although intracellular Mn and Mn-induced p-IGFR/AKT are significantly different. This might suggest that these decreases in total protein and mRNA may actually be independent of intracellular Mn and subsequent kinase signaling, or that Mn levels are differentially affected in cytoplasmic and nuclear cellular compartments. Future studies should pharmacologically or genetically investigate this mechanism further and will likely need to measure mRNA transcription, stability, and translation. Together, our data demonstrate that Mn²⁺ is able to modulate insulin/IGF signaling via multiple avenues directly related to the IR and IGFR, and thereby impacts downstream p-AKT levels. Furthermore, it is possible that cellular Mn²⁺ levels could be further regulating this signaling via means which we have not examined in this study, such as receptor internalization, Mn²⁺-induced phosphatase activity, or interactions with autophagy or the proteasome. Thus, to fully understand the complete coordination between Mn²⁺ homeostasis and IGF signaling, all of these distinct modalities of IGF signaling regulation (transcription, mRNA stability, and translation) must be assessed in the future.

Mn²⁺-induced AKT signaling was first reported years ago, specifically in the context of toxic exposures, but the mechanism of this effect remained unknown (4). Our recent study demonstrates that the majority of Mn²⁺-induced AKT is dependent on PI3K (28) (which also uses Mn²⁺ as a cofactor), however, PI3K and AKT can be activated via dozens of routes including receptor tyrosine kinases, integrins, cytokines, and G protein-coupled receptors (102, 103). Given the extensive crosstalk between Mn²⁺ and IGF observed in this study, we posited that Mn²⁺-induced AKT signaling is mediated by IR/IGFR. We found that several IR/IGFR inhibitors with non-overlapping off-target kinases (Fig 5A) could inhibit the vast majority of Mn²⁺-induced p-IGFR and p-AKT across every cell model tested (with the exception of HEK293 cells)— achieving ~50% inhibition of Mn²⁺-induced p-AKT at approximate IC50 concentrations and >70% inhibition at higher concentrations (Fig 5, 6, 8, Supp Fig 6A–C). By correlating the degree of p-IGFR inhibition to p-AKT inhibition [(% of p-AKT inhibition) / (% of p-IGFR/IR inhibition)], we estimate that >70–80% of Mn²⁺-induced p-AKT is dependent on p-IGFR in most cell lines (Supp Fig 6 B–C). Our observations strongly suggest that the vast majority of Mn²⁺-induced AKT signaling in many cell types is mediated via interactions between Mn²⁺ and the insulin/IGF receptors themselves, rather than other upstream or downstream effectors.

Our data show that Mn²⁺ synergistically works with IR/IGFR to promote downstream AKT signaling, and that this is most likely due to the effects of intracellular, not extracellular, Mn²⁺ (Fig 9). Together, with the previously discussed inhibitor data (Fig 5, 8), we can narrow the initial site of action to the intracellular kinase domains of IGFR/IR (as these receptor-kinases do not require an upstream kinase for activation). As Mn²⁺ is a known cofactor for several other kinases, we postulated that Mn²⁺ is binding the kinase domain of IR/IGFR to stimulate receptor autophosphorylation and propagate signaling downstream. Our results are consistent with cell free in vitro experiments demonstrating the ability of Mn²⁺ to act as a cofactor for these receptor and increase kinase activity in a defined biochemical system (61, 62). Future studies using structural crystallography or computational assessments could be performed to confirm that Mn²⁺ binds this intracellular region, but this will be challenging given the limited utility of available methods to detect Mn²⁺-binding of proteins in living systems. Previous studies have shown that Mn²⁺ can stimulate insulin production itself to promote IGF signaling (51, 59). This study does not directly assess whether Mn²⁺ can increase insulin/IGF production, though our data suggest the mechanism of the Mn²⁺-induced effect on IR/IGFR is derived from intracellular interactions, not interactions with the extracellular, ligand-binding, regions of the receptor (Fig 9). As the vast majority of Mn²⁺-induced p-AKT is regulated by IR/IGFR, and Mn²⁺ is able to increase maximal AKT activity even under saturating concentrations of IGF, this also supports the hypothesis that Mn²⁺ is not acting at the ligand-binding domain or increasing the ligands themselves in our conditions/cell lines. However, we cannot exclude that Mn²⁺ may be able to promote insulin/IGF production or extracellular binding in addition to intracellular stimulation of the receptor itself.

