Selenoprotein P Regulates Synaptic Zinc and Reduces Tau Phosphorylation

Selenoprotein P (SELENOP1) is a selenium-rich antioxidant protein involved in extracellular transport of selenium (Se). SELENOP1 also has metal binding properties. The trace element Zinc (Zn2+) is a neuromodulator that can be released from synaptic terminals in the brain, primarily from a subset of glutamatergic terminals. Both Zn2+ and Se are necessary for normal brain function. Although these ions can bind together with high affinity, the biological significance of an interaction of SELENOP1 with Zn2+ has not been investigated. We examined changes in brain Zn2+ in SELENOP1 knockout (KO) animals. Timm-Danscher and N-(6-methoxy-8-quinolyl)-p-toluenesulphonamide (TSQ) staining revealed increased levels of intracellular Zn2+ in the SELENOP1−/− hippocampus compared to wildtype (WT) mice. Mass spectrometry analysis of frozen whole brain samples demonstrated that total Zn2+ was not increased in the SELENOP1−/− mice, suggesting only local changes in Zn2+ distribution. Unexpectedly, live Zn2+ imaging of hippocampal slices with a selective extracellular fluorescent Zn2+ indicator (FluoZin-3) showed that SELENOP1−/− mice have impaired Zn2+ release in response to KCl-induced neuron depolarization. The zinc/metal storage protein metallothionein 3 (MT-3) was increased in SELENOP1−/− hippocampus relative to wildtype, possibly in response to an elevated Zn2+ content. We found that depriving cultured cells of selenium resulted in increased intracellular Zn2+, as did inhibition of selenoprotein GPX4 but not GPX1, suggesting the increased Zn2+ in SELENOP1−/− mice is due to a downregulation of antioxidant selenoproteins and subsequent release of Zn2+ from intracellular stores. Surprisingly, we found increased tau phosphorylation in the hippocampus of SELENOP1−/− mice, possibly resulting from intracellular zinc changes. Our findings reveal important roles for SELENOP1 in the maintenance of synaptic Zn2+ physiology and preventing tau hyperphosphorylation.

