Pharmacological Npt2a Inhibition Causes Phosphaturia and Reduces Plasma Phosphate in Mice with Normal and Reduced Kidney Function

Significance Statement

Hyperphosphatemia is common in the later stages of CKD and treatment options are limited to dietary phosphate restriction and oral phosphate binders. The sodium-phosphate cotransporter Npt2a, which mediates a large proportion of phosphate reabsorption in the kidney, might be a good therapeutic target for new medications for hyperphosphatemia. The authors show that pharmacologic inhibition of Npt2a in mice not only causes a dose-dependent phosphaturia, reductions in plasma phosphate levels, and suppression of parathyroid hormone, but also increases urinary excretion of sodium, chloride, and calcium. It does this without affecting urinary potassium excretion, flow rate, or pH. The results show for the first time that a novel Npt2a inhibitor has potential as a treatment for kidney disease-related hyperphosphatemia.

Visual Abstract

Abstract

Background

The kidneys play an important role in phosphate homeostasis. Patients with CKD develop hyperphosphatemia in the later stages of the disease. Currently, treatment options are limited to dietary phosphate restriction and oral phosphate binders. The sodium-phosphate cotransporter Npt2a, which mediates a large proportion of phosphate reabsorption in the kidney, might be a good therapeutic target for new medications for hyperphosphatemia.

Methods

The authors assessed the effects of the first orally bioavailable Npt2a inhibitor (Npt2a-I) PF-06869206 in normal mice and mice that had undergone subtotal nephrectomy (5/6 Nx), a mouse model of CKD. Dose-response relationships of sodium, chloride, potassium, phosphate, and calcium excretion were assessed in response to the Npt2a inhibitor in both groups of mice. Expression and localization of Npt2a/c and levels of plasma phosphate, calcium, parathyroid hormone (PTH) and fibroblast growth factor 23 (FGF-23) were studied up to 24-hours after Npt2a-I treatment.

Results

In normal mice, Npt2a inhibition caused a dose-dependent increase in urinary phosphate (ED₅₀ approximately 21 mg/kg), calcium, sodium and chloride excretion. In contrast, urinary potassium excretion, flow rate and urinary pH were not affected dose dependently. Plasma phosphate and PTH significantly decreased after 3 hours, with both returning to near baseline levels after 24 hours. Similar effects were observed in the mouse model of CKD but were reduced in magnitude.

Conclusions

Npt2a inhibition causes a dose-dependent increase in phosphate, sodium and chloride excretion associated with reductions in plasma phosphate and PTH levels in normal mice and in a CKD mouse model.

Inorganic phosphate (P_i) is an important mineral within the body. The intestine and kidneys play important roles in P_i homeostasis, whereas bone acts as a storage compartment.¹ Absorption of P_i from the diet occurs in the small intestine, with a large proportion being via the type 2 sodium-dependent phosphate cotransporter Npt2b (SLC34A2), which accounts for approximately 50% of the overall intestinal P_i absorption.² In the proximal tubule of the kidney, the type 2 sodium-dependent phosphate cotransporters Npt2a (SLC34A1), Npt2c (SLC34A3), and type 3 sodium-dependent phosphate cotransporter Pit2 (SLC20A2) mediate P_i reabsorption.³ Fractional P_i excretion is between 5% and 20% depending on dietary P_i content. Parathyroid hormone (PTH) and fibroblast growth factor 23 (FGF-23) are the primary regulators of renal P_i excretion, both mediating the retrieval of Npt2a/c from the luminal membrane.⁴ The contribution of Npt2a and Npt2c for renal P_i transport has been delineated in gene-targeted mice and mutations of P_i transporters in humans. Studies in Npt2a knockout mice (Npt2a^−/−) identified that Npt2a contributes to 70% of renal P_i reabsorption.⁵ Npt2c^−/− mice do not have a P_i phenotype but exhibited hypercalcemia (only at 4 weeks of age), hypercalciuria, and increased 1,25-dihydroxyvitamin D₃ levels.⁶ In contrast, kidney-specific Npt2c knockout mice did not show any differences in urinary P_i and calcium excretion, or plasma levels of P_i, calcium, PTH, or FGF-23 levels.⁷ Npt2a/c double knockout mice display a blend of both phenotypes: hypophosphatemia, more severe phosphaturia, increased 1,25-dihydroxyvitamin D₃ levels, and hypercalciuria.⁸ Humans with heterozygous mutations in the Npt2a gene presenting with urolithiasis or bone demineralization have hypophosphatemia and urinary P_i loss.⁹ Single nucleotide deletions, compound heterozygous deletions, and missense mutations in the Npt2c gene are the cause of hereditary hypophosphatemic rickets with hypercalciuria.^10–12 In mice, gene deletion of another P_i transporter, Pit2, does not affect urinary P_i and calcium excretion, plasma P_i, calcium, PTH, or FGF-23 levels when fed a control diet.¹³

