Combination treatment with tenapanor and sevelamer synergistically reduces urinary phosphorus excretion in rats.

The majority of patients with chronic kidney disease (CKD) receiving dialysis do not reach target serum phosphorus concentrations, despite treatment with phosphate binders. Tenapanor is a non-binder, sodium/hydrogen exchanger isoform 3 (NHE3) inhibitor that reduces paracellular intestinal phosphate absorption. This pre-clinical study evaluated the effect of tenapanor and varying doses of sevelamer carbonate on urinary phosphorus excretion, a direct reflection of intestinal phosphate absorption. We measured 24-hour urinary phosphorus excretion in male rats assigned to groups dosed orally with vehicle or tenapanor (0.3 mg/kg/day) and provided a diet containing varying amounts of sevelamer (0-3% w/w). We also evaluated the effect of the addition of tenapanor or vehicle on 24-hour urinary phosphorus excretion to rats on a stable dose of sevelamer (1.5% w/w). When administered together, tenapanor and sevelamer decreased urinary phosphorus excretion significantly more than either tenapanor or sevelamer alone across all sevelamer dose levels. The Bliss statistical model of independence indicated that the combination was synergistic. A stable sevelamer dose (1.5% w/w) reduced mean (±standard error of the mean) urinary phosphorus excretion by 42±3% compared with vehicle; together, tenapanor and sevelamer reduced residual urinary phosphorus excretion by an additional 37±6% (P < 0.05). While both tenapanor and sevelamer reduce intestinal phosphate absorption individually, administration of tenapanor and sevelamer together results in more pronounced reductions in intestinal phosphate absorption than if either agent is administered alone. Further evaluation of combination tenapanor plus phosphate binder treatment in patients receiving dialysis with hyperphosphatemia is warranted.


INTRODUCTION
Serum phosphorus concentrations are normally maintained within a relatively narrow physiological range, principally through the regulation of renal phosphate excretion (1). Impaired or absent excretion of phosphate via urine contributes to hyperphosphatemia, which is common in patients with chronic kidney disease (CKD) receiving dialysis (2). Generally, dialysis fails to adequately control serum phosphorus concentrations in patients who have little or no residual kidney function (3). Intestinal phosphate absorption is linearly dependent on the lumen to serum phosphorus concentration gradient and does not saturate even at high luminal phosphate concentrations (4)(5)(6)(7). Furthermore, intestinal phosphate absorption in patients receiving dialysis is similar to that in healthy individuals, and it is increased further by calcitriol or active vitamin D analogs that are commonly used to manage patients with secondary hyperparathyroidism (8). The combination of impaired renal excretion, limited dialytic clearance, and sustained intestinal absorption of phosphate explains the high prevalence of hyperphosphatemia in patients with CKD receiving dialysis. Current medical management of patients with hyperphosphatemia aims to reduce intestinal phosphate absorption by combining dietary phosphorus restriction, dialysis treatment (9), and administration of oral phosphate binders (1). Restricting dietary phosphorus intake can modestly reduce the severity of hyperphosphatemia, although adherence is challenging and typically poor, in part owing to the widespread use of phosphorus-containing additives and preservatives in processed foods (10). Phosphate binders are modestly effective, partly due to variable adherence to treatment because of a high pill burden, large pill size, dosing frequency, and adverse gastrointestinal effects (11,12). Phosphate binders, taken with meals, bind dietary phosphate in the gut, reducing the concentration gradient for paracellular phosphate absorption, thereby reducing serum phosphorus concentrations. Hyperphosphatemia has been consistently associated with mortality, cardiovascular events, left ventricular hypertrophy, fracture, and progression of chronic kidney disease in patients not already receiving dialysis (13)(14)(15)(16). Therefore, additional therapeutic approaches are required to optimize serum phosphorus management in patients with hyperphosphatemia.
Tenapanor is an investigational, nonbinder, inhibitor of sodium and phosphate absorption that is being developed to treat hyperphosphatemia in patients with CKD receiving dialysis (17,18). Tenapanor has a unique mechanism of action and acts locally in the gut to inhibit sodium/hydrogen exchanger isoform 3 (NHE3) (19). This results in the tightening of the epithelial cell junctions, significantly reducing the paracellular uptake of phosphate, the primary pathway of phosphate absorption, and thereby reduces serum phosphorus concentrations (20). NHE3 is expressed on the luminal surface throughout the small intestine and proximal colon, and functions as a transporter to import luminal sodium in exchange for a cellular proton (21). We recently reported that inhibition of NHE3 by tenapanor reduces intestinal phosphate absorption by decreasing paracellular phosphate permeability-a result of intracellular proton retention that modulates the tight junction to increase transepithelial electrical resistance (20).
Gastrointestinal phosphate absorption is mediated by at least two distinct mechanisms. NaPi2b (SLC34A2), the predominant sodium-dependent phosphate transporter in the gut, mediates the bulk of transcellular intestinal phosphate absorption (22) and plays an important role in maintaining systemic phosphate and bone homeostasis (22)(23)(24)(25). However, NaPi2b has a high affinity for phosphate and saturates owing to the high concentrations of phosphate in the gut lumen (26,27). Paracellular phosphate flux through tight junction complexes, driven by the electrochemical phosphate gradient, is quantitatively the most important mechanism of intestinal phosphate absorption under typical conditions of phosphate availability (5). Paracellular phosphate flux in the intestine is determined by the combination of the prevailing electrochemical phosphate gradient and concurrent paracellular phosphate permeability (28). Therefore, the combination of tenapanor, which reduces paracellular phosphate permeability (20), and a phosphate binder, such as sevelamer, which complexes with luminal phosphate to reduce the concentration gradient affecting phosphate absorption (29), offers the potential to additively or synergistically reduce phosphate uptake.
The objective of this study was to investigate the effect of the combination of tenapanor and the phosphate-binding agent sevelamer carbonate on intestinal phosphate absorption in rats. In healthy rats (and humans), urinary phosphorus excretion reflects intestinal phosphate absorption under physiological conditions (9,(30)(31)(32). We therefore used urinary phosphorus excretion as a marker of intestinal phosphate absorption.

