Acute manganese exposure impairs glutamatergic function in a young mouse model of Alzheimer’s disease

Abstract

Manganese (Mn) is an essential metal that serves as a cofactor for metalloenzymes important in moderating oxidative stress and the glutamate/glutamine cycle. Mn is typically obtained through the diet, but toxic overexposure can occur through other environmental or occupational exposure routes such as inhalation. Mn is known to accumulate in the brain following exposure and may contribute to the etiology of neurodegenerative disorders such as Alzheimer’s disease (AD) even in the absence of acute neurotoxicity. In the present study, we used in vitro primary cell culture, ex vivo slice electrophysiology and in vivo behavioral approaches to determine if Mn-induced changes in glutamatergic signaling may be altered by genetic risk factors for AD neuropathology. Primary cortical astrocytes incubated with Mn exhibited early rapid clearance of glutamate compared to saline treated astrocytes but decreased clearance over longer time periods, with no effect of the AD genotype. Further, we found that in vivo exposure to a subcutaneous subacute, high dose of Mn as manganese chloride tetrahydrate (3 × 50 mg/kg MnCl₂ • 4(H₂O) over 7 days) resulted in increased expression of cortical GLAST protein regardless of genotype, with no changes in GLT-1. Hippocampal long-term potentiation was not altered in APP/PSEN1 mice at this age and neither was it disrupted following Mn exposure. Mn exposure did increase sensitivity to seizure onset following treatment with the excitatory agonist kainic acid, with differing responses between APP/PSEN1 and control mice. These results highlight the sensitivity of the glutamatergic system to Mn exposure. Experiments were performed in young adult APP/PSEN1 mice, prior to cognitive decline or accumulation of hallmark amyloid plaque pathology and following subacute exposure to Mn. The data support a role of Mn in pathophysiology of AD in early stages of the disease and support the need to better understand neurological consequences of Mn exposure in vulnerable populations.

1. Introduction

Multiple environmental risk factors, including neurotoxic metal exposure, can exacerbate neuropathology or increase the risk of developing Alzheimer’s disease (AD). Manganese (Mn), readily accumulates within the brain and significantly higher levels of Mn of have been found in the brains of AD patients compared to healthy control patients, highlighting a potential understudied pathological mechanism in AD (Chen et al., 2016; Srivastava and Jain, 2002; Tong et al., 2014). Mn is one of many essential physiological trace metals serving a range of metabolic and cellular functions. Mn is required for essential biological functions including macromolecular metabolism and serves as a cofactor for many metalloenzymes such as arginase, superoxide dismutase, glutamine synthetase, and pyruvate carboxylase. Further, Mn serves functions in numerous tissues for cellular energy, mechanisms of oxidation of free radicals, digestion, and bone growth (Aschner et al., 2005; Chen et al., 2018; Roth et al.). Humans are rarely deficient in Mn, however, tolerable Upper Level Intake Levels (ULs) for Mn in healthy individuals have been identified by the Food and Nutrition Board of the National Academies, above which could result in higher plasma levels and risk of neurotoxicity.

Toxicity can occur by environmental exposures including ingestion of Mn-rich water sources, dermal exposure such as through pesticides, or inhalation of polluted air found close to manganese mines or around industrial areas. Chronic toxic exposure to Mn results in neurotoxicity and the development of a Parkinsonian-like syndrome known as Manganism which features similar movement and neurological symptoms as seen in Parkinson’s disease (Guilarte, 2010; Neal and Guilarte, 2013). In Manganism, Mn accumulates preferentially in some brain regions such as the striatum and globus pallidus in the basal ganglia and within the cortex and this accumulation has a positive linear relationship with neurological symptom progress (Criswell et al., 2019; Racette et al., 2017). The extent to which Mn accumulation in specific brain regions is implicated in other neurodegenerative diseases, such as the hippocampus and cortical areas in AD, and whether this accelerates or exacerbates existing pathology has yet to be fully elucidated.

Mn neurotoxicity and AD share several underlying mechanisms including oxidative stress, mitochondrial dysfunction and decreased mitochondrial respiration, apoptosis, and neuroinflammation (Milatovic et al., 2007, 2009). Dysregulated glutamatergic signaling, and particularly glutamate clearance, is also observed in both AD and following Mn exposure and may underlie behavioral phenotypes in both cases. Rapid clearance of glutamate is critical to protect neurons from excitotoxicity and this occurs predominantly via the glutamate aspartate transporter (GLAST) and glutamate transporter-1 (GLT-1) in rodents, (EAAT (Excitatory Amino Acid Transporter) 1 and 2, respectively, in humans). Impaired clearance mechanisms may include change in both expression and activity of GLT-1 and GLAST. Excitotoxicity is understood to be a major contributor to neuronal cell death in AD. Lower GLT-1 expression and activity in brain tissue correlates with neurodegeneration and clinical signs of patients with AD (Anderson et al., 2001; Li et al., 1997; Masliah et al., 1996a), with similar decreased expression reported in murine models of AD (Hefendehl et al., 2016). Mn exposure has also been shown to decrease expression of GLAST and GLT-1 in astrocytes both in vivo and in vitro (Erikson and Aschner; Hazell and Norenberg, 1997; Johnson et al., 2018a; Pfalzer et al., 2020a Sidoryk-Wgrzynowicz et al., 2009), and directly drives excitotoxicity from disruption of glutamatergic signaling (Brouillet et al., 1993; Erikson and Aschner, 2002; Fitsanakis et al., 2006; Karki et al., 2015; Sidoryk-Wegrzynowicz, 2014). Excitotoxicity can disrupt the excitatory/inhibitory balance within neural circuits and result in aberrant epileptiform activity (Asadollahi et al.; Li et al., 2009; Vossel et al., 2017). Subclinical epileptiform activity occurs more frequently in AD populations compared to cognitively normal age-matched groups and within AD populations it correlates with accelerated cognitive decline as measured by mini-mental state exams (Vossel et al., 2013, 2016, 2017). Critically, case studies of humans exposed to either chronic or acute high doses of Mn report both accumulation in the brain and seizure activity (Komaki et al., 1999; Nemanich et al., 2021).

