Memo1 gene expression in kidney and bone is unaffected by dietary mineral load and calciotropic hormones

Abstract Mediator of cell motility 1 (MEMO1) is a ubiquitously expressed modulator of cellular responses to growth factors including FGF23 signaling, and Memo1‐deficient mice share some phenotypic traits with Fgf23‐ or Klotho‐deficient mouse models. Here, we tested whether Memo1 gene expression is regulated by calciotropic hormones or by changing the dietary mineral load. MLO‐Y4 osteocyte‐like cells were cultured and treated with 1,25(OH)2‐vitamin D3. Wild‐type C57BL/6N mice underwent treatments with 1,25(OH)2‐vitamin D3, parathyroid hormone, 17β‐estradiol or vehicle. Other cohorts of C57BL/6N mice were fed diets varying in calcium or phosphate content. Expression of Memo1 and control genes was assessed by qPCR. 1,25(OH)2‐vitamin D3 caused an acute decrease in Memo1 transcript levels in vitro, but not in vivo. None of the hormones tested had an influence on Memo1 transcripts, whereas the assessed control genes reacted the expected way. Dietary interventions with calcium and phosphate did not affect Memo1 transcripts but altered the chosen control genes’ expression. We observed that Memo1 was not regulated by calciotropic hormones or change in mineral load, suggesting major differences between the regulation and physiological roles of Klotho, Fgf23, and Memo1.


| INTRODUCTION
Memo1 is an evolutionary conserved protein in all kingdoms of life that has shown intracellular expression in cytoplasma and nucleus (Haenzi et al., 2014;Moor, Haenzi, et al., 2018;Schlatter et al., 2012). A conditional and inducible knockout mouse model with postnatal deletion of exon 2 of the Memo1 gene has resulted in a syndrome of aging and premature death with traits such as elevated calcemia, elevated FGF23 and 1,25(OH) 2 -vitamin D 3 , bone disease, lung emphysema, atrophy of subcutaneous fat, insulin hypersensitivity, and renal insufficiency (Haenzi et al., 2014;Moor, Ramakrishnan, et al., 2018). This phenotype significantly overlaps with phenotypes of mouse models deficient in KLOTHO (Kuro-o et al., 1997) or FGF23 (Shimada et al., 2004), two proteins which have tremendously reshaped our understanding of the regulation of calcium and phosphate metabolism by the kidney and bone and to a lesser extent also the intestine (Hu, Shiizaki, Kuro-O, & Moe, 2013;Moor & Bonny, 2016).
In addition, evidence from cell culture experiments investigating co-localization and phosphorylation status of adaptor proteins suggested that mediator of cell motility 1 DOI: 10.14814/phy2.14410 (MEMO1) protein participates in and modulates a signaling cascade involving FGF23 and the FGFR (Haenzi et al., 2014).

| Animal experiments
C57BL/6N mice were obtained from Janvier. Mice were held in a conventional animal facility with up to six animals per cage and they were fed a standard laboratory chow (Kliba Nafnag TS3242; calcium 1%, phosphorus 0.65%, magnesium 0.23%, vitamin D 1,600 IU/kg, vitamin A 27,000 IU/kg, vitamin E 150 mg/kg, protein 18.8%, crude fat 5.6%, crude fiber 3.5%, lysine 1.1%; KLIBA) unless specified otherwise and were kept on 12/12 (experimental) or 14/10 (breeding) light-dark cycles. All animal experimental protocols were approved by the State Veterinary Service of the Canton de Vaud, Switzerland. For all mouse studies, sample sizes were considered based on previous results in our laboratory.

| Hormone injections
All hormone injections were performed in the home cage of the mice after random treatment allocation of individual ear-marked mice. For 1,25(OH) 2 -vitamin D 3 treatment, male C57BL/6N mice aged 13-15 weeks were subcutaneously injected with 2 μg/kg body weight 1,25(OH) 2 -vitamin D 3 (Sigma D1530) in ethanol 1% in NaCl 0.9%. The dose and application route was the same as previously used in our laboratory. Control mice were injected with 1% (v/v) ethanol in NaCl 0.9%. Mice were sacrificed 6 hr after injection.
For estradiol treatment, male C57BL/6N mice aged 16 weeks received one daily subcutaneous injection of 15 μg 17β-Estradiol (Sigma E8875) in ethanol 0.1% (v/v) in NaCl 0.9% for five consecutive days and were sacrificed 4 hr after the last injection. The dose per body weight and the drug application route was derived from the literature, as shown Van Abel et al. (2002) to induce Trpv5 expression. Control mice were subcutaneously injected with 0.1% ethanol in NaCl 0.9% for 5 days.