Animals
In total, 102 male Sprague-Dawley rats at approximately 8 wk of age (Envigo, Livermore, CA), weighing $250 g on arrival, were housed two per cage in micro-isolator cages for a 48-h acclimation period before being assigned to study groups. Animals were housed in a temperature-(65-75 F) and humidity-(35-55%) controlled facility utilizing a reversed 12:12-h light-dark cycle (0700-1900 dark cycle). For the duration of the study, rats had ad libitum access to standard rodent chow spiked with an additional 0.4% inorganic phosphate [1:1 sodium:potassium salt, 1.1% (wt/wt) total phosphate content, Harlan Teklad, Madison, WI]. The majority of phosphate in normal rodent diets is present as phytates, which is not readily available for absorption (33). Therefore, supplemental phosphate was added to achieve a level of orally bioavailable phosphate equivalent to "high-phosphate" diets used in other studies (34,35). Rats also had access to deionized drinking water throughout the study. We recorded body weights at study initiation and daily for the remainder of the study. Animals were maintained in accordance with the Guide for the Care and Use of Laboratory Animals (36). The study was conducted under a protocol approved by the Institutional Animal Care and Use Committee (Ardelyx AUP 2013-01).

Study Design
We conducted two studies. Figures 1 and 2 show overviews of the respective study designs. In study 1 (Fig. 1), we randomly assigned rats into eight groups. Each group received either vehicle (0.01% Tween 80) or tenapanor (0.15 mg/kg) via twice-daily oral gavage, in combination with one of four diets containing different concentrations of sevelamer [0%, 0.75%, 1.5%, and 3% (wt/wt)], for 11 consecutive days. During this treatment period, we dosed animals while pair-housed in their home cages for the first five days before being transferred into individual metabolic cages on treatment day 6 for the remainder of the study. Following two days of acclimation to the metabolic cages (days 6-7), we performed daily measurements of 24-h food intake, water intake, and urinary volume over the final four days of the treatment period (days 8-11). We obtained a terminal blood sample by cardiac puncture under isoflurane anesthesia on the final day of treatment (day 11) to measure serum sodium and inorganic phosphorus (Axcel, Alfa Wassermann, West Caldwell, NJ) and to allow calculation of the renal clearance of sodium and phosphorus using standard methods. We measured urinary sodium and phosphorus concentrations by ion chromatography and calculated 24-h urinary sodium and phosphorus excretion by multiplying ion concentration by 24-h urine volume. We normalized urinary ion excretion to daily dietary ion intake to account for daily fluctuations in food intake.
In study 2 ( Fig. 2), we randomly assigned rats into two groups, each receiving a diet containing 0% or 1.5% (wt/wt) sevelamer for 6 days. Rats that received the 0% sevelamer diet were re-randomized into five groups receiving diets containing either 0% or 1.5% (wt/wt) sevelamer, in combination with vehicle (0.01% Tween 80), or different doses of tenapanor (0.15 or 0.5 mg/kg) via twice-daily oral gavage, for 7 days. Rats that received the 1.5% sevelamer diet in the first treatment period were re-randomized into four groups receiving diets containing either 0%, 0.75%, or 1.5% (wt/wt) sevelamer, in combination with vehicle (0.01% Tween 80), or different doses of tenapanor (0.15 or 0.5 mg/kg) via twice-daily oral gavage, for 7 days. On both the sixth and final treatment day (day 13), we recorded 24-h urinary phosphorus excretion measurements.

