Advertisement
Research Article
Transparent Process
Open Access

Human genetic variants in SLC39A8 impact uptake and steady-state metal levels within the cell

Wen-An Wang, Andrea Garofoli, Evandro Ferrada, Christoph Klimek, Barbara Steurer, Alvaro Ingles-Prieto, Tanja Osthushenrich, Aidan MacNamara, Anders Malarstig, Tabea Wiedmer, Giulio Superti-Furga  Correspondence email
Published 30 January 2025. DOI: 10.26508/lsa.202403028

Abstract

The human SLC39A8 (hSLC39A8) gene encodes a plasma membrane protein SLC39A8 (ZIP8) that mediates the specific uptake of the metals Cd2+, Mn2+, Zn2+, Fe2+, Co2+, and Se4+. Pathogenic variants within hSLC39A8 are associated with congenital disorder of glycosylation type 2 (CDG type II) or Leigh-like syndrome. However, numerous mutations of uncertain significance are also linked to different conditions or benign traits. Our study characterized 21 hSLC39A8 variants and measured their impact on protein localization and intracellular levels of Cd2+, Zn2+, and Mn2+. We identified four variants that disrupt protein expression, five variants with high retention in the endoplasmic reticulum, and 12 variants with localization to the plasma membrane. From the 12 variants with plasma membrane localization, we identified three with complete loss of detectable ion uptake by the cell and five with differential uptake between metal ions. Further in silico analysis on protein stability identified variants that may affect the stability of homodimer interfaces. This study elucidates the variety of effects of hSLC39A8 variants on ZIP8 and on diseases involving disrupted metal ion homeostasis.

Introduction

Metal ions are vital for ∼80% of all cellular biochemical processes and play a role structurally and chemically in over a third of the organisms’ proteome (1, 2). Many of these are essential metal ions that act as cofactors to a variety of proteins involved in metabolic pathways and transcriptional processes. Despite the importance of essential metals, their excessive concentration can result in cellular toxicity (3, 4, 5). Therefore, it is crucial to maintain metal ion homeostasis through catalysis, compartmentalization, and transport, as both deficiencies and overexposures have implications for health and disease.

The homeostasis of metal ions within cells is maintained by various proteins, such as the zinc-responsive (Zrt)– and iron-responsive transporter (Irt)–like protein (ZIP) family, a group of 14 cation transporters responsible for the transport of a variety of metal ions across cellular membranes (6, 7). Specifically, the human SLC39A8 (hSLC39A8) gene encodes SLC39A8 (ZIP8), a plasma membrane (PM) transporter that is responsible for the influx of a variety of divalent cations into the cell. ZIP8 is broadly expressed across various tissues (https://www.proteinatlas.org/ENSG00000138821-SLC39A8/tissue), but is most highly expressed in the lung, duodenum, kidney, endometrium, placenta, breast, and tonsil. In addition, ZIP8 is present in specific cell types involved in the immune response within the central nervous system (8, 9, 10). The initial discovery of ZIP8 involved its role as a transporter of zinc (Zn2+), but it is now known to play a role in the cellular uptake of a broader spectrum of divalent metals, including cadmium (Cd2+), manganese (Mn2+), iron (II) (Fe2+), and cobalt (Co2+) (6, 11, 12). Specifically, ZIP8 was proposed to be an electroneutral cotransporter of a divalent metal ion with two bicarbonate anions (HCO3−)2 or one bicarbonate anion (HCO3−) and one selenite anion (HSeO3−) (13, 14). The uptake of HSeO3− into the cell therefore also means that ZIP8 is an importer of selenium (Se4+) (14).

Although the differential substrate affinity of ZIP8 and the metabolic consequences are complex and require further investigation, the functional transport of Mn2+ is known to play a critical role in human pathophysiology. It has been shown that Mn2+ is an essential cofactor for an array of enzymes, participating in many physiological processes, and is particularly important in the function and development of dopaminergic neurons (15, 16). Disruptions in Mn2+ homeostasis have been associated with several diseases. Specifically, missense mutations in the hSLC39A8 gene have been linked to two severe conditions associated with Mn2+ deficiency: Leigh-like syndrome and congenital disorder of glycosylation (CDG). Leigh syndrome is an early-onset progressive neurodegenerative disorder characterized by defects in mitochondrial energy production caused by decreased activity of the key mitochondrial enzyme manganese superoxide dismutase (17, 18, 19). CDG is a severe developmental disorder linked to a defect in the processing of protein-bound glycans. CDG type 2 (CDG type II) is the most common type associated with hSLC39A8 genetic alterations and involves manganese-dependent β-1,4-galactosyltransferase, an enzyme essential for protein glycosylation (20, 21). Moreover, reports on additional genetic variants implicate the transporter in several other diseases. The missense variant Ala391Thr (rs13107325) has been associated with extensive pleiotropic effects across several biological traits including blood pressure, body mass index, serum levels of N-terminal pro-B-type natriuretic peptide, total cholesterol, and triglyceride levels (10, 22). Furthermore, Ala391Thr is associated with several neuropsychiatric conditions such as schizophrenia and Parkinson’s disease (10, 23), as well as cardiovascular diseases, liver disorders, and Crohn’s disease (24, 25, 26). The extensive pleiotropic effects of this alteration may be attributed to the diverse cellular roles of Mn2+, the wide range of substrates transported by ZIP8, and the combined role of the transporter within specific tissues. In addition to the transport of Mn2+, the cellular uptake of Zn2+ by ZIP8 is critical for immune cell function and it has been shown that Zn2+ plays a pivotal role in the myeloid-mediated immune response to pneumococcal pneumonia and in lung epithelial cells during tumour necrosis factor–induced cytotoxicity (27, 28). Although the significance of hSLC39A8 in both health and disease is widely recognized, the precise mechanism by which genetic variants cause the loss of protein function, and subsequently how this loss affects metal ion transport leading to disease, remains unclear.

The aim of this work was to study the effect of hSLC39A8 genetic variants on the protein’s impact on metal ion levels within the cell. We selected 21 genetic variants of the hSLC39A8 gene, based on their relevance to human disease, and variant effect predictions (VEPs). We characterized the effect of these variants experimentally on protein expression, cellular localization, and cellular metal ion uptake, and computationally on dimerization stability. Ultimately, our goal is to highlight the impact of these mutations on the diverse spectrum of human diseases, which has the potential to hint at additional therapeutic avenues for hSLC39A8 pathologies.

Results

Selection of variants for functional characterization

The selection of the variants to analyse in the study required a thorough survey and careful choice of different properties. Our study focused on the canonical isoform of the human SLC39A8 (hSLC39A8) gene (GenBank Accession NM_001135146.2), which is composed of nine exons, is ubiquitously expressed, and encodes a transmembrane protein spanning 460 amino acids. We performed a literature review and compiled several hSLC39A8 genetic variants from publicly available databases to assess their clinical and biological relevance (see the Materials and Methods section). As of January 2024, our search leads to 232 reported missense variants; of these, a large majority were unclassified variants, four were reported as pathogenic/likely pathogenic, four had conflicting interpretations of pathogenicity, five variants were classified as benign/likely benign, and 24 were classified as variants of uncertain significance (VUS) (https://re-solute.eu/knowledgebase/gene/SLC39A8). Next, we performed in silico methods, by compiling publicly available VEP to assess the pathogenicity of all possible missense variants within the hSLC39A8 coding region (Table 2).

For the final list of 21 variants that we chose, we incorporated several considerations (Tables 1 and 2). We included an early termination variant, Glu31* (V1), to serve as a loss-of-function genetic alteration control; 11 variants (V2-6, V8, V9, V13, V14, V16, and V19) that were reported from patients diagnosed with either CDG type II or Leigh-like disease; two variants (V7 and V21) reported with statistical association with non-pathogenic phenotypes; five VUS (V10-12, V15, and V17) with a potential pathogenic role based on the VEP scoring; and, lastly, two variants, Asn389Ser (V18) and Ala391Ser (V20), chosen because of their spatial proximity to the highly pleiotropic variant Ala391Thr (V19). To investigate the effects that the above-selected variants have on ZIP8 function, we generated stable HEK293 cell lines with the doxycycline-inducible overexpression of WT or variant ZIP8 (HEK SLC39A8WT and HEK SLC39A8V1-V21) (Fig 1A).

View this table:
Table 1.

Selection of 21 variants of the hSLC39A8 gene for functional characterization.

View this table:
Table 2.

Summary of results for the 21 variants of the hSLC39A8 gene characterized within this study.

Figure 1.
Figure 1. Subcellular localization of ZIP8 variants.

(A) Workflow schematic of the study on 21 single variants of SLC39A8, highlighted in black on the topology of the SLC39A8/ZIP8 transporter. (B) Representative immunofluorescence images of HEK SLC39A8WT cells stained with anti-HA (αHA) antibody for ZIP8, cell mask for plasma membrane (PM), and anti-calreticulin (αCALR) antibody for the endoplasmic reticulum (ER). Representative of 81–135 images. (C) Quantification of subcellular localization of ZIP8WT and ZIP8V1-21. A Pearson correlation coefficient was calculated between the immunofluorescence signals of each respective variant as detected by anti-HA antibody versus the PM (cell mask) or ER (anti-calreticulin) markers. Bars show the standard error of the mean over at least four replicates. Variants are grouped based on cluster formation with the colour indicating the impact of the variant on protein expression and localization. Representative images are presented in the right panel to display the different impacts on protein expression and localization (WT, ER, no expression).

Variant impact on cellular expression and localization

The ZIP8 transporter is predominately located at the PM, where it functions as an importer of divalent ions into the cell (8, 11). Mutations within the hSLC39A8 gene that result in a non-functional transporter may also derive from its effect on cellular expression or PM localization. Therefore, as an initial assessment, we characterized the expression and subcellular localization of WT and variant ZIP8 through immunofluorescence imaging of HEK SLC39A8WT and HEK SLC39A8V1-V21 cells (Fig 1A). WT and variant ZIP8 were visualized with the C-terminal HA tag and costained with markers for the nucleus, endoplasmic reticulum (ER), and PM (Figs 1B and S1). We further quantified the intensity of signal overlap between ZIP8 and the PM or between ZIP8 and the ER into Pearson’s correlation coefficient (rPM and rER) for each variant (Fig 1C). As expected, ZIP8WT colocalized adequately with both the PM (rPM = 0.56 ± SD = 0.10) and the ER (rER = 0.68 ± SD = 0.09). We arbitrarily classified these variants into three categories based on the clustering of rPM versus rER (Fig 1C): WT-like PM localization, high ER retention (relative to WT), and no expression. As ZIP8V1 is an early truncation mutation, it lacks the C-terminal HA tag for immunofluorescence detection and its localization was marked as inconclusive because of the lack of an appropriate conclusion. Based on the results, which we complemented with immunoblot analysis (Fig S2), we found three variants (V2, V5, and V19) that resulted in a lack of ZIP8 expression (Fig S1A), five variants (V3, V4, V6, V12, and V14) that lead to high ER retention (Fig S1B), and 12 variants (V7-11, V13, V15, V16-18, V20, and V21) that displayed PM localization comparable to WT (Fig S1C and D).