Selenoprotein P (SELENOP1) is a selenium-rich antioxidant protein involved in extracellular transport of selenium (Se). SELENOP1 also has metal binding properties. The trace element Zinc (Zn 2+ ) is a neuromodulator that can be released from synaptic terminals in the brain, primarily from a subset of glutamatergic terminals. Both Zn 2+ and Se are necessary for normal brain function. Although these ions can bind together with high affinity, the biological significance of an interaction of SELENOP1 with Zn 2+ has not been investigated. We examined changes in brain Zn 2+ in SELENOP1 knockout (KO) animals. Timm-Danscher and N-(6-methoxy-8-quinolyl)-p-toluenesulphonamide (TSQ) staining revealed increased levels of intracellular Zn 2+ in the SELENOP1 −/− hippocampus compared to wildtype (WT) mice. Mass spectrometry analysis of frozen whole brain samples demonstrated that total Zn 2+ was not increased in the SELENOP1 −/− mice, suggesting only local changes in Zn 2+ distribution. Unexpectedly, live Zn 2+ imaging of hippocampal slices with a selective extracellular fluorescent Zn 2+ indicator (FluoZin-3) showed that SELENOP1 −/− mice have impaired Zn 2+ release in response to KCl-induced neuron depolarization. The zinc/metal storage protein metallothionein 3 (MT-3) was increased in SELENOP1 −/− hippocampus relative to wildtype, possibly in response to an elevated Zn 2+ content. We found that depriving cultured cells of selenium resulted in increased intracellular Zn 2+ , as did inhibition of selenoprotein GPX4 but not GPX1, suggesting the increased Zn 2+ in SELENOP1 −/− mice is due to a downregulation of antioxidant selenoproteins and subsequent release of Zn 2+ from intracellular stores. Surprisingly, we found increased tau phosphorylation in the hippocampus of SELENOP1 −/− mice, possibly resulting from intracellular zinc changes. Our findings reveal important roles for SELENOP1 in the maintenance of synaptic Zn 2+ physiology and preventing tau hyperphosphorylation.
Keywords: Selenoprotein P, zinc, Alzheimer's disease, tau, selenium INTRODUCTION Within the body, selenium (Se) functions primarily in the form of selenocysteine (Sec), the 21st amino acid, which is incorporated into members of the selenoprotein family (1-3). Selenoprotein P (SELENOP1) is a selenium-rich protein with 10 Sec residues that transports Se in serum from liver to the brain and other organs (4). SELENOP1 is present in the cerebral spinal fluid (CSF) and in the choroid plexus, which releases CSF (5,6), and in glial cells (7). SELENOP1 has also been described in brain neurons (8,9), which may be the targets of Se transport. SELENOP1 KO mice have reduced brain selenium and reduced levels of antioxidant selenoproteins such as glutathione peroxidases 1 and 4 (GPX1 and GPX4) (10). Mice with the SELENOP1 gene deletion have deficient hippocampal synaptic function and deficits in spatial learning and long-term potentiation (LTP), a model for learning and memory (11). SELENOP1 is increased in the brain and CSF in Alzheimer's disease (5,8,12) and associated with both Alzheimer's and Parkinson's pathology (8,13).
Se and Zn 2+ are both essential trace elements required for proper brain function. Selenium deficiency correlates with impaired cognitive and motor function (14,15), while Zn 2+ deficiency correlates with decreased nerve conduction and impaired cognitive performance (16). Alzheimer's disease is associated with increased brain Zn 2+ levels (17). Zn 2+ can increase tau phosphorylation (18,19), which contributes to the formation of neurofibrillary tangles, a hallmark of Alzheimer's disease (20). However, studies have yet to address the biological relevance of the interaction between these elements despite their high affinity for each other and their importance in brain function. Figure 1A shows SELENOP1's two functional, glycosylated domains: (1) a Se-rich C-terminal domain with 9 S residues, (2) an N-terminal domain with 1 S (U) in U-x-x-C redox motif, 2 histidine-rich metal binding sites (located at residue 204-217 and residue 244-250) and a heparin binding site (4, 21). SELENOP1 also has an N-terminal signal peptide for extracellular secretion, which is cleaved in the Golgi (22).
We hypothesized that SELENOP1, as a metal-binding protein, could have a role in brain Zn 2+ homeostasis. In this study, we investigated whether SELENOP1 influences brain Zn 2+ by evaluating changes in hippocampal Zn 2+ in SELENOP1 KO animals. Here we report that deletion of SELENOP1 alters levels of chelatable Zn 2+ and prevents the release of synaptic Zn 2+ in mouse hippocampus. These findings indicate an important additional role for SELENOP1 in the regulation of zinc in the brain. They could also have important implications for the treatment of disorders where Zn 2+ physiology is impaired.

Animals
Mice were group housed on a 12-h light cycle and provided food and water ad libitum. All animals in this study were maintained on diets containing adequate selenium (∼0.25 ppm) and zinc (∼80 ppm). All mice used were 3-6 months of age and included both male and female mice as indicated. All animal procedures were approved by the University of Hawaii Institutional Animal Care and Use Committee. SELENOP1 −/− mice were obtained from the laboratory of Dr. Raymond Burk at Vanderbilt University. The mutant mice were backcrossed to C57BL/6J for at least ten generations with C57BL/6J mice from Jackson Laboratories to ensure congenic strains. Breeding of SELENOP1 +/− mice generated littermates of SELENOP1 +/+ and SELENOP1 −/− pups, which were used in this study in addition to SELENOP1 +/− mice. Genomic DNA extracted from mice tails was used for genotyping PCR using specific primers (forward-ACCTCAGC AATGTGGAGAAGCC, reverse-TGCCCTCTGAGTTTAGC ATTG for wild-type, and reverse-GATGATCTGGACG AAGAGCATCA for SELENOP1 −/− ). Products were run on a 1.5% DNA agarose gel with a SYBR Safe DNA gel stain (Invitrogen) and genotypes were confirmed under UV light.

Timm-Danscher Zn 2+ Labeling
The Danscher modification of Timm's zinc stain ("neo-Timm's") was used to label intracellular chelatable zinc (i.e., not tightly bound to proteins or other molecules) (23). Deeply anesthetized mice received intraperitoneal (IP) injections of 20 mg/kg sodium selenite (2 mg/ml in normal saline). After 2 h, mice were perfused with saline followed by 4% paraformaldehyde (PFA) in saline. Brains were postfixed overnight in 4% PFA, dehydrated serially in 10% and 30% sucrose, and mounted in optimal cutting temperature compound (OCT) for cryostat sectioning. After thoroughly washing in PBS, sections were developed with the IntenSE M silver Enhancement kit (Amersham International) according to the manufacturer's protocol and as previously described (24).

