Loss of Function of Slc20a2 Associated with Familial Idiopathic Basal Ganglia Calcification in Humans Causes Brain Calcifications in Mice

Familial idiopathic basal ganglia calcification (FIBGC) is a neurodegenerative disorder with neuropsychiatric and motor symptoms. Deleterious mutations in SLC20A2, encoding the type III sodium-dependent phosphate transporter 2 (PiT2), were recently linked to FIBGC in almost 50 % of the families reported worldwide. Here, we show that knockout of Slc20a2 in mice causes calcifications in the thalamus, basal ganglia, and cortex, demonstrating that reduced PiT2 expression alone can cause brain calcifications.


Introduction
Familial idiopathic basal ganglia calcification (FIBGC), which formerly was called "Fahr's disease" and now also is called primary familial brain calcification, is characterized by symmetrical calcifications of the basal ganglia and other brain regions such as the thalamus and cerebellum (Sobrido et al. 2004). The neuropsychiatric and motor symptoms are very heterogeneous, including dementia, psychosis, parkinsonism, dystonia, and migraine. Brain computed tomography has led to more frequent detection of brain calcification, but FIBGC is a rare genetic disease usually with an autosomal dominant inheritance, and there are now clear genetic culprits. Wang et al. (2012) showed that deleterious mutations in SLC20A2, encoding the type III sodium-dependent inorganic phosphate (P i ) transporter PiT2, were found in affected families from China, Brazil, and Spain. Thereafter, several other kindred carrying SLC20A2 mutations were reported in various countries (Schottlaender et al. 2012;Hsu et al. 2013;Lemos et al. 2013;Nicolas et al. 2013;Zhang et al. 2013). The mutation type varies, and some lead to truncated mRNA that is most likely degraded, and some lead to PiT2 proteins that are shown or predicted to be unable to transport P i Pedersen 2002, 2011;Wang et al. 2012;Hsu et al. 2013;Zhang et al. 2013). Mutations in SLC20A2 can, however, not account for all cases of FIBGC, and mutations in the PDGFRB gene, which encodes a receptor for members of the platelet-derived growth factor family, have also been linked to the disease (Nicolas et al. 2013). It is therefore of great importance to evaluate to which degree the load of PiT2 alone can contribute to the development of brain calcification. To investigate the role of PiT2 in FIBGC, we established Slc20a2 homozygous knockout mice. We here report on 19-week-old Slc20a2 knockout mice with calcifications in the thalamus, basal ganglia, and brain cortex.

Animals
Transgenic C57BL/6N Tac mice (C57BL/6N Tac -Slc20a2 tm1a-(EUCOMM)Wtsi /Ieg (EM: 05549)) were obtained from the European Mouse Mutant Archive, Germany. The strain is heterozygous for the knockout cassette, L1L2-PGK-P, inserted in Slc20a2 after exon 2 and flanking exon 3. It introduces splice acceptor and SV40 polyadenylation sequences in Slc20a2 between the second and third coding exons (http://www. knockoutmouse.org/martsearch/project/24503). The mice were housed in the Health Faculty Animal Facility at Aarhus University, Denmark. All mice were fed the same standard diet ad libitum and handled according to the Danish Animal Welfare guidelines. Two 19-week-old mice homozygous for the knockout cassette and two 22-week-old wild-type (wt) mice from the same heterozygous breeding pair (as controls) were sacrificed. The male and female, which were homozygous for the knockout cassette, showed similar patterns of calcifications in their brains, while no calcifications were detected in the male and female wt mice.
Genotyping DNA was purified from tail tissue by overnight incubation at 58°C in Laird's buffer (Laird et al. 1991) (200 mM NaCl, 100 mM Tris-HCl pH 8.5, 5 mM EDTA pH 8.0, 0.2 % SDS) with proteinase K (no. EO0491; Thermo Scientific). The DNA was then precipitated in 2-propanol and redissolved in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). All mice were genotyped by PCR using the Phusion Hot Start II High-Fidelity DNA Polymerase (no. F-549S; Thermo Scientific). Three primers were used: positioned upstream of, downstream of, and inside the L1L2-PGK-P cassette (LAR3), respectively. The sequences of the primers were as follows: The PCR products were separated by agarose gel electrophoresis using the GeneRuler 100 bp Plus DNA Ladder (Fermentas) as a size marker. The wt and knockout products are 523 and 443 bp, respectively.