Figure S1.
Figure S1. Subcellular localization of ZIP8 variants.

Representative immunofluorescence images of HEK SLC39A8WT and HEK SLC39A8V1-21 cells stained with anti-HA (αHA) antibody, cell mask (PM), anti-calreticulin (ER) antibody, and DAPI (nuclear) for ZIP8. Representative of 81–135 images for WT and 4–20 images for variants. The displayed scale bars are 25 μm in length. (A) Images of variants with disrupted expression (V1, V2, V5, and V19). (B) Images of variants with high ER retention (V3, V4, V6, V12, and V14). (C) Images of variant at the PM with WT-like activity (WT, V9, V15, V20, and V21). (D) Images of variant at the PM with disrupted transport activity (V7, V8, V10, V11, V13, V16, V17, and V18).

Figure S2.
Figure S2. Immunoblot analysis of ZIP8 variants.

Representative immunoblots of HEK SLC39A8WT and HEK SLC39A8V1-21 cell lysates probed with anti-HA for SLC39A8 and anti-GAPDH. Cells were induced for ZIP8 protein expression for 24 h, and two biological replicates were performed per variant. *indicates ZIP8 monomers (∼75 kD), and **indicates ZIP8 dimers (∼200 kD).

Variant impact on the cellular uptake of Cd2+, Zn2+, and Mn2+

The differential affinities of ZIP8 for the different metal substrates that it transports have been characterized in several studies (6, 34). Affinities for human ZIP8 have been reported for Cd2+ (Km = ∼1.1 ± 0.2 μM), Mn2+ (Km = ∼1.44 ± 0.39 μM), Zn2+ (Km = ∼1.3 ± 0.3 μM), and Fe2+ (Km = ∼23 ± 4.6 μM) (6, 34). To start the evaluation of the impact of ZIP8 mutations on the cellular uptake of metal ions, we first assessed Cd2+ for its high affinity to ZIP8 (6, 12, 35). We used the FLIPR Calcium-5 Assay kit to measure Cd2+ uptake (36, 37) and performed an initial CdCl2 titration in HEK SLC39A8WT cells. We assessed an EC50 of 2.64 μM (95% Cl: 2.13, 3.44) for the HEK SLC39A8WT cells (Fig 2A). Next, we measured Cd2+ uptake in HEK SLC39A8WT and HEK SLC39A8V1-V21 cells, using a set CdCl2 concentration of 2 μM, based on the EC50 (Fig 2B). Of the 21 variants, nine (V7, V9-11, V15, V17, V18, V20, and V21) exhibited WT-like cellular uptake of Cd2+ and the remaining 12 variants (V1-6, V8, V12-14, V16, and V19) did not show any measurable response (Fig 2B, and Table 2). As expected, although most variants with PM localization showed detectable Cd2+ uptake, all variants resulting in no ZIP8 expression or high ER retention did not exhibit a response. Notably, only three variants with PM localization failed to show measurable Cd2+ uptake (Table 2).

Figure 2.
Figure 2. Transport activity of ZIP8 variants.

(A) (Left) Representative recordings of FLIPR Calcium-5 dye fluorescence in response to CdCl2 (0–64 μM) addition in HEK SLC39A8WT cells. The fluorescence signal was normalized to basal, before substrate addition (FΔ/F). (Right) The area under the curve (AUC) from the recordings (left) is averaged and plotted against the CdCl2 concentrations to determine an EC50 of 2.64 μM (95% Cl: 2.13, 3.44). Average of two biological replicates. (B) Cadmium transport activity in HEK SLC39A8WT and HEK SLC39A8V1-21 cells. The activity defined by the area under the curve (AUC) from the variant-expressing cells was normalized to that of HEK SLC39A8WT cells induced with doxycycline (WT+DOX). Green indicates WT-like activity, and grey indicates no activity. At least three biological replicates were performed per variant. An ordinary one-way ANOVA test was used to determine statistical significance, and multiple comparisons were analysed against WT−DOX (grey) and WT+DOX (green) (ns, not significant; ****P < 0.0001). (C) Representative fluorescence (FITC 488 nm) signal of Zinpyr-1 in response to ZnCl2 (0–100 μM) in addition to WT+DOX and WT−DOX cells. Mean fluorescence intensities are indicated on the left of the plot, and the dotted line indicates the mean intensity of WT+DOX cells treated with 0 μM of ZnCl2. Representative of two biological replicates. (D) Zinc transport activity in HEK SLC39A8WT and HEK SLC39A8V1-21 cells. The activity defined by median fluorescence (FMedian) from the variant-expressing cells was normalized to that of HEK SLC39A8WT cells induced with doxycycline (WT+DOX). Green indicates WT-like activity, light green indicates decreased activity, and grey indicates no activity. Three biological replicates were performed per variant. An ordinary one-way ANOVA test was used to determine statistical significance, and multiple comparisons were analysed against WT−DOX (grey) and WT+DOX (green) (ns, not significant; *P < 0.0332; **P < 0.0021; ***P < 0.0002; ****P < 0.0001). (E) Per cent survival (%) of HEK SLC39A8WT cells, induced and non-induced with doxycycline, in response to MnCl2 (0–1 mM) treatment for 48 h. Bars show the standard error of the average over two to four biological replicates. (F) Manganese transport activity in HEK SLC39A8WT and HEK SLC39A8V1-21 cells. The transport activity defined by the percentage of cell death is graphed. Green indicates WT-like activity, light green indicates decreased activity, and grey indicates no activity. Three biological replicates were performed per variant. An ordinary one-way ANOVA test was used to determine statistical significance, and multiple comparisons were analysed against WT−DOX (grey) and WT+DOX (green) (ns, not significant; *P < 0.0332; **P < 0.0021; ***P < 0.0002; ****P < 0.0001).

To determine whether variants with PM localization had an impact on substrate specificity, we measured the cellular uptake of Zn2+ and Mn2+ for the 12 WT-like PM localization variants (V7-11, V13, V15, V16, V17, V18, V20, and V21). As a negative control, we included the early termination variant Glu31* (V1). The cellular uptake of Zn2+ was measured using a Zn2+-sensitive cytoplasmic Zinpyr-1 dye (38, 39). To identify the best signal window of the Zn2+ uptake between ZIP8 variants, we measured the fluorescence of Zinpyr-1 in induced and uninduced HEK SLC39A8WT cells exposed to previously reported ZnCl2 concentrations (40, 41) and determined 100 μM ZnCl2 to be a suitable concentration (Fig 2C). As expected, the early truncation Glu31* (V1) variant did not result in Zn2+ uptake (Fig 2D). Of the remaining 12 PM localizing variants tested, seven (V9-11, V15, V20, and V21) showed WT-like uptake, whereas three variants (V8, V13, and V16) did not have a measurable response to ZnCl2, and two variants (V7 and V18) had diminished uptake activity (Fig 2D and Table 2).

For the measurement of the Mn2+ uptake by the cell, we leveraged the cytotoxic properties of Mn2+ overexposure to establish a cell viability assay as an indirect approach (3, 42). We first performed a MnCl2 titration in induced and uninduced HEK SLC39A8WT cells and measured the percentage of cell survival after 48 h. We determined an LC50 of 0.048 μM (95% Cl: 0.04 to 0.07) for HEK SLC39A8WT cells (Fig 2E). Next, we measured the survival of the variant-expressing cells exposed to several MnCl2 concentrations and recorded the percentage of cell death at 0.1 μM of MnCl2, a concentration selected for the best signal window for the Mn2+ cell viability assay between ZIP8 variants (Fig 2F). In congruence with the Zn2 uptake assay, the early truncation Glu31* (V1) variant did not show Mn2+ uptake activity (Fig 2F and Table 2). Also consistent with the Zn2+ uptake data, four variants (V9, V15, V20, and V21) were associated with WT-like uptake activity, whereas three variants (V8, V13, and V16) showed no response, and two variants (V7 and V18) showed decreased uptake (Fig 2F and Table 2). In contrast to the Zn2+ uptake data, three variants (V10, V11, and V17) with WT uptake for Zn2+ showed diminished uptake for Mn2+ (Fig 2F and Table 2).

ICP-MS and Zn2+ and Mn2+ competition

To validate key findings from the uptake assays of Cd2+, Zn2+, and Mn2+, we performed inductively coupled plasma mass spectrometry (ICP-MS) for a set of variants (V1, V7, V9, V10, V11, V17, V18, and V19) (Fig 3A–C). Consistent with the uptake results, we saw increased accumulation of intracellular Cd2+, Zn2+, and Mn2+ for V7, V9, V10, V11, V17, and V18, and not for V1 and V19 (Fig 3A–C). Notably, there were some discrepancies between the uptake assays and the ICP-MS results. Specifically, we did not see a decreased accumulation of Zn2+ and Mn2+ for V7 and V18, there was a decrease for all ions tested in V10, and there was a decrease in Zn2+ but not in Mn2+ for V17 (Fig 3A–C, Table 2).

Figure 3.
Figure 3. ICP-MS and competitive transport between Zn2+ and Mn2+ of ZIP8 variants.

(A, B, C) Levels of intracellular Cd2+, Zn2+, and Mn2+ were measured with ICP-MS in HEK SLC39A8WT and HEK SLC39A8V1-21 cells after 48 h of doxycycline induction and 24 h of incubation with 0.5 μM CdCl2. Green indicates WT-like activity, light green indicates decreased activity, and grey indicates no activity. Three biological replicates were performed per variant. An ordinary one-way ANOVA test was used to determine statistical significance, and multiple comparisons were analysed against WT+DOX (green) (ns, not significant; *P < 0.0332; **P < 0.0021; ***P < 0.0002; ****P < 0.0001). (D) Zinc and manganese competitive transport in HEK SLC39A8WT and HEK SLC39A8V1-21 cells. The concentration of MnCl2 needed to inhibit 50% of zinc transport (IC50) is graphed. Green indicates WT-like, dark green indicates higher than WT, and grey indicates less than WT. Three biological replicates were performed per variant. An ordinary one-way ANOVA test was used to determine statistical significance, and multiple comparisons were analysed against induced (green) ZIP8WT expression (ns, not significant; **P < 0.0021; ****P < 0.0001).