Given these similarities between AD pathology and consequences of Mn neurotoxicity, we hypothesize that Mn exposure in the context of AD-related neuropathology can lead to glutamatergic dyshomeostasis and increased susceptibility to epileptiform activity. The present study used the APP/PSEN1 mouse model of AD and wild-type littermate controls to assess the effects of acute Mn exposure on glutamatergic neurotransmission. We used both in vitro and in vivo approaches of acute Mn exposure to increase temporal resolution of the specific role and response of astrocytic glutamate clearance following exposure using the in vitro model, and to see the behavioral and physiological consequences of these cellular responses using the in vivo model. Our results suggest that even limited Mn exposure leads to accumulation in brain tissue, alters glutamate clearance dynamics in primary cortical astrocytes, increases susceptibility to seizure onset and severity, but does not induce changes in hippocampal long-term potentiation (LTP). Importantly, we report these changes in young adult animals prior to onset of significant amyloid-plaque or other AD-related neuropathology. These data suggest both that glutamate dysregulation is an early part of AD pathogenesis, at least in rodent models, and that Mn-exposure may drive increased susceptibility to AD genetic risk factors even in young adult animals.

2. Methods

2.1. Animals

Animal husbandry and genotyping has been previously described (Pfalzer et al., 2020b). The current study used APP^SWE/PSEN1^ΔE9 transgenic mice on the C57BL/6J background (APP/PSEN1) and their non-transgenic wild-type (WT) littermates. Mice were weaned at 21–28 days and housed in groups of 2–5 of a single sex in a temperature- and humidity-controlled housing room on a 12:12 light:dark cycle with food and water available ad libitum. Mice were used on post-natal day 4 for cell culture studies or aged 10–12 weeks for all other experiments. APP/PSEN1 mice at this age produce increased amounts of amyloid beta oligomers but do not exhibit amyloid plaque formation and do not exhibit cognitive decline until approximately 8 months-of-age (Gengler et al., 2010a; Radde et al., 2006). Approximately equal numbers of male and female mice were used in each genotype-treatment group. Genotypes were established through DNA from tail snip at weaning and confirmed following euthanasia. All protocols were approved by the Vanderbilt University Institutional Animal Care and Use Committee. All experiments were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

2.2. Culturing Primary Astrocytes

Primary astrocytes were isolated from day 4 postnatal mouse brain as previously described (Schildge et al., 2013). Mouse pups were sacrificed by decapitation after being sprayed with 70% ethanol. The rain was removed and placed into dissecting dish filled with Hanks’ Buffered Salt Solution (HBSS) (Corning, cat # 21–21-CV) on ice, and additional dissection procedures were performed under a dissection microscope. Cortical tissue was isolated and meninges were carefully removed. Cortical tissue was cut into small pieces and transferred into a 5 mL microcentrifuge tube and digested with 0.5% Trypsin-EDTA (Gibco, cat # 15–400-054) in a humidified incubator at 37°C and 5% CO₂ for 30 min, shaking every 5–10 minutes and then neutralized with astrocyte medium (DMEM (Gibco, cat # 11995–065) supplemented with 10% FBS (Sigma, cat # F2442) and 1x Penicillin/Streptomycin (Gibco, cat # 15140-122) at 3 °C for 5 min. Cells were then centrifuged at 100x g for 5 minutes and the supernatant was decanted. The dissociated tissue was resuspended using the same astrocyte medium described above and the remaining tissue was triturated using a P1000 pipette and then a P200 pipette to dissociate cells into a single cell suspension. Cells were plated on poly-L-lysine (Sigma, cat # 25988-63-0) coated T25 flasks (Thermo Scientific, cat # 156367) and housed in a humidified incubator at 37°C in and 5% carbon dioxide (CO₂) until ~80% confluent. The medium was changed every 2–3 days and dishes were shaken at 180 rpm for a minimum of 30 minutes after reaching ~50% confluence to remove microglia and generate an astrocyte-enriched culture.

2.3. Glutamate uptake assay

Glutamate uptake by cultured primary murine astrocytes was determined as previously described (Hedegaard et al., 2020) using an enzymatic glutamate detection kit (Sigma, MAK004). Cells were first grown to confluency in T25 flask coated with poly-L lysine as described above. To prepare cells for transfer into poly-L-lysine coated 6-well plates, culture medium was removed from flasks, flasks were treated using 0.25% trypsin with EDTA (Gibco, cat # 25200072) for 5 minutes to detach cells, and single-cell suspensions were created through trituration. Cells were resuspended in cell culture medium and plated at equal cell densities with cells from each animal plated into multiple wells for within-subject control conditions for the assay. Cell plates were incubated until a minimum of 80% confluency was achieved in a humidified incubator at 37°C and 5% CO₂. Twenty-four hours prior to quantification, media was removed and replaced with either untreated media or media containing 100μM MnCl₂ • 4(H₂O). Following a 24 hour incubation the media was removed and wells were treated with HBSS for 10 minutes to allow cells to equilibrate before the addition of 50 μM L-glutamate diluted in HBSS. Media samples were collected after 5, 10, 15, 30, 45 and 60 minutes and kept on ice.