General Methods (Studies 1 and 2)
Dosing solution and diet formulation.
Dosing solutions containing tenapanor (Ardelyx, Fremont, CA) at a concentration of 0.03 mg/mL or 0.1 mg/mL were made fresh each week by adding the appropriate weight of tenapanor to 0.1% Tween 80 vehicle. The dose volume was 5 mL/kg for all groups. The phosphate-spiked standard rodent diet was weighed out into mixing bowls. We added the appropriate amount of sevelamer carbonate (Renvela, Sanofi Genzyme, Cambridge, MA) to the powdered diet [0%, 0.75%, 1.5%, or 3% (wt/wt) sevelamer] in the mixing bowl, placed the bowl on a stand mixer (KitchenAid), and set to speed "Stir-2" for 10 min.

Dose administration.
We dosed animals twice daily (bid) orally (po), immediately before lights out (0700) and during the middle of dark cycle (1400) with a standard gavage needle (75 mm, 18 gauge). We removed animals from the holding or metabolic cages for dosing and then immediately returned them to the cage of origin.

Urinary ion analysis by ion chromatography.
We analyzed urine samples for sodium and phosphorus content using an ion chromatography system (Thermo Fisher Scientific ICS-3000 or ICS-5000, Ardelyx, Fremont, CA) coupled with conductivity detectors. We performed chromatographic separation of cations using an IonPac CS12A-MS (Thermo Fisher), 2 Â 100 mm, 8.5 mm or IonPac CS12A (Thermo Fisher), 2 Â 250 mm, 8.5 mm analytical columns with an isocratic elution using 25 mM methanesulfonic acid. We performed chromatographic separation of anions using an IonPac AS18 (Thermo Fisher) 2 Â 250 mm, 7.5 mm analytical column with an isocratic elution using 35 mM potassium hydroxide. We interpolated concentrations from a standard curve prepared in Milli-Q water for each analyte ion based on retention time and peak area.

Statistical Analysis
All data sets for each group are presented as means ± SE We compared experimental groups using one-way ANOVA followed by Tukey's post hoc test to correct for multiple comparisons and to enable individual group comparisons. Statistical significance between groups is indicated in the figures and was compared with the vehicle group.
We used the Bliss model of independence (37) to statistically determine if the combination of tenapanor and sevelamer was independent, synergistic, or antagonistic. The model compares the predicted combination response calculated using the complete additivity of probability theory based Treatment period (rats were fed respective diets for 11 days) Terminal blood samples were obtained to measure serum phosphorus/sodium concentrations n = 6 rats/group Group 2: tenapanor 0.15 mg/kg bid po + sevelamer 0% (w/w) Group 6: tenapanor 0.15 mg/kg bid po + sevelamer 0.75% (w/w) Group 7: tenapanor 0.15 mg/kg bid po + sevelamer 1.5% (w/w) on the observed effect of each individual agent administered alone, with the observed effect of combination treatment (37). For these analyses, we used the mean urinary phosphorus excretion over the entire 4-day collection period for each treatment group to ensure that all collected data points were included in the assessment of combination effects.