Heavy metals, including Mn²⁺, have specific proteins to which they bind, leading to activation of specific biological responses. However, excessive accumulation of cations can incur broad, non-specific activation of specific signaling cascades due to dyshomeostasis and toxicity. For instance, accumulation of many heavy metals will promote ROS production, broadly activating several signaling cascades (104, 105). In this study, we provide evidence that Mn²⁺-induced IGF signaling is due to a highly specific interaction between Mn²⁺ and insulin/IGF receptors themselves, not merely a shared effect, such as ROS, by other cations (Fig 2A–C). This is supported by a previous study in which Mn²⁺-induced p-AKT in the striatum was not blocked by antioxidant treatment, consistent with a ROS-independent mechanism (30). The concentrations we chose are the highest concentrations that these cells can be exposed to these cells for 24hrs without incurring detectable cell death (79). While Mn²⁺ can clearly activate p-AKT at sub-cytotoxic concentrations at 24hrs or during sub-acute, 7-day exposures (Fig 2A–F), it is possible that other metals could activate p-AKT at supra-toxic concentrations due to non-specific toxicity. A previous study in our lab observed that Cd could induce p-AKT at 50μM, which is a cytotoxic concentration of Cd (79). In this study, we used 20μM Cd which did not induce p-AKT. We posit that cytotoxic levels of Cu, Ni, Zn and other metals may also promote AKT signaling as this has been previously shown, though we did not detect any increase p-AKT induction by sub-toxic concentration of these metals (106–110).

Mn²⁺ and Mg²⁺ can compete for binding and activation of the same protein kinases. In fact, the vast majority of kinases in the human body are Mn²⁺- and/or Mg²⁺-dependent, though more often Mg²⁺-dependent (111). Since DMEM-based media contains high physiological concentrations of Mg²⁺ to sustain enzymatic activity, we could not directly test whether Mg²⁺ could stimulate p-AKT alone. However, even in the presence of apparent saturating concentrations of Mg²⁺, Mn²⁺ was still able to incur a ~3–4 fold increase in p-AKT after 24hrs in WT cells (Fig 1D,2B). Previous studies in a biochemically defined system reported the activation of the insulin receptor in the presence of Mg²⁺ or Mn²⁺+Mg²⁺. Our data demonstrate that Mn²⁺ can increase the maximal activity of the insulin/IGF receptors compared to Mg²⁺ alone in a living biological system as well. In this way, intracellular Mn²⁺ could modulate maximal IGF signaling in a biological setting, even when sufficient Mg²⁺ is present in the cell. In the case of HD, reduced intracellular Mn²⁺ may cripple maximal IGF signaling and contribute to the AKT-related defects observed in HD, discussed more below.

Mn²⁺ is an essential cofactor for a variety of enzymes (1). Other studies, including our own, show that Mn²⁺ can incur activity of specific kinases and pathways that are not broadly shared by other metals (2, 3, 14, 57, 61, 63, 103, 112, 113) (Fig 2A–C). However, Mn²⁺ is not broadly recognized as a signaling molecule itself. We know that Mn²⁺ binds to and promotes activity of specific kinase pathways, but few studies have examined how biologically-relevant intracellular Mn²⁺ may help regulate cellular kinase activities, facilitating signaling cascades within the cell. Additionally, most studies have only examined the effect of Mn²⁺-induced signaling at toxic concentrations of Mn²⁺—here we demonstrate that Mn²⁺-induced signaling can also occur at sub-cytotoxic concentrations. This suggests biologically-relevant concentrations of intracellular Mn²⁺ play a role in the normal signaling homeostasis of dozens of pathways. The discovery of a Mn²⁺-specific chelator or small molecules that could manipulate intracellular Mn²⁺ could be incredibly beneficial in further elucidating the necessity of Mn²⁺ in cell signaling.