Cells were then lifted with trypsin/EDTA, washed, pelleted and subsequently lysed in 500 μL ice-cold RIPA buffer (Sigma, cat # R0278) containing proteinase inhibitors (Roche, cat # 04693132001). Cell debris was removed by centrifuging the lysate at 13,500 RPM for 5 minutes at 4°C. The enzymatic assay was conducted in 96-well plate format, by adding 100 μL reaction mix to each 50 μL media sample or diluted glutamate standard. The plate was incubated for 30 minutes at 37°C, protected from light and then the absorbance was measured at 450 nm on a plate-reader (BioTek Synergy H4 Hybrid Reader). All values were normalized to protein content, determined by running an accompanying bicinchoninic acid (BCA) protein detection assay on the cell lysates (Thermo Fisher Pierce, cat # PI23227). The amount of glutamate uptake by the cultured primary astrocytes was calculated by subtracting the glutamate concentration remaining in the media, from the glutamate available from the previous timepoint. Results are expressed as glutamate uptake by astrocytes over time, per amount of protein.

2.4. Acute Manganese Injections

Acute Mn injections were performed as previously described (Pfalzer et al., 2020a). In brief, manganese chloride tetrahydrate (MnCl₂ • 4(H₂O), Fisher Scientific) was dissolved in deionized water as a 1% stock solution, filtered through a 0.2 μm membrane, and administered at 50 mg/kg body weight. Saline (0.9%) was used as a control injection. Treatment was administered at a volume of 5 mL/kg injected subcutaneously in the left or right (alternating) inguinal area using an insulin syringe (27 G, ½ inch). The 50 mg/kg MnCl₂ • 4(H₂O) dose corresponds to a dose of 13.8 mg/kg Mn²⁺. Mice were exposed acutely (1-week exposure) at 10–12 weeks of age using a previously adapted exposure paradigm (Dodd et al., 2005) known to significantly increase brain Mn (Pfalzer et al., 2020a). Mice were injected on days 1, 4, and 7 prior to dissection on day 8 (24 hours after the final injection, Fig. 2), sacrifice and LTP experiments (Fig. 3). or kainic acid (KA) exposure experiments (Fig. 4).

Figure 2:

Mn accumulates in peripheral and neural tissues resulting in protein expression changes. (A) Experimental timeline. APP/PSEN1 and wild-type control mice were given three subcutaneous injections of MnCl₂ • 4(H₂O) at 50mg/kg or saline over the course of one week. Mn accumulation was observed in treated mice in (B) cortex, (C) hippocampus and (D) liver. (E) Cortical DMT1 expression did not differ among treatment and genotype groups although male Mn-treated mice had greater expression than female Mn-treated APP/PSEN1 mice (P = 0.038). (F) Mn treatment had no effect on hippocampal DMT1 expression. (G) Cortical GLAST expression was elevated in Mn treated mice. (H) There was no significant contribution of Mn treatment to hippocampal GLAST expression. GLT-1 expression was unchanged in either (I) cortex or (J) hippocampus. B-D: saline-treated WT, n = 11–12; Mn-treated WT, n = 14–15, saline-treated APP/PSEN1, n = 9–11, and Mn-treated APP/PSEN1, n = 12. E-J: saline-treated WT, n = 9–10; Mn-treated WT, n = 11–13, saline-treated APP/PSEN1, n = 9–11, and Mn-treated APP/PSEN1, n = 8–11.

*p<0.05; ***p<0.001 main effect of Mn treatment by 2 genotype × 2 Mn treatment ANOVA. Data shown are mean +/− standard error of the mean, WT: wild-type, GLT-1: glutamate transporter 1, GLAST: glutamate aspartate transporter 1, DMT-1: Divalent metal transporter 1.

Figure 3.

Evaluation of long-term potentiation (LTP) following exposure to Mn. (A) Stimulation and recording setup in ex vivo hippocampal slices. (B) Representative excitatory post-synaptic potential (EPSP) traces after first pulse prior to (thin line) and post (thick line) TBS. (C) Input-output curves of EPSP (i) and fiber volley (ii) with increasing stimulation. (D) Fiber volley to EPSP ratio. (E) Fiber volley ratio pre/post TBS. (F) EPSP slope percent change from baseline in WT (i) and APP/PSEN1 (ii) treated with Mn or saline and (G) final EPSP slope percent change from baseline in last 5 minutes post TBS. (H) Paired pulse ratio prior to and post-TBS. n’s = 7–10 slices from 4–6 animals. TBS – theta burst stimulation.

Figure 4:

Mn increases seizure susceptibility and severity following kainic acid (KA). All data collected within 20min time window post KA injection. (A) Latency to seizure onset measured as time to first head-bob following KA was faster in Mn-treated mice. (B) Number of head-bobs did not vary significantly according to group. (C) Number of head-bobs occurring per 5-minute interval showed a significant increase in Mn-treated APP/PSEN1 mice. (D) Number of subjects within each group to have a tonic-clonic seizure. (E) Time immobile quantified as absence of movement detected on force plate was lower in APP/PSEN1 mice. (F) Average motion index of groups over-time. Dark center line represents the group average and shading represents upper and lower standard error mean bounds. Time to first recorded tonic-clonic seizure in any mouse (12.5 minutes) indicated by grey shaded area (G) Summation of the recorded average motion index over time period prior to first tonic-clonic seizure in a subject (AUC for first 12.5 minutes, gray box in G). Saline-treated WT, n = 12; Mn-treated WT, n = 13, saline-treated APP/PSEN1, n = 13, and Mn-treated APP/PSEN1, n = 12. *p<0.05 treatment main effect, ^#p<0.05 genotype main effect, ^{†, ‡}p<0.05 Mn-treated APP/PSEN1 compared to saline-treated APP/PSEN1 and Mn-treated WT, respectively. Error bars are standard error of the mean, WT: wild-type, TC: tonic-clonic, abu: arbitrary units, AUC: area under the curve

2.5. Tissue Collection

Mice were anesthetized using isoflurane prior to cervical dislocation and subsequent cardiac perfusion with a 1X phosphate-buffered saline solution. Following perfusion, the brain was removed and the cortex and hippocampus were isolated individually and tissue was frozen on dry ice and stored at −80°C until used for biochemical analyses.