Study 1
Effects on body weight and food intake and calculating the administered sevelamer dose.
Body weight (Supplemental Fig. S1; all Supplemental Material is available at https://doi.org/10.5281/zenodo.3738955) and food intake (Supplemental Fig. S2) were not significantly affected by tenapanor, sevelamer, or combination treatment compared with control. Because sevelamer was administered as a diet admixture, the actual dose of sevelamer delivered in the food was calculated based on food consumption (Supplemental Fig. S3). There was a linear increase in sevelamer dose with increasing sevelamer content in the food; the sevelamer dose was stable over time and co-administration of tenapanor did not significantly affect the dose of sevelamer consumed in the food. As a result, the dose of sevelamer was well matched between tenapanor-and non-tenapanor-treated groups consuming sevelamer, allowing direct comparison of the combination effect. The sevelamer dose administered on the final day of the study, along with the conversion to human equivalent dose based on body surface area conversion (correction factor 7), is summarized in Table 1.
The recommended human starting dose of sevelamer is 2.4-4.8 g/day; therefore, on a dose conversion basis, the 0.75% sevelamer dose in rats is roughly equivalent to the recommended human starting dose of sevelamer.
Effect on urinary phosphorus excretion.
The 24-h urinary phosphorus excretion over the final four days of treatment is shown in Fig. 3A, with the final treatment day shown in Fig. 3B   compared with vehicle control. Sevelamer significantly and dose dependently decreased urinary phosphorus excretion. When used together, tenapanor and sevelamer (dosedependently) decreased urinary phosphorus excretion such that combination reductions were significantly larger than either tenapanor or sevelamer alone, across all sevelamer dose levels. Figure 3C shows normalization of urinary phosphorus excretion to dietary phosphorus intake, to account for fluctuations in daily intake over the final four days of treatment, with the final treatment day shown alone in Fig. 3D for clarity. This analysis confirms the enhanced urinary phosphorus-lowering effects of the tenapanor and sevelamer combination and demonstrates that variations in dietary phosphorus intake are not responsible for the observed findings. Table 2 summarizes results of the Bliss model of independence. The observed combination treatment reduction in urinary phosphorus excretion was larger than the predicted reduction based on the single-agent effects in accordance with the Bliss model, an indicator of synergy, for tenapanor with all sevelamer doses. The difference in the observed (64.8% reduction) from predicted (47.2% reduction) when tenapanor and sevelamer were used together, was the most pronounced at the lowest [0.75% (wt/wt)] sevelamer dose, which suggests that tenapanor and sevelamer act synergistically to reduce urinary phosphorus excretion, especially at lower sevelamer doses.
Effect on urinary sodium excretion.  Fig. 4B for clarity. Tenapanor treatment significantly decreased urinary sodium excretion compared with vehicle control, whereas sevelamer alone did not significantly affect urinary sodium excretion. Tenapanor also significantly reduced urinary sodium excretion compared with control when co-administered with all doses of sevelamer. The 0.75% and 1.5% (wt/wt) doses of sevelamer did not significantly affect the magnitude of the sodium-lowering effect of tenapanor. However, the highest dose of sevelamer [3% (wt/wt)] modestly but significantly attenuated the sodiumlowering effect of tenapanor. Figure 4C shows results following normalization of urinary sodium excretion to dietary sodium intake, to account for any fluctuations in daily sodium intake over the final 4 days of treatment, with the final treatment day shown alone in Fig. 4D for clarity. These findings demonstrated that the effects of tenapanor on urinary sodium excretion are not confounded by variations in sodium intake. The Bliss model confirmed that the observed lowering of urinary sodium excretion with the combination of tenapanor and 3% (wt/wt) sevelamer is less than the predicted urinary sodium lowering based on individual agent treatment ( Table 2), indicating that the high dose of sevelamer [3% (wt/wt)], equivalent to $20 g/day of sevelamer in humans (fivefold higher than the approved dose), attenuates the sodium-lowering pharmacodynamic effect of tenapanor.
Effects on renal clearance of phosphorus and sodium.
The effect of tenapanor, sevelamer, and the combination of tenapanor and sevelamer on renal clearance of phosphorus and sodium is shown in Fig. 5, A and B, respectively. Tenapanor and sevelamer alone significantly decreased renal Values are means ± SE for sevelamer dose. Ã 60-kg human; rat to human surface area dose correction factor: 1/7.  clearance of phosphorus, reflective of the appropriate renal response if there is decreased intestinal phosphate absorption. Used together, tenapanor and sevelamer further and dosedependently reduced renal phosphorus clearance such that the combined reductions were significantly larger than either tenapanor or the equivalent dose of sevelamer alone (Fig. 5A). Tenapanor alone significantly decreased renal sodium clearance, reflective of the appropriate renal response to conserve sodium if there is decreased intestinal sodium absorption (Fig.  5B). Sevelamer alone had no effect on renal sodium handling because it does not affect intestinal sodium absorption. Renal sodium clearance was not significantly affected by sevelamer at 0.75% or 1.5% (wt/wt) when combined with tenapanor; however, compared with tenapanor alone, combination therapy with tenapanor and 3% (wt/wt) sevelamer resulted in a significantly higher renal sodium clearance, suggesting an attenuated sodium-lowering pharmacodynamic effect of tenapanor when combined with the highest dose of sevelamer.