Treatment of hyperphosphatemia is limited to a few options: dietary P_i restriction, oral P_i binders,¹⁴ sodium-hydrogen exchanger isoform 3 (NHE3) inhibitors (currently not approved),¹⁵ and nicotinamide.¹⁶ Recently, an orally nonabsorbable Npt2b inhibitor was developed with the vision of inhibiting intestinal P_i uptake.¹⁷ Unfortunately, this compound was ineffective in reducing plasma P_i levels in healthy volunteers or patients with ESKD.¹⁷ The aim here was to examine for the first time the in vivo role of an orally absorbable, selective Npt2a inhibitor (Npt2a-I), PF-06869206. The effect of this Npt2a-I on urinary P_i and calcium excretion, plasma levels of P_i and calcium, as well as the effect on PTH and FGF-23 levels was determined under normal conditions and in a model of CKD. The data demonstrate for the first time that pharmacologic Npt2a inhibition causes dose-dependent phosphaturia associated with decreased plasma P_i and PTH levels. Pharmacologic Npt2a inhibition could be a valuable tool to increase renal P_i excretion to treat conditions associated with hyperphosphatemia.

Methods

All animal experimentation was conducted in accordance with the Guide for Care and Use of Laboratory Animals (National Institutes of Health, Bethesda, MD) and was approved by the local Institutional Animal Care and Use Committee (reference 3338R). Male C57BL/6J mice were purchased from the Jackson Laboratories (000664; Bar Harbor, ME). Mice were housed under a 12-hour light/dark cycle in isolated, ventilated cages with free access to standard rodent chow (TD.2018; 0.7% phosphate, 1% calcium; Envigo, Madison, WI) and tap water. Experiments were performed on 3- to 5-month-old mice.

Short-Term Metabolic Cage Experiments

Mice were randomized to acute application, via oral gavage (1% of body weight), of vehicle (5% DMSO, 5% Cremophor EL, 90% sterile water) or Npt2a-I (PF-06869206, 0.3–300 mg/kg; Pfizer Worldwide Research & Development, Cambridge, MA). The synthesis of PF-06869206 and pharmacokinetic data were described recently.¹⁸ The reported t_1/2 after 1 mg/kg intravenous administration was 45 minutes.¹⁸ After their bladders were emptied (urine was obtained by picking up the mice to elicit reflex urination and holding them over a clean petri dish for sample collection), the mice were given vehicle or Npt2a-I by oral gavage and placed in metabolic cages for quantitative urine collection over 3 hours without access to food or water.¹⁹ Urine was analyzed as described below.

Twenty-Four-Hour Metabolic Cage Experiments

Mice were randomized to acute application, via oral gavage, of vehicle or Npt2a-I (30 mg/kg). Because mice underwent several rounds of acute metabolic cage experiments, mice were not further acclimatized for 24-hour metabolic cage experiments. After their bladders were emptied, mice were placed in metabolic cages at 9:00 for quantitative urine collection for 24 hours with free access to food and water. Urine was collected at 3, 6, 12, and 24 hours.²⁰ After each collection period, the bladders were emptied, as described above, to assure complete collection. Urine was analyzed as described below.

Blood Chemistry after Ntp2a Inhibition

To determine time-dependent changes in plasma P_i, calcium, PTH, and FGF-23 in response to Npt2a inhibition, vehicle or Npt2a-I (30 mg/kg by oral gavage) were administered. C57Bl/6J mice were briefly anesthetized with isoflurane and blood withdrawn from the retro-orbital plexus at baseline (0 minutes), after 3 and 24 hours. Each time, 40 µl of blood was collected. In time-course experiments, blood was collected in a similar manner at baseline (0 minutes) after 1, 2, 3, and 24 hours. Only 20 µl of blood was collected for determination of plasma P_i. After centrifugation, plasma was analyzed as described below. In another cohort of mice, plasma concentrations of sodium, chloride, potassium, pH, and bicarbonate were determined at baseline and 3 hours after administration of vehicle or Npt2a-I using a blood gas analyzer (OPTIMedical, Roswell, GA).

Western Blotting and Immunohistochemistry

Another cohort of mice was randomized for acute application of vehicle or Npt2a-I (30 mg/kg by oral gavage) and anesthetized by ketamine/xylazine after 3 or 24 hours. The left kidney was immediately removed and total kidney homogenates or plasma membrane–enriched samples (by centrifugation at 17,000 × g) were prepared for Western blotting.^21–23 Proteins were transferred to polyvinylidene difluoride membranes and immunoblotted with Npt2a,^20,23 Npt2c,^20,23 total NHE3,^20,24 or phosphorylated serine residue 552–NHE3 (Novus Biologicals, Littleton, CO). Chemiluminescent detection was performed with ECL Plus (Amersham, Piscataway, NJ). Signal intensity in specific bands was quantified using Image Studio Lite (Qiagen, Hilden, Germany) densitometry analysis. Proteasome 20S was used as loading control (Abcam, Cambridge, United Kingdom). The right kidney was perfused in vivo with 4% paraformaldehyde in PBS and fixed overnight in the same solution. Kidney sectioning and labeling were performed as previously described.^22,23,25