2.6. Mn Levels

Mn concentrations were measured by inductively coupled plasma mass spectrometry (ICP-MS) from a protein lysate (described below in western blot section) using the methods previously described (Pfalzer et al., 2020b). Briefly, 50 uL protein lysate was digested at a 1:1 dilution in ultrapure nitric acid (70%) for 24–48 hours at 70°C (water bath); 100 uL of digested tissue was further diluted to 2% nitric acid with DI water and Mn analyzed. Sample concentrations were calculated relative to a standard curve which was on the same day as samples.

2.7. Western Blot

Protein lysates were prepared by homogenizing frozen tissue using a Bullet Blender STORM with associated eads (Next Advance, cat # ZROB05) in 250 μL Pierce RIPA lysis uffer (Thermo Scientific, cat # 89900) with protease and phosphatase inhibitors [cOmplete^™ EDTA-free phosphatase inhibitor cocktail (Roche, 04693132001), 1:100 phosphatase inhibitor cocktail 3 (MilliporeSigma, cat. # P0044), and 1mM sodium orthovanadate per 10mL of RIPA buffer. Protein lysate was collected following centrifugation at 12,000g for 5 minutes. Protein concentration was measured using standard BCA assay protocol (Pierce BCA Protein Assay Kit, Thermo Scientific, cat # 23225). Samples were denatured with NuPAGE LDS sample buffer (Thermo Scientific, cat # NP0007) and NuPAGE reducing agent (Thermo Scientific, cat # NP0009), and samples were loaded into Bolt^™ 4–12% Bis-Tris Plus gels (Thermo Scientific, cat # NW04120BOX) at 10 μg protein per well. Samples were transferred to nitrocellulose mem ranes using the iBlot2^™ system (Thermo Scientific, cat # IB23001). Following transfer, membranes were blocked for 1 hour with 5% nonfat milk or 5% bovine serum albumin (BSA, Sigma cat # A3059) in tris-buffered saline with 0.1% Tween-20 (TBST). Blots were incubated in one of the following antibodies in 5% milk in TBST [Guinea Pig anti-GLT-1 [1:4000], EMD Millipore Corp cat # AB1783; Rabbit anti-GLAST [1:1000], Novus cat # NB100–1869; mouse anti-GFAP [1:5000], EMD Millipore Corp cat #MAB360; mouse anti-actin [1:1000], Santa Cruz Biotechnology cat# sc-47778] or 5% BSA in TBST [Rabbit anti-DMT-1 [1:1000], EMD Millipore Corp ABS983] at 4°C on a rotator overnight. Following primary antibody incubation, blots were washed with TBST and then incubated for 2 hours shaking at room temperature in appropriate horseradish peroxidase (HRP) conjugated secondary antibody [HRP-anti-rabbit [1:2000], Promega cat # W401B; HRP-anti-guinea pig [1:5000], Sigma cat # A7289; HRP-anti-mouse [1:2000], Promega cat # W402B]. Protein bands were visualized using chemiluminescence (Western Lighting Plus ECL, Perkin Elmer, cat # 103E001EA). Membranes were stripped (Restore Stripping Buffer, Thermo Fisher Scientific, cat # 21059) before probing for subsequent proteins. Mice representing each sex, genotype, and treatment group were present on each blot. Protein bands were quantified using ImageJ (imagej.nih.gov). Each protein band was normalized to its own actin internal loading control, and then to average WT saline treated mouse for that blot. Each reported sample value represents the average value from 2–3 technical replicates for each biological replicate (mouse).

2.8. Long-Term Potentiation

Hippocampal slices were prepared from 10–12-week-old mice treated with Mn or saline. Mice were briefly anesthetized with isoflurane and sacrificed by decapitation. The brain was quickly removed into an ice-cold solution of sucrose-rich artificial cerebrospinal fluid (aCSF) containing 85 mM NaCl (MilliporeSigma, cat #S9888), 2.5 mM KCl (MilliporeSigma, cat #P3911), 1.25 mM NaH₂PO₄ (MilliporeSigma, cat #S0751), 25 mM NaHCO₃ (Millipore Sigma, cat #S6014), 75 mM sucrose (Millipore Sigma, cat #S0389), 25 mM glucose (Millipore Sigma, cat #G8270), 10 μM DL-APV (Abcam, ab120271) (NMDA antagonist), 100 uM kynurenate (Hellobio, HB0362), 0.5 mM sodium L-ascorbate (Millipore Sigma, A7631), 0.5 mM CaCl₂ (Millipore Sigma, cat #C4901), and 4 mM MgCl₂ (Millipore Sigma, cat #M8266) oxygenated and equilibrated with 95%O₂/5%CO₂ and titrated to a pH of 7.4. The brain was mounted on a slicing stage ventral side up, and 300 μm horizontal slices were prepared using a Leica VT-1200S vibratome (Leica Biosystems) in sucrose-aCSF. Slices from both hemispheres were micro-dissected before transferring to a holding chamber containing sucrose-aCSF warmed to 32°C that slowly returned to room temperature over the course of 30 minutes. Slices were then transferred to a second holding chamber containing room temperature aCSF containing 125 mM NaCl, 2.4 mM KCl, 1.2 mM NaH₂PO₄, 25 mM NaHCO₃, 25 mM glucose, 2 mM CaCl₂, and 1 mM MgCl₂ oxygenated and equilibrated with 95%O₂/5%CO₂ and titrated to a pH of 7.4. Slices were maintained under these conditions until transferred into a submerged recording chamber (Scientifica SliceScope Pro 2000, Scientific UK) flowing warmed (to approximately 30°C with an in-line heater) oxygenated aCSF at a flow rate of approximately 6 ml/min.