| Mouse dissection
For euthanasia, mice were intraperitoneally injected with 0.1 mg/gBW of ketamine (Ketanarkon 100 Vet., Streuli) and 0.02 mg/gBW of xylazine (Rompun, Bayer), followed by terminal exsanguination by orbital puncture under full anesthesia and/or by cervical dislocation. Organs were collected, kidneys were cut in half, and organs were snap frozen in liquid nitrogen immediately, followed by storage at −80°C until further use.

| RNA extraction
RNA was extracted using TRI reagent (Applied Biosystems by Life Technologies) according to manufacturer's instructions. RNA pellets were dried and dissolved in RNase-free H 2 O. RNA concentration was measured photometrically using NanoDrop (NanoDrop 2000, Thermo Fisher Scientific). RNA A260/A280 ratio was assessed and each RNA sample was visualized on a 1% agarose gel.
RNA was reverse transcribed to cDNA using the PrimeScript RT reaction kit (Takara Bio Inc). RNA input quantities per sample were 1-2 μg for bone, 500 ng for kidney or 1 μg of MLO-Y4 RNA. The resulting cDNA mix was diluted 2-12x depending on tissue type.

| qPCR
For quantitative gene transcript expression analysis, 2 μL of cDNA was used for SYBR Green qPCR (Applied Biosystems by Life Technologies) on a 7500 Fast machine (Applied Biosystems). Samples were run in triplicate in 20 μL total volume for each gene, and actin or GAPDH was used for normalization. Melting curves were obtained for every run. Program settings were: 95°C during 20 s, 40 cycles (95°C 3 s, 60°C 30 s), and for melting curve stage: 95°C 15 s, 60°C 1 min, rising at 1% ramp speed to 95°C (15 s), and 60°C 15 s. Data were analyzed using the delta-delta CT method. Primers were ordered from Microsynth (Switzerland) and sequences are shown in Table 1. All amplified products were visualized on agarose gels.

| Data analysis
Human Memo1 promoter sequences were analyzed in silico using the UCSC Genome Browser and Serial Cloner 2.6.1.
Data from experiments with two independent groups were analyzed by t test or Mann-Whitney U test. For comparison of three groups, Kruskal-Wallis test was used with Bonferroni's Multiple Comparison posttest. All statistical analyses were conducted using GraphPad PRISM 5.03. Two-sided p < .05 were considered significant.

| RESULTS
Memo1-deficient mice resemble by some traits Klotho mutant and FGF23 KO mice (Haenzi et al., 2014), and the promoters of Klotho and FGF23 harbor regulatory sequences that can be bound by vitamin D receptors (Forster et al., 2011;Orfanidou et al., 2012). For these reasons, we determined the regulation of Memo1 gene expression by minerals and calciotropic hormones. We have previously shown that MLO-Y4 osteocyte-like cell line expressed Memo1 transcripts and protein (Moor, Ramakrishnan, et al., 2018).

| Memo1 is not regulated by dietary calcium
Next, we determined the effect of varying dietary calcium content for 7 days on Memo1 gene expression. RNA was obtained from a previous experiment performed in our lab (366). In the kidney (Figure 3a) and in the tibia (Figure 3b) of these mice exposed to three different calcium-containing diets (0.17%, 0.82%, and 1.69%), no change in Memo1 gene expression was observed. A 2.5-fold increase in gene expression of Casr encoding the calcium-sensing receptor in the bone upon dietary calcium restriction serves as an experimental control for the dietary intervention and was reported for the samples we used in (O'Seaghdha et al., 2013).

