Partial reduced Pi transport function of PiT-2 might not be sufficient to induce brain calcification of idiopathic basal ganglia calcification

Idiopathic basal ganglia calcification (IBGC) is a rare intractable disease characterized by abnormal mineral deposits, including mostly calcium in the basal ganglia, thalamus, and cerebellum. SLC20A2 is encoding the phosphate transporter PiT-2 and was identified in 2012 as the causative gene of familial IBGC. In this study, we investigated functionally two novel SLC20A2 variants (c.680C > T, c.1487G > A) and two SLC20A2 variants (c.82G > A, c.358G > C) previously reported from patients with IBGC. We evaluated the function of variant PiT-2 using stable cell lines. While inorganic phosphate (Pi) transport activity was abolished in the cells with c.82G > A, c.358G > C, and c.1487G > A variants, activity was maintained at 27.8% of the reference level in cells with the c.680C > T variant. Surprisingly, the c.680C > T variant had been discovered by chance in healthy members of an IBGC family, suggesting that partial preservation of Pi transport activity may avoid the onset of IBGC. In addition, we confirmed that PiT-2 variants could be translocated into the cell membrane to the same extent as PiT-2 wild type. In conclusion, we investigated the PiT-2 dysfunction of four SLC20A2 variants and suggested that a partial reduced Pi transport function of PiT-2 might not be sufficient to induce brain calcification of IBGC.


Results
Variants and clinical manifestation. Clinical manifestations. The clinical manifestations are summarized in Table 1. A positive family history of IBGC was present in 3 families.
Familial cases. Case 1 (in family 1). The proband ( Fig. 1a; I-1) was a 64-year-old man with parkinsonism. Resting tremor in his leg was recognized at the age of 63. He showed hyperreflexia in his extremities and mild rigidity in his lower extremities. However, he apparently did not present cognitive impairment and signs of cerebellar ataxia. His CT images revealed severe calcification at the bilateral basal ganglia and faint calcification at the cerebellum dentate nucleus (total calcification score = 18) ( Fig. 1b; I-1). The autopsy report of his elder brother also revealed similar calcification at the age of 79. This autopsy report and family-genetic studies will be published in detail elsewhere. Genetic studies were performed in this family, following the obtainment of written informed consent. Surprisingly, the novel SLC20A2 variant (c.680C > T, p.A227V) in exon 6 was accidentally found in his wife, son, and daughter, but not in the proband himself ( Supplementary Fig. 1a). His wife, son, and daughter had no calcification in the brain ( Fig. 1b; I-2, II-1, and II-2) and showed no abnormalities in the neurological examination.
Familial cases. Case 2 (in family 2). The patient was an 81-year-old woman and had been referred to the department of neurology for evaluation because of colon cancer. She had suffered from right frontal subcortical hemorrhage 3 years before. She presented with left superficial sensory impairment after the stroke and postural hand tremors. Her CT images revealed severe calcification at the bilateral basal ganglia, mild calcification at the cerebellum dentate nucleus, and faint calcification at the pineal bodies, the choroid plexus, and subcortical white matter (Fig. 1c). Her DNA study identified the variant (c.82G > A, p.D28N) in exon 2 ( Supplementary Fig. 1b). This variant was previously reported 17,18 . According to her medical records, her only daughter suffered from schizophrenia and showed calcification at the bilateral globus pallidus (not shown). Her DNA analysis has not been performed yet.
Familial cases. Case 3 (in family 3). The proband and her variant in SLC20A2 have been reported in our previous study 30 . Briefly, this 63-year-old woman presented with bilateral drop feet and gait disturbance due to peripheral neuropathy. The neurological examination showed mild cognitive impairment (MMSE 21), cerebellar ataxia, and spasticity in lower extremities. Her CT images revealed severe calcification at the bilateral basal ganglia and subcortical area, and faint calcification at the thalamus, but none at the cerebellum dentate nucleus (total calcification a I II  www.nature.com/scientificreports www.nature.com/scientificreports/ score = 20, #3 in Fig. 2 of ref. 30 ). Her DNA study showed a variant (c.358G > C, p.G120R) in exon 3. Her son had similar clinical manifestations and CT images (Fig. 1d) and carried the same variant in exon 3.
