Animal experiments

Six-week-old male Wistar rats were purchased from Japan SLC (Hamamatsu, Japan). For all experiments, rats aged 6–8 weeks were used. All the rats had ad libitum access to water and a regular chow diet containing 1.07% calcium and 0.83% phosphorus (MF diet, Oriental Yeast, Tokyo, Japan). All animal experiments were done in the morning without fasting. When performing invasive procedures, the rats were anesthetized with an intraperitoneal injection of medetomidine, midazolam, and butorphanol and were placed on a heated table to maintain body temperature between 36 and 38 °C. The Animal Committee of Osaka University approved all experiments (approval number: 01-046-005), and the study was performed under standard guidelines regarding the use of animals in scientific experiments and was reported in accordance with ARRIVE guidelines.

Administration of radiolabeled sodium phosphate through the duodenum

For Duo group rats, a 24-gauge catheter (SR-FS2419, TERUMO, Tokyo, Japan) was inserted into the duodenum 1 cm anorectally from the gastric pylorus under open surgery. A Na₂HPO₄/NaH₂PO₄ solution (phosphate 80 mM, sodium 154 mM, pH 7.4, 1.0 ml/animal) containing trace amounts of radiolabeled phosphate (³²P) (1.0 × 10^–5 µmol/ml, 3700 Bq/ml, Perkin Elmer, MA, USA) was administered through the catheter. To prevent dehydration, the surgical wound was covered with gauze soaked in saline solution warmed to 36–38 °C. Blood and tissue samples were solubilized and color quenched by incubating them in a solution containing 1.0 ml of 60% perchloric acid (Nacalai Tesque, Kyoto, Japan) and 0.4 ml of 10% hydrogen peroxide (Wako) at 60 °C for 3 h. The levels of ³²P were measured using Cerenkov counting (AccuFLEX LSC 7400, Hitachi Health care Manufacturing, Chiba, Japan).

Sodium phosphate administration through the portal vein or inferior vena cava

Effects of phosphate administration through the portal vein or inferior vena cava (referred to as phosphate intervention) were evaluated. To access blood vessels for the phosphate injections, a catheter (SR-FS2419) was inserted into the inferior vena cava or portal vein. To avoid dehydration, normal saline (1.0 ml/animal) was injected before the phosphate interventions. The surgical wound was covered similarly to the Duo group. If group names do not contain ³²P, a Na₂HPO₄/NaH₂PO₄ solution (phosphate 20 mM, sodium 154 mM, pH 7.4, 1.0 ml/animal) was administered through the inferior vena cava and portal vein in the groups whose group name contains IVC and groups whose group name contains PV, except for the PV-100 and PV-500 groups, respectively. Normal saline (1.0 m/animal) was administered through the portal vein in the groups whose group name contains Ctrl. A Na₂HPO₄/NaH₂PO₄ solution containing ³²P (phosphate 20 mM, ³²P 3700 Bq/ml, sodium 154 mM, pH 7.4, 1.0 ml/animal) was administered in the groups whose group name contains IVC-³²P and the groups whose group name contains PV-³²P similarly as in the IVC and PV groups, respectively. Blood samples were collected from the inferior vena cava.

Blood and urine collection procedures

For the Duo group, catheters (SR-FS2419) were inserted into the inferior vena cava and portal vein to collect blood samples. For the other groups, i.e., those receiving phosphate interventions through blood vessels, a 24-gauge catheter was inserted into the inferior vena cava to collect blood samples. At the indicated time points, 0.1 ml of blood was discarded, followed by 0.1 ml of blood collection. After each blood collection procedure, 0.2 ml of saline was administered through the catheters. Blood samples were centrifuged in heparin tubes for 15 min at 1200 × g to obtain plasma samples. To obtain serum samples, blood samples were allowed to clot at room temperature and then centrifuged for 15 min at 1200 × g. Urine samples were collected by puncturing the bladder, followed by centrifugation for 15 min at 1200 × g. Plasma, serum, and urine samples were stored at − 80 °C until analysis. FEPi was calculated using the formula: (urine phosphate (mg/dl) × serum creatinine (mg/dl)) × 100/(serum phosphate (mg/dl) × urine creatinine (mg/dl)). Ccr was calculated using the formula: (urine creatinine (md/dl) × urine volume (ml/min)) / serum creatinine (mg/dl).

Pretreatments

Rats with various pretreatments received the phosphate interventions. Groups were named according to the combination of pretreatment and phosphate intervention types. The details of each group’s intervention are summarized in Supplementary Table S1 online.

A bilateral nephrectomy (BNX) was performed by removing both kidneys after ligating the renal artery, renal vein, and urinary tract. Partial hepatectomy (PH) was achieved by surgically removing the median and left lateral lobes. A Kupffer cell depletion model or sinusoidal endothelial cell dysfunction model was prepared by intraperitoneally injecting gadolinium chloride (GdCl₃) (40 mg/kg, Sigma‒Aldrich) or monocrotaline (MCT) (200 mg/kg, Sigma‒Aldrich) dissolved in 1.0 ml of normal saline, respectively. To inhibit D1-like dopamine receptor, SCH23390 dissolved in 1.0 ml of normal saline was intraperitoneally injected (500 µg/kg BW, Sigma‒Aldrich). Vehicle-pretreated rats received intraperitoneal injections of normal saline (1.0 ml/animal). Bilateral renal denervation (RDN) was performed by surgical denervation with a topical application of 10% phenol dissolved in ethanol. Hepatic nerves were ablated (HDN rats) by applying a 10% phenol solution to the hepatic artery, portal vein, and common bile duct using a cotton swab.