The AKT response to Mn²⁺ has also been associated with neuroprotection, as a way to stimulate pro-growth pathways (36, 42, 114–118). Here we observed that Mn²⁺ increased glucose uptake in STHdh HD cells only (Fig 10). While this increase was modest, it was also genotype-dependent, occurring only in a model of HD which exhibits reduced Mn²⁺ uptake and only at concentrations of Mn²⁺ which significantly increase p-AKT. We hypothesize, that under serum-deprived exposures (i.e. HBSS), Mn+IGF treatment may increase glucose uptake to a greater magnitude, perhaps even in WT cells, as Mn+IGF induces ~30-fold increases in p-AKT under these conditions (see Fig 1A), as opposed to 24hr exposures in media used in these experiments (see Fig 1D). This increase in glucose uptake is consistent with a role for AKT in promoting PFK1 phosphorylation and GLUT4 translocation to stimulate glucose uptake (102, 119). Since these glucose uptake experiments were done in serum-containing media, it is likely that glucose uptake (and upstream signaling) are at maximal homeostatic levels in WT cells (i.e. saturated with ligand, normal Mn uptake). We hypothesize that HD cells (possibly because of reduced insulin/IGF levels and/or reduced Mn-induced homeostasis/signaling) are unable to reach this maximal level. When cells are treated with Mn, this either 1) restores normal Mn bioavailability and downstream Mn-induced signaling or 2) compensates for a baseline, phenotypic decrease in insulin/IGF-induced p-AKT to increase glucose uptake. This suggests that Mn²⁺, particularly in cases of reduced Mn²⁺ bioavailability, can stimulate neuroprotective processes, such as glucose uptake, which are impeded in neurodegenerative diseases like HD. This also demonstrates that Mn²⁺-induced IGF signaling serves a role in mediating responsive, downstream biological processes, not merely just a broad response to higher intracellular Mn²⁺. A number of studies have detected clear metabolic defects in glucose metabolism in HD 18F FDG PET imaging studies. Glucose hypometabolism has been detected in manifest HD patients in striatum and cortex and associated with motor and cognitive dysfunction, respectively(100, 120–123). Additionally, these defects have been found in patients prior to neuronal loss, suggesting glucose metabolism may be an early HD phenotype(124–126). Similarly, our studies have detected decreased Mn²⁺ uptake in vivo only in prodromal (premanifest) mice. Given the results from our study, future studies should investigate whether Mn²⁺ uptake defects in vivo correlate temporally with perturbation in glucose uptake in vivo and whether Mn²⁺ treatment can remedy this.

Humbert and colleagues discovered that mutant HTT can be phosphorylated at Ser⁴²¹ specifically by AKT, and this results in robust amelioration of cellular pathology, increasing cell survival and reducing aggregate accumulation (114, 127). Since then, other studies have examined the neuroprotective capacity of AKT stimulation more closely (84, 128, 129). Ahmed et al found that loss of IMPK, a mediator of PI3K and AKT activity, is disrupted in models of HD. However, overexpression of IMPK stimulated AKT activity and rescued cellular pathology and motor abnormalities in HD models, consistent with the neuroprotective potential of AKT. Similarly, we hypothesized, that in the context of reduced Mn²⁺ bioavailability (14, 79), suboptimal Mn²⁺ levels impair activation of IR/IGFR, reducing activity of AKT downstream and concurrently reducing the neuroprotective, pro-growth signaling that AKT provides to combat disease pathology. Of particular interest, Rego and colleagues have elucidated the neuroprotective effects of IGF treatment in cell and mouse models of HD, including upregulation of HTT Ser⁴²¹ phosphorylation (65–67, 85, 130, 131). Additionally, Hiney and colleagues have shown that Mn²⁺ administration upregulates IGFR/AKT/Rheb/mTOR in vivo, which have been targeted to combat HD pathology in separate studies, providing a proof-of-principle that in vivo Mn²⁺ administration alone can potentiate IGF signaling in the brain (10, 132). Our study suggests Mn²⁺ can increase the maximal activity of the IGF signaling axis in living cells, even under saturating concentrations of IGF and Mg²⁺ (Fig 1D, 2A–C). Given the synergistic co-regulation of IGF and Mn²⁺ presented in this study, we hypothesize that co-treatment of HD models with Mn²⁺+IGF may incur the greatest AKT stimulation and thus, increase the neuroprotective potential and therapeutic benefit of IGF-1 treatment alone.