Sporadic cases. Case 4. The patient was an 83-year-old woman who suffered from neurogenic syncope but showed no other neurological abnormalities, such as dementia and extrapyramidal and cerebellar signs. Her CT images revealed moderate calcification at the bilateral basal ganglia and the cerebellum dentate nucleus and faint calcification at the hippocampus (Fig. 1e). Her DNA study revealed the same variant as found in Case 2 ( Supplementary Fig. 1b). No family member presented with neurological symptoms or calcification in the brain as far as we did investigate. Case 2 and Case 4 live in the Toyama Prefecture in Japan. However, no blood relationship was found between their families in the detailed family survey.
Sporadic cases. Case 5. The patient had been reported in a case report written in Japanese 31 . Briefly, this 43-year-old woman suffered from repetitive psychosis, including delusions, hallucinations, disorganized speech, and grossly disorganized behavior. Her symptoms were always relieved with a small amount of risperidone (1.5-2.0 mg/day) within 1 week. Her CT images revealed severe calcification at the bilateral basal ganglia and moderate calcification at the cerebellum dentate nucleus (total calcification score = 24, Fig. 1f). Her DNA study revealed a variant (c.1487G > A, p.C496Y) in exon 8 ( Supplementary Fig. 1c). There were no other family members presenting with similar neurological symptoms or calcification in the brain as far as we surveyed.

Establishment and confirmation of PiT-2 expression in Flp-In Chinese hamster ovary (CHO) cells.
Among the four SLC20A2 variants described in the present study (Fig. 1g), all can influence PiT-2 function based on the PolyPhen-2 software analysis (data not shown). We constructed Chinese hamster ovary (CHO) cells stably expressing PiT-2-WT and PiT-2 variants D28N, G120R, A227V, and C496Y using the Flp-In system. Protein expression was confirmed by western blots (Fig. 2a). We confirmed that proteins were expressed at the same level in all Flp-In CHO cells. However, an extra band appeared under the PiT-2 band in the PiT-2-G120R group that we assumed to be a protein degradation product due to the unstable variant.
Phosphoric acid transport activities were significantly maintained in the PiT-2-A227V variant that did not show IBGC pathology. We measured Pi transport activities of Flp-In CHO cells using the radioactive tracer 32 P-H 3 PO 4 to evaluate whether PiT-2 missense variants had an effect on Pi transport activities. The Pi transport activities were decreased to the mock level in PiT-2-D28N, G120R, and C496Y cells in www.nature.com/scientificreports www.nature.com/scientificreports/ comparison with that in WT cells (Fig. 2b). However, in PiT-2-A227V cells, the Pi transporter activity significantly increased compared with mock or other variant-expressing cells. In addition, PiT-2-A227V variant showed no IBGC pathology, such as mineralization. Therefore, the Pi transport activity of PiT-2 is a contributing factor causing IBGC pathology.

PiT-2 missense variants translocated to the membrane evaluated by PiT-2 virus receptor activity.
In a previous report, PiT-2 virus receptor activity was measured to evaluate protein folding and membrane translocation of PiT-2 26 . Here, we evaluate the effects of PiT-2 variants on PiT-2 folding and membrane translocation using A-MuLV receptor functions. The green fluorescence protein (GFP) gene was introduced into the viral vector to co-express GFP protein in Flp-In CHO cells infected with A-MuLV (Fig. 3a). The number of GFP fluorescent cells was counted by FACS analysis in 10,000 cells, and the viral receptor function was calculated as the ratio of GFP-positive cells to 10,000 cells. Cells were photographed using a fluorescence microscope before FACS analysis (Fig. 3b). PiT-2 viral receptor functions were significantly higher in PiT-2-D28N and did not change in PiT-2-G120R cells as compared with WT, respectively (Fig. 3c,d). On the other hand, viral receptor functions were reduced by half in PiT-2-A227V and C496Y cells (Fig. 3c,d).