Our uptake assays revealed three variants with WT-like cellular uptake for Zn2+ but a lower cellular uptake for Mn2+ (Fig 2D and F and Table 2). We therefore explored the impact of variants on substrate specificity between Zn2+ and Mn2+. We performed a competition assay using Zinpyr-1 to measure Zn2+ uptake with a set of 40 μM ZnCl2, a concentration that provided the best signal window with the addition of MnCl2, and a titration of MnCl2 to determine the concentration of Mn2+ required to interfere with Zn2+ uptake. We selected all the variants with WT or decreased transport activity for both Zn2+ and Mn2+ and included a PM localizing, no response variant (V16) as a negative control. We quantified the IC50 of MnCl2 for each variant and compared it with the IC50 values obtained in uninduced HEK SLC39A8WT cells (364.9 ± 216.9) and induced HEK SLC39A8WT cells (798.9 ± 159.4) (Fig 3D). As anticipated, V16 exhibited an IC50 value of MnCl2 statistically similar to that observed with no ZIP8 expression. MnCl2 IC50 values from eight variants (V7, V9, V10, V15, V17, V18, V20, and V21) showed no statistical difference compared with ZIP8WT, suggesting that these variants did not alter substrate specificity (Fig 3D and Table 2). In contrast, V11 displayed a significantly higher IC50 than ZIP8WT, indicating increased uptake of Zn2+ over Mn2+ (Fig 3D and Table 2). To further elucidate the mechanistic basis for the observed differences, we investigated the effects of these variants on the structure of ZIP8, specifically in the WT dimer configuration.

Structure prediction of ZIP8 suggests homodimerization

Several members of the ZIP family, including ZIP8, form multimers (43, 44, 45, 46). To investigate whether variants affect multimer formation, we used AlphaFold to predict and compare the structures of ZIP8 as a monomer versus a dimer, and as a dimer versus a trimer. Each modelling step was performed five times, and structural quality was assessed using in silico stability predictions and protein–protein interaction scoring methods. Building upon previous findings (8, 46, 47), our models revealed that the dimer exhibits significantly greater stability in comparison with the monomer configuration. In addition, we used a protein–protein interaction score (pDockQ score) that accounts for the quality of the complex (48, 49). Such a score favoured the dimer over the trimer model. Moreover, the extracellular domain of the dimeric structure revealed high structural similarity to an experimentally characterized ZIP4 dimer (50). We therefore propose the most probable WT configuration of the functional ZIP8 protein of dimeric nature (Fig 4A).

Figure 4.
Figure 4. Structure model of SLC39A8 dimer and mutations critical for dimer stability.

(A) AlphaFold model of the SLC39A8 dimer. Monomers are shown in blue or green. Strongly destabilizing residues in the first percentile of the stability distribution are shown in red. (B) Comparison of thermodynamic changes caused by single mutations in the monomer versus the dimer (inset). Distribution of single mutation effects on dimer stability, calculated as described in Equation (1). The upper fifth percentile contains 442 most destabilizing substitutions distributed over 105 positions. The first percentile contains 87 most destabilizing single amino acid substitutions distributed over 17 sites (highlighted in red). (C) Dimer-destabilizing mutations distribute in clusters. We plotted the upper fifth percentile of the most destabilizing mutations over the length of the protein detecting a heterogeneous distribution of variants per position.

The dimer formation suggests that some mutations that can be mapped within the interaction interface of the two ZIP8 monomers may affect stability of the dimer and lead to a non-functional transporter. To identify mutations of this type, we used in silico estimations of changes in thermodynamic stability upon a single amino acid substitution in the dimer versus the monomer configuration (Fig 4B). We reasoned that mutations causing a strong change in stability in the dimer, but not in the monomer, could reveal amino acids essential for the stability of the dimer. Using this strategy, we explored the upper fifth percentile of the most destabilizing amino acid substitutions, which amounted to 442 mutations with a ΔΔG larger than 2.4 kcal/mol, distributed over 105 positions. The analysis revealed clusters of positions more prone to dimer destabilization (Fig 4C). We found that at residues 29–36, 101–110, 193–209, 306–311, 344–351, and 397–409, as many as 5 to 10 different amino acid substitutions were destabilizing. Although none of our selected variants were the exact mutations in the upper fifth percentile list, a few variants were located within the destabilizing clusters. In particular, these included V1 (Glu31*) and V2 (Val33Met) in the first 29–36 cluster; V7 (Asp193Asn), V8 (Phe203Ser), and V9 (Gly204Cys) in the 193–209 cluster; V11 (Thr308Met) in the 306–311 cluster; and V16 (Gly350Arg) in the 344–351 cluster. Furthermore, V3 (Gly38Arg) was two residues away from Val36, with a peak number of eight destabilizing mutations, and V11 (Thr308Met) was three residues away from Asp311 with a peak number of 16 destabilizing mutations. Our analysis therefore identified the protein residues likely to be essential for ZIP8 dimerization and attributed potential variant effects within our study to dimer destabilization.

Discussion

A major challenge in modern biomedicine arises from the necessity to decipher the functional significance of the numerous possible genetic variants present in an average genome and already existing in the human population (51, 52). It is particularly important to focus on genes associated with human diseases, especially those implicated at the interface between the human system and the environment as they could lead to beneficial interventions. The hSLC39A8 gene serves as an example, encoding the SLC39A8 protein, commonly known as ZIP8, a membrane transporter responsible for the cellular uptake of essential divalent metal ions such as Cd2+, Mn2+, and Zn2+. Our study investigated the expression, localization, cellular ion uptake, and dimerization properties of 21 hSLC39A8 genetic variants to better understand their corresponding pathogenicity.

Upon comparing our findings with previously reported literature, we found that consistent with our results, V3 (Gly38Arg) and V6 (Cys113Ser) were both shown in HeLa cells to have high ER expression and disrupted Mn2+ uptake (6). Our in silico homodimer analysis revealed that Gly38Arg is in proximity to sites potentially involved in homodimerization, located in the extracellular domain at the dimerization interphase. Therefore, this mutation may destabilize dimerization and contribute to the observed variant effect on ER retention. A previous study in HeLa cells found WT-like uptake of Cd2+ for both variants and decreased Se4+ uptake for V3 but did not examine the impact of the mutations on localization (12). Although there is this discrepancy, V3 and V6 were both reported separately in multiple individuals with CDG type II and Leigh-like syndrome diseases, respectively, and the current clinical classification coincides with our classification of both variants as pathogenic (Tables 1 and 2). In addition, we examined V5, an arginine substitution at residue 113, that was reported in one individual with CDG type II but, because of the lack of strong evidence, was classified as VUS. Our study reveals that this mutation is pathogenic because of diminished protein expression, leading to a lack of Cd2+ uptake (Table 2). Indeed, the cysteine at residue 113 is highly conserved in mammals (18) and the importance of the cysteine at residue 113 is also illustrated by the AlphaMissense predictions, as both V5 and V6 were assigned highly pathogenic scores (Table 2).

Several hSLC39A8 genetic variants co-occur in patients, and therefore, it is important to study the variants separately to distinguish their individual impact and contribution to disease manifestations. V2 (Val33Met) was initially discovered in a patient with two other variants, V9 (Gly204Cys) and V13 (Ser335Thr). A subsequent study looked only at characterizing the impact of all three variants together on ZIP8 in HeLa cells (6, 21, 31). For the first time, we experimentally characterized these variants separately in cells to discern their individual impact on ZIP8. Whereas V2 was clinically uncategorized because of conflicting interpretations, we now provide experimental evidence that it represents a loss-of-function mutation caused by loss of protein expression, classifying it as a pathogenic variant (Table 2). Furthermore, according to our analysis, this variant is located within a region potentially involved in homodimerization. Therefore, this mutation may lead to dimer destabilization, which could contribute to the observed effect on protein expression. For V9 and V13, which were both clinically uncategorized because of conflicting interpretations and lack of evidence, respectively, we now provide experimental evidence that V9 is benign, whereas V13 is pathogenic (Table 2). Our findings are consistent with another study showing V9 with WT-like Cd2+ cellular accumulation and V13 with loss of Cd2+ accumulation (12). V9 may therefore be a neutral passenger variant, and the cause of CDG type II in this patient may be attributed to the other two mutations Val33Met and Ser335Thr.

V4 (Ser44Trp) and V16 (Gly350Arg) were also discovered together in two siblings with SLC39A8-CDG and Leigh-like syndrome phenotypes and have been unclassified clinically (29). Our study provides evidence that both variants are pathogenic variants, with the Ser44Trp mutation causing ER retention and the Gly350Arg mutation causing disruption in ion uptake (Table 2). Supporting this, we observed that V16 falls within a potential cluster of residues vulnerable to mutations that destabilize dimerization and may prevent functional dimer formation. This could explain the lack of ion uptake despite proper expression and PM localization. V14 (Ile340Asn) was another variant discovered in co-occurrence with V3 (Gly38Arg) but has not yet been studied individually and is currently clinically uncategorized (6, 21, 31). We showed that the Ile340Asn substitution caused ER retention of the transporter and loss of cellular uptake of Cd2+, and should therefore be classified as pathogenic (Table 2).

The highly pleiotropic and highly frequent (allele frequency of 4.47 × 10−2) V19 (Ala391Thr) was discovered in a patient with SLC39A8-CDG and has been reported in numerous GWAS to be highly associated with various biological phenotypes and disorders (10, 16, 29, 33). The expression of the Ala391Thr variant in cells has produced conflicting results. In terms of expression levels, it has been shown in HEK293 cells to express to a similar level as WT (53) or to a lesser extent (54). In CHO cells (55) and MDCK cells (56), the expression and PM localization of V19 were not different from WT. Functionally, the Ala391Thr variant had decreased Cd2+ uptake in HEK293 cells (53), a lower Zn2+ uptake in CHO cells (55), and no difference in MDCK cells (56). In the latter study, the effect of the mutation was only revealed upon preincubation of MDCK cells with Mn2+ and Zn2+, as Ala391Thr prevented a manganese-dependent increase in PM levels of ZIP8 and Zn2+-dependent reduction in PM levels of ZIP8 (13, 56). However, looking at the data, one could argue that these ions may affect the expression levels of ZIP8 as well (56). The effect of Mn2+ or Zn2+ preincubation and ZIP8 expression has not been investigated in cells other than MDCKs and could potentially be the underlying cause for the differences seen in expression or function.