Field excitatory postsynaptic potentials (fEPSPs) were recorded by stimulating along the Schaffer collaterals in the stratum radiatum and recording the response from the Cornu ammonis 1 (CA1) region of the hippocampus at a rate of 0.05 Hz. An input-output relationship was determined for each slice, plotting the peak amplitude of the fiber volley against the slope of the fEPSP. For LTP experiments, the baseline recordings used a stimulus intensity that produced ~40% of the maximum response and were recorded for at least 20 minutes before tetanizing the slice. LTP was induced using theta-burst stimulation (5 bursts at 100 Hz, repeated at 5 Hz over 5 seconds, with each tetanus including four of these burst trains separated by 10 seconds, totaling 100 bursts). Experiments in which the fiber volley amplitude changed by >20% post-tetanus were discarded. Recordings were continued for at least 60 mins post-tetanus. The magnitude of LTP was measured in the last 5 minutes of the 60 minutes post tetanus recording. There were 7–10 slices in each group from 4–6 mice.

2.9. Seizure induction and severity scoring

Kainic Acid (KA; Sigma, cat # K0250) was dissolved in 0.9% physiological saline and administered at 20mg/kg with an administration volume of 10mL/kg delivered intraperitoneally (i.p.). Following KA administration, mice were immediately placed in a sound attenuating chamber with video and motion recording abilities (Med Associates). Mice were observed continuously for 20 minutes prior to, and post KA dose and seizure behavior was quantified and reported using a modified Racine Scale (Lüttjohann et al., 2009) with the following scores and associated behaviors: 1- immobility and/or flattening, 2- forelimb and/or tail extension; rigid posture, 3- Repetitive movements; head-bobbing, 4- rearing and falling, 5- continuous rearing and falling; barrel rolling, 6- severe tonic-clonic seizures. Head-bobbing is defined here as involuntary myoclonic jerks of the neck that results in a drastic forward motion of the head. Following KA dose, onset and quantity of head-bobs were quantified.

2.10. Statistics

Data are reported as mean ± S.E.M. unless otherwise noted. Analyses were completed in GraphPad Prism 8 and SPSS v27. We did not anticipate sex differences due to subcutaneous Mn exposure, but data were first analyzed with sex as an independent variable to confirm the lack of sex differences using a 3-way ANOVA (2 sexes × 2 treatments × 2 genotypes). Where sex was not a significant contributor to differences, then sexes were combined within groups and analyzed together. Glutamate uptake over specified time periods were first calculated as the area under the curve for individual samples and then the calculated areas were analyzed by two-way univariate ANOVA (2 genotypes × 2 treatments). For all dependent variables, unless otherwise noted, two-way univariate ANOVA (2 genotypes × 2 treatments) with or without the additional repeated measure of time were conducted with appropriate post-hoc follow-up tests. Analyses relating to LTP were analyzed using a repeated measure (RM) two-way univariate ANOVA to account for changes pre- and post-TBS. The threshold for statistical significance for all tests was p < 0.05. Statistical outliers were identified by the ROUT method with a Q coefficient of 1% and subsequently removed from analyses.

3. Results

3.1. Mn modifies glutamate clearance dynamics in primary cortical astrocytes

Primary cortical astrocytes were treated with either saline control or 100 μM manganese chloride tetrahydrate (MnCl₂ • 4H₂O) for 24 hours prior to addition of 50 μM L-glutamate for 1 hour. Samples of growth media containing a starting amount of 50 μM L-glutamate were taken over the course of 1 hour at 5, 10, 30, and 60 minutes and quantified as remaining glutamate normalized to total protein at the end of the time course (Fig. 1A). We observed a dynamic change over the course of glutamate uptake between Mn and saline treated groups. Mn treated groups showed significantly increased reuptake of glutamate within the first 10 minutes compared to control groups of both APP/PSEN1 and WT groups (Treatment F_1,22 = 16.32, P = 0.0005, Genotype F_1,22 = 1.92, P = 0.18, Interaction F_1,22 = 0.387, P = 0.54, Fig. 1B). In contrast, Mn treated groups had decreased glutamate reuptake at later time-points (Treatment F_1,22 = 5.053, P = 0.035, Genotype F_1,22 = 0.0129, P = 0.91, Interaction F_1,22 = 0.581, P = 0.453, Fig. 1C). Total glutamate uptake over the 60-minute period did not differ between groups (Treatment F_1,22 = 2.13, P = 0.158, Genotype F_1,22 = 0.000475, P = 0.983, Interaction F_1,22 = 1.19, P = 0.287). These data show that pre-treatment of primary astrocytes with Mn results in early, rapid clearance of glutamate compared to a delayed response in saline treated astrocytes, suggesting that short term in vitro treatment with Mn alters functional dynamics of glutamate clearance.

Figure 1:

Mn alters glutamate clearance dynamics in primary cortical astrocytes. (A) Glutamate uptake in cultured astrocytes at 5, 10, 15, 30, and 60 minutes following addition of 50 μM L-glutamate. (B) Mn significantly increased glutamate clearance within the first ten minutes. (C) Mn decreased glutamate uptake between 30 and 60 minutes. Cumulative glutamate uptake calculated as area under the curve (AUC) for values gathered between zero and ten minutes (B) or 30 and 60 mins (C). *,*** P<0.05, 0.001 significant main effect of manganese treatment following 2 genotype × 2 Mn treatment ANOVA. Data shown are mean +/− standard error of the mean.