Methods

Immortalized cell culture

The immortalized, murine striatal cell lines (STHdh^Q7/Q7 and STHdh ^Q111/Q111 were obtained from Coriell Cell Repository (Cambden, NJ). STHdh striatal cells were cultured in Dulbecco’s Modified Eagle Medium [D6546, Sigma-Aldrich, St. Louis MO] supplemented with 10% FBS [Atlanta Biologicals, Flowery Branch, GA], 2 mM GlutaMAX (Life Technologies, Carlsbad, CA), Penicillin-Streptomycin, 0.5 mg/mL G418 Sulfate (Life Technologies, Carlsbad, CA), MEM non-essential amino acids solution (Life Technologies, Carlsbad, CA), and 14mM HEPES (Life Technologies, Carlsbad, CA). They were incubated at 33°C and 5% CO₂. Cells were passaged before reaching greater than 90% confluency, and were never passaged past the recommended 14^th passage. The cells were split by trypsinization using 0.05% Trypsin-EDTA solution (Life Technologies, Carlsbad, CA) incubated for five minutes. One day prior to exposure, cells were plated in the appropriate cell culture plate type at 8×10⁴ cells/mL for WT and 1×10⁵ cell/mL for HD (as these HD cells grow slightly slower than WT counterparts). For some experiments, STHdh cells underwent serum deprivation in HBSS prior to exposures. For these, STHdh cells were plated the same as above. The day after, cells were wash 3X with HBSS and incubated for one hour in HBSS. After this, treatments were added to the cells in HBSS for another 3 hours before lysates were collected for western blot.

For 1 week, low-dose Mn²⁺ exposures, STHdh cells were exposed to 0, 1, 5, or 10μM Mn²⁺, 10 or 100nM IGF 24 hours after plating. The cells were split once, 1:4, midway through the week once they were near confluency. Exposure was continued with fresh media. Cells were harvested on day 7 at the same time of day as the original exposure day. Because of the original plating density difference, the density of Q7 and Q111 cells at the end of the 7 days was approximately equal.

NIH-3T3 and HEK293 cells were a generous gift from the Tansey lab. Cells were maintained and plated with high-glucose DMEM (Corning 0013CV) with penicillin-streptomycin. For experiments, they were plated at 100,000 cells/mL. After 48 hours, cell media was aspirated, cells were washed 1X with HBSS, and then fresh HBSS (without growth factors) was added for another hour to begin growth factor deprivation. After 1 hour of deprivation, media with IGF and/or Mn²⁺ and inhibitors was added to the cells for an additional 3 hours before protein lysates were collected.

