Gentamicin causes endocytosis of Na/Pi cotransporter protein (NaPi-2)

Kidney International, Vol. 59 (2001), pp. 1024–1036

Gentamicin causes endocytosis of Na/Pi cotransporter protein (NaPi-2) VICTOR SORRIBAS,1 NABIL HALAIHEL,1 KRISHNA PUTTAPARTHI, THOMAS ROGERS, ROBERT E. CRONIN, ANA ISABEL ALCALDE, JOSE ARAMAYONA, MANUEL SARASA, HUAMIN WANG, PAUL WILSON, HUBERT ZAJICEK, and MOSHE LEVI University of Zaragoza, Zaragoza, Spain, and the Dallas VA Medical Center and The University of Texas Southwestern Medical Center at Dallas, Dallas, Texas, USA

Gentamicin causes endocytosis of Na/Pi cotransporter protein (NaPi-2). Background. Renal toxicity is a major side-effect of aminoglycoside antibiotics and is characterized by an early impairment in proximal tubular function. In a previous study, we have shown that gentamicin administration to the rat causes an early impairment in sodium gradient-dependent phosphate (Na/Pi) cotransport activity. The purpose of our current study was to determine the molecular mechanisms of the impairment in Na/Pi cotransport activity, specifically the role of the proximal tubular type II Na/Pi cotransporter. Methods. Rats were treated for one, two, and three days with two daily injections of 30 mg/kg body weight gentamicin or the vehicle. Results. Gentamicin caused a progressive decrease in superficial cortical apical brush-border membrane (SC-BBM) Na/Pi cotransporter activity (856 ⫾ 93 in control vs. 545 ⫾ 87 pmol/mg BBM protein in 3-day gentamicin, P ⬍ 0.01). Western blot analysis showed a parallel and progressive decrease in SCBBM Na/Pi cotransporter protein abundance, a 50% decrease after one day of treatment, a 63% decrease after two days of treatment, and an 83% decrease after three days treatment with gentamicin. In contrast, gentamicin treatment had no effect on Na/Pi cotransport activity or Na/Pi cotransporter protein abundance in BBM isolated from the juxtamedullary cortex (JMC-BBM). Immunofluorescence microscopy showed a major decrease in the expression of Na/Pi cotransporter protein in the apical membrane of the proximal convoluted tubule, with progressive intracellular accumulation of Na/Pi protein. Colocalization studies showed that in gentamicin-treated rats, Na/Pi protein was colocalized in the early endosomes and especially in the lysosomes. Northern blot analysis of cortical RNA interestingly showed no reduction in Na/Pi cotransporter mRNA abundance even after three days of gentamicin treatment.

1

Drs. Sorribas and Halaihel contributed equally to this study.

Key words: early endosomes, lysosomes, axial hetergeneity, phosphatidylinositol, renal toxicity, type II Na/Pi cotransport activity, nephrotoxicity. Received for publication June 9, 2000 and in revised form September 22, 2000 Accepted for publication September 28, 2000

 2001 by the International Society of Nephrology

Conclusion. We conclude that gentamicin inhibits Na/Pi cotransport activity by causing a decrease in the expression of the type II Na/Pi cotransport protein at the level of the proximal tubular apical BBM and that inhibition of Na/Pi cotransport activity is most likely mediated by post-transcriptional mechanisms.

Proximal tubular dysfunction is a central feature of aminoglycoside nephrotoxicity. In humans and in experimental animals treated with aminoglycoside antibiotics, the increase in the urinary excretion of glucose, amino acids, and luminal brush-border membrane (BBM)-bound enzymes occurs before a reduction in glomerular filtration rate (GFR) [1–4]. We have previously shown that treatment of rats with gentamicin results in an early decrease in proximal tubular luminal BBM sodium gradientdependent phosphate transport (Na/Pi cotransport) and pH gradient-dependent sodium transport (Na/H exchange) activities, prior to a reduction in GFR [5]. The purpose of the present study was to characterize the cellular and molecular mechanisms involved in the gentamicin-induced decrease in renal proximal tubular phosphate transport and specifically to determine a role for the type II Na/Pi cotransport (NaPi-2) protein and mRNA in this process. Our results indicate that treatment of rats with gentamicin causes a progressive decrease in superficial cortical (SC-BBM) but not juxtamedullary cortical (JMC-BBM) Na/Pi cotransport activity, which is associated with a progressive decrease in SC-BBM NaPi-2 protein abundance but not NaPi-2 mRNA abundance. Furthermore, gentamicin causes a progressive translocation of NaPi-2 protein from the apical BBM to the late endosomal/ lysosomal compartment. Our results therefore suggest that gentamicin causes a progressive decrease in Na/Pi cotransport activity by post-transcriptional mechanisms, most likely involving the progressive endocytosis of NaPi-2 protein from the apical BBM.