3.2. Mn accumulates in cortex and hippocampus, altering expression of glutamate reuptake transporters.

To determine if these glutamate reuptake changes observed in astrocytes in vitro concurred with changes in glutamatergic signaling in vivo, male and female APP/PSEN1 mice and wild-type littermates aged 10–12 weeks were given three Mn injections (as 50mg/kg MnCl₂ • 4H₂O) (Fig. 2A). Cortical, hippocampal, and liver tissues were isolated 24 hours following the final injection of saline or Mn. Elemental Mn, quantified by ICP-MS, accumulated significantly in cortex (Treatment F_1,44 = 147.1, P < 0.0001, Genotype F_1,44 = 0.479, P = 0.492, Interaction F_1,44 = 1.38, P = 0.245, Fig. 2B), hippocampus (Treatment F_1,45 = 117.6, P < 0.0001, Genotype F_1,45 = 0.29, P = 0.593, Interaction F_1,45 = 0.293, P = 0.591, Fig. 2C), and liver (Treatment F_1,44 = 43.3, P < 0.0001, Genotype F_1,44 = 1.203, P = 0.279, Interaction F_1,44 = 1.43, P = 0.239, Fig. 2D) following treatment. In cortex and liver, accumulation of Mn was similar in males and females. However, in hippocampal tissue, APP/PSEN1 female mice accumulated significantly less Mn compared to APP/PSEN1 males (Sex * Treatment F_1,41 = 7.44; P = 0.009; Mn treated APP/PSEN Male vs Female P = 0.0348, not shown). Divalent metal transporter-1 (DMT1) is a membrane-bound transporter responsible for the influx of divalent metal ions including Mn and iron (Fe) and is expressed on neurons and astrocytes. Male mice had significantly more DMT1 protein compared to females in cortex (Sex_1,34 F = 16.38, P = 0.0003; Genotype * Treatment Interaction F_1,34 = 5.74, P = 0.022), and the difference was observed more strongly in APP/PSEN1 mice (P=0.038) (Fig. 2E). DMT1 expression in the hippocampus was similar across all groups (Fs < 2.44, Ps > 0.13) (Fig. 2F). As DMT1 is a known transporter of both Mn and Fe, the amount of elemental Fe was quantified in the same samples (Supplemental Fig. S1). Only cortical tissue showed changes in the quantity of Fe where Mn treated mice had less Fe than saline treated. The decrease in Fe was more evident in Mn treated APP/PSEN1 mice compared to WT (P=0.007) (Treatment F_1,40 = 6.54, P = 0.015, Genotype F_1,40 = 1.51, P = 0.227, Treatment * Genotype F_1,40 = 6.95, P = 0.012).

Protein expression of GLT-1 and GLAST were quantified by western blot. Exposure to Mn increased GLAST expression in cortical tissue compared to mice treated with saline (Treatment F_1.36 = 4.42, P = 0.043, Genotype F_1.36 = 0.000134, P = 0.991, Interaction F_1.36 = 0.068, P = 0.796, Fig. 2G). No changes were observed in hippocampal GLAST expression (Treatment F_1.38 = 1.004, P = 0.323, Genotype F_1.38 = 0.000088, P = 0.993, Interaction F_1.38 = 0.230, P = 0.634, Fig. 2H) or in GLT-1 expression in either cortex (Treatment F_1.35 = 1.80, P = 0.188, Genotype F_1.35 = 0.427, P = 0.518, Interaction F_1.35 = 0.097, P = 0.757, Fig. 2I) or hippocampus (Treatment F_1.41 = 0.0001, P = 0.992, Genotype F_1.41 = 0.243, P = 0.625, Interaction F_1.41 = 2.07, P = 0.158, Fig. 2J). Representative images for all western blots are shown in Supplemental Figure 2.

3.3. Short-term Mn accumulation does not alter long-term potentiation in young mice.

To determine potential impacts of acute accumulation of Mn on hippocampal synaptic plasticity we measured long-term potentiation by recording local field excitatory post-synaptic potentials (EPSPs) in ex vivo hippocampal slices prior to and following theta-burst stimulation (Fig. 3A–B). Fiber volley amplitude was higher in APP/PSEN1 mice compared to WT (Genotype F_1,23 = 2.04, P = 0.031, Treatment F_1,23 = 0.10, P = 0.754, Interaction F_1,23 = 0.763, P = 0.391), however, the excitatory post-synaptic potential (EPSP) response and fiber volley to EPSP relationship within slices were normal (Treatment and Genotype, F’s<3.41, p’s>0.081) (Fig. 3C–E). LTP was unchanged between Mn and saline treated mice, and between APP/PSEN1 and WT mice (Genotype F_1,29 = 3.94, P = 0.057, Treatment F_1,29 = 0.321, P = 0.575, Interaction F_1,29 = 0.617, P = 0.438) (Fig. 3F–G). Paired pulse ratio, which reflects presynaptic changes in the vesicular readily releasable pool, was also quantified and neither Mn treatment or APP/PSEN1 genotype resulted in changes (Genotype F_1,29 = 1.02, P = 0.321, Treatment F_1,29 = 0.235, P = 0.632, Interaction F_1,29 = 0.024, P = 0.878) (Fig. 3H).