Study 2
Effect on urinary phosphorus excretion.
In rats acclimated to sevelamer [1.5% (wt/wt)] during the run-in period, the addition of vehicle did not affect urinary phosphorus excretion, whereas the addition of tenapanor (0.15 mg/kg bid) reduced residual urinary phosphorus excretion by 37 ± 6% (P < 0.05) (Fig. 6B). The final level of urinary phosphorus excretion was similar if tenapanor (0.15 mg/kg bid) and sevelamer [1.5% (wt/wt)] were initiated together following the run-in period on control diet (24 ± 4 mg/day) or if tenapanor was added following a 6-day acclimation to sevelamer (21 ± 2 mg/day). In addition, the initiation of tenapanor treatment (0.15 mg/kg bid) in rats already acclimated to phosphate binder [1.5% (wt/wt) sevelamer] allowed for a 50% reduction in binder dose [0.75% (wt/wt) sevelamer] and maintained a similar quantity of urinary phosphorus excretion to those achieved by the higher dose of binder alone (Fig. 6B). Finally, a higher dose of tenapanor (0.5 mg/kg bid) significantly reduced residual urinary phosphorus excretion by 37 ± 6% (P < 0.05) even when the phosphate binder was discontinued (Fig. 6B).

DISCUSSION
Hyperphosphatemia is common in patients with CKD receiving dialysis owing to sustained intestinal absorption of phosphate despite an impaired ability to excrete phosphate by the kidneys; dialytic clearance of phosphate is typically insufficient to maintain serum phosphorus concentrations  within or near the population reference range (1,2,8).
Despite the availability of multiple phosphate binders, control of hyperphosphatemia in patients with CKD receiving dialysis remains poor, and the population-attributable risks of death, cardiovascular events, and fracture exceed those related to comorbid conditions, including diabetes and hypertension, and other complications of CKD, including anemia and "under-dialysis" (lower than targeted Kt/V urea ).
Tenapanor is an investigational, nonbinder, phosphate and sodium absorption inhibitor that has been shown to significantly lower serum phosphorus concentrations in patients receiving maintenance dialysis (17). Tenapanor, via its inhibition of sodium absorption, has been evaluated in a pilot trial examining effects on interdialytic weight gain in patients receiving hemodialysis (38). In a series of randomized clinical trials in patients with constipation-predominant irritable bowel syndrome (IBS-C) (39,40), tenapanor improved IBS-C symptoms and was generally well tolerated, resulting in regulatory approval for that indication. Paracellular phosphate flux through tight junction complexes, determined by the electrochemical phosphate gradient and paracellular phosphate permeability, is quantitatively the most important mechanism of intestinal phosphate absorption under typical conditions of phosphate availability (5).
Tenapanor, a small-molecule phosphate absorption inhibitor, reduces intestinal phosphate absorption by reducing paracellular phosphate permeability (20), whereas sevelamer, a phosphate binder, complexes with luminal phosphate to reduce the concentration gradient of phosphate in the gut (29).
In this study, we demonstrated that both tenapanor and sevelamer work individually in rats, and act synergistically to reduce urinary phosphorus excretion. These findings suggest enhanced inhibition of intestinal phosphate absorption if both mechanisms (e.g., tenapanor reducing the paracellular absorption of phosphate by inhibiting NHE3 and sevelamer reducing the concentration gradient for phosphate absorption by binding luminal phosphate) are combined and support the potential for improved control of hyperphosphatemia with the combination of tenapanor and sevelamer. Using the Bliss model of independence, we determined that the combination of tenapanor and sevelamer yielded synergistic phosphate-lowering effects even at the highest sevelamer dose (a dose equivalent to 20 g per day in humans), well in excess of doses used in clinical practice. Indeed, the enhanced inhibition of intestinal phosphate absorption observed with the combination of tenapanor and sevelamer was the most prominent at lower, clinically relevant sevelamer doses (41). This interpretation was also confirmed when urinary phosphorus excretion was normalized to dietary phosphorus intake.
Paracellular phosphate flux is quantitatively the most important overall mechanism of intestinal phosphate absorption. However, transcellular sodium-dependent phosphate transporters, which are upregulated in response to treatment with sevelamer (25) and display high phosphate affinity (26), also contribute to overall intestinal phosphate absorption and likely remain intact in the presence of the sevelamer and tenapanor combination contributing to some residual phosphate absorption.
It is noteworthy that in this study, we aimed to assess the effects of combined tenapanor and sevelamer administration on intestinal phosphate absorption in healthy rats rather than examining any effects of the combination in a rat model with impaired kidney function. It should be acknowledged that use of urinary phosphorus may be seen a limitation of our methodology, as it is not a direct measure of phosphate absorption. In healthy rats, phosphate homeostasis is achieved by a complex interplay of renal, intestinal, and endocrine mechanisms (42). These mechanisms have the potential to affect urinary phosphorus concentrations in response to altered gastrointestinal phosphate absorption. However, precise measurement of fecal phosphorus can be challenging owing to the higher incidence of loose stools associated with tenapanor treatment (19). Previous phosphate-balance studies involving phosphate binders (43) or an analogue of tenapanor (19) conducted in rats show, however, that decreases in urinary phosphorus are associated with increases in fecal phosphorus content (a direct measure of intestinal absorption), and provide additional support for urinary phosphorus as an accepted marker for intestinal phosphate absorption in healthy rats (9,(30)(31)(32). Furthermore, tenapanor treatment has shown minimal effects on levels of phosphate-regulating hormones (FGF23, parathyroid hormone, and vitamin D) in healthy rats receiving a regular phosphate diet (20). Our overall results are consistent with data reported from the AMPLIFY trial (44) (NCT03824587), in which treatment with tenapanor and phosphate binders (including sevelamer) in combination significantly reduced serum phosphorus concentrations relative to phosphate binders alone (45). The ongoing NORMALIZE trial (46) (NCT03988920) will further evaluate the efficacy and safety of tenapanor and phosphatebinder combination treatment in patients receiving maintenance dialysis (peritoneal and hemodialysis).
In summary, in rats fed standardized high-phosphate diets, tenapanor was effective at reducing urinary phosphorus excretion (reflecting intestinal phosphate absorption), and when used together with the phosphate binder sevelamer, tenapanor significantly augmented the reduction in urinary phosphorus excretion. Formal analysis suggested a synergistic, rather than an additive, effect. Combined tenapanor and phosphate binder therapy could help a larger proportion of patients achieve target serum phosphorus concentrations. For the minority of patients who achieve control of hyperphosphatemia, the addition of tenapanor to any phosphate binder regimen could enhance control of serum phosphorus with a lower pill burden and cause a reduction in phosphate binder-associated adverse effects. Further evaluation of combination tenapanor plus phosphate binder treatment in patients receiving dialysis with hyperphosphatemia is warranted.

GRANTS
This study was funded by Ardelyx. Medical writing support was provided by Steven Inglis (PhD) and Andrew Liew (PhD) of Oxford PharmaGenesis (Melbourne, VIC, Australia) and funded by Ardelyx.

DISCLOSURES
A.J.K. was affiliated with Ardelyx at the time this study was conducted and has since become an employee of Chinook Therapeutics. Z.J. was affiliated with Ardelyx at the time this study was conducted and has since become an employee of Anwita Biosciences. A.Q. was affiliated with Ardelyx at the time this study was conducted and has since become an employee of Senti Biosciences. J.K., C.F., P.K., and D.P.R. are employees and have ownership interest in Ardelyx. G.M.C. is a consultant to and has equity ownership interest in Ardelyx.