Quantitative Reverse Transcription PCR Analysis
Cultured mouse tail fibroblasts were lysed using lysis/binding buffer containing guanidinium thiocyanate provided with the Fig. 1 Genotyping and Slc20a2 mRNA levels. a Genotyping of breeders heterozygous for the knockout cassette (−/+) and of homozygous (−/−) and wt (+/+) offspring. Black arrow, 500-bp size marker. White arrows, 523bp PCR product derived from wt allele (upper) and 443-bp PCR product derived from allele with inserted knockout cassette (lower). b Tail fibroblasts from a Slc20a2 knockout mouse (−/−) and three wt mice were lysed, and Slc20a2 mRNA levels were evaluated by qRT-PCR using primers positioned downstream of the knockout cassette. Slc20a2 mRNA levels relative to B2M are shown. Error bars represent standard deviations RNAqeous®-4PCR Kit (Applied Biosystems). RNA was purified from the lysates using the RNAqeous®-4PCR Kit as described by the manufacturer and quantified by UV spectrometry. Immediately after purification, it was reverse transcribed to cDNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). For quantitative PCR analysis, the following TaqMan® Gene Expression Assays (Applied Biosystems) were employed: Mm00660204 (Slc20a2) and, as endogenous control, Mm00437762 (beta-2microglobulin (B2M)). The individual quantitative PCR (qPCR) reactions contained 10 μl TaqMan Fast Universal PCR Master Mix, 1 μl TaqMan Gene Expression Assay (Slc20a2 FAM), 1 μl TaqMan Gene Expression Assay (B2M VIC), and 8 μl cDNA (approximately 10 ng). The PCR cycles employed were as follows: 95°C for 10 min and 40 cycles of 95°C for 1 s and 60°C for 20 s. The relative mRNA levels were calculated as described in Hellemans et al. (2007). Each qPCR reaction was set up as triplicates.

Histology
Brains were fixed in 4 % paraformaldehyde (formalin 10 % buffered (VWR)) for at least 2 days and then embedded in paraffin. Histological stainings were performed on 4-μm sections Calcifications were also found in other brain regions, as seen by Von Kossa staining of the basal ganglia (f) and cortex (g) (scale bars, 100 μm) from formalin-fixed, paraffin-embedded tissue. The sections were deparaffinized in xylene and rehydrated.
For detection of calcium, sections were stained with the Von Kossa staining technique. The sections were incubated 20 min in 10 % silver nitrate, rinsed in sterile water, followed by 2 min in 5 % pyrogallol. The sections were rinsed again in sterile water and, afterwards, were placed in 5 % sodium thiosulfate for 5 min. For counterstaining, eosin 2.1 % was used.
For immunohistochemical staining, the sections were emerged in H 2 O 2 for 10 min to abolish endogenous peroxidase activity. Antigen retrieval was performed by boiling in TEG (10 mM Tris-HCl, 0.5 mM EGTA, pH 9.0) in a microwave oven for 15 min. F4/80 was detected using a rat antibody to murine F4/80 (1:100; AbD Serotec, Bio-Rad). As secondary antibody, a rabbit antibody to rat immunoglobulins (IgG1, IgG2a, and IgG2b) (1:200, E0468; Dako) was used. The sections were then exposed to a goat antibody to rabbit immunoglobulins conjugated with a peroxidase-labeled polymer (Dako EnVision®+ System-HRP Labelled Polymer Anti-Rabbit, K4003; Dako). The presence of peroxidase activity was visualized by incubation with DAB+ (K3468; Dako). Lastly, the sections were stained with Mayer's hematoxylin (Ampliqon), to visualize cell nuclei. In between each step, the sections were washed in TNT buffer (Fagron).
Hematoxylin-eosin (HE) staining was performed using a standard protocol. Periodic acid-Shiff (PAS) staining was performed using a standard protocol; Mayer's hematoxylin was used as counterstain.