PiT-2 membrane translocation was also shown in PiT-2 missense variants evaluated by HiBiT System. We evaluated the membrane migration of PiT-2 variants more deeply using the HiBiT System in accordance with a previous report 23 . Membrane translocation rates of PiT-2 were measured in cells transiently expressing HiBiT-tagged PiT-2-WT or variants to investigate the effect of PiT-2 missense variants on membrane translocation. HiBiT is a small tag protein consisting of 11 amino acids that binds to LgBiT and emits luminescence by the substrate. Since LgBiT is cell membrane impermeable, protein can be quantified on the membrane surface (Fig. 4a). Membrane translocation rate was calculated as the ratio of membrane protein expression level to total protein expression level. The membrane translocation rate was significantly higher in PiT-2-D28N cells than PiT-2-WT cells (Fig. 4b). In contrast, there was no significant difference of membrane translocation rate in other PiT-2 variant-expressing cells in comparison with PiT-2-WT cells (Fig. 4b).
PiT-2 topology model and variant distributions. A PiT-2 topology model was created using the transmembrane protein display software (TOPO2). The region of I 11 -L 161 (N-PD1131) and the region of V 492 -V 640 (C-PD1131) were emphasized with orange and blue, respectively. Variants reported here are shown in red. Whereas PiT-2-D28N, G120R, and C496Y variants were found in the PD1131 region, the PiT-2-A227V variant was not. (Fig. 5).

Discussion
We have previously reported six new variants in SLC20A2 (four missense variants, one nonsense variants, and one frameshift variants). The four missense variants were predicted to have deleterious damages of protein function in the in silico analysis using PolyPhen-2 4 . In addition, missense variants showed impairment in the Pi transport activity in CHO cells stably expressing them and in endothelial cells developed from patient's iPSCs 23 .
In this study, we have reported two new missense variants in SLC20A2 (c.680C > T, p.A227V, and c.1487G > A, p.C496Y) with their clinical manifestation. One new missense variant was found by chance in the clinically non-affected persons of the family 1 genetic study. Four variants (including the variant c.680C > T, p.A227V) were predicted to cause deleteriousness of protein function based on the PolyPhen-2 result. We performed a functional study using Flp-In CHO cells expressing PiT-2 variants.
Interestingly, the Pi transport activity in PiT-2-A227V cells was maintained at 27.8% of the level of PiT-2-WT cells, whereas in D28N, G120R, and C496Y variants, the activities were lower than 10% of the control level. These results indicate that PiT-2 dysfunction caused by variants leads to the pathogenesis of IBGC and suggest that a partial decrease in Pi uptake might not be sufficient for the onset of IBGC.
PiT-2 is a 12-pass transmembrane protein and sodium-dependent phosphate transporter of type III 11 . A single nucleotide variant of SLC20A2 from IBGC patients can cause loss of Pi transport activity 4,6,15 . The PD1131 region in the N-terminus and the C-terminus of PiT-2 was reported as essential for Pi transport activity 25,26 , and each single variant of SLC20A2 reported previously 6 was found in the PD1131 region. Among the variants studied, PiT-2-D28N and G120R are located in the N-PD1131 region on the N-terminal side, and PiT-2-C496Y is located in the C-PD1131 region on the C-terminal side. Three variants of this study influenced the Pi transport activity and manifested typical clinical features resulting from phosphorus homeostasis impairment. On the other hand, the single nucleotide variant A227V, found in the IBGC patient's families presenting no IBGC clinical features, showed 27.8% of the normal Pi transport activity.
A previous report had shown that deletion of the L183-R254 region of PiT-2, encompassing the A227V variant position, showed a diminished the Pi transport activity 26 . The differences between A227V variant and L183-R254 deletion 26 are likely derived from the different mechanisms of their transporter activities.
Interestingly, the retention of the Pi transport activity in the A227V variant was correlated with no clinical manifestations and no findings in brain CT images, suggesting that a partial decrease in Pi uptake might not be sufficient for the onset of IBGC.