Compounding this inconsistency, Ala391Thr was classified as benign according to ClinVar and predicted as benign by AlphaMissense (Table 2). Prediction algorithms like AlphaMissense are trained based on a structural context and are limited in their ability to address more physiological properties such as the effect of a variant on interactions with other proteins. Unlike most studies, we found that the Ala391Thr substitution caused a loss of uptake of all ions tested because of impaired protein expression, eliciting the classification of this variant as pathogenic in our study (Table 2). The difference in our results to others may be due to the expression system. When we substituted the same alanine at residue 391 to a serine instead of a threonine (V20), the resulting substitution produced a ZIP8 transporter with proper PM localization and uptake of Cd2+, Zn2+, and Mn2+ in the cell (Table 2). Furthermore, the substitution of another nearby residue, Asn389, to a serine (V18) also resulted in proper expression and PM localization (Table 2). The pathogenic property of V19 is therefore unique and warrants further studies.

V7 (Asp193Asn) and V21 (Leu449Phe) were both included in our study because of their previous associations with non-pathogenic phenotypes, and our aim was to contextualize possible biological rationales for such associations. Specifically, the Asp193Asn substitution has been associated with changes in speech and psychomotor speed, whereas the Leu449Phe mutation correlates with changes in grey matter volume and alcohol consumption (57). The Asp193Asn substitution was predicted to be benign by AlphaMissense, and our study presents the variant as benign as well because of the observed WT-like accumulation of both Zn2+ and Mn2+ with ICP-MS (Table 2). With our uptake assays, we saw a decreased uptake for both Zn2+ and Mn2+; this may be due to the differences in the experimental setup including higher concentrations of ions exposed and shorter duration of incubation time. The decreases seen with the uptake assays in V7 may be attributed to the variant location at the dimer interface, but whether this is associated with the correlating phenotype requires further investigation. The Leu449Phe variant is highly frequent (allele frequency of 1.08 × 10−2), and consistent with the AlphaMissense scoring of benign, we observed WT-like localization and cellular uptake for all three ions tested. The distribution of trace elements in the brain is organized within distinct regions, with higher concentrations of iron, copper, zinc, and selenium in the grey matter regions (58). Therefore, a comprehensive characterization of the transport activity for all ZIP8 substrates would be interesting, as the findings could potentially elucidate the correlated phenotypes of Leu449Phe.

The remainder of the variants, V8 (Phe203Ser), V10 (Trp306Cys), V11 (Thr308Met), V12 (Ile322Thr), V15 (Ile340Met), and V17 (Gln364Arg), are all classified as VUS or uncharacterized because of conflicting evidence (Table 1). Other than V12 (Ile322Thr) that affected the PM localization of ZIP8 and is pathogenic, the other variants were all detectable on the PM (Table 2). V15 demonstrated WT-like cellular uptake of all the ions tested and is benign in our study (Table 2). V8 (Phe203Ser) lacked cellular uptake for all tested substrates, and consistent with AlphaMissense, it is pathogenic in our study (Table 2). Given that the Phe203 residue is located in a region potentially crucial for homodimer formation, the Phe203Ser substitution may disrupt proper dimerization, thereby contributing to its pathogenicity. V10 (Trp306Cys) exhibited decreased cellular uptake of Mn2+ and decreased measures of accumulated Cd2+, Zn2+, and Mn2+, and was therefore, in accordance with AlphaMissense, likely pathogenic within our study (Table 2). V11 (Thr308Met) exhibited WT-like uptake of Cd2+ and Zn2+ but diminished Mn2+ cellular uptake (Table 2). Further validation through a competition assay revealed a higher IC50 of Mn2+ is required to compete with Zn2+ transport, compared with ZIP8WT, and is likely pathogenic within our study (Fig 3, Table 2). Both the Trp306 and Thr308 residues are in proximity to potential sites involved in homodimer formation, suggesting that the variant phenotypes may be for this reason. In addition, Thr308 is adjacent to Cys310, which, when mutated to alanine, has been shown to increase the affinity for Zn2+ over other ion substrates (34). The phenotypic consequences of this substitution and its impact on affinity, as well as its association with disease, remain to be elucidated in the human population. Lastly, V17 (Gln364Arg) showed WT-like uptake but decreased accumulated levels of Zn2+, and decreased uptake but WT-like accumulated levels of Mn2+ and is therefore likely pathogenic within this study (Table 2).

The use of an overexpression system with cell-based assays to study transporters bypasses the need to use highly specialized techniques, such as patch clamp, radiolabelled isotopes, or reconstituted liposomes. However, using these assays to infer transport activity has its limits. Specifically, within this study, we are limited by the affinity and kinetics of the dyes used (e.g., FLIPR-5, Zinpyr-1), and the long duration of certain experiments (e.g., cell toxicity, ICP-MS), and do not translate into direct transport function because of other cellular complexities. We therefore combined a number of different experimental approaches with computational analysis and VEPs in this study. Ultimately, we characterized 21 variants in the protein-coding region of the hSLC39A8 gene, encoding SLC39A8 (ZIP8), and as a result, we now suggest new clinical classifications for 16 variants with uncategorized or VUS clinical status. By employing both experimental and computational approaches, we display the possibility of investigating the functional consequence of poorly characterized genetic variants within human transporter genes. As more analyses such as ours are performed, we expect improvements in the variant effect prediction algorithms to better serve clinical interpretations of disease-causing mutations.

Materials and Methods

Variant selection

Variants were collated based on two approaches: a review of the literature, and a programmatic mining of publicly available datasets to curate a collection of variants with annotated clinical and biological impacts. Datasets were obtained from ClinVar (https://www.ncbi.nlm.nih.gov/clinvar) (59), data freeze 01.02.2022, Genebass (https://app.genebass.org/, version 0.7.8-alpha) (60), Open Targets Genetics (https://www.opentargets.org/, version 6) (61), the IEU Open GWAS Project collection of GWAS (https://gwas.mrcieu.ac.uk/, version 5.9.0) (62 Preprint), and LitVar (https://www.ncbi.nlm.nih.gov/research/litvar2/) (63). The mutation nomenclature used in this study was based on the coding region of the solute carrier family 39, member 8 (SLC39A8) transcript (ENSG00000138821), and GenBank Accession NM_001135146.2 (cDNA), following the recommendations of the Human Genome Variation Society (HGVS; https://www.hgvs.org/mutnomen); and the cDNA numbering of nucleotide variants, according to journal guidelines (https://www.hgvs.org/mutnomen).

Using these approaches and other considerations (in silico prediction of thermodynamic stability and variant effect, inclusion of negative and positive controls, and clinical significance), we selected 21 variants within the coding region of the hSLC39A8 gene. We prioritized variants linked to interesting phenotypes, predicted to be highly destabilizing, classified clinically as uncertain significance (VUS) or uncategorized, and highly frequent in the human population. For convenience, variants were annotated numerically from the N to C terminus (Table 1). All variants are summarized in Table 1, and if applicable, we include the minor allele frequency obtained from the Genome Aggregation Database (gnomAD) across all populations (https://gnomad.broadinstitute.org, version 3.1.2) (64 Preprint) and the Regeneron Genetics Center (RGC) Million Exome data (https://www.regeneron.com/science/genetics-center) (65 Preprint).

Structure prediction and identification of key residues for multimerization

We used AlphaFold2 to explore the folding of WT SLC39A8/ZIP8 protein (66). Because multimerization has been described for many members of the ZIP family (67, 68, 69, 70), we studied ZIP8 as a dimer and trimer, using AlphaFold multimer (71 Preprint). To assess model quality, we used the per-residue local distance difference test (pLDDT), reported by AlphaFold. To identify residues essential for multimerization, we used FoldX (72) to estimate changes in thermodynamic stability caused by a mutation in the monomeric (ΔΔGm(i)) and dimeric (ΔΔGd(i)) forms. To identify destabilizing mutations, we estimated the changes in stability because of dimerization for each possible mutation as follows:Dimer stability=ΔΔGd(i)−2ΔΔGm(i)=ΔGd(i)−2ΔGm(i)−(ΔGd(wt)−2ΔGm(wt)).(1)

Mutations causing significant changes in the stability of the monomer compared with the dimer were identified as the upper fifth percentile of the distribution of stabilities.

VEPs

To assess the effect of missense variants on the WT ZIP8, we used AlphaMissense (73), a state-of-the-art prediction method based on deep learning. This method uses both sequence and structure information to classify variants based on a continuous score ranging from 0.0 to 1.0. Pathogenic variants (P) have scores ranging between 0.56 and 1.00; benign variants (B) have scores ranging between 0.0 and 0.34; and otherwise, variants are classified as ambiguous or uncertain (U). We used the precalculated scores reported by Reference (73).

Cell line generation

HEK293 cell lines stably overexpressing the WT or variant human SLC39A8 (hSLC39A8) genes were generated in a doxycycline-inducible manner using the Jump In T-REx HEK293 system (#A15008; Thermo Fisher Scientific) as previously described (74). These cells are referred to as HEK SLC39A8WT and HEK SLC39A8V1-V21 throughout the article. The hSLC39A8 gene harbouring the WT or variant sequences was codon-optimized and carries C-terminal STREPII and HA tags (#132177; Addgene) (generated by GenScript). The WT or the variant hSLC39A8 gene was cloned into a JTI R4 CMV-TO MCS pA vector backbone (#A15004; Thermo Fisher Scientific) and introduced into the Jump In parental cell line using Lipofectamine 3000 Transfection Reagent (#L3000001; Thermo Fisher Scientific) according to the manufacturer’s protocol. The transfected cells were selected over a 14-d period with G418 Geneticin (A1720-5G; Sigma-Aldrich) and then tested for mycoplasma contamination. All cells were cultured in DMEM (#D5796; Sigma-Aldrich) supplemented with 10% FCS (#S1810-500; BioWest), with 1% penicillin and 1% streptomycin (#P4333; Sigma-Aldrich) at 37°C with 5% CO2 and 95% humidity.