| Memo1 is not regulated by dietary phosphate
We investigated the influence of different systemic phosphate loads on Memo1 expression. We showed that different dietary phosphate contents (0.2%, 0.8%, 1.5%) did not significantly affect Memo1 transcript levels in kidney (Figure 4a) or in the tibia (Figure 4b). As an experimental control gene we used renal transcripts of Slc34a3 encoding sodium-dependent phosphate transporter type 2c (NaPi2c). Renal Slc34a3 was increased, as expected, under low phosphate and decreased under high phosphate diets (Figure 4c).

| Memo1 unchanged by PTH
To determine the effect of PTH on Memo1, human PTH fragments 1-34 were subcutaneously injected to wild-type mice, and the animals were euthanized after 2 hr. Memo1 gene expression in kidney (Figure 5a) or in tibia (Figure 5b) remained unchanged upon PTH treatment. Transcripts of Cyp27b1, the gene coding for the renal vitamin D activating enzyme cytochrome P450 27b1 were increased upon PTH compared to NaCl 0.9%-treated controls (Figure 5c).

| Memo1 unchanged by estradiol
As sex hormones exert effects on both renal calcium transport proteins (van Abel et al., 2002) and FGF23 in the bone (Carrillo-Lopez et al., 2009), we tested if Memo1 is a target gene induced by estradiol. We subcutaneously injected 17β-estradiol once daily over 5 days. This induced the expression of the control gene F3 encoding coagulation factor III (Figure 6c), but gene expression of Memo1 in the kidney (Figure 6a) and in the bone (Figure 6b) both remained unchanged compared to mice injected with vehicle.

F I G U R E 2 Memo1 transcripts
were not changed by 1,25(OH) 2 -vitamin D 3 treatment in wild-type mice. Memo1 transcripts were assessed in kidney (a) and tibia (b) and were indifferent 6 hr after 1,25(OH) 2 -vitamin D 3 injection, whereas renal Cyp24a1 transcripts were over 10-fold increased in comparison to vehicle control (c). Fgf23 gene expression in tibia was increased by 1,25(OH) 2 -vitamin D 3 compared to vehicle (d). n = 5 per condition, *p < .05 (Mann-Whitney U test) F I G U R E 3 Memo1 transcript levels were not influenced by varying dietary calcium contents. Memo1 transcripts in kidney (a) and tibia (b) were not significantly affected by different dietary calcium intakes; n = 5 per diet (Kruskal-Wallis tests). Calcium-sensing receptor increased 2.5fold in bone of mice on 0.17% calcium diet published in (O'Seaghdha et al., 2013) using the same samples serves as an experimental control To summarize, we found a small but significant decrease in Memo1 expression upon 1,25(OH) 2 -vitamin D 3 exposure in vitro, but we failed to detect any major regulation of Memo1 transcript abundance upon mineral load or calciotropic hormone treatment in vivo.