PTH Administration

Mice were randomized to either acute application of PTH vehicle (sterile water, 2 µl/g body wt, intraperitoneally) with Npt2a-I vehicle (see above, oral gavage), PTH (200 µg/kg body wt, intraperitoneally; GenScript, Piscataway, NJ)^23,26,27 with Npt2a-I vehicle, PTH vehicle with Npt2a-I, or PTH with Npt2a-I. PTH or PTH vehicle were applied 30 minutes before Npt2a-I or Npt2a-vehicle application. After their bladders were emptied, the mice were placed in metabolic cages for quantitative urine collection over 3 hours without access to food or water. Urine was analyzed as described below.

CKD Model

A 5/6 (subtotal) nephrectomy (Nx) model was used to mimic CKD. Surgery was performed at the Jackson Laboratories. The surgery follows a two-stage procedure. First, around each pole of the left kidney, at its one-third position, two ligatures are placed and the poles excised. One week thereafter, a right-side Nx is performed. One week after recovery, mice were received at the University of South Florida. Fluid and food intake were determined in their home cages daily, and body weight was determined every other day. Eight weeks after the last surgery, mice were used for experiments (short-term metabolic cage experiments as well as blood collections, as described above). GFR was determined in conscious mice using plasma kinetics of FITC-Sinistrin after a single intravenous injection 10 weeks after surgery.^28,29

Analysis of Plasma and Urine Samples

All clinical chemistry was performed using commercially available assays that were modified to work with small volumes.^20,28 Phosphate and calcium in plasma and urine were determined using Inorganic Phosphorous Reagent and Calcium Arsenazo III Reagent, respectively (Pointe Scientific, Canton, MI). Urinary chloride was analyzed by the thiocyanide method (Stanbio Laboratory, Boerne, TX). Urinary sodium and potassium concentrations were measured by flame photometry (BWB Technologies, Newbury, United Kingdom). Urinary glucose and creatinine were determined using Infinity Glucose Hexokinase and Infinity Creatinine Liquid Stable Reagent (Thermo Fisher Scientific), respectively. Urinary l-amino acids were determined using a commercial assay (MAK002; Sigma-Aldrich). PTH (Quidel, San Diego, CA) and intact FGF-23 (Quidel) were determined according to the manufacturer instructions. Urinary pH was determined using a pH electrode (9810BN; Thermo Fisher Scientific).

Statistical Analyses

The data are expressed as mean±SEM. Unpaired and paired t tests were performed, as appropriate, to analyze for statistical differences between and within groups. One-way and two-way ANOVA, repeated measure where applicable, were used for comparison of several experimental curves and procedures followed by the Dunnett, Newman–Keuls and Tukey post hoc tests (all data analyzed via GraphPad Prism, San Diego, CA, and SigmaPlot, San Jose, CA). P<0.05 was considered statistically significant.

Results

Npt2a Inhibition Causes Dose-Dependent Phosphaturia

Administration of Npt2a-I via oral gavage caused a dose-dependent increase in urinary P_i excretion from 27±6 nmol/min in response to vehicle to a maximum of 150±14 nmol/min at 300 mg/kg (50% effective dose of 21 mg/kg) over a 3-hour period (Figure 1A). In addition, Npt2a inhibition resulted in a dose-dependent increase in urinary calcium excretion from 2.8±0.4 to 8.5±0.4 nmol/min (Figure 1C), urinary sodium excretion from 113±24 to 551±48 nmol/min (Figure 1E), and urinary chloride excretion from 90±24 to 413±22 nmol/min (Figure 1G). Similar results were obtained when minerals and electrolytes were related to urinary creatinine (Figure 1, B, D, F, and H). In contrast, urinary potassium excretion and urinary potassium/creatinine ratios (Figure 2, A and B), urinary glucose excretion and urinary glucose/creatinine ratios (Figure 2, C and D), urinary l-amino acid excretion and urinary l-amino acid/creatinine ratios (Figure 2, E and F), urinary flow rate (Figure 2G), and urinary pH (Figure 2H) did not show a clear dose dependence and there was no significant difference between the response to vehicle and the maximum response observed at 300 mg/kg. The percentage changes of urinary excretions are shown in Supplemental Figure 1, A–H.

Figure 1.