Immunofluorescence imaging and analysis

HEK SLC39A8WT and HEK SLC39A8V1-V21 cells were seeded (at 1 × 104 cells/well) into 96-well plates previously coated with poly-L-lysine for 16 h and induced for protein expression with doxycycline (1 μg/ml) for 24 h. For ZIP8 staining against the PM marker CellBrite Red (#30023; Biotium), cells were fixed with 4% formaldehyde in PBS (#D8537; Sigma-Aldrich) for 10 min at RT. For ZIP8 staining against the endoplasmic reticulum (ER) marker calreticulin, cells were fixed with ice-cold methanol for 10 min at −20°C and rehydrated with PBS for 10 min at RT. Cells were then blocked with blocking buffer (0.5% BSA and 0.15% glycine in PBS) for 1 h at RT, and primary antibody staining was performed overnight at 4°C. The following primary antibodies were used and diluted in blocking buffer: anti-HA (rat) (#47877600; Roche) at 1:1,000 dilution and anti-calreticulin (rabbit) (#ab92516; Abcam) at 1:250 dilution. The next day, cells were washed with blocking buffer and stained with secondary antibody for 1 h at RT. The following secondary antibodies were used and diluted at 1:500 in blocking buffer: Alexa Fluor 488 goat anti-rat (#A-11006; Thermo Fisher Scientific) and Alexa Fluor 568 (#A-11011; Thermo Fisher Scientific) goat anti-rabbit (Invitrogen). Cells were then washed with blocking buffer, and stained with DAPI (#D9542; Sigma-Aldrich) and, if applicable, CellBrite Red according to the manufacturer’s protocol. Lastly, cells were washed with PBS before imaging on the Opera Phenix (PerkinElmer) high-throughput microscope (40x magnification). Image processing and analysis were performed as previously described (74). Briefly, 4–20 images per variant cell line and 81–135 images for the WT cell line were processed and analysed using ImageJ. Pearson’s correlation coefficient for the PM (rPM) or the ER (rER) was calculated per WT or variant image. The membranes were immunoblotted with mouse anti-HA (H6533; Sigma-Aldrich) and mouse anti-GAPDH (sc-365062; Santa Cruz). Bound antibodies were detected with horseradish peroxidase–conjugated secondary goat anti-mouse antibodies using ECL (Thermo Fisher Scientific).

Immunoblot analysis

HEK SLC39A8WT and HEK SLC39A8V1-V21 cells were seeded in six-well plates and grown to confluency before harvesting for sample preparation. ZIP8 expression was induced with doxycycline (1 μg/ml) for 24 h before harvest. Cells were washed with PBS before harvesting with a Nonidet P-40 lysis buffer (50 mM Hepes, pH 7.4, 250 mM NaCl, 5 mM EDTA, 1% NP-40, 1x Roche EDTA-free protease inhibitor cocktail [Sigma-Aldrich]) for 30 min on ice. Lysates were precleared with centrifugation (16,000g, 15 min, 4°C), and proteins were quantified with the Bradford assay (Bio-Rad), according to the manufacturer’s protocol.

Cd2+ uptake using the FLIPR Calcium-5 assay

For the measurement of Cd2+ uptake by HEK SLC39A8WT and HEK SLC39A8V1-V21 cells, we used the FLIPR Calcium-5 Assay kit (#R8185; Molecular Devices) as previously described (36). Briefly, cells were seeded (at 1 × 104 cells/well) into 96-well plates previously coated with poly-L-lysine for 24 h, and then, if applicable, protein expression was induced with doxycycline (2 μg/ml) for 24 h. The next day, cellular media were exchanged with the FLIPR Calcium-5 dye, previously prepared according to the manufacturer’s protocol, for a 30-min incubation at 37°C with 5% CO2. The fluorescence signal (λEx = 470–495 nm; λEm = 515–575 nm) was then recorded with the SpectraMax i3X plate reader (Molecular Devices).

For the measurement of Cd2+ uptake, fluorescent signals were recorded before substrate addition at two reads per minute for 5 min. The recording was then paused for cadmium chloride (CdCl2) substrate addition and resumed at two reads per minute for 30 min (Fig 2A). To quantify Cd2+ uptake for each concentration of CdCl2 addition, the recordings were normalized to the measurements before substrate addition and the area under the curve (AUC) from 10 to 30 min post-substrate addition was calculated using the trapezoidal rule (Fig 2A). For CdCl2 titration, the calculated AUC for Cd2+ uptake was then plotted against the concentration of added CdCl2 and an EC50 of 2.64 μM (95% Cl: 2.13, 3.44) for ZIP8WT was determined based on a non-linear regression curve fitting method (GraphPad). For determining Cd2+ uptake for each WT or variant ZIP8, the AUC for Cd2+ uptake was measured and calculated in each cell line with a set CdCl2 concentration of 2 μM (Fig 2B).

Zn2+ uptake using Zinpyr-1

To measure Zn2+ uptake, we used the intracellular Zn2+-sensitive dye Zinpyr-1 as previously described (38, 39). Briefly, HEK SLC39A8WT and HEK SLC39A8V1-V21 cells were seeded in a six-well plate (at 6 × 105 cells/well) for 24 h and protein expression was induced with doxycycline (1 μg/ml) for 24 h. The next day, cells were incubated with 50 μM of Zinpyr-1 in experimental buffer consisting of HBSS (#14175095; Gibco) and 25 mM Hepes (#15630056; Gibco) for 30 min at 37°C. Cells were then harvested by trypsin (#T3924; Sigma-Aldrich) and incubated with different concentrations of zinc chloride (ZnCl2) in experimental buffer for 10 min at 37°C before washing and resuspending the cells in flow cytometry buffer (10% FBS in PBS) for flow cytometry analysis. Fluorescence was measured using BD LSRFortessa Cell Analyzer (λEx = 492 nm, λEm = 527 nm) and analysed with FlowJo software.

Zn2+ and Mn2+ competition assay

For the Zn2+ and Mn2+ competition assay, cells were plated (at 1 × 104 cells/well) in a 96-well plate format previously coated with poly-L-lysine for 24 h, and then, if applicable, protein expression was induced with doxycycline (1 μg/ml) for 24 h. The next day, cells were incubated with Zinpyr-1 as described above. Cells were then washed with experimental buffer and co-incubated with ZnCl2 (40 μM) and MnCl2 (0 to 2.56 mM) for 10 min at 37°C. Zinpyr-1 fluorescence was measured with Opera Phenix (PerkinElmer) (λEx = 492 nm, λEm = 527 nm), and the images were analysed with ImageJ.

ICP-MS

Cells were plated on 15-cm plates and grown to confluency before harvesting for sample preparation. ZIP8 expression was induced with doxycycline (1 μg/ml) for 48 h before harvest, and cells were incubated with 0.5 μM CdCl2 for 24 h before harvest. Cells were harvested with trypsin and washed with PBS before pellets were weighed and frozen with liquid nitrogen. Samples were then sent to BOKU Core Facility Mass Spectrometry for analysis.

Chemicals used at the facility: Water for solution preparation was purified employing an Ultra Clear basic reverse osmosis system (SG Wasseraufbereitung und Regenerierstation GmbH) and subsequently sub-boiled (MLS, Leutkirch). Nitric acid (p.a.), hydrogen peroxide 30% (Suprapur), and a certified multi-element standard solution VI were purchased from Merck (KGaA). Nitric acid was double-sub-boiled and stored in a perfluoroalkoxy-polymer bottle.

Instrumentation used: A MW7000 from Anton Paar was used for microwave-assisted acid digestion with nitric acid/hydrogen peroxide. An iCAP quadrupole ICP-MS (Thermo Fisher Scientific) was used for elemental analysis. The system was operated in kinetic energy discrimination mode; that is, the reaction/collision cell was filled with helium (4.8 ml min−1). Cadmium was detected at m/z 111, and zinc was detected at m/z 66. Multi-element standards were employed to calibrate the system for quantification.

Sample preparation and experimental setup: All materials that came into contact with the samples, blanks, and standards were precleaned by soaking for 24 h in 10% and 24 h in 1% nitric acid followed by flushing with ultra-pure water and drying under a class 100 laminar flow bench. One replicate (n = 1) per sample was digested and measured. Each sample for digestion is around 50–100 mg and digested with 500 μl nitric acid and 30 μl H2O2. Nitric acid blanks (2% [vol/vol]) were measured between the sample runs for instrumental blank control.

Quantification was performed using multilevel external calibration with multi-element standards and internal standardization via indium.

Mn2+ uptake using Mn2+ cell toxicity

For quantification of Mn2+ uptake, we used the cell death properties of Mn2+ observed in HEK293 cells with an overexpression of the ZIP8 transporter (Fig 2E). HEK SLC39A8WT and HEK SLC39A8V1-V21 cells were seeded (at 1 × 104 cells/well) into 96-well plates previously coated with poly-L-lysine for 24 h, and then, if applicable, protein expression was induced with doxycycline (1 μg/ml) for 24 h. Cells were then incubated with differing concentrations of manganese chloride (MnCl2) in culture media for 48 h, and cellular viability was then assessed with CellTiter-Glo Luminescent Cell Viability Assay (#G7570; Promega) according to the manufacturer’s protocol. The luminescence was measured with the SpectraMax i3X plate reader (Molecular Devices).

Ion uptake classification

For each ion uptake assay, we characterized the cellular uptake for each ion and classified each variant into WT, decreased, or no response. This was based on the comparison of uptake in HEK SLC39A8V1-V21 cells with the uptake in either doxycycline-induced (for ZIP8WT activity) or uninduced (for the absence of ZIP8WT activity) HEK SLC39A8WT cells. The WT category is defined as when the variant-expressing cell shows significantly higher uptake than the uninduced HEK SLC39A8WT cells and not significantly different than induced HEK SLC39A8WT cells. The decreased category is defined as when the variant-expressing cell shows significantly lower uptake than the induced HEK SLC39A8WT cells and significantly higher uptake than the uninduced HEK SLC39A8WT cells. As well, the decreased category is also true for when the variant-expressing cell shows no statistical difference in uptake with the induced HEK SLC39A8WT cells nor the uninduced HEK SLC39A8WT cells. Lastly, the no response category is defined as when the variant-expressing cell shows significantly lower uptake than the induced HEK SLC39A8WT cells and not significantly different than uninduced HEK SLC39A8WT cells.

Acknowledgements

We would like to thank the laboratory of Giulio Superti-Furga for all the fruitful discussions and Andrè Ferdigg for the original idea to use Mn2+ toxicity as a transport assay. We thank Dr. Chan Xiong and Ing. Andreas Brandstätter for conducting MS experiments. The MS equipment was kindly provided by the BOKU Core Facility Mass Spectrometry. This work was partly performed by the RESOLUTE (https://re-solute.eu/) and REsolution (https://re-solute.eu/resolution) consortia. Plasmids are available through Addgene (https://www.addgene.org/depositor-collections/re-solute/). RESOLUTE has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 777372. This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA. REsolution has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 101034439. This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA. Continuation of the work after the funding period occurred via the Austrian Academy of Sciences. This article reflects only the authors’ views, and neither IMI nor the European Union and EFPIA are responsible for any use that may be made of the information contained therein.

Author Contributions

  • W-A Wang: conceptualization, data curation, formal analysis, validation, investigation, visualization, methodology, and writing—original draft, review, and editing.

  • A Garofoli: data curation, investigation, and writing—original draft, review, and editing.

  • E Ferrada: data curation, software, formal analysis, validation, investigation, and writing—review and editing.

  • C Klimek: validation, investigation, and methodology.

  • B Steurer: conceptualization, investigation, visualization, methodology, and writing—review and editing.

  • A Ingles-Prieto: conceptualization and writing—review and editing.