1024

Sorribas et al: Gentamicin and NaPi-2

1025

METHODS

Enzyme activity measurements

Study groups

The activities of maltase (a marker of proximal convoluted tubule BBM) and ␥-glutamyl transferase (a marker of proximal straight tubule BBM) were measured as previously described [5, 6] in SC and JMC homogenates and SC and JMC BBM to determine the purity of BBM isolated from control and gentamicin-treated rats.

Studies were performed on male Sprague-Dawley rats (Charles River Laboratories, Kingston, NY, USA). The rats were virus antibody titer-free and weighed between 225 and 250 g. Experiments were started after animals were adapted on a control diet (0.6% Ca, 0.6% Pi, Teklad) for seven days. The animals then received gentamicin sulfate, 30 mg/kg body weight, administered subcutaneously twice a day, or vehicle (0.9% NaCl) for a one-, two-, or three-day period. At the end of the treatment period, eight rats in each experimental group were anesthetized with intraperitoneal pentobarbital. The kidneys were then rapidly removed; one half of one kidney was used for RNA isolation, and the other three halves were used for BBM isolation. In additional experiments, four rats in each experimental group were anesthetized and perfused in vivo for immunofluorescence microscopy for NaPi-2 protein, as described later in this article. Isolation of cortical homogenate and brush-border membranes Thin slices from the rat kidney SC and JMC were cut at 4⬚C and homogenized with a Polytron homogenizer in a buffer consisting of 300 mmol/L mannitol, 5 mmol/L ethylene glycol-bis (-aminoethyl ether) -N, N, N· N·tetraacetic acid, 1 mmol/L phenylmethylsulfonyl fluoride, 16 mmol/L N-2-hydroxyethylpiperazine-N· -2 ethanesulfonic acid (HEPES), and 10 mmol/L Tris (hydroxymethyl) aminomethane (Tris), pH 7.50. Aliquots were saved for (1) enzyme activity measurement, (2) Western blotting, and (3) measurement of gentamicin content. From the resulting SC and JMC homogenates, apical BBM vesicles were prepared by differential centrifugation after Mg2⫹ aggregation as previously described [6]. Briefly, to each 15 mL of homogenate in isolation buffer, 0.54 mL of 1.0 mol/L MgCl2 in 21 mL of water was added. After 20 minutes of shaking, the homogenate was centrifuged at 2790 ⫻ g for 15 minutes. The pellet resulting from the first Mg precipitation step was saved and labeled as non-BBM membranes, and an aliquot was saved for Western blotting. The supernatant was subjected to another round of Mg precipitation, and the resultant supernatant was centrifuged at 40,000 ⫻ g for 30 minutes. The resulting pellet corresponding to BBM was resuspended in a buffer of 300 mmol/L mannitol, 16 mmol/L HEPES, and 10 mmol/L Tris, pH 7.50, and was aliquoted for simultaneous measurements of (1) enzyme activity, (2) transport activity, (3) protein electrophoresis and Western blotting, (4) lipid composition, and (5) gentamicin binding. Protein concentrations of the cortical homogenate, non-BBM, and BBM fractions were determined by the method of Lowry et al [7].

Brush-border membrane transport activity measurements Brush-border membrane Na/Pi cotransport activity measurements were performed in freshly isolated BBM vesicles by the radiotracer uptake of 100 ␮mol/L K2H32PO4 (Dupont-NEN Research Products, Boston, MA, USA) and an inwardly directed sodium gradient (150 mmol/L NaCl) followed by rapid filtration [8]. Uptake was terminated after 10 seconds, representing the initial linear rate. Na/glucose and Na/proline cotransport measurements were also performed in a similar manner by radiotracer uptake of [3H]-d-glucose or [3H]-l-proline (Dupont-NEN), respectively. Brush border membrane SDS-gel protein electrophoresis and Western blot analysis Brush border membranes were denatured for two minutes at 95⬚C in 2% sodium dodecyl sulfate (SDS), 10% glycerol, 0.5 mmol/L ethylenediaminetetraacetic acid (EDTA), 95 mmol/L Tris-HCl, pH 6.8 (final concentrations). Ten micrograms of BBM protein/lane were separated on 9% polyacrylamide gels according to the method of Laemmli [9] and then electrotransferred onto nitrocellulose paper [10]. After blockage with 5% nonfat milk powder with 1% Triton X-100 in Tris-buffered saline (20 mmol/L, pH 7.3), Western blots were performed with antiserum against the C-terminal amino acid sequence of NaPi-2 at a dilution of 1:5000. Primary antibody binding was visualized using enhanced chemiluminescence (Pierce, Bradford, IL, USA), and the signals were quantitated in a Phosphor Imager with chemiluminescence detector and densitometry software (Bio-Rad, Richmond, CA, USA). The protein blots were also probed with antiserum against ␤-actin (Sigma). Measurement of brush-border membrane lipid composition Lipids from BBMs were extracted by the method of Bligh and Dyer [11]. To determine individual phospholipid polar head group species, an aliquot of the lipid extract was applied to thin-layer chromatography plates (Silica Gel 60; E. Merck, Darmstadt, Germany). Individual phospholipids, including sphingomyelin, phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and phosphatidylinositol, were separated by two-dimensional thin layer chromatography (TLC) [12]. Phospho-

