Generation of Fam111a
^−/− mice

The C57BL/6 N Fam111a mouse strains used for this research project were created from embryonic stem (ES) cell clone 19174A-G2, 19174A-A1 generated by Regeneron Pharmaceuticals, Inc. This ES cell clone was subsequently used by the KOMP Repository (https://www.komp.org/) and the Mouse Biology Program (www.mousebiology.org) at the University of California Davis to produce chimera mice. Methods used to create the Velocigene-targeted alleles have been reported previously⁵⁶. The non-conditional allele Fam111a^{tm1(KOMP)Vlcg} was generated by insertion of ZEN-UB1 cassette targeting chromosome 19 in ES cells, which results in the deletion of 4,553 nucleotides (12,584,048–12,588,600; GRCm38.p6) corresponding to the entire protein coding region of Fam111a. ES cell clones were selected based on PCR genotyping and then blastocysts were microinjected into C57BL/6 N background mice. Balb/c donors were used for microinjection to produce the chimera. Sperm rederivation was performed to produce Fam111a^+/− animals. Fam111a^−/− and Fam111a^+/+ mice were generated by breeding Fam111a^+/− mice.

The following primer pair was used for genotyping the wild type (Fam111a^+/+) allele: forward primer (5′-AGCCAAAATGAGCTGTAAGAAGC-3′) and reverse primer (5′-GTGTCATTGTCCCATTCAAC-3′). The following primers were used for genotyping the presence of the ZEN-UB1 cassette: forward primer (5′-TCATTCTCAGTATTGTTTTGCC-3′) and reverse primer (5′-TGAACTTTCTCACATCGTAGC-3′).

Ethics and approval

Animal studies were carried out at the Radboud Institute for Molecular Life Sciences, Nijmegen, The Netherlands. All methods and protocols performed in this study were approved by the local Ethics committee of the Radboud University Nijmegen (RU DEC 2015-0112-006) and the national Ethics committee of the Dutch Central Commission for Animal Experiments (AVD103002016382). This study was performed in compliance with the ARRIVE guidelines. The methods were in accordance with relevant guidelines/regulations, as follows. Fam111a^+/+ mice were used as controls and individual mice were considered experimental units. Sample size was calculated as n = 7 per group based on the outcome measurement of serum magnesium, using a Java Applet⁵⁷. We did not make use of any exclusion criteria and no animals were excluded, but one heterozygous animal died before the study. Randomisation was applied in housing and sample collection. Blinding was performed by giving all animals a unique number, so that the genotypes were not known to the researchers. All genotypes were randomly divided over the cages using the website https://www.random.org/sequences/. Sampling was done per cage. Researchers were blinded to the genotype until the first analyses of results. Mice were housed with maximally 6 animals in standard cages (mouse Eurostandard type IIL) and during the experiments they were individually housed for 48 h in metabolic cages (24 h of acclimatisation, 24 h for actual experiments), after which they were placed back into their own cages. Cage enrichment (bedding, nesting material and igloo) was used in the standard cages, but not the metabolic cages. Specific humane endpoints were defined as follows: (1) The animal has visible signs of discomfort, such as weight loss (more than 15% of the initial weight), hunchback, bad fur and/or bad movement pattern, (2) The animal loses more than 20% weight during housing in a metabolic cage, (3) Specific for the low serum magnesium value: the animal has visible muscle cramps, the animal shows lack of movement, the animal shows a severe abnormal walking pattern (wobbling), the animal has severe exorotation of the heel (indication of muscle weakness) or the animal is not able to hold on to the cage (indication of muscle weakness). These endpoints were monitored daily through weighing and visual monitoring.

Metabolic cages and tissue sampling

Mice (8–10 weeks old) were placed in metabolic cages and left to acclimatise for 24 h. Mice were kept on a normal diet, containing 1% calcium (w/w), 0.22% magnesium (w/w), 0.7% phosphate (w/w) and 1000 IU/kg vitamin D₃. During the next 24 h, the mice were weighed and urine and faeces were collected. In addition, water and food consumption were measured. The following day, the mice were anaesthetised by 4% (v/v) isoflurane inhalation, serum/blood was collected and the mice were sacrificed. Kidney, thyroid glands, duodenum and colon were collected and snap frozen in liquid nitrogen and stored at -80ºC until further analysis.

Serum electrolyte and PTH concentrations

Serum and urine PO₄³⁻, Na⁺ and K⁺ were determined by the clinical laboratory of the Radboud University Medical Center using an automated system (Abbott Diagnostics, Hoofddorp, The Netherlands). Mg²⁺ concentrations were determined by xylidil blue colorimetric assay kit (Roche/Hitachi, Tokyo, Japan) and read at a wavelength of 600 nm on a BioRad plate reader (BioRad, California, USA). Serum and urinary Ca²⁺ concentrations were measured using a chromogen-based colorimetric assay and read at a wavelength of 570 nm (Sigma Aldrich, Zwijndrecht, The Netherlands). The values obtained were calibrated using a Precinorm standard solution (Precinorm U, Roche, Switzerland). Serum albumin was measured using the BCG albumin assay kit (Sigma-Aldrich, Saint Louis, USA) according to the manufacturer’s recommendation and read at a wavelength of 620 nm on a BioRad plate reader (BioRad, California, USA). PTH serum concentrations were measured using the mouse PTH 1-84 ELISA kit (cat. No 60-2305, Quidel, CA, USA) according to the manufacturer’s instructions.