| DISCUSSION
MEMO1 is expressed in the kidney where it plays an intrarenal role in the regulation of calcium transporters (Moor, Haenzi, et al., 2018). In the bone, MEMO1 is expressed in all cell types (Moor, Ramakrishnan, et al., 2018), but its precise bone-specific function remains elusive.
Here we tested the hypothesis whether Memo1 is regulated by key players in mineral homeostasis such as calciotropic hormones or dietary calcium or phosphate. As a readout, we chose Memo1 gene expression in an osteocyte-like cell line, and in bone and kidney tissues. For each intervention, an experimental control gene was assessed and revealed effects similar as shown before by others.
We observed that Memo1 gene expression was diminished in the osteocyte-like cells upon 1,25(OH) 2 -vitamin D 3 F I G U R E 4 Memo1 transcript levels were not affected by varying dietary phosphate contents. Memo1 transcripts abundance in kidney (a) and in tibia (b) were not significantly changed by dietary phosphate contents. Neither bone Memo1 (b). Renal Slc34a3 transcripts were used as an experimental control gene and were affected by dietary phosphate content (c). N = 5 per condition; *p < .05 (Kruskal-Wallis test with Dunn's multiple comparisons correction) F I G U R E 5 Memo1 gene expression remained unchanged upon PTH treatment in wild-type mice. Human PTH1-34 or NaCl 0.9% vehicle was injected 2 hr prior to dissection and transcripts of Memo1 in kidney (a) and tibia (b) were unchanged between experimental conditions. Transcripts of Cyp27b1 (c) were increased by PTH1-34. n = 6 per condition; ns, not significant; *p < .05 (Mann-Whitney U test). PTH, parathyroid hormone treatment. However, in bone and tissues, we failed to detect any effect on Memo1 by all interventions that we studied. This shows a major difference between Memo1 and the most studied contributors to mineral homeostasis. As examples in the kidney, Type II sodium-dependent phosphate cotransporters are regulated by dietary phosphate supply (Bourgeois et al., 2013) and gene expression of Trpv5 encoding a renal calcium transport protein is controlled by 1,25(OH) 2 -vitamin D 3 (Hoenderop et al., 2001). As examples in the bone, Fgf23 expression is stimulated by 1,25(OH) 2 -vitamin D 3 (Liu et al., 2006) or PTH (Kawata et al., 2007), while dietary phosphate restriction or renal phosphate-wasting disorders reduce Fgf23 expression (Ansermet et al., 2017;Schlingmann et al., 2016;Vervloet et al., 2011). Even intravenous calcium loading in rats increased Fgf23 expression in bone and hormone concentrations in the serum (Shikida et al., 2018).
This study contains some limitations: The current interventions were confined to the analysis of gene expression, but we did not directly assess Memo1 promoter activity using a reporter construct. Such an approach would more sensitively discriminate and would allow validating putative response elements in the Memo1 promoter. In addition, our in silico analysis of the presumed promoter sequence did not allow base mismatches compared to known response elements. However, we argue that a physiologically relevant the regulation of Memo1 gene expression, if present, should have been visible using the experimental approaches that were undertaken.
Further, we have assessed a single but physiologically reasonable time point, and only a narrow selection of tissues and cells. In addition to bone and kidney, the intestine would be another major turnover place for minerals. Memo1 expression and potential regulation in healthy intestine have not been investigated so far. In colorectal cancer cells Memo1 promoter activity is increased in response to the transcription factors Aryl hydrocarbon receptor/ Aryl hydrocarbon receptor nuclear-translocator complex, indicating some intestinal disease relevance (Bogoevska et al., 2017).
Another limitation is the fact that we investigated only mice of male sex as to simplify experimental planning, reduction of mice numbers used, and as to reproduce the hormone injection protocols in the cited references, including the estradiol injection protocol (van Abel et al., 2002). Future experiments should be conducted with both sexes independently to allow the detection of sex-specific effects.
Finally, as Memo is a redox enzyme with incompletely understood reaction partners and substrates (MacDonald et al., 2014), an assessment of posttranslational regulation such as by the oxidative modification of MEMO1 protein, subcellular localization, or changes in its putative enzymatic activity may be helpful to investigate a regulation of Memo1.
To conclude, besides a minor effect in bone cells stimulated with 1,25(OH) 2 -vitamin D 3 , we did not detect a major regulation of Memo1 gene expression upon minerals and calciotropic stimuli in bone and kidney, two organs relevant for mineral homeostasis. Further studies inquiring the regulation of this and similar genes may contribute to the understanding of the regulation of mineral homeostasis in health and renal and bone diseases.

ACKNOWLEDGMENTS
OB and MBM's work was supported by the Swiss National Science Foundation through the special program NCCR F I G U R E 6 Memo1 transcripts remained unchanged upon 17β-estradiol treatment. Memo1 transcripts assessed by qPCR in kidney (a) and tibia (b) were unchanged after five daily subcutaneous injections of 17β-estradiol compared to vehicle. Renal gene expression of tissue factor F3 was increased by 17β-estradiol (c). n = 4 estradiol and n = 5 control condition. ns, not significant; *p < .05 (Mann-Whitney U test) Kidney.CH and by unrestricted grants from the Association pour l'Information et la Recherche sur les maladies rénales Génétiques (AIRG)-Suisse and from the Novartis Foundation. The authors are thankful to Candice Stoudmann and Finola Legrand for assisting with an experiment.