Primary Rat Astrocyte cell culture

Primary cortical astrocytes were obtained according previous published protocols. Briefly, cortices of newborn Sprague-Dawley rats were carefully dissected and the meninges were removed(133, 134). Next, the cells were dissociated with a multi extraction protocol using dispase and DNAse and plated at 10,000 cells/cm² on poly-l-lysine coated dishes. The cells were maintained in minimum essential medium (MEM) supplemented with 10% horse serum, 100 U/ml penicillin and 100 μg/ml streptomycin. The first media exchanged occurred 24 hours after plating the cells, and after that once every 3–4 days. After 7–8 days in culture, the cells reached confluence and the astrocytes were used for the respective experiments.

hiPSC cell culture

Three control and HD-patient derived hiPSC-lines were differentiated into striatal islet-1-positive neuroprogenitor cells as previously described (14). HD patient mutant alleles were 58, 66, and 70 CAG repeats. All hiPSC lines were confirmed to be pluripotent (PluriTest- Expression Analysis, Durham, NC) and to have normal karyotypes (Genetics Associates, Nashville, TN). Additionally, a subset of cells was fixed with 4% paraformaldehyde for 15 minutes and immunocytochemistry was performed to ensure all cultures expressed the striatal marker Islet-1.

For experiments, striatal neurprogenitor cells were plated at 300,000 cells/mL at day 10 of differentiation. At day 11, purmorphamine and rock inhibitor-containing media was replaced with fresh media containing purmorphamine (.65μM) for 24hrs. For 3-hour exposures, cells were washed with HBSS 3X and then cells underwent growth factor/insulin deprivation in media without N2 supplement (which contains the insulin for the differentiation media) for 3 hours. This was followed by exposing the cells in N2 supplement-free media with Mn²⁺, inhibitors, and IGF-1 for another 3 hours prior to protein collection. Other N2 components include human transferrin, progesterone, putrescine, and selenite.

In order to test the effects of intracellular vs extracellular Mn²⁺, cells were exposed to 200 or 500 μM Mn²⁺ for 3hrs in -N2 media 24hrs after removing rock inhibitor containing media. After, cells were washed 3 times with HBSS and then replaced with IGF-containing -N2 media for another 3 hours prior to protein collection. -N2 media without IGF was added to half the wells to ascertain the effect of IGF alone.

Inhibitors/growth factors/metals

Inhibitors/growth factors were used at the following concentrations: recombinant human IGF-1 (R+D systems Cat# 291-G1)= .1–10μM, recombinant human EGF protein (R+D systems Cat# 236-EG), BMS-53692436924 (SelleckChem)= 100nM-1μM, BMS-7548075480754807 (SelleckChem)= 2nM, Linsitinib (SelleckChem)= 100nm-1μM, NVP-AEW541 (SellekChem)= 100nM-1μM, and LY294002 (Tocris)= 7μM.

The following metallic compounds were used as sources for the metal exposures: MnCl₂ · 4H₂O (Sigma Aldrich, St. Louis, MO) for Mn²⁺, FeCl₃ (Sigma Aldrich, St. Louis, MO) as Fe³⁺, : CuCl₂ · 2H₂O (Alfa Aesar, Ward Hill, MA) for Cu²⁺, MgCl₂ · 6H₂O (Alfa Aesar, Ward Hill, MA) for Mg²⁺, ZnCl₂ (Acros Organics, Morris, NJ) for Zn²⁺, CdCl₂ · H₂O (for Cd, NiCl₂ · 6H₂O (Alfa Aesar, Ward Hill, MA) for Ni²⁺, and CoCl₂ ∙ 6H₂O (MP Biomedicals, Solon, OH) for Co²⁺ exposures.