  • T Osthushenrich: data curation, investigation, methodology, and writing—review and editing.

  • A MacNamara: conceptualization and supervision.

  • A Malarstig: conceptualization and supervision.

  • T Wiedmer: conceptualization, supervision, project administration, and writing—review and editing.

  • G Superti-Furga: conceptualization, supervision, funding acquisition, and writing—review and editing.

Conflict of Interest Statement

G Superti-Furga is the scientific founder and a shareholder of Proxygen and Solgate, the latter a SLC-focused company. The rest of the authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this study.

  • Received September 2, 2024.
  • Revision received January 15, 2025.
  • Accepted January 15, 2025.
Creative Commons logoCreative Commons logohttps://creativecommons.org/licenses/by/4.0/

This article is available under a Creative Commons License (Attribution 4.0 International, as described at https://creativecommons.org/licenses/by/4.0/).

References

  1. 1.↵
    1. Waldron KJ,
    2. Robinson NJ
    (2009) How do bacterial cells ensure that metalloproteins get the correct metal? Nat Rev Microbiol 7: 25–35. doi:10.1038/nrmicro2057
    OpenUrlCrossRefPubMed
  2. 2.↵
    1. Shi W,
    2. Chance MR
    (2011) Metalloproteomics: Forward and reverse approaches in metalloprotein structural and functional characterization. Curr Opin Chem Biol 15: 144–148. doi:10.1016/j.cbpa.2010.11.004
    OpenUrlCrossRefPubMed
  3. 3.↵
    1. Milatovic D,
    2. Zaja-Milatovic S,
    3. Gupta RC,
    4. Yu Y,
    5. Aschner M
    (2009) Oxidative damage and neurodegeneration in manganese-induced neurotoxicity. Toxicol Appl Pharmacol 240: 219–225. doi:10.1016/j.taap.2009.07.004
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Tchounwou PB,
    2. Yedjou CG,
    3. Patlolla AK,
    4. Sutton DJ
    (2012) Heavy metal toxicity and the environment. Exp Suppl 101: 133–164. doi:10.1007/978-3-7643-8340-4_6
    OpenUrlCrossRef
  5. 5.↵
    1. Genchi G,
    2. Sinicropi MS,
    3. Lauria G,
    4. Carocci A,
    5. Catalano A
    (2020) The effects of cadmium toxicity. Int J Environ Res Public Health 17: 3782. doi:10.3390/ijerph17113782
    OpenUrlCrossRef
  6. 6.↵
    1. Choi EK,
    2. Nguyen TT,
    3. Gupta N,
    4. Iwase S,
    5. Seo YA
    (2018) Functional analysis of SLC39A8 mutations and their implications for manganese deficiency and mitochondrial disorders. Sci Rep 8: 3163. doi:10.1038/s41598-018-21464-0
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Polesel M,
    2. Ingles-Prieto A,
    3. Christodoulaki E,
    4. Ferrada E,
    5. Doucerain C,
    6. Altermatt P,
    7. Knecht M,
    8. Kuhn M,
    9. Steck AL,
    10. Wilhelm M, et al.
    (2023) Functional characterization of SLC39 family members ZIP5 and ZIP10 in overexpressing HEK293 cells reveals selective copper transport activity. Biometals 36: 227–237. doi:10.1007/s10534-022-00474-6
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Wang CY,
    2. Jenkitkasemwong S,
    3. Duarte S,
    4. Sparkman BK,
    5. Shawki A,
    6. Mackenzie B,
    7. Knutson MD
    (2012) ZIP8 is an iron and zinc transporter whose cell-surface expression is up-regulated by cellular iron loading. J Biol Chem 287: 34032–34043. doi:10.1074/jbc.M112.367284
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. GTEx Consortium
    (2013) The genotype-tissue expression (GTEx) project. Nat Genet 45: 580–585. doi:10.1038/ng.2653
    OpenUrlCrossRefPubMed
  10. 10.↵
    1. Costas J
    (2018) The highly pleiotropic gene SLC39A8 as an opportunity to gain insight into the molecular pathogenesis of schizophrenia. Am J Med Genet B Neuropsychiatr Genet 177: 274–283. doi:10.1002/ajmg.b.32545
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Nebert DW,
    2. Liu Z
    (2019) SLC39A8 gene encoding a metal ion transporter: Discovery and bench to bedside. Hum Genomics 13: 51. doi:10.1186/s40246-019-0233-3
    OpenUrlCrossRefPubMed
  12. 12.↵
    1. Liang ZL,
    2. Tan HW,
    3. Wu JY,
    4. Chen XL,
    5. Wang XY,
    6. Xu YM,
    7. Lau ATY
    (2021) The impact of ZIP8 disease-associated variants G38R, C113S, G204C, and S335T on selenium and cadmium accumulations: The first characterization. Int J Mol Sci 22: 11399. doi:10.3390/ijms222111399
    OpenUrlCrossRef
  13. 13.↵
    1. Liu Z,
    2. Li H,
    3. Soleimani M,
    4. Girijashanker K,
    5. Reed JM,
    6. He L,
    7. Dalton TP,
    8. Nebert DW
    (2008) Cd2+ versus Zn2+ uptake by the ZIP8 HCO3--dependent symporter: Kinetics, electrogenicity and trafficking. Biochem Biophys Res Commun 365: 814–820. doi:10.1016/j.bbrc.2007.11.067
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. McDermott JR,
    2. Geng X,
    3. Jiang L,
    4. Gálvez-Peralta M,
    5. Chen F,
    6. Nebert DW,
    7. Liu Z
    (2016) Zinc- and bicarbonate-dependent ZIP8 transporter mediates selenite uptake. Oncotarget 7: 35327–35340. doi:10.18632/oncotarget.9205
    OpenUrlCrossRef
  15. 15.↵
    1. Avila DS,
    2. Puntel RL,
    3. Aschner M
    (2013) Manganese in health and disease. Met Ions Life Sci 13: 199–227. doi:10.1007/978-94-007-7500-8_7
    OpenUrlCrossRefPubMed
  16. 16.↵
    1. Mealer RG,
    2. Jenkins BG,
    3. Chen CY,
    4. Daly MJ,
    5. Ge T,
    6. Lehoux S,
    7. Marquardt T,
    8. Palmer CD,
    9. Park JH,
    10. Parsons PJ, et al.
    (2020) The schizophrenia risk locus in SLC39A8 alters brain metal transport and plasma glycosylation. Sci Rep 10: 13162. doi:10.1038/s41598-020-70108-9
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Lake NJ,
    2. Compton AG,
    3. Rahman S,
    4. Thorburn DR
    (2016) Leigh syndrome: One disorder, more than 75 monogenic causes. Ann Neurol 79: 190–203. doi:10.1002/ana.24551
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Riley LG,
    2. Cowley MJ,
    3. Gayevskiy V,
    4. Roscioli T,
    5. Thorburn DR,
    6. Prelog K,
    7. Bahlo M,
    8. Sue CM,
    9. Balasubramaniam S,
    10. Christodoulou J
    (2017) A SLC39A8 variant causes manganese deficiency, and glycosylation and mitochondrial disorders. J Inherit Metab Dis 40: 261–269. doi:10.1007/s10545-016-0010-6
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Anagianni S,
    2. Tuschl K
    (2019) Genetic disorders of manganese metabolism. Curr Neurol Neurosci Rep 19: 33. doi:10.1007/s11910-019-0942-y
    OpenUrlCrossRefPubMed
  20. 20.↵
    1. Boycott KM,
    2. Beaulieu CL,
    3. Kernohan KD,
    4. Gebril OH,
    5. Mhanni A,
    6. Chudley AE,
    7. Redl D,
    8. Qin W,
    9. Hampson S,
    10. Küry S, et al.
    (2015) Autosomal-recessive intellectual disability with cerebellar atrophy syndrome caused by mutation of the manganese and zinc transporter gene SLC39A8. Am J Hum Genet 97: 886–893. doi:10.1016/j.ajhg.2015.11.002
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Park JH,
    2. Hogrebe M,
    3. Grüneberg M,
    4. DuChesne I,
    5. von der Heiden AL,
    6. Reunert J,
    7. Schlingmann KP,
    8. Boycott KM,
    9. Beaulieu CL,
    10. Mhanni AA, et al.
    (2015) SLC39A8 deficiency: A disorder of manganese transport and glycosylation. Am J Hum Genet 97: 894–903. doi:10.1016/j.ajhg.2015.11.003
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Pickrell JK,
    2. Berisa T,
    3. Liu JZ,
    4. Ségurel L,
    5. Tung JY,
    6. Hinds DA
    (2016) Detection and interpretation of shared genetic influences on 42 human traits. Nat Genet 48: 709–717. doi:10.1038/ng.3570
    OpenUrlCrossRefPubMed
  23. 23.↵
    1. Li M,
    2. Wu DD,
    3. Yao YG,
    4. Huo YX,
    5. Liu JW,
    6. Su B,
    7. Chasman DI,
    8. Chu AY,
    9. Huang T,
    10. Qi L, et al.
    (2016) Recent positive selection drives the expansion of a schizophrenia risk nonsynonymous variant at SLC39A8 in Europeans. Schizophr Bull 42: 178–190. doi:10.1093/schbul/sbv070
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Lin W,
    2. Li D,
    3. Cheng L,
    4. Li L,
    5. Liu F,
    6. Hand NJ,
    7. Epstein JA,
    8. Rader DJ
    (2018) Zinc transporter Slc39a8 is essential for cardiac ventricular compaction. J Clin Invest 128: 826–833. doi:10.1172/JCI96993
    OpenUrlCrossRef
  25. 25.↵
    1. Liu L,
    2. Geng X,
    3. Cai Y,
    4. Copple B,
    5. Yoshinaga M,
    6. Shen J,
    7. Nebert DW,
    8. Wang H,
    9. Liu Z