3.4. Mn exposure increases seizure susceptibility following injection of the excitatory agonist kainic acid

Young adult male and female APP/PSEN1 and wild-type control mice were given subcutaneous injections of Mn as 50 mg/kg MnCl₂ • 4H₂O as above, followed by a single intraperitoneal injection of kainic acid (KA, 20 mg/kg), an excitatory agonist (Fig. 4). Immediately following the single KA dose, mice were placed in a sound attenuating chamber that collects sensitive movement data and analysis paired with video recordings. KA-induced seizure-related behaviors were quantified using a modified Racine scale. Time to first head-bob (an involuntary myoclonic jerk that results in a drastic forward motion of the head; 3 on the modified Racine scale) was used to represent latency to seizure onset. Mn treated mice had a shorter latency to first head-bob compared to saline treated mice following KA (Treatment F_1,46 = 5.158, P = 0.028, Genotype F_1,46 = 0.113, P = 0.739, Interaction F_1,46 = 0.319, P = 0.575, Fig. 4A). The total number of head-bobs at the end of the 20-minute recording period was not affected by Mn treatment (Treatment F_1,46 = 3.08, P = 0.086, Genotype F_1,46 = 2.98, P = 0.096, Interaction F_1,46 = 3.104, P = 0.085, Fig. 4B). However, 2 of 11 saline-treated APP/PSEN1 and 3 of 10 Mn-treated APP/PSEN1 experienced tonic-clonic seizures after KA administration (Fig. 4D). This more severe seizure activity decreases the likelihood of observing the milder head-bob behavior. We therefore quantified number of head-bobs per 5-minute interval only during the first 15 mins, during which time only 1 short seizure was observed at ~12.5 minutes. We found that overall APP/PSEN1 mice and Mn-treated mice had greater head-bob rates (Genotype F_1,46 = 4.474, P = 0.0399, Treatment F_1,46 = 4.706, P = 0.0353), and that the effect of Mn treatment was greatest in APP/PSEN1 mice at 10 minutes post KA (Interaction F_1,46 = 3.419, P = 0.037, Fig. 4C). Freezing behavior (1 on the Racine scale) was lower overall in APP/PSEN1 mice compared to wild-type littermates, with no further effect of Mn treatment (Genotype F_1,45 = 6.05, P = 0.018, Treatment F_1,45 = 2.94, P = 0.093, Interaction F_1,45 = 0.728, P = 0.398, Fig. 4E). An average motion index was calculated as a more sensitive metric to determine extent of movement over the course of the 20-minute recording period (Fig. 4F). When considering only the time window prior to any subject first entering a tonic-clonic seizure (time 0–12.5 minutes), Mn treatment decreased motion overall, whereas APP/PSEN1 mice moved more (Treatment F_1,45 = 4.73, P = 0.035, Genotype F_1,45 = 6.82, P = 0.012, Interaction F_1,45 = 3.98, P = 0.052, Fig. 4G). These findings are consistent with Mn exposure conferring sensitivity to the excitatory agonist kainic acid, possibly through the disruption of glutamatergic signaling, resulting in quicker seizure onset and increased seizure-related behaviors following exposure. The higher average motion in APP/PSEN1 is consistent with the finding of APP/PSEN1 mice spending less time immobile and the trend towards having more head-bobbing behavior combined with increased likelihood of tonic-clonic seizures.

4. Discussion

With this study we undertook to investigate potential mechanisms underlying the relationship between Mn exposure and one aspect of AD neuropathology. We showed that Mn dynamically alters glutamate clearance in primary cortical astrocytes with opposing effects at early and later times of glutamate exposure. The early, more rapid clearance of glutamate suggests a possible cellular priming in response to Mn exposure. A dynamic response to Mn exposure was also seen in vivo where Mn-treated mice had decreased latency to seizure onset indicating increase sensitivity to excitatory agonists such as kainic acid. Specifically, Mn-treated APP/PSEN1 mice had an increased rate of head-bobs at earlier time-points, suggesting a genotype driven sensitivity to Mn exposure even at this early age prior to AD pathology accumulation in these mice. Further confirmation of in vivo impacts of Mn on glutamatergic functioning in APP/PSEN1 and wild-type mice comes from the observed upregulation of glutamate reuptake transporter protein, GLAST, in cortical tissue. Consistent with the absence of changes in glutamate reuptake transporter protein expression within the hippocampus of Mn treated mice, we did not observe any alterations in hippocampal long-term potentiation (LTP) following Mn exposure under these age and treatment conditions.

Astrocytic expression of glutamate reuptake transporters GLT-1 and GLAST has been shown to decrease in AD following Mn exposure (Johnson et al., 2018b; Li et al., 1997; Pajarillo et al., 2021), independently. Our present study provides more sensitive temporal resolution to glutamate uptake showing a dynamic response compared to previous reports (Figure 1). Following a 24-hour incubation with Mn, we observed a rapid clearance of glutamate that decreased with time resulting in significantly less clearance compared to saline control by 1 hour. The results at later time periods are more consistent with previous studies, but the earlier exaggerated changes suggest a mechanism of cellular priming for rapid clearance and an inability to attenuate clearance under prolonged glutamate exposures. Future studies quantifying the expression of GLT-1 and GLAST in these primary cortical astrocytes following incubation with Mn would better elucidate a specific mechanism underlying this functional change of impaired clearance in vitro. Interestingly, our in vivo GLT-1 and GLAST protein expression data did not replicate previous studies that reported decreased expression of GLAST and GLT-1 (Johnson et al., 2018b; Pajarillo et al., 2021). The increase in cortical GLAST expression could be one possible mechanism through which cells were able to manifest the earlier rapid increase in glutamate uptake in our primary cortical astrocytes. The discrepancy between our in vivo expression of GLT-1 and GLAST with a previous study could be due to our study having a shorter exposure period, or exposure route, as well as age and genetic background of the animals (Johnson et al., 2018b). An exposure period more consistent with chronic Mn exposure, as experienced in occupational and environmental settings (Lucchini et al., 1999; Mattison et al., 2017; Roels et al., 1987), may indeed result in a decrease in expression of these transporters as reported previously. Future studies examining the molecular responses to a temporal titration of Mn exposure would provide valuable information regarding the transcriptional and translational changes in GLT-1, GLAST, and DMT-1 following varying Mn exposure paradigms. Nevertheless, it is important to note that even acute exposures to toxic levels of Mn can result in glutamatergic dyshomeostasis.