Short-term Npt2a inhibition causes a dose-dependent increase in urinary phosphate, calcium, sodium, and chloride excretion. Response to acute oral (p.o.) application of Npt2a-I (0.3–300 mg/kg via oral gavage) or vehicle in short-term (3 hour) metabolic cage experiments. (A) Dose-dependent increase in urinary phosphate excretion in absolute terms (50% effective dose [ED₅₀] of 21 mg/kg), and (B) when corrected by urinary creatinine. (C and D) Similar effects were observed for calcium, (E and F) sodium, and (G and H) chloride. Data are expressed as mean±SEM. n=6–14 for 0.3–100 mg/kg; n=3 for 300 mg/kg.

Figure 2.

Short-term Npt2a inhibition does not result in a clear dose-dependent effect on urinary potassium, glucose, or L-amino acid excretion, urinary flow rate, or urinary pH. Response to acute oral (p.o.) application of Npt2a-I (0.3–300 mg/kg via oral gavage) or vehicle in short-term (3 hour) metabolic cage experiments. (A) A clear dose-dependent increase in urinary potassium excretion in absolute terms, and (B) when corrected by urinary creatinine, was absent. Similar results were obtained for urinary glucose (C) and L-amino acid excretion (E) in absolute terms and when corrected by urinary creatinine (D and F, respectively). There was also no effect on urinary flow rate (G) or urinary pH (H). Data were analyzed by two-way ANOVA followed by the Dunnett multiple comparisons test and are expressed as mean±SEM. n=6–14 for 0.3–100 mg/kg; n=3 for 300 mg/kg. *P<0.05 versus vehicle.

Changes in Plasma Electrolytes and Minerals, PTH, and FGF-23 after Npt2a Inhibition

Next, we determined if the increased urinary excretion of P_i, calcium, sodium, and chloride in response to the Npt2a-I affects the plasma concentrations of electrolytes and minerals as well as PTH and FGF-23. A time course of Npt2a inhibition (30 mg/kg) on plasma P_i levels identified a rapid decrease in plasma P_i (−0.6 mmol/L) after 1 hour (Figure 3A) that started to recover after 3 hours and returned to relatively normal levels after 24 hours. This fits well with the expected t_1/2 of the Npt2a-I (approximately 45 minutes).¹⁸ We studied the 3-hour time point in all other acute experiments. In another set of experiments, compared with baseline, vehicle application (Figure 3B) did not change plasma P_i concentrations after 3 or 24 hours. Once again, Npt2a-I (Figure 3B) caused a decrease in plasma P_i concentrations after 3 hours followed by an increase after 24 hours.

Figure 3.

Npt2a inhibition decreases plasma phosphate and PTH levels without affecting plasma calcium. Plasma phosphate significantly decreased (A) after 1 hour followed by (A and B) an increase after 24 hours. Acute oral (p.o.) vehicle application did not significantly affect plasma phosphate. (C) Plasma calcium levels were not affected by vehicle or Npt2a-I. (D) PTH significantly decreased after 3 hours followed by an increase after 24 hours. Vehicle application did not significantly affect PTH levels. (E) Intact FGF-23 levels significantly decreased after 3 hours followed by a significant increase after 24 hours in response to vehicle application. In contrast, the decrease after 3 hours was absent in response to Npt2a-I; however, a small but significant increase was present after 24 hours. Data were analyzed (A) by two-way ANOVA and (B–E) repeated measures two-way ANOVA, followed by the Tukey multiple comparisons test, and are expressed as mean±SEM. n=6–18. *P<0.05 versus vehicle, ^#P<0.05 versus previous time point.

Neither vehicle nor Npt2a-I application (Figure 3C) changed plasma calcium concentrations after 3 or 24 hours. Compared with baseline, vehicle application (Figure 3D) did not change plasma PTH concentrations after 3 or 24 hours. In contrast, Npt2a-I (Figure 3D) caused a decrease after 3 hours, followed by an increase at 24 hours. Compared with baseline, vehicle application (Figure 3E) caused a small but significant decrease of intact FGF-23 concentrations in plasma after 3 hours, followed by an increase after 24 hours. In contrast, Npt2a-I (Figure 3E) did not affect intact FGF-23 concentrations after 3 hours and remained significantly higher compared with vehicle, but subsequently caused a significant increase at 24 hours. Changes in plasma P_i, calcium, PTH, and FGF-23 are shown in Supplemental Figure 2, A–D. Despite dose-dependent natriuresis and urinary chloride excretion, no significant changes in plasma sodium and chloride were observed in response to Npt2a inhibition (Supplemental Figure 2, E and F). In addition, the responses of plasma potassium, pH, and bicarbonate were not significantly affected by Npt2a inhibition compared with vehicle (Supplemental Figure 2, G–I).