Histology (haematoxylin & eosin)

Tissues were fixed for at least 24 h in 4% (v/v) formalin. Samples were dehydrated in ethanol, embedded in paraffin and cut into sections of 5 μm. The sections were deparaffinised in xylene for 10 min and then rehydrated in graded concentrations of ethanol (100%, 95%, 75%, 50% v/v). The sections were incubated 5 min in haematoxylin (Merck, Amsterdam, Netherlands) and differentiated in tap water after which the sections were stained with eosin (Merck) and dehydrated in in graded concentrations of ethanol (50%, 70%, 96%, 100% (v/v)). Slides were cleared in xylene and mounted using Pertex mounting medium (VWR, Amsterdam, The Netherlands). Tissue slices were imaged using a Zeiss Axiocam 503 color camera mounted onto a Zeiss Axio Imager M2.

RT-qPCR

Total RNA was extracted from kidney and intestinal tissues using TRIzol reagent (Invitrogen, Bleiswijk, The Netherlands) according to the manufacturer’s protocol. The isolated RNA was treated with DNase (Promega, Madison, WI, USA) to prevent contamination from genomic DNA. RNA concentrations and quality were checked using the Nano-Drop 2000c Spectrophotometer (Thermo Fisher Scientific, Breda, Netherlands). To generate cDNA, 1.5 µg RNA was reverse transcribed using Moloney murine leukemia virus (M-MLV) reverse transcriptase (Invitrogen, Breda, Netherlands) at 37 °C for 1 h. Gene expression levels were quantified by SYBR-Green (BioRad, Hercules, CA, USA) on a CFX96 real-time PCR detection system (BioRad) and normalised for Gapdh expression. Relative mRNA expression was analysed using the Livak method (2^−ΔΔCT) and annotated as fold change of expression compared to the Fam111a^+/+ group. Primer sequences are provided in the Supplementary Table S1.

RNAseq

Total RNA was extracted from kidney tissue using TRIzol reagent (Invitrogen) according to manufacturer’s protocol and described earlier. RNA samples from Fam111a^+/+ (n = 4) and Fam111a^−/− (n = 3) male mice were sent for RNA sequencing. Quality control and RNAseq were performed by BaseClear B.V. (Leiden, The Netherlands). Counts of transcripts per sample were generated and quantified using Salmon and expressed as transcripts per million⁵⁸. The reference transcriptome was downloaded from the BioMart repository at Ensembl (Cambridge, UK). The raw and normalized count data were further analysed using R (Vienna, Austria). Raw and processed data files are available in the Gene Expression Omnibus database (GSE196466).

Micro-CT

Femurs were dissected and fixed in 4% (v/v) formalin for 24 h and subsequently kept in 70% (v/v) ethanol. Femurs were scanned using the Skyscan 1076 in vivo X-ray computed tomographer (Bruker microCT, Kontich, Belgium) with a voxel size of 8.88 μm and an exposure time of 2300 ms as described previously⁴⁹. Briefly, X-ray power was set at 40 kV and tube current at 250 mA. Beam hardening (20%) was reduced using a 1 mm aluminum filter, ring-artefacts was set at 5 and an average of three photos (frame averaging) at each angle (0.8°) were taken to generate the final images. Cortical bone parameters were performed in the diaphyseal cortex which included a scan area of 0.45 mm in the femoral center. Trabecular bone measurements were obtained by scanning the distal metaphysis, a total scan area of 1.35 mm from the distal growth plate to the femoral center.

Immunohistochemistry

Immunohistochemistry was performed as previously described⁵⁹. Briefly, co-staining for parv and NCC were performed on 4 μm sections of formalin fixed and paraffin embedded kidney samples. Sections were incubated with blocking buffer (TSA fluorescence system, Perkin Elmer) to prevent unspecific antibody binding. Sections were incubated overnight at 4 °C in rabbit anti-parvalbumin (1:200, Swant, Bio Connect, Switzerland) and sheep anti-NCC (1:200, MRC) in blocking buffer. For detection, sections were incubated with Alexa Fluor-conjugated secondary antibodies (Thermo Scientific, Amsterdam, The Netherlands). Cell nuclei were stained with DAPI (1:20,000) for 10 min at room temperature. Images were taken with an AxioObserver camera and stitched with AxioVision software (Zeiss, Sliedrecht, The Netherlands). NCC or parv positive regions of interest were obtained by thresholding and total positive area per kidney slice was measured using the analyse particle function in Fiji⁶⁰.

Statistical analysis

Data are expressed as mean ± SD. Analysis was done using Graphpad Prism v8 (San Diego, CA, USA). For statistical analysis of three groups, the assumption of equal variances was tested using the Brown-Forsythe test and normality was tested using the Shapiro–Wilk test. For two groups, the assumption of equal variances was tested using an F test. If assumptions were not met, data was first logarithmithically transformed to achieve equal normality. Then, for three groups one-way ANOVA was used to test for differences in the data, followed by a Tukey’s multiple comparison test in case the null hypothesis was rejected. In the case of two groups, an unpaired t-test was used. Differences with a p value of < 0.05 were considered statistically significant. For the RNAseq, additionally, a false discovery rate (FDR) ≤ 0.05 was used to determine statistical significance.