Western blot

Cells were washed once with ice cold PBS and then scraped from wells of a 6-well plate directly into 100ul, ice-cold RIPA buffer containing protease inhibitor (Sigma-Aldrich, St. Louis, MO) and phosphatase inhibitor cocktails 2 & 3 (Sigma, Sigma-Aldrich, St. Louis, MO). Cell lysates were centrifuged at 4°C for 10 minutes at 20,000 g. Protein concentration was quantified using the BCA assay (Peirce Technologies). Samples were mixed with 5x SDS loading buffer containing 1% 2-mercaptoethanol and boiled for 5 minutes. Fifteen μg of protein was loaded for each sample onto a 4–20% pre-cast gel SDS-PAGE gel (BioRad, Hercules, CA) and run at 90V for 120 minutes. The protein bands were then transferred onto nitrocellulose membranes using iBlot Gel Transfer Device (Life Technologies). The remaining gel was stained with Coomassie Blue stain (Biorad, Hercules, CA). Since the gels retained ~1/3 of the original protein after transferring with the iBlot, we imaged the stained gel on the Li-Cor Odyssey Imaging System and quantified the intensity entire lane from ~150–20 kDa. This value was used to normalize the quantification of the immunostained bands. The membrane was blocked in Odyssey Blocking Buffer for one hour prior to the addition of the primary antibodies. The primary antibodies were diluted 1:1000 in Odyssey Blocking Buffer containing 0.1% TWEEN and incubated overnight. After washing three times for 5 minutes in TBST, membranes were incubated with secondary antibodies at 1:10,000 (LiCor, Lincoln, NE) for 1 hour. Following three, 5-minute washes in TBST, membranes were imaged using the Li-Cor Odyssey Imaging System, and quantification was performed using Image Studio Lite (LiCOR, Lincoln, NE). In Figure 6A, membranes were visualized using CL-XPosure film (Thermo Scientific, #34090) and quantified via ImageJ.

pIR/IGFR Tyr¹¹³⁵ (3918), pan IR/IGFR (9750), p-AKT Ser⁴⁷³ (4060), p-AKT Thr³⁰⁸ (2965), pan AKT (2920), and pan S6 (2317), p-S6 Ser^235/236 (2211) antibodies were purchased from cell signaling technologies and used at 1:1000 dilution except pIR/IGFR was used at 1:500. Note: Unless stated, all western blots examined p-AKT at the Ser⁴⁷³ residue. All blots aside from Figure 6A were normalized by total protein (Coomassie). Figure 6A was quantified at p-AKT over actin. Pan protein expression was quantified separately in the Supplemental Figure 1, 4 as it did not significantly change under our exposure paradigms. Additionally, a direct comparison showing lack of a significant differences between phosphorylated expression of proteins normalized to pan protein expression or to Coomassie (total protein) for STHdh and Islet cells can be found in Supplemental Figure 7.

Note: Many of our experiments were performed to maximize the number of conditions/genotypes/cell lines/exposure time and, thus most replicates of western blots included conditions that were not relevant for specific figure panels. Thus, some western blot images were cut to exclude these non-relevant conditions. However, any western blots that are grouped together for any specific figure panel come from a single blot and are only cut/copied to exclude unneeded conditions/replicates for simplicity and space.

Cellular Fura-2 Manganese Uptake Assay (CFMEA) was performed as described previously(86).

PC12 cell HTT induction and differentiation

HTT PC12 cells were purchased from Coriell Repositories (CH00285, CH00287, CH00289). Undifferentiated cells were plated at 75,000 cell/mL in DMEM F12 media with 4.5 g/L glucose, L-glutamine, and sodium pyruvate with 10% fetal bovine serum, 5% horse serum, 250ug/mL zeocin (Invitrogen), 100ug/mL G418, and 1% penstrep. Cells were plated on collagen-coated plates at 75,000 cells/mL. Cells were induced with 5μM ponasterone A (Invitrogen) and differentiated with 50ng/uL neural growth factor (NGF- Cell signaling technologies) for 6 days prior to exposures. On the 6^th day, exposures were added for an additional 24 hours with ponasterone A. Cells were imaged by microscopy for neurite outgrowth and expression of RFP-HTT, assuring induction and differentiation of cultures.