Effects of Npt2a Inhibition on Npt2a and Npt2c Protein Abundance and Localization

The effects of Npt2a-I on Npt2a and Npt2c total protein abundance and plasma membrane levels were studied at 3 and 24 hours after Npt2a inhibition. At 3 hours after Npt2a inhibition, Npt2a membrane abundance was increased 1.6-fold compared with vehicle (Figure 4A). Total Npt2a abundance was not affected at the 3-hour time point (Figure 4B). After 24 hours, Npt2a membrane abundance was not significantly different between vehicle and Npt2a-I (Figure 4A). In contrast, total Npt2a abundance was 50% lower in response to Npt2a-I compared with vehicle application (Figure 4B). The abundance of Npt2c (membrane and total) was not affected by Npt2a inhibition compared with vehicle at either time point (Figure 4, C and D). Immunohistochemistry (Figure 5) detected Npt2a staining predominantly in the microvilli of the proximal tubule and no visible changes were observed in localization or labeling intensity 3 or 24 hours after vehicle administration. In contrast, labeling intensity of Npt2a was visibly weaker 24 hours after Npt2a inhibition, consistent with Western blot data. At the level of light microscopy, clear differences in membrane-associated Npt2a 3 hours after Npt2a-I were not apparent. No visible differences were observed for Npt2c in response to vehicle administration or Npt2a inhibition, consistent with the Western blotting data. Because Npt2a-I acts on the proximal tubule and causes a dose-dependent natriuresis, we determined the effects of Npt2a inhibition on the major sodium reabsorption pathway in this segment, NHE3. The abundance of NHE3 and phosphorylation of serine residue 552 (an inhibitory site) was not affected by Npt2a inhibition compared with vehicle at any time point examined (Supplemental Figure 3, A–D).

Figure 4.

Npt2a inhibition affects Npt2a but not Npt2c. Western blot analysis of Npt2a and Npt2c abundances in membrane fractions and total homogenates of kidneys from mice treated (30 mg/kg) with vehicle (V) or Npt2a inhibitor (I). (A) Npt2a membrane abundance was significantly increased 3 hours after Npt2a-I treatment and not different from vehicle treatment after 24 hours. (B) Total Npt2a abundance was unaffected after 3 hours and significantly reduced after 24 hours. In contrast, no effect was observed at any time point in (C) membrane fractions or (D) total homogenate Npt2c abundance. Data were analyzed by unpaired t test and are expressed as mean±SEM, n=5–6. *P<0.05 versus vehicle.

Figure 5.

Qualitative immunohistochemical labeling identifies differences in Npt2a after 24 hours but not in Npt2c. Representative immunohistochemistry images after 3 and 24 hours from mice treated with vehicle or the Npt2a-I. (A) In vehicle-treated mice, Npt2a was detected throughout the proximal tubule starting in the early (S1) segment. Npt2a staining was predominantly localized to the apical microvilli. (B) Three hours after Npt2a-I treatment, Npt2a staining was not visibly different compared with vehicle treatment. (C and D) Twenty-four hours after treatment with the Npt2a-I, Npt2a staining was clearly weaker compared with vehicle. (E–H) Npt2c labeling was found throughout the proximal tubule in apical microvilli but staining intensity or distribution of Npt2c was not visibly different in vehicle- or Npt2a-I–treated mice after 3 or 24 hours. Arrows indicate regions magnified in the insets. n=5–6.

Twenty-Four-Hour Urinary Phosphate and Calcium Excretion in Response to Npt2a Inhibition

To establish a potential circadian profile of urinary P_i and calcium excretion in response to Npt2a-I, 24-hour metabolic cage experiments were performed. Following vehicle application via oral gavage (Figure 6A), urinary P_i excretion was 22±3 nmol/min from 9:00 to 12:00, which subsequently significantly increased from 12:00 to 15:00 (approximately threefold) and from 15:00 to 21:00 (approximately twofold). Subsequently, from 21:00 to 9:00 urinary P_i excretion significantly decreased (approximately twofold). A similar pattern of P_i excretion has been described before in humans.³⁰ Confirming the previous results of Npt2a application during a 3-hour period (Figure 1A), in this study the Npt2a-I at 30 mg/kg caused significant phosphaturia (95±10 nmol/min) from 9:00 to 12:00 (Figure 6A). Urinary P_i excretion decreased in the subsequent period from 12:00 to 15:00 (approximately twofold) to levels similar to vehicle application (Figure 6A). Excretion during subsequent collecting periods was similar to vehicle application. Because plasma P_i increased at 24 hours (Figure 3, A and B), but circadian urinary P_i excretion was identical between groups after the initial 3-hour period, this suggests that increased intestinal P_i uptake or release from other cells in the body contributes to the re-establishment of plasma P_i levels.

Figure 6.