    (2018) Hepatic ZIP8 deficiency is associated with disrupted selenium homeostasis, liver pathology, and tumor formation. Am J Physiol Gastrointest Liver Physiol 315: G569–G579. doi:10.1152/ajpgi.00165.2018
    OpenUrlCrossRef
  26. 26.↵
    1. Nakata T,
    2. Creasey EA,
    3. Kadoki M,
    4. Lin H,
    5. Selig MK,
    6. Yao J,
    7. Lefkovith A,
    8. Daly MJ,
    9. Graham DB,
    10. Xavier RJ
    (2020) A missense variant in SLC39A8 confers risk for Crohn’s disease by disrupting manganese homeostasis and intestinal barrier integrity. Proc Natl Acad Sci U S A 117: 28930–28938. doi:10.1073/pnas.2014742117
    OpenUrlAbstract/FREE Full Text
  27. 27.↵
    1. Pyle CJ,
    2. Akhter S,
    3. Bao S,
    4. Dodd CE,
    5. Schlesinger LS,
    6. Knoell DL
    (2017) Zinc modulates endotoxin-induced human macrophage inflammation through ZIP8 induction and C/EBPβ inhibition. PLoS One 12: e0169531. doi:10.1371/journal.pone.0169531
    OpenUrlCrossRefPubMed
  28. 28.↵
    1. Hall SC,
    2. Smith DR,
    3. Dyavar SR,
    4. Wyatt TA,
    5. Samuelson DR,
    6. Bailey KL,
    7. Knoell DL
    (2021) Critical role of zinc transporter (ZIP8) in myeloid innate immune cell function and the host response against bacterial pneumonia. J Immunol 207: 1357–1370. doi:10.4049/jimmunol.2001395
    OpenUrlAbstract/FREE Full Text
  29. 29.↵
    1. Bonaventura E,
    2. Barone R,
    3. Sturiale L,
    4. Pasquariello R,
    5. Alessandrì MG,
    6. Pinto AM,
    7. Renieri A,
    8. Panteghini C,
    9. Garavaglia B,
    10. Cioni G, et al.
    (2021) Clinical, molecular and glycophenotype insights in SLC39A8-CDG. Orphanet J Rare Dis 16: 307. doi:10.1186/s13023-021-01941-y
    OpenUrlCrossRefPubMed
  30. 30.
    1. Riley LG,
    2. Cowley MJ,
    3. Gayevskiy V,
    4. Minoche AE,
    5. Puttick C,
    6. Thorburn DR,
    7. Rius R,
    8. Compton AG,
    9. Menezes MJ,
    10. Bhattacharya K, et al.
    (2020) The diagnostic utility of genome sequencing in a pediatric cohort with suspected mitochondrial disease. Genet Med 22: 1254–1261. doi:10.1038/s41436-020-0793-6
    OpenUrlCrossRefPubMed
  31. 31.↵
    1. Park JH,
    2. Hogrebe M,
    3. Fobker M,
    4. Brackmann R,
    5. Fiedler B,
    6. Reunert J,
    7. Rust S,
    8. Tsiakas K,
    9. Santer R,
    10. Grüneberg M, et al.
    (2018) SLC39A8 deficiency: Biochemical correction and major clinical improvement by manganese therapy. Genet Med 20: 259–268. doi:10.1038/gim.2017.106
    OpenUrlCrossRefPubMed
  32. 32.
    1. Zang ZS,
    2. Xu YM,
    3. Lau ATY
    (2016) Molecular and pathophysiological aspects of metal ion uptake by the zinc transporter ZIP8 (SLC39A8). Toxicol Res 5: 987–1002. doi:10.1039/c5tx00424a
    OpenUrlCrossRef
  33. 33.↵
    1. Haller G,
    2. McCall K,
    3. Jenkitkasemwong S,
    4. Sadler B,
    5. Antunes L,
    6. Nikolov M,
    7. Whittle J,
    8. Upshaw Z,
    9. Shin J,
    10. Baschal E, et al.
    (2018) A missense variant in SLC39A8 is associated with severe idiopathic scoliosis. Nat Commun 9: 4171. doi:10.1038/s41467-018-06705-0
    OpenUrlCrossRefPubMed
  34. 34.↵
    1. Jiang Y,
    2. Li Z,
    3. Sui D,
    4. Sharma G,
    5. Wang T,
    6. MacRenaris K,
    7. Takahashi H,
    8. Merz K,
    9. Hu J
    (2023) Rational engineering of an elevator-type metal transporter ZIP8 reveals a conditional selectivity filter critically involved in determining substrate specificity. Commun Biol 6: 778. doi:10.1038/s42003-023-05146-w
    OpenUrlCrossRef
  35. 35.↵
    1. He L,
    2. Girijashanker K,
    3. Dalton TP,
    4. Reed J,
    5. Li H,
    6. Soleimani M,
    7. Nebert DW
    (2006) ZIP8, member of the solute-carrier-39 (SLC39) metal-transporter family: Characterization of transporter properties. Mol Pharmacol 70: 171–180. doi:10.1124/mol.106.024521
    OpenUrlAbstract/FREE Full Text
  36. 36.↵
    1. Franz MC,
    2. Simonin A,
    3. Graeter S,
    4. Hediger MA,
    5. Kovacs G
    (2014) Development of the first fluorescence screening assay for the SLC39A2 zinc transporter. J Biomol Screen 19: 909–916. doi:10.1177/1087057114526781
    OpenUrlCrossRefPubMed
  37. 37.↵
    1. Montalbetti N,
    2. Simonin A,
    3. Dalghi MG,
    4. Kovacs G,
    5. Hediger MA
    (2014) Development and validation of a fast and homogeneous cell-based fluorescence screening assay for divalent metal transporter 1 (DMT1/SLC11A2) using the FLIPR tetra. J Biomol Screen 19: 900–908. doi:10.1177/1087057114521663
    OpenUrlCrossRefPubMed
  38. 38.↵
    1. Alker W,
    2. Schwerdtle T,
    3. Schomburg L,
    4. Haase H
    (2019) A zinpyr-1-based fluorimetric microassay for free zinc in human serum. Int J Mol Sci 20: 4006. doi:10.3390/ijms20164006
    OpenUrlCrossRef
  39. 39.↵
    1. Moskovskich A,
    2. Goldmann U,
    3. Kartnig F,
    4. Lindinger S,
    5. Konecka J,
    6. Fiume G,
    7. Girardi E,
    8. Superti-Furga G
    (2019) The transporters SLC35A1 and SLC30A1 play opposite roles in cell survival upon VSV virus infection. Sci Rep 9: 10471. doi:10.1038/s41598-019-46952-9
    OpenUrlCrossRef
  40. 40.↵
    1. Watt NT,
    2. Griffiths HH,
    3. Hooper NM
    (2013) Neuronal zinc regulation and the prion protein. Prion 7: 203–208. doi:10.4161/pri.24503
    OpenUrlCrossRefPubMed
  41. 41.↵
    1. Alker W,
    2. Haase H
    (2021) Comparison of free zinc levels determined by fluorescent probes in THP1 cells using microplate reader and flow cytometer. Biol Trace Elem Res 199: 2414–2419. doi:10.1007/s12011-020-02355-w
    OpenUrlCrossRef
  42. 42.↵
    1. Stanwood GD,
    2. Leitch DB,
    3. Savchenko V,
    4. Wu J,
    5. Fitsanakis VA,
    6. Anderson DJ,
    7. Stankowski JN,
    8. Aschner M,
    9. McLaughlin B
    (2009) Manganese exposure is cytotoxic and alters dopaminergic and GABAergic neurons within the basal ganglia. J Neurochem 110: 378–389. doi:10.1111/j.1471-4159.2009.06145.x
    OpenUrlCrossRefPubMed
  43. 43.↵
    1. Liu MJ,
    2. Bao S,
    3. Gálvez-Peralta M,
    4. Pyle CJ,
    5. Rudawsky AC,
    6. Pavlovicz RE,
    7. Killilea DW,
    8. Li C,
    9. Nebert DW,
    10. Wewers MD, et al.
    (2013) ZIP8 regulates host defense through zinc-mediated inhibition of NF-κB. Cell Rep 3: 386–400. doi:10.1016/j.celrep.2013.01.009
    OpenUrlCrossRefPubMed
  44. 44.↵
    1. Kambe T,
    2. Tsuji T,
    3. Hashimoto A,
    4. Itsumura N
    (2015) The physiological, biochemical, and molecular roles of zinc transporters in zinc homeostasis and metabolism. Physiol Rev 95: 749–784. doi:10.1152/physrev.00035.2014
    OpenUrlCrossRefPubMed
  45. 45.↵
    1. Hu J
    (2021) Toward unzipping the ZIP metal transporters: Structure, evolution, and implications on drug discovery against cancer. FEBS J 288: 5805–5825. doi:10.1111/febs.15658
    OpenUrlCrossRef
  46. 46.↵
    1. Verouti SN,
    2. Pujol-Giménez J,
    3. Bermudez-Lekerika P,
    4. Scherler L,
    5. Bhardwaj R,
    6. Thomas A,
    7. Lenglet S,
    8. Siegrist M,
    9. Hofstetter W,
    10. Fuster DG, et al.
    (2022) The allelic variant A391T of metal ion transporter ZIP8 (SLC39A8) leads to hypotension and enhanced insulin resistance. Front Physiol 13: 912277. doi:10.3389/fphys.2022.912277
    OpenUrlCrossRef
  47. 47.↵
    1. Aydemir TB,
    2. Liuzzi JP,
    3. McClellan S,
    4. Cousins RJ
    (2009) Zinc transporter ZIP8 (SLC39A8) and zinc influence IFN-gamma expression in activated human T cells. J Leukoc Biol 86: 337–348. doi:10.1189/jlb.1208759
    OpenUrlCrossRefPubMed
  48. 48.↵
    1. Basu S,
    2. Wallner B
    (2016) DockQ: A quality measure for protein-protein docking models. PLoS One 11: e0161879. doi:10.1371/journal.pone.0161879
    OpenUrlCrossRefPubMed
  49. 49.↵
    1. Bryant P,
    2. Pozzati G,
    3. Elofsson A
    (2022) Author correction: Improved prediction of protein-protein interactions using AlphaFold2. Nat Commun 13: 1694. doi:10.1038/s41467-022-29480-5
    OpenUrlCrossRef
  50. 50.↵
    1. Zhang T,
    2. Sui D,
    3. Hu J
    (2016) Structural insights of ZIP4 extracellular domain critical for optimal zinc transport. Nat Commun 7: 11979. doi:10.1038/ncomms11979
    OpenUrlCrossRefPubMed
  51. 51.↵
    1. ENCODE Project Consortium
    (2012) An integrated encyclopedia of DNA elements in the human genome. Nature 489: 57–74. doi:10.1038/nature11247
    OpenUrlCrossRefPubMed
  52. 52.↵
    1. Brookes AJ,
    2. Robinson PN
    (2015) Human genotype-phenotype databases: Aims, challenges and opportunities. Nat Rev Genet 16: 702–715. doi:10.1038/nrg3932