qRT-PCR for IR/IGFR/IRS2

STHdh Q7/Q7 and Q111/Q111 cells were plated in 6-well plates and treated with Mn²⁺ and/or IGF-1. After 24 hours, cells were lysed in 1mL of TRIzol® Reagent (CatNo. 15596) and stored at -80°C until use. Total RNA was extracted using TRIzol® Reagent, according to the manufacturer’s User Guide (Invitrogen, Carlsbad, CA). RNA was DNAse I treated, following the NEB DNase I Reaction Protocol (M0303) (New England BioLabs® Inc, Ipswich, MA). cDNA was generated on Bio-Rad Laboratories’ S1000™ Thermal Cycler, using 50μM Random Hexamers (P/N 100026484, Invitrogen, Carlsbad, CA), 50U MuLV Reverse Transcriptase (P/N 100023379, Applied Biosystems, Foster City, CA), 10mM dNTPs (CatNo. N0446S, NEB, Ipswich, MA), 20U RNAse, Inhibitor (CatNo. N8080119, Applied Biosystems, Foster City, CA), 25mM MgCl₂ (P/N100020476, Applied Biosystems, Foster City, CA), 10x PCR Buffer II (RefNo. 4486220 Applied Biosystems by Life Technologies, Austin, TX), and 1ug of extracted RNA, and using the following cycling times: 25°C for 10 min, 42°C for 60 min, 99°C for 5 min, 5°C for 5 min, 4°C forever. Q-RT-PCR was performed on Eppendorf’s Mastercycler® epGradientS Realplex², using KAPA SYBR® FAST qPCR Master Mix (2X) Universal according to the manufacturer’s recommendations (KM4101, KAPA Biosystems, Wilmington, MA). Sequences of primers used are as follows: mpgk1_R2 AAAGGCCATTCCACCACCAA, mpgk1_F2 GCTATCTTGGGAGGCCTAA, Irs2 forward, GTCCAGGCACTGGAGCTTT, Irs2 reverse, GCTGGTAGCGCTTCACTCTT, IGFR For Primer GCTTCTGTGAACCCCGAGTATTT, IGFR Rev Primer TGGTGATCTTCTCTCGAGCTACCT, IR For Primer TTTGTCATGGATGGAGGCTA, IR Rev Primer CCTCATCTTGGGGTTGAACT.

Glucose uptake assay

Glucose Uptake-Glo Assay was conducted per manufacturer instructions (Promega, Madison, WI). Q7 and Q111 cells were cultured after plating for 16–24 hours in 96-well tissue plates. 24 hours prior to the Glucose Uptake-Glo Assay, the media was removed and replaced with media containing 0, 50, or 100μM Mn²⁺. The, the media was removed and the wells washed with 100μl PBS. 50μl of prepared 1mM 2DG was added per well, plates were briefly shaken, and incubated for 10 minutes at room temperature. The, 25μl of Stop Buffer was added and the plate briefly shaken again. Thereafter, 25μl of Neutralization Buffer was added to each well and the plate shaken briefly. This was followed by addition of 100μl of 2DG6P Detection Reagent to each well, the plate was briefly shaken and then incubated for 30 minutes at 33 degrees (the normal temperature for these cell lines). Luminescence was recorded using a 0.3–1 second integration on a microplate reader (BioTek Synergy H1, Winooski, VT).

Statistical analysis

GraphPad Prism version 8.0.2 was used. Most data are shown by fold change to increase clarity for the reader. However, all data that was analyzed by one-way or two-way ANOVA was converted to log values to allow for a normal distribution for the statistical analysis. For any data comparing values versus a unitary value (=1) on a graph- raw, unnormalized data was converted to log values and underwent ANOVA analysis to maintain normal distribution. Data was then often plotted as normalized values for ease of review. If data sets were significant by ANOVA, post-hoc multiple comparisons tests were used to determine specific differences within data sets. For analyses across all possible treatments/genotypes, Tukey’s test was performed. For analyses comparing data back to a specific condition only, Dunnet’s or Sidak’s tests were performed. For the majority of data sets, paired analyses were used as all samples were collected as full sets. For the data sets which compared two points specifically, a paired student’s t-test was used as indicated.