Npt2a inhibition–induced phosphaturia is only present in the first 3 hours after application. Response to acute oral (p.o.) application of Npt2a-I (30 mg/kg via oral gavage, n=9) or vehicle (n=11) in 24-hour metabolic cages. Experiments were started at 9:00. Urine was collected in 3-hour intervals until 15:00, a 6-hour interval until 21:00, and another 12-hour interval until 9:00 the next day. (A) Acute oral vehicle application showed a rise in urinary phosphate excretion until 12 hours after the start of the experiment. In the last 12 hours, urinary phosphate excretion significantly decreased. The Npt2a-I caused phosphaturia in the first 3 hours, but urinary phosphate excretion decreased in the subsequent period to levels similar with vehicle application. Excretion during subsequent collecting periods was similar to vehicle application. (B) Vehicle application showed the highest urinary calcium excretion in the first 3 hours, urinary calcium excretion was significantly lower in the remaining collecting periods. The Npt2a-I caused calciuria in the first 3 hours, but subsequently urinary calcium excretion followed the pattern of vehicle application. (C) Pretreatment with PTH (n=10) caused significantly greater urinary phosphate excretion compared with vehicle (n=6), with application of PTH plus Npt2a-I (n=10) showing an additive effect compared with Npt2a-I alone (n=10). (D) Calcium excretion increased in response to PTH, but no difference was observed when combined with the Npt2a-I. Data were analyzed by one-way ANOVA followed by the Tukey multiple comparisons test and are expressed as mean±SEM. *P<0.05 versus vehicle, ^#P<0.05 versus 0–3 hours, ^$P<0.05 versus PTH plus vehicle, ^†P<0.05 versus vehicle plus Npt2a-I.

Urinary calcium excretion in response to vehicle application (Figure 6B) was 2.0±0.3 nmol/min from 9:00 to 12:00, which subsequently significantly decreased from 12:00 to 15:00 (approximately 6-fold) and remained at this level from 15:00 to 9:00. Npt2a-I at 30 mg/kg caused significant calciuria (9.7±0.8 nmol/min) from 9:00 to 12:00 (Figure 6B). In the subsequent collecting period, calcium excretion dropped to levels similar to vehicle and followed a similar pattern for the remainder of the experimental period (Figure 6B).

To limit the amount of Npt2a in the brush border membrane, mice were treated with PTH (200 µg/kg) 30 minutes before an acute 3-hour metabolic cage experiment. This dose was previously shown to cause significant Npt2a internalization.^23,26,27 In contrast to vehicle application, PTH caused significantly higher urinary P_i (Δ approximately 25 nmol/min) and calcium (Δ approximately 5 nmol/min) excretion (Figure 6, C and D). Application of Npt2a-I after PTH sustained the effect seen with PTH alone (Δ approximately 25 nmol/min) and resulted in the highest urinary P_i excretion (Figure 6C). This effect was not seen for urinary calcium excretion (Figure 6D).

Effect of Npt2a Inhibition in a Model of CKD

Ten days after surgery, body weight in 5/6 Nx mice was significantly lower compared with sham mice. In subsequent weeks, both groups significantly gained body weight; however, 5/6 Nx mice gained significantly more weight so that the difference between groups was lost at 21 days after surgery (Supplemental Figure 4A). The increase in body weight was associated with a significantly higher average food intake in 5/6 Nx mice compared with sham mice (Supplemental Figure 4B). At 10 weeks after surgery, GFR was approximately 66% lower in 5/6 Nx mice compared with sham mice (160±13 versus 469±36 µl/min, P<0.05; Supplemental Figure 4C). Consistent with previous reports,^31,32 5/6 Nx mice were polyuric, as evidenced by a 50% lower urine osmolality and an approximately twofold higher fluid intake compared with sham mice (Supplemental Figure 4, D and E). This was associated with a more acidic urine (Δ 1.6 pH units) in 5/6 Nx mice, a predictor of CKD,³³ and lower hematocrit (Supplemental Figure 4, F and G). Urinary P_i/creatinine ratio in spontaneously voided urine was not significantly different between sham and 5/6 Nx mice (9.7±1 versus 10.3±1 mmol/mmol, NS). In sham mice, administration of Npt2a-I via oral gavage caused a dose-dependent increase in urinary P_i excretion from 14±1 nmol/min in response to vehicle to 140±12 nmol/min at 300 mg/kg (50% effective dose of 21 mg/kg) over a 3-hour period (Figure 7, A and B). Sham mice showed dose dependency of urinary calcium (Figure 7, C and D), sodium, and chloride excretion as well as urinary flow rate (Supplemental Figure 5, A–E). Urinary potassium and glucose excretion as well as urinary pH did not show a clear dose dependence (Supplemental Figure 5, F–K). Mice with 5/6 Nx showed a higher urinary P_i excretion in response to vehicle compared with sham mice (Figure 7, A and B), consistent with a higher fractional P_i excretion in CKD.³⁴ Increasing doses of Npt2a-I in 5/6 Nx mice also showed a dose dependence for urinary P_i excretion; however, the maximum effect was reduced compared with sham mice (Figure 7, A and B). Similar effects were observed for urinary calcium excretion (Figure 7, C and D). Urinary excretions of sodium and chloride, as well as urinary flow rate in response to Npt2a-I, did show a significant dose-dependence difference between sham and 5/6 Nx mice (Supplemental Figure 6, A–E). Urinary excretions of potassium and glucose as well as urinary pH were not dose dependent (Supplemental Figure 6, F–K). Because urinary P_i excretion in response to Npt2a-I at 30 mg/kg was not as pronounced compared with sham mice, we studied the effect of 100 mg/kg on the time course of plasma P_i. Both sham and 5/6 Nx mice had significantly decreased plasma P_i (−0.8±0.1 and −0.5±0.1 mmol/L, P<0.05, versus baseline and versus sham) 1 hour after Npt2a-I administration, which was slightly but significantly less pronounced in 5/6 Nx mice (Figure 7E). Both groups remained at this level up to 3 hours after administration and plasma P_i subsequently recovered after 24 hours. Compared with baseline, vehicle application (Figure 7F) did not change plasma PTH concentrations after 3 or 24 hours in sham or 5/6 Nx mice. Both sham and 5/6 Nx mice showed a significant decrease in plasma PTH (−63±22 and −70±12 pg/ml, P<0.05 versus baseline, respectively) followed by an increase after 24 hours (Figure 7F) after Npt2a inhibition.

Figure 7.

Npt2a inhibition in mice with 5/6 Nx causes a dose-dependent increase in urinary phosphate excretion and a decrease in plasma phosphate and PTH levels. Response to acute oral (p.o.) application of Npt2a-I (0.3–300 mg/kg) or vehicle in short-term (3-hour) metabolic cage experiments in sham and 5/6 Nx mice. Sham mice show dose-dependent increases in (A) urinary phosphate and (C) calcium excretion in absolute terms and (B and D) when corrected by urinary creatinine. Mice with 5/6 Nx show dose-dependent urinary phosphate and calcium excretion but the maximum effects are reduced. The dose of 300 mg/kg in 5/6 Nx mice was not included in the analysis because mice appeared lethargic after application. Application of Npt2a-I (100 mg/kg) caused a decrease in (E) plasma phosphate and (F) PTH in both sham and 5/6 Nx mice. Data were analyzed by one-way ANOVA followed by Newman–Keuls multiple comparisons test and are expressed as mean±SEM. n=6–14 for 0.3–100 mg/kg; n=4 for 300 mg/kg. *P<0.05 versus vehicle, ^§P<0.05 versus sham, ^#P<0.05 versus previous time point.

Discussion

These studies demonstrate for the first time that in vivo pharmacologic Npt2a inhibition, via an orally absorbable Npt2a-I, causes phosphaturia in mice. This is consistent with studies of Npt2a^−/− mice for a major role of Npt2a in P_i reabsorption. Acutely, Npt2a inhibition also results in increased renal calcium, sodium, and chloride excretion, alongside significant reductions in plasma P_i and PTH. Under conditions of reduced kidney function (5/6 Nx model), Npt2a inhibition had similar effects, but they were slightly reduced in magnitude.

Although the role of Npt2a in humans remains controversial,^35,36 these data from Npt2a inhibition in mice are consistent with certain phenotypic features of individuals who carry specific mutations in the SLC34A1 gene. For example, a homozygous single nucleotide variation in SLC34A1 is associated with hypercalcemia, hypercalciuria, hypophosphatemia, and nephrocalcinosis,³⁷ and a heterozygous mutation causes hypophosphatemia, phosphaturia, urolithiasis, or bone demineralization.⁹ Another study identified a homozygous mutation in SLC34A1 causing renal Fanconi syndrome and hypophosphatemic rickets.³⁸ Whether these differences relate to changes in the P_i-transporting activity^9,37 of Npt2a and/or trafficking defects^38–40 is unknown. Of note, other proximal tubule transport pathways, including uptake of glucose or amino acids, appeared to not be affected by Npt2a inhibition. Our data also provide evidence, albeit indirectly, that the Npt2a-I (PF-06869206) specificity affects Npt2a but not Npt2c in vivo. As described, patients with SLC34A1 mutations may also present with calcium-P_i stones or nephrocalcinosis.^11,37,39 However, although stone formation is facilitated by alkaline urinary pH,⁴¹ our studies did not identify substantial changes in urinary or blood pH caused by Npt2a-I. It remains to be determined if long-term inhibition of Npt2a results in stone formation, nephrocalcinosis, or changes in bone mineralization.

Approximately 60%–70% of filtered calcium is passively reabsorbed in the proximal tubule as a consequence of primary reabsorption of sodium and water.⁴² Our data demonstrating that Npt2a inhibition caused a dose-dependent calciuria are consistent with, but do not prove, that inhibition of proximal tubular reabsorption of sodium via Npt2a causes a secondary reduction in reabsorption of calcium; however, the mechanism(s) remain elusive. Another possible player is PTH. Following Npt2a-I, phosphaturia and a reduction of plasma P_i resulted in a significant reduction of plasma PTH, similar to dietary P_i restriction.²³ A reduction in plasma P_i was previously seen with phosphonoformic acid (a nonselective Npt2 inhibitor) in clearance experiments.⁴³ These reduced PTH levels might further contribute to reduced calcium reabsorption in the distal convoluted tubule via transient receptor potential vanilloid 5 (TRPV5) channels⁴⁴ that are activated by PTH.⁴⁵ However, despite calciuria and changes in plasma PTH levels, plasma calcium levels were not significantly altered. This implies that P_i alone can regulate PTH release via unknown mechanism(s) given that plasma calcium levels remained unchanged. The latter is somewhat surprising given the drastic changes in PTH. Reasons for unchanged calcium might include: (1) stimulation of 1,25-dihydroxyvitamin D₃ production, stimulated by hypophosphatemia, resulting in enhanced intestinal calcium absorption; and/or (2) inhibition of Npt2a in osteoclasts,⁴⁶ where Npt2a is localized at the basolateral membrane which could have secondary consequences on bone resorption.

Why does pretreatment with PTH result in greater phosphaturia when combined with Npt2a inhibition than either PTH or Npt2a-I alone? PTH administration at the used dose is well described to retrieve Npt2a from the brush border membrane,^23,26,27 resulting in less Npt2a available for P_i reabsorption. Thus, at the dose used in this experiment (30 mg/kg), more drug was available to inhibit the remaining transporters, resulting in a phosphaturic effect in the range of the 300 mg/kg dose (Figure 1A). However, it cannot be excluded that early effects of PTH on Npt2c are also contributing to the increased urinary P_i excretion after 3 hours.

Because Npt2a transports three sodium ions together with P_i, it is not surprising that Npt2a inhibition causes natriuresis. Npt2a inhibition results in a dose-dependent inhibition of chloride reabsorption because chloride reabsorption in the proximal tubule is sodium dependent. Because the observed dose-dependent sodium and chloride excretion after Npt2a inhibition did not result in changes in NHE3 phosphorylation or abundance, this indirectly indicates that the major sodium transport pathway in the proximal tubule is not mediating this response. However, downstream effects in other tubule segments cannot be completely excluded. Along those lines, why urinary potassium excretion is not affected by Npt2a inhibition remains to be determined.

Our results employing pharmacologic inhibition of Npt2a are consistent with previously described roles of the transporter from studies in Npt2a^−/− mice and humans with certain mutations in the SLC34A1 gene. However, whether Npt2a inhibition in humans is a valuable treatment for conditions with impaired P_i homeostasis remains to be determined considering Npt2c in humans is hypothesized to have a larger role in renal P_i reabsorption.^10,47 Are effects of Npt2a inhibition preserved in CKD? Data regarding a possible role of Npt2 inhibition in CKD are controversial. In vitro studies on brush border membrane vesicles from 5/6 Nx rats showed that phosphonoformic acid had a similar inhibitory potency on Npt2 activity compared with brush border membrane vesicles from sham rats.⁴⁸ In contrast, the responsiveness of the proximal tubule to PTH or, potentially, to drugs acting on the tubule are impaired in CKD. Our studies demonstrate that, although most of the effects seen in control mice after Npt2a-I are preserved after 5/6 Nx, they are attenuated. We appreciate that the model used in this study represents a mild form of CKD in that plasma P_i was normal and vehicle application resulted in a higher urinary P_i excretion compared with sham. It remains to be seen if Npt2a inhibition is still efficacious in conditions where hyperphosphatemia is present, e.g., in severe CKD, acute tumor lysis syndrome, rhabdomyolysis, hemolysis, hyperthermia, profound catabolic stress, or acute leukemia.

Disclosures

None.

Funding

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases grant 1R01DK110621 (to Dr. Rieg) and American Heart Association Transformational Research Award 19TPA34850116 (to Dr. Rieg). Funding (to Dr. Fenton) was provided by the LeDuq Foundation, the Danish Medical Research Council, and the Novo Nordisk Foundation. Mr. Xue was supported by an American Heart Association predoctoral fellowship (18PRE33990236) and Dr. Thomas by a postdoctoral fellowship (19POST34400026).

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2018121250/-/DCSupplemental.

Supplemental Figure 1. Percent change in urinary parameters in response to Npt2a inhibition.

Supplemental Figure 2. Change in plasma parameters in response to Npt2a inhibition.

Supplemental Figure 3. Npt2a-inhibition does not affect NHE3 and pS552-NHE3.

Supplemental Figure 4. Pathophysiological consequences of 5/6 nephrectomy (Nx).

Supplemental Figure 5. Comparison of Npt2a inhibitor dose-response effects between Sham and 5/6 nephrectomy (Nx) mice.