In this study, although the SLC20A2 variants found in IBGC diminished the Pi transport activity, the patients carried heterozygous variants, and both wild-type PiT-2 and variant PiT-2 are theoretically expressed in the body. The mechanism by which SLC20A2 variants lead to IBGC has not yet been elucidated. SLC20A2 variants could lead to either haploinsufficiency or dominant negative effects. Based on the haploinsufficiency hypothesis, half of SLC20A2 expression would decrease the Pi transport activity 32 . On the other hand, a study suggested the dominant negative effect of the SLC20A2 variants encoding D28N, H502A, and E575K on the Pi transport activity 33     www.nature.com/scientificreports www.nature.com/scientificreports/ of patients with IBGC, especially SLC20A2 variants, suggest the disease progresses with the dyshomeostasis of Pi transport in the brain.
The originally reported PiT-2 function was as a retrovirus receptor on the cell surface [26][27][28][29] . However, the correlation between viral receptor activity and Pi transport remains unclear. We assessed the gamma-retroviral receptor function and found that although viral receptor functions were altered by variants of PiT-2, the activity varied between 50% and 160% of WT levels in all variants. Furthermore, the translocation of PiT-2 to the cell surface was assessed using the HiBiT technology to measure the membrane transfer rate of WT PiT-2 and PiT-2 variants. The membrane translocation rate of PiT-2-D28N was significantly increased, whereas those of G120R, A227V, and C496Y were similar to WT. Although the virus receptor function mediated by D28N seems related to membrane surface relocation, the relationships among virus receptor function, Pi transport activity, and/or clinical symptoms remain to be elucidated.
In addition, PiT-2 was reported to have additional functions, including sensing extracellular Pi concentrations and modulating neurite outgrowth 34,35 . However, the relationship between these functions and IBGC pathology has not been recognized. The neurite outgrowth is thought to be associated with loop 7, which is a large intracellular domain of PiT-2 34  In conclusion, we have reported two novel pathogenic variants in PiT-2 and showed their associated dysfunction using CHO cells. Interestingly, in the A227V variant, the Pi transport activity partially remained, and no clinical manifestations of IBGC were found in the carriers. On the other hand, this suggests that the partial increase of PiT-2 activity may lead to improved manifestation in patients with IBGC.

Materials and Methods
Subjects and samples. We collected clinical information on patients with IBGC in a nationwide study supported by a grant for research on intractable diseases from the Ministry of Health, Labor and Welfare of Japan. The diagnostic criteria were described previously 4 . Briefly, patients with causative biochemical abnormalities, including calcium, inorganic phosphate, and intact parathyroid hormone (iPTH), were excluded. The genetic survey and all experiments on human blood samples were approved by the Ethics Committees of Gifu University and Gifu Pharmaceutical University and performed in accordance with Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan, and Ethical Guidelines for Human Genome/Gene Analysis Research in Japan. After written informed consent was obtained, peripheral blood samples were collected. This study was registered to the UMIN Clinical Trials Registry approved by International Committee of Medical Journal Editors (UMIN000030100). Sequence analysis. Genomic DNA was extracted from the patient's blood samples using DNA Quick II (Genomic DNA Isolation kit; DS Pharma Biomedical Co., Ltd.). PCR amplification was performed using the primers whose sequences are listed in Supplementary Table 1. Primers were designed from the NCBI Probe database. For PCR, 0.4 µL template DNA, 0.4 µL of 10 µM forward and reverse primers, 1.6 µL dNTP, 2.0 µL of 10 × Ex Taq buffer, and 0.1 µL Ex Taq HS were mixed with 15.1 µL sterile water. The thermal cycling conditions were as follows: preamplification denaturation (1 cycle), 94 °C for 30 s; amplification (total of 30 