    OpenUrlCrossRefPubMed
  53. 53.↵
    1. Zhang R,
    2. Witkowska K,
    3. Afonso Guerra-Assunção J,
    4. Ren M,
    5. Ng FL,
    6. Mauro C,
    7. Tucker AT,
    8. Caulfield MJ,
    9. Ye S
    (2016) A blood pressure-associated variant of the SLC39A8 gene influences cellular cadmium accumulation and toxicity. Hum Mol Genet 25: 4117–4126. doi:10.1093/hmg/ddw236
    OpenUrlCrossRefPubMed
  54. 54.↵
    1. Melia JMP,
    2. Lin R,
    3. Xavier RJ,
    4. Thompson RB,
    5. Fu D,
    6. Wan F,
    7. Sears CL,
    8. Donowitz M
    (2019) Induction of the metal transporter ZIP8 by interferon gamma in intestinal epithelial cells: Potential role of metal dyshomeostasis in Crohn’s disease. Biochem Biophys Res Commun 515: 325–331. doi:10.1016/j.bbrc.2019.05.137
    OpenUrlCrossRef
  55. 55.↵
    1. Tseng WC,
    2. Reinhart V,
    3. Lanz TA,
    4. Weber ML,
    5. Pang J,
    6. Le KXV,
    7. Bell RD,
    8. O’Donnell P,
    9. Buhl DL
    (2021) Schizophrenia-associated SLC39A8 polymorphism is a loss-of-function allele altering glutamate receptor and innate immune signaling. Transl Psychiatry 11: 136. doi:10.1038/s41398-021-01262-5
    OpenUrlCrossRefPubMed
  56. 56.↵
    1. Nishito Y,
    2. Fujishiro H,
    3. Nagamatsu S,
    4. Kambe T
    (2024) Reduced Mn uptake of pleiotropic ZIP8 SNP is caused by its loss of Mn-responsive accumulation on the cell-surface. Biosci Biotechnol Biochem 88: 1019–1026. doi:10.1093/bbb/zbae076
    OpenUrlCrossRef
  57. 57.↵
    1. Liu M,
    2. Jiang Y,
    3. Wedow R,
    4. Li Y,
    5. Brazel DM,
    6. Chen F,
    7. Datta G,
    8. Davila-Velderrain J,
    9. McGuire D,
    10. Tian C, et al.
    (2019) Association studies of up to 1.2 million individuals yield new insights into the genetic etiology of tobacco and alcohol use. Nat Genet 51: 237–244. doi:10.1038/s41588-018-0307-5
    OpenUrlCrossRefPubMed
  58. 58.↵
    1. Duflou H,
    2. Maenhaut W,
    3. De Reuck J
    (1989) Regional distribution of potassium, calcium, and six trace elements in normal human brain. Neurochem Res 14: 1099–1112. doi:10.1007/BF00965616
    OpenUrlCrossRefPubMed
  59. 59.↵
    1. Landrum MJ,
    2. Lee JM,
    3. Benson M,
    4. Brown GR,
    5. Chao C,
    6. Chitipiralla S,
    7. Gu B,
    8. Hart J,
    9. Hoffman D,
    10. Jang W, et al.
    (2018) ClinVar: Improving access to variant interpretations and supporting evidence. Nucleic Acids Res 46: D1062–D1067. doi:10.1093/nar/gkx1153
    OpenUrlCrossRefPubMed
  60. 60.↵
    1. Karczewski KJ,
    2. Solomonson M,
    3. Chao KR,
    4. Goodrich JK,
    5. Tiao G,
    6. Lu W,
    7. Riley-Gillis BM,
    8. Tsai EA,
    9. Kim HI,
    10. Zheng X, et al.
    (2022) Systematic single-variant and gene-based association testing of thousands of phenotypes in 394,841 UK Biobank exomes. Cell Genom 2: 100168. doi:10.1016/j.xgen.2022.100168
    OpenUrlCrossRefPubMed
  61. 61.↵
    1. Ghoussaini M,
    2. Mountjoy E,
    3. Carmona M,
    4. Peat G,
    5. Schmidt EM,
    6. Hercules A,
    7. Fumis L,
    8. Miranda A,
    9. Carvalho-Silva D,
    10. Buniello A, et al.
    (2021) Open Targets genetics: Systematic identification of trait-associated genes using large-scale genetics and functional genomics. Nucleic Acids Res 49: D1311–D1320. doi:10.1093/nar/gkaa840
    OpenUrlCrossRef
  62. 62.↵
    1. Elsworth B,
    2. Lyon M,
    3. Alexander T,
    4. Liu Y,
    5. Matthews P,
    6. Hallett J,
    7. Bates P,
    8. Palmer T,
    9. Haberland V,
    10. Smith GD, et al.
    (2020) The MRC IEU OpenGWAS data infrastructure. BioRxiv. doi:10.1101/2020.08.10.244293 (Preprint posted August 10, 2020).
    OpenUrlAbstract/FREE Full Text
  63. 63.↵
    1. Allot A,
    2. Peng Y,
    3. Wei CH,
    4. Lee K,
    5. Phan L,
    6. Lu Z
    (2018) LitVar: A semantic search engine for linking genomic variant data in PubMed and PMC. Nucleic Acids Res 46: W530–W536. doi:10.1093/nar/gky355
    OpenUrlCrossRefPubMed
  64. 64.↵
    1. Chen S,
    2. Francioli LC,
    3. Goodrich JK,
    4. Collins RL,
    5. Wang Q,
    6. Alföldi J,
    7. Watts NA,
    8. Vittal C,
    9. Gauthier LD,
    10. Poterba T, et al.
    (2022) A genome-wide mutational constraint map quantified from variation in 76,156 human genomes. BioRxiv. doi:10.1101/2022.03.20.485034 (Preprint posted March 21, 2022).
    OpenUrlAbstract/FREE Full Text
  65. 65.↵
    1. Sun KY,
    2. Bai X,
    3. Chen S,
    4. Bao S,
    5. Kapoor M,
    6. Zhang C,
    7. Backman J,
    8. Joseph T,
    9. Maxwell E,
    10. Mitra G, et al.
    (2023) A deep catalog of protein-coding variation in 985,830 individuals. BioRxiv. doi:10.1101/2023.05.09.539329 (Preprint posted May 10, 2023).
    OpenUrlCrossRef
  66. 66.↵
    1. Jumper J,
    2. Evans R,
    3. Pritzel A,
    4. Green T,
    5. Figurnov M,
    6. Ronneberger O,
    7. Tunyasuvunakool K,
    8. Bates R,
    9. Žídek A,
    10. Potapenko A, et al.
    (2021) Highly accurate protein structure prediction with AlphaFold. Nature 596: 583–589. doi:10.1038/s41586-021-03819-2
    OpenUrlCrossRefPubMed
  67. 67.↵
    1. Bin BH,
    2. Fukada T,
    3. Hosaka T,
    4. Yamasaki S,
    5. Ohashi W,
    6. Hojyo S,
    7. Miyai T,
    8. Nishida K,
    9. Yokoyama S,
    10. Hirano T
    (2011) Biochemical characterization of human ZIP13 protein: A homo-dimerized zinc transporter involved in the spondylocheiro dysplastic ehlers-danlos syndrome. J Biol Chem 286: 40255–40265. doi:10.1074/jbc.M111.256784
    OpenUrlAbstract/FREE Full Text
  68. 68.↵
    1. Taylor KM,
    2. Hiscox S,
    3. Nicholson RI,
    4. Hogstrand C,
    5. Kille P
    (2012) Protein kinase CK2 triggers cytosolic zinc signaling pathways by phosphorylation of zinc channel ZIP7. Sci Signal 5: ra11. doi:10.1126/scisignal.2002585
    OpenUrlAbstract/FREE Full Text
  69. 69.↵
    1. Taylor KM,
    2. Muraina IA,
    3. Brethour D,
    4. Schmitt-Ulms G,
    5. Nimmanon T,
    6. Ziliotto S,
    7. Kille P,
    8. Hogstrand C
    (2016) Zinc transporter ZIP10 forms a heteromer with ZIP6 which regulates embryonic development and cell migration. Biochem J 473: 2531–2544. doi:10.1042/BCJ20160388
    OpenUrlAbstract/FREE Full Text
  70. 70.↵
    1. Ahern ME,
    2. Bafaro EM,
    3. Cowan A,
    4. Dempski RE
    (2019) Quantifying the oligomeric state of hZIP4 on the surface of cells. Biochemistry 58: 1705–1708. doi:10.1021/acs.biochem.9b00131
    OpenUrlCrossRef
  71. 71.↵
    1. Evans R,
    2. O’Neill M,
    3. Pritzel A,
    4. Antropova N,
    5. Senior A,
    6. Green T,
    7. Žídek A,
    8. Bates R,
    9. Blackwell S,
    10. Yim J, et al.
    (2022) Protein complex prediction with AlphaFold-Multimer. BioRxiv. doi:10.1101/2021.10.04.463034 (Preprint posted March 10, 2022).
    OpenUrlAbstract/FREE Full Text
  72. 72.↵
    1. Kiel C,
    2. Serrano L,
    3. Herrmann C
    (2004) A detailed thermodynamic analysis of ras/effector complex interfaces. J Mol Biol 340: 1039–1058. doi:10.1016/j.jmb.2004.05.050
    OpenUrlCrossRefPubMed
  73. 73.↵
    1. Cheng J,
    2. Novati G,
    3. Pan J,
    4. Bycroft C,
    5. Žemgulytė A,
    6. Applebaum T,
    7. Pritzel A,
    8. Wong LH,
    9. Zielinski M,
    10. Sargeant T, et al.
    (2023) Accurate proteome-wide missense variant effect prediction with AlphaMissense. Science 381: eadg7492. doi:10.1126/science.adg7492
    OpenUrlCrossRefPubMed
  74. 74.↵
    1. Ferrada E,
    2. Wiedmer T,
    3. Wang WA,
    4. Frommelt F,
    5. Steurer B,
    6. Klimek C,
    7. Lindinger S,
    8. Osthushenrich T,
    9. Garofoli A,
    10. Brocchetti S, et al.
    (2024) Experimental and computational analysis of newly identified pathogenic mutations in the creatine transporter SLC6A8. J Mol Biol 436: 168383. doi:10.1016/j.jmb.2023.168383
    OpenUrlCrossRefPubMed
Back to top
Download PDF
Email Article
Citation Tools
SLC39A8 variants and metal uptake
Wen-An Wang, Andrea Garofoli, Evandro Ferrada, Christoph Klimek, Barbara Steurer, Alvaro Ingles-Prieto, Tanja Osthushenrich, Aidan MacNamara, Anders Malarstig, Tabea Wiedmer, Giulio Superti-Furga
Life Science Alliance Jan 2025, 8 (4) e202403028; DOI: 10.26508/lsa.202403028
Twitter logo Facebook logo Mendeley logo
Issue Cover

In this Issue

Volume 8, No. 4
April 2025
Advertisement

Jump to section

Subjects

Related Articles

  • No related articles found.

Cited By...

  • No citing articles found.

More in this TOC Section

Similar Articles