Biochemical measurement

Levels of phosphate, calcium, and creatinine were measured using each assay kit (Wako, Osaka, Japan). Plasma i-PTH and i-FGF23 concentrations were measured using Rat BioActive Intact PTH ELISA (Immutopics, CA, USA) and FGF23 ELISA kits (Kainos, Tokyo, Japan), respectively. Serum samples were sent to SRL (Tokyo, Japan) to measure serum 1,25VD levels.

Antibodies

Antibodies against specific molecules were obtained as follows: CD68 (catalog: ab25212, Abcam Cambridge, UK), hepatic sinusoid endothelial cells (SE-1) (catalog: NB110-68095, Novusbio, Colorado, USA), Alexa Fluor 555 donkey anti-rabbit IgG antibody (catalog: A-31572, Invitrogen), endothelial nitric oxide synthase (eNOS) (catalog: #9572, Cell Signaling, Danvers, USA), phospho-eNOS Thr495 (catalog: #9574, Cell Signaling), and horseradish peroxidase (HRP)-conjugated 2^nd Ab for WB (catalog: P0448, DAKO, Santa Clara, USA).

A polyclonal antibody against sodium-phosphate cotransporter 2a (NaPi-2a) was raised in rabbits by immunization with rat NaPi-2a peptide (MMSYSERLGGPAVSP)²⁵.

Histology and immunostaining

Tissues were fixed with 4.0% paraformaldehyde in phosphate-buffered saline (PBS) (pH 7.4) for 6 h, cryoprotected with 30% sucrose in PBS for two days, and then embedded in optimal cutting temperature compound (Sakura, Tokyo, Japan). Freshly prepared 0.1% sodium borohydride (catalog: 452882-5G, Sigma‒Aldrich, MO, USA) in Tris-buffered saline (TBS) was applied for 30 min to reduce background autofluorescence as needed. Tissue sections (10 µm) were incubated with a blocking solution (1.5% horse serum (Vector laboratories, California, USA) in PBS or 3% bovine serum (Sigma‒Aldrich, MO, USA) in PBS) followed by primary antibody reactions overnight at 4 °C. The sections were washed three times in PBS and then incubated with Alexa Fluor 555-conjugated secondary antibodies at room temperature for 1 h. FITC-phalloidin (Invitrogen, 0.066 µM) was added to the secondary antibody solution where indicated. Again, the sections were washed three times in PBS and mounted with VectaShield (Vector Laboratories, California, USA). The dilution ratios of the primary and secondary antibodies were as follows: CD68 1:150, SE-1 1:150, NaPi-2a 1:200, and Alexa Fluor 555 1:200. Immunofluorescence images were obtained using confocal microscopy (FV-1000D or SpinSR10, Olympus, Tokyo, Japan).

Western blot analyses

Proteins were extracted from the aorta using the RIPA Lysis Buffer system (Santa Cruz, Texas, USA). Extracted protein samples were diluted in 2 × electrophoresis sample buffer (Santa Cruz) containing dithiothreitol. Each sample was subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS‒PAGE) using a 10% acrylamide gel and electroblotted onto a polyvinylidene difluoride (PVDF) membrane (Amersham, Germany)²⁶. The membrane was cut prior to hybridization to reduce the amount of antibodies used. Then, the membrane was incubated with a blocking solution (5% bovine serum albumin (BSA) in TBS containing 0.1% Tween 20 (TBS-T)) followed by anti-eNOS or anti-phospho-eNOS Thr495 antibody reactions overnight at 4 °C. The membranes were washed three times in TBS-T and then incubated with secondary antibodies at room temperature for 1 h. Again, the membranes were washed three times in TBS-T. Signals were detected by enhanced chemiluminescent reagents (Bio-Rad Laboratories). Images were obtained by using ChemiDoc Touch (Bio-Rad Laboratory). The dilution ratios of antibodies were as follows: eNOS 1:1000, phospho-eNOS Thr495, 1:1000, and HRP secondary antibody 1:3000. Full immunoblot images are shown in Supplementary Figs. S11–S12 online.

Analysis of calciprotein particles (CPPs)

To measure ³²P incorporation levels to the CPPs in the BNX rats, serum was processed as described previously²⁷. Fifty-microliter aliquots of serum in 1.5-ml tubes were centrifuged at room temperature at 16,000 × g for 2 h. After discarding the supernatant, each tube was rinsed with 100 µl of 150 mM NaCl solution and centrifuged for 5 min at 16,000 g. The resultant pellets were solubilized, and the color was quenched in the same way as above. The levels of ³²P in the resultant solutions were measured using Cerenkov counting (AccuFLEX LSC 740).

Statistical analysis

A t-test evaluated significant differences between two groups. Multiple-group comparisons were evaluated using Dunnett’s test. Repeated measures of ANOVA were applied where indicated. Statistical analyses were performed using GraphPad Prism 8.0 software (GraphPad Software, San Diego, CA, USA) or JMP 16.1 software (SAS Institute, Cary, NC, USA). A P value less than 0.05 was considered statistically significant. Bonferroni correction was applied for multiple comparisons where indicated.