Results
To establish Slc20a2 homozygous knockout mice, we used the Slc20a2 tm1a(EUCOMM)Wtsi allele on a C57BL/6N Tac background obtained from the European Mouse Mutant Archive. The targeting vector introduced a splice acceptor and SV40 polyadenylation sequences between the second and third coding exons in Slc20a2, which leads to a premature stop of transcription. The truncated RNA lacks eight downstream coding exons and does not code for any functional gene. Breeding of heterozygous Slc20a2 tm1a(EUCOMM)Wtsi mice resulted in viable mice homozygous for the mutation (Fig. 1a) in a Mendelian ratio. To test for knockout of Slc20a2, we used primers positioned downstream of the knockout cassette and quantified the expression level by a quantitative reverse transcription (RT)-PCR (qRT-PCR) analysis. As seen in Fig. 1b, the mouse homozygous for the knockout cassette shown in Fig. 2 has no expression of Slc20a2 downstream of the cassette. In sections from the brain, we found extensive, bilateral irregular aggregates of calcified spheres in the thalamus using Von Kossa staining (Fig. 2a). The spheres were about 10 μm in diameter, and the size of the aggregates was about 50-100 μm (Fig. 2c, d). A few calcifications were also found in the basal ganglia (Fig. 2f) and in the cortex (Fig. 2g). There were no calcifications observed in the cerebellum (not shown). No calcifications were detected in wt controls (Fig. 2b). Tissue reaction to the calcification was demonstrated by immunohistochemistry showing F4/80 positive microglia surrounding the individual elements of the calcified aggregates (Fig. 2e). No other lesions were observed in the brain, and there was no additional focal increase in microglia. The calcifications in FIBGC are believed to initiate in or around blood vessels (Sobrido et al. 2004). Interestingly, PAS staining suggest that this is also the case in the homozygous Slc20a2 knockout mouse (Fig. 3). The results show that knockout of Slc20a2 alone is sufficient to cause FIBGC-like calcifications in mice and support the significance of the linkage between mutations in SLC20A2 and FIBGC.

Discussion
PiT2 has a paralog, PiT1, encoded by the gene SLC20A1. The two ubiquitously expressed P i transporters (Uckert et al. 1998) share some transport characteristics (Bøttger et al. 2006), but differ in others (Bøttger et al. 2006;Byskov et al. 2012). PiT1 is shown to be involved in calcification in the medial layer of blood vessels (Li et al. 2006;Koleganova et al. 2009), and it is well-accepted that the calcification is a cell-mediated process but debated to which degree vascular smooth muscle cells and pericytes contribute to the calcification. Since the calcifications in FIBGC are believed to initiate in or around blood vessels (Sobrido et al. 2004), and the results on the Slc20a2 knockout mouse are in agreement with this, it is tempting to speculate that the mechanism is similar to the one involving PiT1, where vascular smooth muscle cells transdifferentiate to mineralizing osteoblast-like cells, possibly caused by elevated extracellular P i levels (Jono et al. 2000) or by contact with precipitated calcium phosphate products (Sage et al. 2011;Villa-Bellosta et al. 2011). Medial vascular calcification is strongly linked to increased levels of P i in the blood (Covic et al. 2009). In FIBGC, the blood P i levels are within the normal range (Sobrido et al. 2004). However, it is possible that a decrease in functional PiT2 leads to a local increase in extracellular P i . SLC20A2 and Slc20a2 have the highest expressions in the affected areas (Lagrue et al. 2010;Silva et al. 2013), and it is, therefore, likely that PiT2 is especially important for maintaining phosphate homeostasis in the affected areas and that only these experience an increase in extracellular P i concentration. In these areas, the local increase in P i concentration alone might trigger a cell-mediated mineralization process, or it may lead to passive precipitation of calcium phosphate products, which might trigger a cell-mediated mineralization process, thus leading to an escalation of the calcification. However, the actual mechanism for the development of brain calcifications in the close to 50 % of FIBGC-affected families with an established link to mutations in SLC20A2 yet remains to be identified. We therefore expect that the Slc20a2 knockout mouse will develop into an important tool for the study of development of brain calcifications and as a preclinical model for drug testing.