6-Methoxy-8-P-Toluenesulfonamido-Quinoline (TSQ) Stain
TSQ is a Zn 2+ fluorophore that binds intracellular chelatable Zn 2+ in a 2:1 ligand-to-metal ratio that results in increased fluorescence emission at 490 nm in response to excitation at 360 nm (25). Serial sagittal cryosections (10 µm) of brain hemispheres were mounted on positively charged microscope slides. The slides were immersed with 4.5 µM TSQ (Enzo Lifesciences, UltraPure) in 140 mM sodium barbital and 140 mM sodium acetate buffer (pH 10) for 90 s, as previously described (26) and washed in 0.1% NaCl. TSQ-stained sections were imaged using DAPI filter settings (200 ms, monochrome with a 5x objective). The mean fluorescence intensity of the hippocampal CA1 stratum oriens and stratum radiatum, CA3 mossy fibers, and hilar region were measured with ImageJ software (NIH). The background was measured in unstained areas within lateral ventricles and subtracted from mean TSQ signals. Schematic of SELENOP1 structure. The N-terminal region has 1 S (U) in a redox domain, a heparin-binding site and a zinc transporter domain with two metal-binding sites. The C-terminal region has 9 S residues for Se transport. Sites of residues predicted to bind zinc using Predzinc are indicated by *. (B) Alignment of human SELENOP1 metal binding region with other proteins containing the zincin motif. (C) Western blot of metal column eluates after applying mouse serum from WT (+/+) or SELENOP1 KO ( −/− ) mice to mini columns with agarose only or agarose bound to Co 2+ , Ni 2+ , and Zn 2+ . Untreated SELENOP1 −/− and WT serum were also added to the blot as a positive control (left lanes). SELENOP1 protein detected with anti-SELENOP1 antibody (1:1000) had a molecular weight of approximately 55 kDa in the wildtype serum, which was not seen in SELENOP1 −/− serum. SELENOP1 was detected in column eluates from wild-type serum applied to all metal columns, but not from columns with agarose only.

Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES)
The metal concentrations within frozen right brain hemispheres and liver samples were processed at the Agricultural Diagnostic Service Center (ADSC) run by the College of Tropical Agriculture and Human Resources (CTAHR), University of Hawaii. Dry ash sample preparations were subjected to acid digest before ICP-OES (0.01 ppm detection limit) to measure total brain and liver metal content (Zn 2+ , Cu 2+ , Fe 2+ ). Water blanks and solution standards were included with each run to calibrate results.

Metal Agarose Column Purification
To observe SELENOP1 binding to biometals, we used mini spin-columns containing high-density agarose beads conjugated with Zn 2+ , Ni 2+ , and Co 2+ (Agarose Bead Technologies) to isolate metal-binding proteins from wildtype and SELENOP1 −/− mouse serum. A metal-free agarose column served as a negative control. Serum samples were diluted 1:100 in PBS and added to the column, gently shaken for 60 min at 4 • C, then spun for 60 s at 800x g to collect flow-through. Columns were washed with increasing concentrations of imidazole (0, 10, and 20 mM) diluted in PBS, and then bound proteins were eluted with 250 mM imidazole diluted in PBS. Eluted proteins were determined with western blot using an anti-SELENOP1 antibody (27).

Live Hippocampal Slice Imaging
To measure stimulus-induced extracellular Zn 2+ accumulation, hippocampal slices were prepared from 3 to 6 month old SELENOP1 KO and wild-type littermate mice as previously described (28). Following slice preparation, slices were acclimated to room temperature and superfused with oxygenated (95 O 2 and 5% CO 2 gas mix) artificial cerebral spinal fluid (ACSF, composition in mM: NaCl 130; KCl 3.5; glucose 10; NaHCO 3 24; NaH 2 PO 4 1.25; MgSO 4 1.5; CaCl 2 2.0) for at least 60 min. The CA1 stratum radiatum region of the slices was imaged with a Zeiss laser-scanning microscope using a 10X objective with the pinhole fully opened at 1 frame/s at 640 × 480 resolution. Cell-impermeant FluoZin-3 at 1.5 µM (Molecular Probes) was added to the ACSF to detect extracellular Zn 2+ accumulation from hippocampal slices in response to the administration of a depolarizing (35 mM) KCl concentration for 60 s. In some experiments, a slow onset Zn 2+ chelator (Ca 2+ -EDTA) was added to remove contaminating Zn 2+ in media and to reduce background fluorescence (29,30). Fluorescence intensities of hippocampal slices in the CA1 stratum radiatum region upon addition of KCl for each slice were expressed as the fluorescence intensity over the fluorescence during baseline (F/F 0 ). The mean area under the curve during application of high K + for signals of SELENOP1 −/− slices were compared to WT to determine if changes in Zn 2+ release were altered.

Statistical Analysis
All statistical analyses were carried out with Graphpad Prism Software with measurements given as means ± SE. Comparisons between treatments, genotypes and sex were performed by student's unpaired t-test and one-way or two-way analysis of variance (ANOVA), with p < 0.05 considered significant. In general, we did not find sex differences in Zn 2+ distribution or release unless shown; otherwise, males and females were averaged together by genotype.

Zn 2+ Binding Properties of SELENOP1
Previous studies described the affinity of SELENOP1 for several metals, including Zn 2+ . However, the biological interaction of SELENOP1 with Zn 2+ is unclear. We investigated potential Zn 2+ -binding domains of SELENOP1. Using the web-based software Predzinc (https://predzinc.bioshu.se/pred/), a web server that predicts zinc-binding proteins and zinc-binding sites from given sequences, we analyzed the SELENOP1 coding sequences for potential zinc-binding sites. We found that a Hisresidue within the SELENOP1 His-rich metal binding domain as well as two other His residues in the C-terminal region are predicted to be Zn 2+ binding motifs, shown in Figure 1A by asterisks ( * ).
A domain search of the Kyoto Encyclopedia of Genes and Genomes (KEGG) database indicates that the human SELENOP1 protein structure has a zinc-binding domain overlapping one of the His-rich regions, which has homology to the ZIP zinc transporter domain. Additionally, within this sequence is a zincin Zn 2+ -binding motif, generally found in the metzincin family of metalloproteases ( Figure 1B) (32).
Because of the predicted Zn 2+ -binding region of SELENOP1, we tested SELENOP1's potential to bind Zn 2+ ions by passing mouse serum through agarose columns bound to Co 2+ , Ni 2+ , or Zn 2+ , or metal-free as a negative control. Western blot of sample elution from each column purification shows that SELENOP1 binds to Co 2+ , Ni 2+ , and Zn 2+ , but not to the metal-free agarose column (Figure 1C). When serum from SELENOP1 −/− mice was used, SELENOP1 immunoreactivity was not detected in eluent from any of the columns. This demonstrates that SELENOP1 is capable of binding different biometals, with a greater amount of binding to Zn 2+ compared with Co 2+ and Ni 2+ .

Elevated Levels of Intracellular Zn 2+ in SELENOP1 -/-Hippocampus
To explore the effects of SELENOP1 on brain Zn 2+ homeostasis, we first evaluated intracellular Zn 2+ levels in SELENOP1 −/− mice using histological methods. Our studies focused on the hippocampus as a zinc-rich region important for learning and memory (33). The Timm-Danscher method (34) revealed a pronounced increase in histologically-detectable Zn 2+ in the hippocampus of SELENOP1 −/− compared with SELENOP1 +/+ hippocampus (Figure 2A). The hilar and CA3 mossy fiber regions had the most intense Zn 2+ labeling in both wild-type and SELENOP1 −/− animals. However, all layers positive for Zn 2+ were increased in KOs, including the stratum radiatum and stratum oriens of the CA1 and CA3 layers.
We further compared bioreactive Zn 2+ levels in SELENOP1 −/− and wildtype mouse hippocampi using TSQ labeling. Quantitation of TSQ fluorescence also revealed significantly higher levels of intracellular Zn 2+ in the CA1 stratum oriens, stratum radiatum, and CA3 mossy fibers of the SELENOP1 −/− animals compared to their control (Figures 2B,C). We also observed a sex difference in CA3, with increased Zn 2+ in mossy fibers in female compared to male hippocampus regardless of genotype. Unstained hippocampal brain sections showed no fluorescence at TSQ wavelengths, indicating that differences were not due to autofluorescence signals. Hematoxylin staining TABLE 1 | Metal content in brain and liver of Sepp1 wildtype (Sepp1 +/+ ) and knockout (Sepp1 −/− ) mice at the age of 3 months (mg/kg of dry tissue, n = 20 each) measured by ICP-OES.  of sections previously used for TSQ labeling showed no morphological differences between SELENOP1 wildtype and KO hippocampi. Based on our findings, SELENOP1 may be regulating Zn 2+ levels directly or through one or more Zn 2+ interacting proteins. As we observed increased levels of intracellular Zn 2+ within the hippocampus of SELENOP1 −/− mice, we then investigated if the total brain Zn 2+ levels are altered by subjecting whole brain hemispheres via ICP-OES to measure total metal content for Zn 2+ , Cu 2+ , and Fe 2+ . There was no significant increase in total brain Zn 2+ levels in SELENOP1 −/− mice ( Table 1). This may indicate that deletion of SELENOP1 results in changes to the distribution of Zn 2+ within the brain rather than an increase in total brain Zn 2+ .

Expression Levels of Zn 2+ -Interacting Proteins
We investigated whether changes in Zn 2+ regulating proteins could explain differences in Zn 2+ . We investigated hippocampal expression of the Zn 2+ transporters ZnT1 and the vesicleassociated ZnT3, as well as the metal storage protein MT3. Western blot indicated that ZnT1 and ZnT3 proteins were unchanged in SELENOP1 −/− animals (Figures 3A,B). However, MT3 protein expression was increased (Figure 3C), suggesting Fluorescence in SELENOP1 −/− hippocampal slices was significantly reduced compared to WT animals. Statistical analysis was performed by Student's unpaired t-test comparing the mean area under the curve for each genotype between 60 and 120 s, *p < 0.05, **p < 0.01. that deletion of the SELENOP1 gene does not affect zinc transport through these major pathways, but rather upregulates expression of the Zn 2+ storage protein. The enhanced expression of MT3 may be a result of a feedback mechanism in response to increased Zn 2+ levels in the SELENOP1 −/− hippocampus.

Zn 2+ Release Is Impaired in SELENOP1 -/-Hippocampus in Response to Neuron Depolarization
Most histologically-detectable Zn 2+ in hippocampal neurons is vesicular (25). We investigated if the increased Zn 2+ levels visualized by TSQ or Timm's staining in SELENOP1 −/− mice results in increased extracellular accumulation following depolarizing stimuli designed to promote vesicular release. We imaged extracellular Zn 2+ accumulation in hippocampal slices with a selective cell-impermeant fluorescent Zn 2+ indicator, Fluozin-3, in the extracellular media. By depolarizing hippocampal cells with the addition of 35 mM KCl, we noticed an increase in fluorescence in slices from wild type mice, indicating release of Zn 2+ into the extracellular space (Figures 4A,B). In contrast, we observed a minimal increase in fluorescence in hippocampal slices from SELENOP1 −/− mice, indicating negligible Zn 2+ release in the SELENOP1 −/− hippocampus.
A high background fluorescence recording from Zn 2+ contaminants in the reagents used could possibly mask Zn 2+ release. To reduce background fluorescence, we also imaged slices with the Zn 2+ chelator, Ca 2+ -EDTA, added to the ACSF. Ca 2+ -EDTA enables removal of basal levels of extracellular Zn 2+ , but its slow kinetics do not prevent the detection of synaptically-evoked Zn 2+ accumulation (35). Even in the presence of Ca 2+ -EDTA, we still observed significantly larger FIGURE 5 | Intracellular Zn 2+ levels are increased by 0 Se and GPX4 inhibitor RSL-3. (A) Relative FluoZin-3 fluorescence of SH-SY5Y cells. Cells exposed to DTDP to release zinc showed increased fluorescence, while cells exposed to PTEN to reduce intracellular zinc had no change in fluorescence. (B) FluoZin-3 fluorescence was increased in cells grown in media with 0 nM Se relative to control media (10 nM), while 100 nM Se did not alter intracellular zinc. (C) Hydrogen peroxide but not GPX1 inhibitor mercaptosuccinate (MCS) increase intracellular zinc. (D) The GPX4 inhibitor RSL-3 increased intracellular zinc, which was reversed when RSL-3 was added with α-tocopherol. Statistical analysis is performed by one-way ANOVA followed by Dunnett's multiple comparison test. *p < 0.05, ***p < 0.005, ****p < 0.0001.

Selenium Deficiency Releases Intracellular Zn 2+
We hypothesized that increased oxidation from Se-deficient conditions could result in reduced intracellular chelation of Zn 2+ . We tested this by measuring the fluorescence of cellpermeable FluoZin-3 in cultured SH-SY5Y cells. As shown in Figure 5A, release of intracellular zinc by DTDP, a thiol oxidizer which liberates intracellular Zn 2+ , significantly increased FluoZin-3 fluorescence. However, the intracellular Zn 2+ chelator TPEN did not reduce FluoZin-3 fluorescence, indicating that free non-chelated Zn 2+ levels were too low to be detectable. We also found that cells grown in 0 Se culture media had significantly increased fluorescence compared with cells grown in our baseline media with 10 nM Se (Figure 5B). Oxidation with H 2 O 2 greatly increased FluoZin-3 fluorescence, demonstrating that oxidation could free Zn 2+ from stores. To test if a reduction in activity of the H2O2-reducing selenoprotein GPX1, which depends on Se for synthesis, could lead to increased intracellular Zn 2+ , we exposed cells to the GPX1 inhibitor mercaptosuccinate (MCS) (Figure 5C). MCS did not alter fluorescence, however, we found that RSL-3, an inhibitor of the phospholipid peroxidase selenoprotein GPX4, increased intracellular Zn 2+ (Figure 5D). This increase was prevented by the vitamin E compound α-tocopherol, which reduces lipid peroxidation. These findings suggest that a deficiency in Se can increase free intracellular Zn 2+ by causing a reduction in GPX4, which results in increased oxidation of lipids.

Elevated Tau Phosphorylation in SELENOP1 Knockout Hippocampus
Zn 2+ promotes tau phosphorylation leading to neurofibrillary tangle formation (18). We questioned whether tau phosphorylation could be altered in the SELENOP1 −/− mouse. We performed western blot analysis to compare specific pTau sites to total tau protein. We found that phosphorylation at threonine 231 and at serine 396 were significantly increased in SELENOP1 −/− mice ( Figure 6A). However, phosphorylation at the serine 214 site was unchanged (Figures 6B,C). Deletion of SELENOP1 −/− thus results in a site-specific increase in tau phosphorylation.

DISCUSSION
Our findings show that deletion of SELENOP1 increased free (chelatable) intracellular Zn 2+ levels. However, release of synaptic zinc was impaired. Although the Zn 2+ storage protein MT3 was elevated without change to the zinc transporters ZnT1 or ZnT3, we showed that selenium deficiency could induce release of Zn 2+ from stores, likely through a decrease of the selenoprotein GPX4 and a subsequent increase in lipid peroxidation. Interestingly, GPX4 helps to prevent neurodegeneration though ferroptosis (36). Lastly, we demonstrated an increase in site-specific phosphorylation of tau. These findings suggest that SELENOP1 plays a role in regulating storage of intracellular Zn 2+ . This role may be important for preventing tau hyperphosphorylation in AD.
Dietary supplementation with selenium in the form of sodium selenate reduces tau phosphorylation to potentially reduce neurofibrillary tangle formation (37,38). Selenate can act as an agonist for protein phosphatase 2A (PP2A), which targets tau phosphorylation (39). Interestingly, Zn 2+ is an inhibitor of PP2A, and may promote neurofibrillary tangle formation (18). Hyper-phosphorylation of tau leads to neurofibrillary tangle formation, and de-phosphorylation by PP2A should reduce tangle formation. However, dietary selenate supplementation also upregulates the expression of selenoproteins (40,41). We have previously reported that a reduction in selenoprotein S can promote tau phosphorylation (42). In a type 2 early clinical trial for selenate supplementation in Alzheimer's disease, selenate could increase brain selenium in some patients, and the increase in brain selenium correlated with lack of decline in performance on the Mini-Mental Status Examination (MMSE) (43). Thus, brain selenoproteins, including SELENOP1, may be important for preventing Alzheimer's pathology.
Zn 2+ metabolism is altered in AD, resulting in abnormally enriched Zn 2+ environments within the AD brain (44). Zincbinding sites on the Aβ peptide result in Zn 2+ mediated aggregation of Aβ and amyloid plaque formation. Furthermore, both the neuroprotective role of SELENOP1 against Aβ toxicity and the role of Zn 2+ in protecting the cell against oxidative damage, could be working together to reduce the levels of Aβ stress observed in AD pathology, however more studies need to be done to further elucidate their contribution to alleviating oxidative stress.
We did not observe an increase in total brain Zn 2+ levels with ICP-OES, suggesting only changes in local Zn 2+ distribution.
The absence of SELENOP1 could result in increased oxidative stress in the brain and lead to Zn 2+ release from MT3 (45), possibly inducing an upregulation of the Zn 2+ storage protein. In the absence of SELENOP1, we observed a seemingly paradoxical impairment of Zn 2+ release despite an overall increase in intracellular chelatable Zn 2+ . This was surprising, since most chelatable Zn 2+ is thought to be localized to synaptic vesicles (46). Overexpression of MT3 in SELENOP1 −/− mice may affect the subcellular distribution of Zn 2+ by limiting the amount of free Zn 2+ available for loading into the synaptic vesicles. The protein Reelin can increase release of a subset of synaptic vesicles (47). This increase is dependent on Reelin binding to ApoER2. SELENOP1 is another ligand for ApoER2 (48,49), and thus may also modulate vesicle release, possibly including zincergic vesicles. A decrease in release of Zn 2+ vesicles could result in a "back-up" of these vesicles, which could contribute to the observed increase in chelatable Zn 2+ . The APOE ε4 allele of the major ligand ApoE for ApoER2, increases risk of Alzheimer's disease (50). Interestingly, the APOE ε4 allele is also associated with decreased selenium in the brain (51), again suggesting a possible role for selenium and selenoproteins in preventing Alzheimer's disease.
We found an interesting sex difference in Zn 2+ levels in the CA3 region of the hippocampus (Figure 2). Female mice generally had higher Zn 2+ levels but with some variability. Previous studies have shown higher Zn 2+ levels in female wildtype mice and AD mouse models (52,53), although these studies did not agree on the age of sex differences. The differences suggest Zn 2+ as a possible explanation for the increased risk of AD in women (54). Researchers recently discovered that estrogen increases hippocampal Zn 2+ , which was cycledependent in female mice (55,55), which may explain the increased Zn 2+ in other studies as well as the variability of the current results.
SELENOP1 contains a putative metal-binding domain that can potentially bind Zn 2+ with a high affinity. Our results demonstrate that SELENOP1 is capable of binding Zn 2+ as well as Co 2+ and Ni 2+ . However, our finding that selenium deficiency and inhibition of GPX4 suggest that the absence of selenium and loss of antioxidant selenoprotein function in the SELENOP1 KO mice is responsible for the increased Zn 2+ levels. The reason for the presence of the functional metal-binding domain of SELENOP1 remains unknown. It is possible that Zn 2+ binding could alter the affinity of SELENOP1 for the ApoER2 receptor, allowing for regulation of Se by excess Zn 2+ . Additionally, the homology of the Zn 2+ binding domain with metalloproteases such as ADAM10 and neprilysin (Figure 1) opens the possibility of an enzymatic role in proteolysis. ADAM10 has a protective role in Alzheimer's disease as a putative α-secretase that promotes non-amyloidogenic cleavage of amyloid precursor protein, preventing amyloid beta formation (56).
The data presented here indicate that SELENOP1 may play a crucial role in the maintenance of brain Zn 2+ . The SELENOP1 gene can affect Zn 2+ metabolism and synaptic release from neuronal synapses. Zn 2+ is increased in Alzheimer's disease and interacts with amyloid beta (17), and can also promote tau phosphorylation (18). Thus, Zn 2+ -binding properties of SELENOP1 could contribute to the association of SELENOP1 with amyloid beta plaques in Alzheimer's disease (8). The SELENOP1-Zn 2+ interaction has potentially important implications in neuronal function and synaptic physiology.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The animal study was reviewed and approved by University of Hawaii IACUC.