cycles), 98 °C for 10 s, Table 1 listed temperature for 30 s, 72 °C for 30 s; and final elongation (1 cycle), 72 °C for 2 min. PCR products were purified with illustra ExoStar (GE Healthcare Life Sciences). The purified PCR products were used for sequencing with an ABI BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) using M13 forward or reverse primers. After ethanol precipitation, automated sequencing was performed on an ABI 3100 Genetic analyzer using both the forward and reverse primers. Variants found in dbSNP, the 1000 Genomes Project database, or the Exome Variant Server database of the National Heart, Lung, and Blood Institute Exome Sequencing Project were excluded. Western blotting. Cells were washed with PBS and lysed in 200 µL RIPA buffer (20 mM Tris-HCl, pH = 7.5, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.1% sodium deoxycholate, 0.1% SDS, 10 µg/mL aprotinin, 10 µg/mL leupeptin, and 1 mM phenylmethylsulfonyl fluoride). Cell extracts were incubated on ice for 20 min and then centrifuged at 14,000 × g for 30 min at 4 °C. Supernatants were collected as protein samples, mixed with SDS sample buffer and incubated overnight at 4 °C. Protein concentration was determined using the Pierce BCA protein Assay Kit (Thermo Fisher Scientific). Proteins were separated by 10% SDS-PAGE for 50 min and transferred to PVDF membrane for 90 min. Blocking was performed with Blocking One (Nacalai Tesque Inc.) for 1 h at room temperature (RT). After blocking, the membrane was incubated at 4 °C overnight with the following mouse monoclonal primary antibodies: ANTI-FLAG M2 (1:1000, Sigma) and β-actin (1:1000, Santa Cruz Biotechnology) dissolved in 5% Blocking One/TBS-T (50 mM Tris-HCl, pH 7.6, 150 mM NaCl, 0.5% Tween-20), and 5% skim milk, respectively. After incubation, the membrane was incubated with the goat anti-mouse secondary antibody conjugated with HRP (1:2000, Santa Cruz Biotechnology) dissolved in 3% Blocking One/TBS-T or 3% skim milk for 60 min at RT. The membrane was incubated in ECL Prime (GE Healthcare) to generate chemiluminescence from HRP antibodies, and chemiluminescence was detected using LAS3000 Mini (Fujifilm). Band density was measured using ImageJ software.
Phosphate uptake assay. All phosphate uptake assays were performed essentially as described 38 . The transport rate was expressed as nmol Pi per minute per mg protein.

Virus infection.
Plat-A cells were seeded in 10 cm dishes and transfected with pMXs-EGFP. Cell culture supernatant was collected and added to the Flp-In CHO cells. Cells were lysed in PBS. The amphotropic receptor function was analyzed by measuring GFP expression cell count rates using a BD FACSVerse flow cytometer (BD Biosciences).
HiBiT extracellular system. CHO cells were seeded in 96-well plates and transiently transfected with pBiT3.1-C-PiT-2 (WT, D28N, G120R, A227V, C496Y, or mock). Translocation rates of PiT-2 were measured using the Nano-Glo HiBiT Extracellular Detection System according to the manufacturer's protocol (Promega). Briefly, HiBiT is a small tag protein of 11 amino acids that binds to LgBiT, and the HiBiT-LgBiT complex emits luminescence upon substrate binding. Because LgBiT is impermeable to the cell membrane, only the HiBiT-LgBiT complex on the cell membrane surface is detected. The membrane translocation rate was calculated as the ratio of the membrane protein expression level to the total protein expression level. The transfection efficiency of each vector was confirmed via the co-transfection of GFP. Nano-Glo HiBiT Extracellular Reagent was prepared in advance and added to the transfected CHO cells. Cells were incubated for 10 min, and luminescence of the membrane protein was measured using the GloMax Multi Detection System (Promega). After measurement, cells were lysed with Triton-X100, and total protein was measured.
Statistical evaluation. Data are given as the mean ± standard error of the mean (SEM). The significance of differences was determined by an analysis of variance (ANOVA). Further statistical analysis for post hoc comparisons was performed using the Bonferroni/Dunn test (Stat View).
Supplementary Information accompanies this paper at doi: