|Home | About | Journals | Submit | Contact Us | Français|
Dietary potassium (K) deficiency is accompanied by phosphaturia and decreased renal brush border membrane (BBM) vesicle sodium (Na)-dependent phosphate (Pi) transport activity. Our laboratory previously showed that K deficiency in rats leads to increased abundance in the proximal tubule BBM of the apical Na-Pi cotransporter NaPi-IIa, but that the activity, diffusion, and clustering of NaPi-IIa could be modulated by the altered lipid composition of the K-deficient BBM (Zajicek HK, Wang H, Puttaparthi K, Halaihel N, Markovich D, Shayman J, Beliveau R, Wilson P, Rogers T, Levi M. Kidney Int 60: 694–704, 2001; Inoue M, Digman MA, Cheng M, Breusegem SY, Halaihel N, Sorribas V, Mantulin WW, Gratton E, Barry NP, Levi M. J Biol Chem 279: 49160–49171, 2004). Here we investigated the role of the renal Na-Pi cotransporters NaPi-IIc and PiT-2 in K deficiency. Using Western blotting, immunofluorescence, and quantitative real-time PCR, we found that, in rats and in mice, K deficiency is associated with a dramatic decrease in the NaPi-IIc protein abundance in proximal tubular BBM and in NaPi-IIc mRNA. In addition, we documented the presence of a third Na-coupled Pi transporter in the renal BBM, PiT-2, whose abundance is also decreased by dietary K deficiency in rats and in mice. Finally, electron microscopy showed subcellular redistribution of NaPi-IIc in K deficiency: in control rats, NaPi-IIc immunolabel was primarily in BBM microvilli, whereas, in K-deficient rats, NaPi-IIc BBM label was reduced, and immunolabel was prevalent in cytoplasmic vesicles. In summary, our results demonstrate that decreases in BBM abundance of the phosphate transporter NaPi-IIc and also PiT-2 might contribute to the phosphaturia of dietary K deficiency, and that the three renal BBM phosphate transporters characterized so far can be differentially regulated by dietary perturbations.
the kidney plays a critical role in the maintenance of phosphate (Pi) homeostasis. Plasma phosphate levels are maintained through regulated reabsorption of filtered Pi along the proximal tubule. Apical entry of Pi into the proximal tubule epithelial cells is mediated by at least two brush border membrane (BBM) sodium-coupled Pi transporters: NaPi-IIa (SLC34A1) and NaPi-IIc (SLC34A3) (6, 9, 19, 20).
NaPi-IIa and NaPi-IIc respond to variations in dietary Pi, parathyroid hormone (PTH), fibroblast growth factor-23, and other factors by changing their abundance in the proximal tubule BBM. However, the time scales of their regulation by dietary Pi and PTH are quite different, with rat and mouse NaPi-IIa downregulation by high-dietary Pi concentrations or PTH and upregulation under low-dietary Pi occurring within 1 h, while NaPi-IIc regulation takes up to four times longer (23, 24). Also, while their overall molecular structure is predicted to be very similar, important differences between these two Na+-Pi cotransporters exist. For example, NaPi-IIa is electrogenic, coupling Pi transport (at physiological pH mainly HPO42−) with the transport of three Na+ ions. In contrast, NaPi-IIc is electroneutral, only transporting two Na+ for every Pi (2). In addition, our laboratory recently determined that NaPi-IIc interacts with some of the same proteins as NaPi-IIa [Na/H exchanger regulating factor (NHERF)-1 and NHERF-3/PDZK1], but not others (Shank2E, PIST) (27).
To date, NaPi-IIa has been considered the main apical proximal tubule Pi transporter, as deduced from studies with NaPi-IIa knockout mice. The NaPi-IIa knockout mice exhibit a 70% increase in phosphaturia, indicating that, at least in mice, NaPi-IIa might be responsible for up to 70% of the renal Pi reabsorption, with NaPi-IIc or other apical Pi transporters accounting for the remaining 30% (3). However, recent research suggests that, in humans, NaPi-IIc might be an important regulator of Pi balance. Indeed, mutations in Npt2c, the gene encoding NaPi-IIc, have been identified as the cause of hereditary hypophosphatemic rickets with hypercalciuria, a life-long phosphate wasting disease (4, 10, 12, 15). These findings, in addition to the studies indicating important differences between NaPi-IIa and NaPi-IIc physiology, have stimulated great interest in further studies of NaPi-IIc regulation.
Here we determined whether NaPi-IIc is also regulated by dietary potassium (K) deficiency. Our laboratory previously showed regulation of NaPi-IIa, NaPi-1 [a type I Na+-Pi cotransporter (5)], and PiT-1 [a type III Na+-Pi cotransporter (8, 29)] in rats fed a K-deficient diet. In particular, we found that K deficiency leads to increased abundance of these Pi transporters in the proximal tubule BBM; however, at the same time, Na+-coupled Pi transport in BBM vesicles derived from K-deficient rats was reduced compared with Na+-Pi-coupled transport in BBM from control rats (32).
We next sought to determine whether an altered lipid environment could explain the decreased Na+-Pi cotransport activity in K-deficient BBM, despite the increased NaPi-IIa abundance. We found that K deficiency is associated with an increase in BBM sphingomyelin, glucosylceramide, and ganglioside GM3 and a decrease in BBM lipid fluidity (32). Using scanning fluctuation correlation spectroscopy and molecular brightness analysis, we further found that NaPi-IIa diffusion in K-deficient BBM was slowed, and NaPi-IIa cluster size increased compared with the transporter's diffusion and clustering in control BBM (11). Decreased diffusion and increased cluster size could both reduce the Pi transport activity of NaPi-IIa and thus contribute to the reduced Na+-coupled Pi transport activity in K-deficient BBM, despite the increased NaPi-IIa BBM abundance.
In this study, we find that dietary K deficiency is associated with a sharp decrease in NaPi-IIc BBM abundance, in both rats and mice. We suggest that this decrease contributes to the observed hypophosphatemia and increased urinary Pi excretion in K deficiency. In addition, electron microscopy indicates redistribution of NaPi-IIc from the BBM microvilli to intracellular vesicles in dietary K deficiency. Finally, we also find that PiT-2, another type III Na-Pi cotransporter, is localized in the apical BBM of the proximal tubule, is regulated by dietary Pi, and is downregulated in K-deficient rats and mice.
Male Sprague-Dawley rats, weighing between 175 and 200 g, were obtained at 6 wk of age from Charles River Laboratories (Wilmington, MA). C57BL/6 mice, weighing between 20 and 25 g, were obtained at 6 wk of age from Jackson Laboratory (Ben Harbor, MA). All animals were housed at the Center for Laboratory Animal Care at the University of Colorado Health Sciences Center and put on a 12:12-h light-dark cycle. At 8 wk of age, the animals were fed a control diet (Teklad Diet no. TD88238, Harlan Laboratories, Madison, WI) or a K-deficient diet (Teklad Diet no. TD95006) for 14 days. The diets were otherwise matched for their phosphorus, calcium, magnesium, sodium, protein, carbohydrate, and fat content. We used 6 rats or 12 mice in each group per experiment, and the experiments were repeated three times. For the dietary Pi studies, animals were fed, for 14 days, a low-Pi diet (0.1% Pi, Teklad diet no. 85010), a control Pi diet (0.6% Pi, Teklad diet no. 84122), or a high-Pi diet (1.2% Pi, Teklad diet no. 85349). The animal studies were approved by the Institutional Animal Care and Use Committee at the University of Colorado. Urine K, phosphorus, and creatinine were measured using commercially available kits from Stanbio Laboratory (Boerne, TX). Plasma creatinine was measured using a 3200 Q trap liquid chromatography-tandem mass spectrometry (Applied Biosystems, Foster City, CA), according to Ref. 25.
All chemicals were obtained from Sigma, except when noted. Two polyclonal rabbit anti-NaPi-IIa antibodies were generated, one by Affinity Bioreagents (Golden, CO) and one by Colorado State University (Fort Collins, CO), and used at 1:10,000 for Western blotting and at 1:250 for immunofluorescence microscopy. The rabbit polyclonal anti-NaPi-IIc antibody was used as described (22). A chicken anti-NaPi-IIc antibody was custom-made by Davids Biotechnologie (Regensburg, Germany), as described (27), and used at 1:1,000 for Western blotting and at 1:100 for immunofluorescence. A rabbit anti-PiT-2 antibody was also from Davids Biotechnologie and generated by injection of the peptide HCKVGSVVAVGWIRSRKA. The PiT-2 antibody was used at 1:1,000 for Western blotting and 1:50 for immunofluorescence. The NHERF-1 antibody was a generous gift from Dr. E. Weinman (University of Maryland, Baltimore, MD) and used at 1:2,500 for Western blotting. The PDZK1 (CLAMP1) antibody was from BD Biosciences (San Jose, CA) and used at 1:5,000 for Western blotting. The EEA1 and Rab5 antibodies were also from BD Biosciences and used at 1:2,000 and 1:250, respectively, for Western blotting.
Rats were anesthetized via an intraperitoneal injection of 50 mg/kg pentobarbital sodium (Pentothal, Abbott Laboratories, Chicago, IL). The renal vessels were clamped before removal of the kidneys. Thin slices from the superficial cortex (SC) were dissected on ice-cooled glass. The SC and juxtamedullary cortex (JMC) kidney slices were homogenized in 15-ml ice-cold isolation buffer consisting of 300 mM mannitol, 5 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 16 mM HEPES, and 10 mM Tris, pH 7.5, and 1 tablet Roche Mini-Complete per 200 ml buffer using a Polytron homogenizer (90 s at 40% power). The BBM were isolated from the homogenate by Mg2+ precipitation, followed by differential centrifugation, as described (7). Briefly, 0.54 ml of 1 M MgCl2 and 21 ml of water were added to each 15 ml of kidney homogenate. After 20 min of shaking, the homogenate was centrifuged at 2,790 g for 15 min. The supernatant was subjected to another round of Mg2+ precipitation, and the resulting supernatant was centrifuged at 40,000 g for 30 min. The resulting BBM pellets were resuspended in a buffer containing 300 mM mannitol, 16 mM HEPES, 10 mM Tris, pH 7.5, and one Mini-Complete tablet (Roche, Indianapolis, IN) per 50 ml buffer. BBM total protein concentration was determined using the bicinchoninic acid assay (Pierce, Rockford, IL).
Mice were anesthetized via an intraperitoneal injection of 50 mg/kg pentobarbital sodium (Pentothal, Abbott Laboratories). After clamping of the renal vessels, the kidneys were removed and thinly sliced. Kidney slices from two mice were combined in 7.5 ml isolation buffer consisting of 15 mM Tris·HCl, pH 7.4, 300 mM mannitol, 5 mM EGTA, and 1 Roche Complete inhibitor tablet per 250 ml buffer. The kidney slices were homogenized using a Potter-Elvejham homogenizer with 8–10 rapid strokes and transferred to a chilled capable tube. Kidney residues remaining on the homogenizer were rinsed off with 10 ml water that was then added to the kidney homogenate. BBM were prepared by a double Mg2+ precipitation analogous to the preparation of rat BBM. For the first Mg2+ precipitation, 300 μl of 1 M MgCl2 were added to the homogenate, and the solution was vortexed and shaken every 5 min for 20 min before centrifugation at 2,500 g for 15 min. The supernatant was subjected to a second Mg2+ precipitation, and, from the resulting supernatant, the BBM was recovered by centrifugation at 38,000 g for 40 min. The BBM was resuspended, and its protein content quantified as above for the rat BBM.
Phosphate transport was measured by radioactive 32Pi uptake in freshly isolated BBM vesicles, as described (32).
BBM proteins (10 or 20 μg total protein) were separated by 7.5 or 10% SDS-PAGE (Criterion, Bio-Rad, Hercules, CA) and transferred onto nitrocellulose membranes (Bio-Rad). Membranes were blocked for 30 min at room temperature with 5% milk in TBST buffer (20 mM Tris·HCl, 150 mM NaCl, 0.5% Tween, pH 7.4) before incubation with primary antibodies diluted in TBTS/milk overnight at 4°C. After four washes with TBST, membranes were incubated with horseradish peroxidase-conjugated goat secondary antibodies (Pierce), diluted 1:10,000 for 1 h at room temperature. Horseradish peroxidase was detected following 2 min of incubation in Supersignal West Pico Chemiluminescent Substrate or Supersignal West Dura Extended Duration Substrate (Pierce), using either film or a charge-coupled device imaging system. Films were scanned using a Bio-Rad imager. Band intensities were quantified using Quantity One or ImageJ software. Membranes were stripped using Restore Stripping buffer (Pierce) and reprobed using a mouse anti-β-actin antibody (Sigma) to confirm equal total protein loading. Densitometry data are presented as average ± SD.
Rat kidney cortex homogenates were cleared of large debris by centrifugation at 250 g for 10 min at 4°C. To obtain total membranes, the supernatant was further centrifuged at 100,000 g for 1 h at 4°C. The membranes were resuspended in a buffer containing 300 mM mannitol, 16 mM HEPES, 10 mM Tris, pH 7.5, and 1 Mini-Complete tablet (Roche, Indianapolis, IN) per 50 ml buffer. Western blotting was performed as above, except that infrared fluorescence detection was also used, using Li-COR IRDye-conjugated secondary antibody and an Odyssey instrument (Li-COR Biosciences, Lincoln, NE).
Total RNA was isolated from kidney samples using the RNeasy Mini Kit from Qiagen (Valencia, CA), and cDNA was synthesized using reverse transcription reagents from Bio-Rad. The mRNA level was quantified using a Bio-Rad iCyCler real-time PCR machine. Cyclophilin was used as internal control, and the amount of RNA was calculated by the comparative threshold cycle method, as recommended by the manufacturer. All of the data were calculated from triplicate reactions. Primer sequences used are as follows: rat NaPi-IIa forward, 5′-GCC ACT TCT TCT TCA ACA TC-3′; rat NaPi-IIa reverse, 5′-CAC ACG AGG AGG TAG AGG-3′; rat NaPi-IIc forward, 5′-TCT TCG CAG TTC AGG TTG-3′; rat NaPi-IIc reverse, 5′-GTG AGT AGT AAG TAG ACA ATG G-3′; rat NHERF-1 forward, 5′-TCA ACA TTC AAA TCA GCA TCA G-3′; rat NHERF-1 reverse, 5′-GAA GAG CAG GGA GTC AGG-3′; rat PDZK1 forward, 5′-AAT CAT CAA GGA CAT AGA ACC-3′; rat PDZK1 reverse, 5′-CCA GCA CCA ACA GAG TAG-3′; rat PiT-2 forward, 5′-GTG GAT GGA ACT CGT CAA G-3′; rat PiT-2 reverse, 5′-CAG GAT GAA CAG CAC ACC-3′; mouse NaPi-IIa forward, 5′-AGA CAC AAC AGA GGC TTC-3′; mouse NaPi-IIa reverse, 5′-CAC AAG GAG GAT AAG ACA AG-3′; mouse NaPi-IIc forward, 5′-CAT CTT CAA CTG GCT CAC-3′; mouse NaPi-IIc reverse, 5′-GGT TAT CAC ACT GCT ATC C-3′; mouse PiT-2 forward, 5′-TGC TCT GCT GTT CGC CTT C-3′; mouse PiT-2 reverse, 5′-TCT CTA ATC TGC CTG CTA TCT TCC-3′. RNA data are presented as average relative levels vs. cyclophilin A ± SD.
Six control and six K-deficient rats were anesthetized, as described above, before retrograde perfusion through the abdominal aorta. The perfusion buffer consisted of 3% paraformaldehyde (diluted from 32% paraformaldehyde ampoules, Electron Microscopy Sciences, Hatfield, PA) in a 6:4 mixture of cacodylate buffer (pH 7.4, adjusted to 300 mosmol with sucrose) and 10% hydroxyethyl starch. The perfusion pressure was 300 mmHg, and 1 ml perfusion buffer was used per gram body weight. After 5 min, the fixative was washed out by perfusion for 5 min with the cacodylate buffer. The kidneys were removed, cut into four to six slabs, embedded in optimum cutting temperature, and frozen in liquid nitrogen. Sections 6 μm thick were cut on a Leica cryostat at −20°C and stored at −20°C until ready to use. Sections were washed once with PBS before blocking for 30 min in staining solution (PBS containing 0.1% Triton X-100 and 5% milk powder) to which 10% goat serum was added. Sections were then incubated overnight at 4°C with primary antibodies diluted in the staining solution. After four washes with PBS containing 0.1% Triton X-100, sections were incubated with secondary goat antibodies (Invitrogen, Carlsbad, CA) diluted 1:250 as well as Alexa 633-phalloidin (Invitrogen) in staining solution for 1 h at room temperature. After four washes with PBS containing 0.1% Triton X-100, the sections were mounted in Vectashield (Vector Laboratories, Burlingame, CA). Fluorescence images were acquired on a Zeiss 510 LSM laser scanning confocal microscope (Carl Zeiss Microimaging, Thornwood, NY).
Adult Sprague-Dawley rat were anesthetized with inhalant isoflurane. The kidneys were preserved by in vivo retrograde aortic perfusion with PBS, followed by 4% paraformaldehyde in 0.1 M lysine buffer containing 0.01 M NaIO4 and 0.15 M sucrose. Samples of renal cortex were dissected and shipped in fixative by overnight courier to the University of Florida College of Medicine Electron Microscopy Core Facility, where they were rinsed in PBS, incubated in 0.1 M NH4Cl for 1 h, and then dehydrated in a graded series of ethanols, infiltrated, embedded in Lowicryl K4M resin (Electron Microscopy Sciences, Ft. Washington, PA), and polymerized under UV light for 24 h at −20°C, followed by ~60 h at room temperature. Ultrathin sections of samples containing well-preserved proximal tubule were mounted on Formvar/carbon-coated nickel grids.
For immunogold labeling, the ultrathin tissue sections were exposed to the primary antibody and then to a colloidal gold-conjugated secondary antibody. For NaPi-IIc localization, the chicken anti-NaPi-IIc primary antibody was diluted 1:100, and the secondary was rabbit anti-chicken IgY conjugated to 5-nm diameter colloidal gold (BB International, Ted Pella, Redding, CA). Unless noted otherwise, all steps were done by floating the grids on droplets of solution at room temperature. The sections were exposed to 0.1 M NH4Cl for 1 h, incubated with 1% BSA in PBS, pH 7.4, for 30 min, washed with incubation solution [0.2% acetylated BSA (Aurion BSA-c, Electron Microscopy Sciences), 10 mM NaN3, in PBS, pH 7.4], and then incubated in a humidified chamber overnight at 4°C with the primary antibody diluted in incubation solution. The sections were then washed with incubation solution and exposed to the secondary antibody diluted in incubation solution for 1.5 h at room temperature. The sections were washed with incubation solution, washed with PBS, postfixed with 1.25% glutaraldehyde in PBS, washed with PBS, and finally washed with glass-distilled water. The sections were air dried over night and counterstained with saturated uranyl acetate. Each group of sections subjected to the immunogold procedure included a control section that was exposed to incubation buffer in place of the primary antibody.
Ultrathin sections were examined using a Hitachi 7600 transmission electron microscope (Hitachi High Technologies America, Pleasanton, CA) equipped with a MacroFire slow-scan charge-coupled device camera (Optronics, Goleta, CA) and AMT Image Capture software (Advanced Microscopy Techniques, Danvers, MA).
Transverse slices of 4% paraformaldehyde-lysine-peridoate-sucrose preserved kidney, 2–3 mm thick, were rinsed in PBS, dehydrated in a graded series of ethanols, and infiltrated and embedded in polyester wax, made from polyethylene glycol 400 distearate (Polysciences, Warrington, PA) and 10% acetyl alcohol. Sections 3 μm thick were mounted on gelatin-coated slides and heated at 37°C overnight. The sections were dewaxed in a graded series of ethanols, rinsed in PBS, and treated with 3% H2O2, followed by 5% normal goat serum in PBS to prevent nonspecific reactivity. The sections were then incubated with the anti-NaPi-IIa primary antibody diluted 1:1,000 overnight at 4°C. The sections were then washed, exposed to anti-rabbit polymer-linked, peroxidase-conjugated secondary antibody (MACH2, Biocare Medical, Concord, CA), washed, and reacted with diaminobenzidine (Vector Laboratories, Burlingame, CA) for 5 min. Finally, sections were washed, dehydrated in graded ethanols followed by xylene, and mounted on glass slides with Eukitt Mounting Medium (Hawthorne, NY).
Sections were photographed using a Nikon LaboPhot-2 microscope equipped with a Nikon DS-5M digital color camera and NIS Elements software (Nikon USA, Melville, NY). Only sections from the same immunohistochemistry experiment, subjected to identical experimental conditions, were used for comparisons of immunolabel between control and K-deficient rats. Digital images for comparisons were collected in a single session with identical manual settings on both the camera and photomicroscope. No corrections or adjustments in contrast, brightness, or color balance were made to the collected images.
Data are expressed as means ± SD. Data were analyzed for statistical significance by unpaired Student's t-test or one-way analysis of variance.
The presence of PiT-2 in the rat proximal tubule BBM and its response to changes in dietary Pi intake, establishing PiT-2 as a third player in rat renal Pi reabsorption, was recently reported (28) and is further illustrated by immunofluorescence data shown in Fig. 1A. In particular, we wish to illustrate the upregulation of PiT-2 at the BBM in all segments of the proximal tubule under dietary Pi deprivation. Under normal (0.6%) dietary Pi, PiT-2 BBM localization overlaps strongly with NaPi-IIc BBM localization at the S1 segments only, while, in the S2 and S3 segments, PiT-2 is not concentrated at the BBM (see e.g., Fig. 3 in Ref. 23 or Fig. 6C). When dietary Pi is chronically low (0.1%), PiT-2 is expressed at the BBM, not only in S1, but also in S2 and S3 (Fig. 1A, top right, red, S3 segments indicated by asterisks), similar to NaPi-IIa (Fig. 1A, top left, red). In contrast, high-dietary Pi (1.2%) dramatically decreases the expression levels of NaPi-IIa and PiT-2 in all segments of the proximal tubule (Fig. 1A, bottom).
Figure 1B establishes PiT-2 as a kidney BBM transporter regulated by dietary Pi intake in mice. BBM were isolated from mice chronically fed either a low-Pi (0.1%) diet, a normal Pi (0.6%) diet, or a high-Pi (1.2%) diet and probed by Western blots for NaPi-IIa, NaPi-IIc, and PiT-2. For the three Na+-coupled Pi transporters, parallel decreases in mouse renal BBM abundance are observed when the dietary Pi content is increased.
A summary of the rat parameters measured in this dietary K deficiency study is presented in Table 1. In addition to decreased body weights but increased kidney weights, rats fed a K-deficient diet for 14 days display hypokalemia and hypophosphatemia and have increased urine Pi output compared with rats fed a control diet.
Total renal cortical BBM was isolated from rats fed either a control diet or a K-deficient diet for 14 days. Cortical BBM Na+-coupled Pi transport activity was decreased by almost 40% in BBM isolated from K-deficient rats compared with control rats (from 800 ± 200 to 480 ± 130 pmol 32P·10 s−1·mg BBM protein−1, P = 0.002). Separation of the BBM proteins by SDS-PAGE followed by Western blotting indicated a significant increase in NaPi-IIa BBM abundance in dietary K deficiency (Fig. 2A), in agreement with our laboratory's earlier studies (32). In contrast, NaPi-IIc BBM abundance in dietary K deficiency was dramatically decreased (Fig. 2B), while BBM PiT-2 abundance was also significantly reduced (Fig. 2C). When total renal cortical membranes were assayed, the observed changes in NaPi-IIa, NaPi-IIc, and PiT-2 paralleled the ones seen in the BBM (Supplemental Fig. S1). (The online version of this article contains supplemental data.)
In addition to the significant decreases in NaPi-IIc and PiT-2 BBM abundance with dietary K deficiency, a small decrease was observed in BBM abundance of the NaPi-IIa and NaPi-IIc interacting proteins NHERF-1 (EBP50) and PDZK1 (NHERF-3) in dietary K deficiency (Fig. 3). Interestingly, we always observed two specific bands for NHERF-1, and only the lower molecular weight band is decreased in dietary K deficiency.
We also separated the rat SC from the JMC and isolated BBM from both. Western blotting revealed an increase in NaPi-IIa in both SC and JMC BBM with dietary K deficiency (Fig. 4A). For both NaPi-IIc and PiT-2, JMC BBM abundance was much smaller compared with SC BBM abundance, in agreement with their observed localization predominantly in the S1 segments of the proximal tubule with a 0.6% Pi diet. However, for both NaPi-IIc and PiT-2, a significant decrease in BBM abundance was observed with dietary K deficiency, in both the SC and the JMC BBM (Fig. 4, B and C).
Measurements of mRNA levels by quantitative real-time PCR indicated a decrease in Npt2a transcripts in dietary K deficiency in both the SC and JMC (Fig. 5A), while the Npt2c and PiT-2 transcript decreased in SC only (Figs. 5, B and C). Thus, whereas for NaPi-IIc and PiT-2 BBM protein levels parallel mRNA levels, for NaPi-IIa posttranslational mechanisms are at play that increase its BBM protein abundance, despite decreases in transcript amount. Transcript levels of NHERF-1 and PDZK1 were not altered in dietary K deficiency (Fig. 5D).
The increased abundance of NaPi-IIa and decreased abundance of NaPi-IIc and PiT-2 in the K-deficient BBM were also observed by immunofluorescence of rat kidney sections. We simultaneously used a rabbit anti-NaPi-IIa antibody, a chicken anti-NaPi-IIc antibody, and Alexa Fluor-labeled phalloidin to localize NaPi-IIa, NaPi-IIc, and F-actin in the same sections. In a parallel section, we then used the rabbit anti-PiT-2 antibody, the chicken anti-NaPi-IIc antibody, and phalloidin to obtain, in addition, the localization and abundance of PiT-2. Figure 6 shows representative images obtained in this manner for parallel sections from a kidney from a control rat (Fig. 6A) and from a rat on a K-deficient diet (Fig. 6B). Whereas the increase in NaPi-IIa BBM staining intensity in K deficiency is subtle, there is a very clear decrease in both the intensity and the number of proximal tubules that show BBM staining for NaPi-IIc or PiT-2. This is also clearly visible in the pseudocolored overlay images shown in Fig. 6C. These images further indicate that, under normal dietary Pi, BBM PiT-2 is localized at the same proximal tubule sites as NaPi-IIc, i.e., S1 fragments only (yellow in NaPi-IIc/PiT-2 overlay image, tubule segments indicated by an asterisk) (23), while NaPi-IIa is more widely distributed and present in S1, S2, and S3 proximal tubule BBM. However, an intracellular pool of PiT-2 is visible in all proximal tubule epithelial cells, and, when animals are chronically fed a low-Pi diet (0.1% Pi) rather than a control Pi diet (0.6% Pi), upregulation of PiT-2 at the BBM of the later proximal tubule segments is visible, resulting in an expression pattern that is similar to NaPi-IIa and encompasses BBM expression in later segments of the proximal tubule (Fig. 1A).
The abundance of NaPi-IIa in the BBM of rats fed a K-deficient or control diet was further evaluated by immunohistochemistry. Figure 7 compares similar fields in the JMC in a control (panel a) and K-deficient (panel b) rat. A clear increase in NaPi-IIa staining is seen in the K-deficient BBM. In addition, an undulating luminal surface is observed in many proximal tubules of the rats fed a K-deficient diet (indicated by asterisks), caused by variations in cell height in neighboring cells.
To determine whether the decrease in NaPi-IIc BBM abundance seen by immunofluorescence microscopy (Fig. 6) and in Western blots (Figs. 2 and and4)4) is accompanied by internalization of the transporter, we used immunogold labeling and electron microscopy. In control rats, the majority of immunogold label for NaPi-IIc was associated with the microvilli in the proximal tubule BBM (Fig. 8a, arrows). Only a few gold particles were present over cytoplasmic vesicles, primarily in the pits at the base of the microvilli (Fig. 8c, arrows) and in small vesicles in the subapical region (Fig. 8c, black arrowheads). Gold particles were only rarely found in endosomes or lysosomes (Fig. 8e). In K-deficient rats, the BBM immunolabel for NaPi-IIc was markedly reduced (Fig. 8b, arrows), and labeling of cytoplasmic vesicles was prevalent (Fig. 8d, black arrowheads). Furthermore, in the K-deficient animals, NaPi-IIc immunolabel was frequently found in cytoplasmic vesicles deep within the proximal tubule cell body (Fig. 8f, black arrowheads), unlike the controls where cytoplasmic label was almost exclusively subapical.
Gold labeling was also found in lysosomes (black arrows) and in small vesicles associated with lysosomes (black arrowheads) more frequently in K-deficient rats (Fig. 9 b) than in the control animals (Fig. 9a).
The specificity of the immunogold labeling is illustrated in Supplemental Fig. S2, which shows the complete absence of label in collecting duct intercalated cells.
We next determined whether mice show the same response to a K-deficient diet as rats. The measured mouse parameters are summarized in Table 2 and indicate hypokalemia and hypophosphatemia. Na+-coupled Pi transport in BBM isolated from mice fed the K-deficient diet for 14 days was reduced by almost 30% compared with the transport measured in BBM from mice fed a control diet (from 211 ± 46 to 155 ± 36 pmol 32P·10 s−1·mg BBM protein−1; P = 0.04). Furthermore, there was a small but insignificant increase in the amount of NaPi-IIa in the K-deficient mouse BBM compared with the control BBM (Fig. 10, A and B). However, the BBM levels of NaPi-IIc and PiT-2 were both significantly decreased in the K-deficient mice compared with the control mice (Fig. 10, A and B). We further found that the mRNA levels of the type II transporters were significantly decreased in mice put on a K-deficient diet for 14 days, but no significant change was observed in the PiT-2 mRNA level (Fig. 10C). Overall, these studies indicate that the responses of rats and mice to a K-deficient diet are very similar with respect to the regulation of the renal BBM Pi transporters.
Studies to date have indicated that renal Pi reabsorption in the proximal tubule is mediated by at least two BBM Na+-coupled Pi transporters: NaPi-IIa and NaPi-IIc. The relative importance of these two Pi transporters in both physiological and disease states is a matter of intense research and debate (30). Earlier studies focused solely on NaPi-IIa, and, from studies in NaPi-IIa knockout mice, it was deduced that NaPi-IIa is responsible for 70% of the total Pi reabsorption, with NaPi-IIc or other Pi transporters accounting for the remaining 30% (3). However, these percentages might not reflect the importance of NaPi-IIc in rat or human Pi reabsorption. Indeed, in humans, mutations in NaPi-IIc and not NaPi-IIa have been identified as the cause of the Pi wasting in hereditary hypophosphatemic rickets with hypercalciuria (4, 10, 12, 15). Our study of the phosphaturia associated with dietary K deficiency in rats and mice further indicates that NaPi-IIc can play an important role in the maintenance of Pi homeostasis since, contrary to NaPi-IIa, its BBM abundance decreases with dietary K restriction (Figs. 2, ,4,4, ,6,6, ,8,8, and and10)10) and thus parallels the observed decrease in renal Na+-coupled Pi transport. In addition, our study further characterized a third apical BBM Pi transporter, PiT-2, which is also regulated differently from NaPi-IIa in diet-induced hypokalemia (Figs. 2, ,4,4, ,6,6, and and1010).
Our earlier studies indicated that dietary K deficiency in rats increases the fractional excretion of Pi, despite increases in the proximal tubule BBM abundance of NaPi-IIa (32). Whereas we previously sought to reconcile these two observations by evoking the altered lipid environment of NaPi-IIa in the K-deficient BBM and its effect on the transporter's activity (10, 26), our present study suggests that an additional regulatory mechanism, leading to decreased NaPi-IIc abundance in K-deficient BBM, could also contribute to the phosphaturia in K deficiency. Indeed, we observed a dramatic decrease in BBM NaPi-IIc with diet-induced hypokalemia, both in rats (Figs. 2, ,4,4, ,6,6, and and8)8) and in mice (Fig. 10). Furthermore, we document the first electron microscopy localization of NaPi-IIc at the proximal tubule brush border and its relocation to deep cytoplasmic vesicles, including endosomes and lysosomes with dietary K-deficiency (Figs. 8 and and99).
It appears likely that the decrease in BBM NaPi-IIc contributes to the decreased Na+-coupled Pi transport activity measured in BBM vesicles from animals on a K-deficient diet. However, to determine the relative contribution of NaPi-IIa and NaPi-IIc to the measured Pi transport activity would require knowledge of their relative expression levels in the rodent BBM. The relative importance of the two transporters was deduced from studies in knockout mice; however, these do not allow accurate determination of protein expression levels, since compensatory mechanisms to improve survival might come into effect. However, future studies of dietary K deficiency in NaPi-IIa or NaPi-IIc knockout mice could more precisely assess the effect of dietary K deficiency on each transporter's expression and activity.
We also found a small decrease in the abundance of one form of the PDZ domain-containing protein NHERF-1, a NaPi-IIa and NaPi-IIc interacting protein (Fig. 2), which might or might not be related to the decreased abundance of NaPi-IIc. This is of interest, since mutations in NHERF-1 were recently identified as a cause of phosphaturia and nephrolithiasis or osteopenia in humans (13, 16).
In addition to decreases in BBM NaPi-IIc, we also observed significant decreases in the BBM abundance of a type III Na+-Pi cotransporter, PiT-2, in dietary K deficiency in both rats and mice (Figs. 2, ,4,4, ,6,6, and and10).10). Type III Na+-Pi cotransporters were originally identified as retroviral receptors (Glvr-1 and Ram-1), but have since been characterized as ubiquitously expressed Pi transporters (PiT-1 and PiT-2, respectively) that were thought to fulfill housekeeping roles in Pi homeostasis (8, 29). Expression of both PiT-1 (26) and PiT-2 (14) in all regions of the kidney has been described. We have focused on PiT-2 as a recent report describes the specific expression of PiT-2 at the apical BBM of renal proximal tubule epithelial cells and its regulation by dietary Pi in the rat (28). Here, we further document the localization of PiT-2 in the proximal tubule BBM of the rat and the downregulation of its BBM abundance by chronic high-dietary Pi (Fig. 1A) and by dietary K deficiency (Figs. 2, ,4,4, and and6).6). We also demonstrate that PiT-2 is localized in mouse BBM, where its abundance is markedly regulated by chronic variations in dietary Pi content (Fig. 1B) or K content (Fig. 10). We note that a low-Pi diet seems to induce apical BBM expression of PiT-2 in the later segments of the proximal tubule, whereas, under normal dietary Pi, the localization of PiT-2 is restricted to the early convoluted and S1 segments (compare Figs. 6C, control diet, and 1A, low-Pi diet).
Our studies also point to the importance of posttranslational regulation in the physiology of the Na+-coupled Pi transporters. Indeed, whereas for NaPi-IIc and PiT-2, both SC mRNA and BBM protein levels are decreased in K deficiency, for NaPi-IIa, there is increased BBM protein abundance but decreased mRNA expression, as determined by quantitative real-time PCR (Fig. 5). Thus, for NaPi-IIa, posttranslational mechanisms are at play that increase its retention in the BBM, and that could include influences of the altered lipid environment, as discussed above.
While dietary K deficiency causes differential regulation of the Na+-Pi transporters by different transcriptional, translational, and posttranslational mechanisms, the relevant signal that mediates these changes is not clear. While these animals develop systemic metabolic alkalosis, they have evidence for intracellular acidosis. Indeed, dietary K deficiency has been shown to be associated with increases in Na+/H+ exchange and Na+-citrate cotransport activity and decrease in Na+-sulfate cotransport activity (17, 18), changes that, together with the decrease in Na+-Pi cotransport activity, have also been reported in models of metabolic acidosis (1, 21). While K deficiency also results in increased stimulation of the renin-angiotensin system, the results in the literature suggest that angiotensin II increases rather than decreases Na+-Pi cotransport activity (31).
In conclusion, our studies of the phosphaturia that accompanies dietary K deficiency have shown that three distinct Pi transporter proteins can contribute in different ways to observed imbalances of Pi homeostasis. While NaPi-IIa was long believed to be the main proximal tubule BBM Pi reabsorber, our studies and others now indicate that NaPi-IIc and PiT-2 might also play important roles in renal Pi reabsorption.
This work was supported by the National Institutes of Health (NIH) (R01 DK066029 to M. Levi and N. P. Barry). S. Y. Breusegem acknowledges the support of a postdoctoral fellowship from the American Heart Association, Pacific Mountain Affiliate (award 0520054Z). J. T. Blaine was supported by a NIH National Research Service Award fellowship, and Y. Caldas was supported by a NIH (3R01 AG026529) minority fellowship.
We thank Dr. E. Weinman (University of Maryland, Baltimore, MD) for the gift of the anti-NHERF-1 antibody.