Mn accumulates in neural tissues following exposure with uptake occurring through metal transporters such as DMT1. There were no changes in DMT1 protein expression following Mn exposure dependent on genotype or treatment (Fig. 2E–F), but APP/PSEN1 females had less DMT1 expression than males following Mn treatment. Sex differences are expected following Mn exposure via dietary exposure due to differing intestinal expression of metal transporters (Felber et al., 2019). However, recent studies have shown sex differences in subacute toxic exposures on peripheral tissues (Richter Schmitz et al., 2019), and our study indicates a potential in neural tissues and particularly in the APP/PSEN1 model. DMT1 is also responsible for the transport of other metals such as iron (Fe). We quantified the concentration of Fe in cortical, hippocampal, and liver tissues (Supplementary Figs. 1 and 3). Fe dyshomeostasis in AD has been observed by its interaction with amyloid beta, contribution to inflammation, and ferroptosis (Peng et al., 2021; Tran et al., 2022). While there is competitive absorption between Mn and Fe in the intestine absorption in the brain is not competitive, therefore, excessive Fe plus Mn could increase risks of neural tissue damage (Chua and Morgan, 1996; Rossander-Hulten et al., 1991). Further studies to understand the interaction between Fe accumulation in AD and Mn exposure are needed as the AD population may be especially vulnerable to Mn exposure.

We did not observe differences in LTP depending on Mn exposure or genotype. Changes in hippocampal LTP are reported to change in APP/PSEN1 mice starting around 8 months of age (Gengler et al., 2010b), and given the young age, 10–12 weeks, we did not expect a change in LTP in saline treated conditions. These lack of changes in LTP were consistent with no change in GLT-1 and GLAST expression in the hippocampus. Future studies looking at hippocampal LTP following a more chronic Mn exposure paradigm should look to see if there is a decrease in LTP consistent with decreases in GLT-1 and GLAST expression seen in these longer exposure studies.

To further interrogate glutamatergic signaling dysfunction by Mn exposure, we treated mice with a low dose of kainic acid (KA), an excitatory agonist that acts on kainite and AMPA receptors. Treatment with KA can result in a seizure response and has been shown to increase glutamate release following administration(Holmes et al., 2015). Consistent with Mn exposure altering glutamate clearance, we found that mice treated with Mn had a shorter latency to seizure onset (Fig. 4A). Glutamatergic dysregulation is observed in AD and results in excitotoxicity leading to neuronal cell death (Kirvell et al., 2006; Masliah et al., 1996b; Scott et al., 2011). The additional stressor of Mn exposure in the AD population, or those who are susceptible to developing AD, could result in worsened pathology. Only APP/PSEN1 mice reached tonic-clonic seizures (Fig. 4D) which reduced time available to accumulate more head-bobs compared to the control mice during this 20-min observation period. When the head-bob rate over 5-minute intervals, in which only 1 mouse had one short seizure, was compared between groups (Fig. 4C), we observed a dynamic response to KA with Mn-treated APP/PSEN1 showing an increased rate at earlier time points. Mn-treated APP/PSEN1 mice had the highest number of head bobs overall and interestingly there may have been a sub-set of animals that were more severely impacted. Future work should address the factors that could predict which mice are most likely to undergo the highest number of head-bobs to determine key mechanistic factors that impact susceptibility to Mn exposure. In addition to earlier and more frequent head-bobs, we would expect more freezing behavior in Mn exposed mice that would be reflected in more time spent immobile. We found that APP/PSEN1 mice spent less time immobile, which was likely due to increased head-bobbing behavior and tonic-clonic seizures, but there were no significant effects of Mn treatment detected (Fig. E). When we quantified the average motion in these mice over a set time prior to any mouse entering a tonic-clonic seizure, we saw that APP/PSEN1 mice have greater movement, correlating with less time immobile, but also that mice exposed to Mn have less movement, which would be expected if Mn induced more seizure behavior (Fig. 4G). Although the APP/PSEN1 genotype appear to show a less pronounced Mn effect relative to wildtype animals in time immobile, again this is possibly driven by trend towards increased myoclonic jerking behavior in the Mn-treated APP/PSEN1 mice. These findings in APP/PSEN1 mice are consistent with previous studies showing that these mice are more sensitive to seizure induction by KA (Warner et al., 2015; Wilcox et al., 2021). Interestingly, mice pre-treated with valproic acid (VPA), an antiepileptic drug that increases GABA bioavailability, rescued impairments in glutamate clearance following Mn exposure (Johnson et al., 2018b). Patients with AD have been found to have higher incidences of subclinical epileptiform activity, and these patients tend to have earlier and accelerated cognitive decline compared to AD patients without this aberrant activity (Asadollahi et al.; Vossel et al., 2013, 2016, 2017). AD patients, a population with existing glutamatergic dyshomeostasis, and those with pre-existing epileptiform activity, are potentially more vulnerable to neurotoxic damage caused by Mn exposure.

5. Conclusion

In this study we sought to determine how and acute, high dose exposure to Mn could modify glutamatergic signaling at the cellular, circuit, and behavioral level, particularly in the AD population. We used a well-established mouse model of AD-related pathology, but at a young age, prior to accumulation of beta-amyloid plaques, to determine if early Mn exposure in this pre-symptomatic phase could identify vulnerabilities to Mn exposure within the AD population. Glutamate handling in AD is known to be impaired at many levels including clearance, and Mn exposure may further disrupt clearance. Further studies to better understand mechanistically how Mn exposure impacts glutamate clearance and its role in AD development and prognosis is needed to protect this growing and vulnerable population to neurotoxic environmental exposures.

Highlights:

• Mn exposure dynamically alters astrocytic glutamate reuptake

• Acute exposure to Mn results in brain tissue accumulation

• Mn exposure increases susceptibility and severity to seizures in APP/PSEN1 mice