1026

Sorribas et al: Gentamicin and NaPi-2 Table 1. Effects of three days of gentamicin administration on BBM enzyme activity

Superficial cortex Maltase CH BBM Enrichment Recovery ␥-Glutamyl transferase CH BBM Enrichment Recovery Justamedullary cortex Maltase CH BBM Enrichment Recovery ␥-Glutamyl transferase CH BBM Enrichment Recovery

Control

Gentamicin

P value

16.5 ⫾ 0.7 183.8 ⫾ 19.0 11.1 ⫾ 0.9 32.0 ⫾ 2.6

16.9 ⫾ 1.1 177.2 ⫾ 6.1 10.5 ⫾ 0.4 30.2 ⫾ 0.4

NS NS NS NS

120.1 ⫾ 5.7 1627.1 ⫾ 78.1 13.6 ⫾ 1.0 39.2 ⫾ 2.9

123.6 ⫾ 6.1 1705.7 ⫾ 83.4 13.8 ⫾ 1.2 39.8 ⫾ 3.4

NS NS NS NS

9.9 ⫾ 0.6 107.9 ⫾ 10.4 10.9 ⫾ 0.7 31.4 ⫾ 2.3

10.1 ⫾ 0.8 112.1 ⫾ 12.1 11.1 ⫾ 0.8 32.0 ⫾ 2.9

NS NS NS NS

258.7 ⫾ 11.8 3621.8 ⫾ 173.8 14.0 ⫾ 1.2 40.3 ⫾ 3.4

255.9 ⫾ 12.6 3531.4 ⫾ 184.9 13.8 ⫾ 1.8 39.8 ⫾ 3.1

NS NS NS NS

Values are means ⫾ SE; N ⫽ 12 rats in each group. Abbreviations are: CH, cortical homogenate; BBM, brush-border membrane. Enzyme activity is expressed as micromoles per hour per milligram CH or BBM protein. Enrichment is defined as the ratio of specific activity in BBM fraction/specific activity in homogenate. Recovery (%) is defined as the ratio of total activity in BBM fraction/total activity in homogenate.

Fig. 1. Effect of gentamicin on sodium gradient-dependent phosphate (Na/Pi) cotransport activity in (A) superficial cortical brush-border membranes (SC-BBM) and (B) juxtamedullary cortical brush-border membranes (JMC-BBM). NA/Pi cotransport activity was determined by sodium gradient-dependent uptake of 32Pi into BBM vesicles. Gentamicin caused a progressive decrease in SC-BBM but not JMC-BBM Na/Pi transport activity. The results are presented as mean SEM. N ⫽ 6 in each group.

lipid content of total and individual phospholipids was determined by measuring phosphorus content by the method of Ames and Dubin [13]. Cortical homogenate gentamicin content measurement Gentamicin content of SC and JMC homogenate samples obtained from gentamicin-treated rats was measured by radioimmunoassay (Rainen gentamicin RIA kit; DuPont, Billerica, MA, USA) [14]. Brush-border membrane gentamicin-binding measurements Binding of [3H]-gentamicin to SC-BBM and JMCBBM isolated from control (that is, nongentamicin treated) rats was performed at 4⬚C. Specific binding was calculated as the difference between total gentamicin binding and nonspecific binding in the presence of 100fold excess nonradioactive gentamicin. Bmax (binding

capacity) and Kd (binding affinity) were calculated by Scatchard plot. RNA isolation, formaldehyde gel electrophoresis, and Northern blot analysis Thin slices from kidney superficial cortex were cut on an ice-cold glass dish and homogenized with a Polytron homogenizer in a denaturazation solution containing 4 mol/L guanidium thiocyanate, 25 mmol/L sodium citrate, pH 7.0, 0.5% sarcosyl, and 0.1 mol/L of 2-mercaptoethanol. Sequentially, 0.1 volumes of 2 mol/L sodium acetate, pH 4.0, 1 volume water-saturated phenol, and 0.2 volumes chloroform-isoamyl alcohol mixture (49:1) were added to the homogenate. Total RNA was isolated as previously described [8, 15, 16], and 10 ␮g of each sample were size fractionated by agarose-formaldehyde gel electrophoresis and transferred to nylon membranes by a vacuum-blotting device (Bio-Rad). Prehybridization (4 hours at 42⬚C) and hybridization (18 h at 42⬚C) of the RNA blots were performed with a buffer consisting of 5 ⫻ SSPE (0.75 mol/L NaCl, 50 mmol/L NaH2PO4, 5 mmol/L EDTA, pH 7.40), 5 ⫻ Denhardt’s solution [0.1% Ficoll 400, 0.1% polyvinylpyrrolidone, and 0.1% bovine serum (albumin fraction V)], 0.1% SDS, 100 ␮g/mL denatured salmon sperm DNA, and 50% deionized formamide [8, 16, 17]. A full-length cDNA probe of NaPi-2 [18] was labeled by random hexamer priming (Pharmacia) using [8-32P] dCTP (Dupont-NEN). Blots

Sorribas et al: Gentamicin and NaPi-2

1027

Fig. 2. Effect of gentamicin on NaPi-2 protein abundance, as determined by Western blotting in (A) SC-BBM, (B) SC homogenate, (C ) SC other, non-BBMs, and (D) JMC-BBM. Gentamicin caused a progressive decrease in SC-BBM NaPi-2 protein abundance, no change in SC-homogenate (that is, total) NaPi-2 protein abundance, and a reciprocal increase in SC other membrane (that is, non-BBM) NaPi-2 protein abundance. In contrast, gentamicin had no effect on JMC-BBM NaPi-2 protein abundance. The results are presented as mean ⫾ SEM. N ⫽ 6 in each group.

were washed twice for 15 minutes in 0.1 ⫻ SSPE with 0.1% SDS at 37⬚C and twice for 15 minutes each time in 0.1 ⫻ SSPE with 0.1% SDS at 50⬚C. Hybridization signals were quantitated by a phosphor imager analyzing system (BioRad, Richmond, CA, USA) and normalized for loading using full-length rat glyceraldehyde-3-phosphate dehydrogenase (GAPDH) [19] and 18S rRNA oligonucleotide probes synthesized as described [20]. Immunofluorescence microscopy Rats were anesthetized with thiopental (Penthotal, 100 mg/kg body weight) and perfused retrograde at a pressure of 1.38 hp through the abdominal aorta. The fixative consisted of 3% paraformaldehyde and 0.05% picric acid in a 6:4 mixture of cacodylate buffer (pH 7.4, adjusted to 300 mosmol with sucrose) and 10% hydroxyethyl starch. After five minutes of fixation, the rats were perfused for an additional five minutes with the cacodylate buffer [21]. Coronal slices of fixed kidneys were snap frozen in liquid propane cooled by liquid nitrogen. Sections 3 ␮m

Fig. 3. Effect of gentamicin on NaPi-2 mRNA abundance, as determined by Northern blotting, in SC. Gentamicin caused no changes in NAPi-2 mRNA abundance. The results are presented as mean ⫾ SEM. N ⫽ 6 in each group.

1028

Sorribas et al: Gentamicin and NaPi-2

Fig. 4. Effect of gentamicin on proximal tubular expression of NaPi-2 (fluorescein, green), as determined by immunofluorescence microscopy. Gentamicin caused a progressive decrease in proximal tubular apical brush-border membrane NaPi-2 protein abundance, while causing a progressive increase in intracellular NaPi-2 accumulation. The results are representative of four different rats studied in each group.

thick were cut at ⫺22⬚C in the cryomicrotome, mounted on chromalum/gelatin-coated glass slides, thawed, and stored in cold phosphate-buffered saline (PBS) buffer until use. For NaPi-2 immunofluorescence staining, sections were preincubated for five minutes at room temperature with 3% milk powder in PBS containing 0.05% Triton X-100. They were then covered overnight at 4⬚C with the

NaPi-2 antibody [22] diluted 1:500 in the preincubation solution. The sections were rinsed three times with PBS prior to incubation for one hour at 4⬚C with the secondary antibody, swine anti-rabbit IgG conjugated to FITC (Dakopatts, Glostrup, Denmark). To perform NaPi-2/early endosome (EEA-1) and NaPi-2/lysosome (lgp-120) double immunofluorescence staining, the sections were covered overnight with the

Sorribas et al: Gentamicin and NaPi-2

1029

Fig. 5. Effect of gentamicin on proximal tubular expression of (a) NaPi-2 and (b) EEA-1, an early endosome marker protein, as determined by immunofluorescence microscopy. Following treatment with gentamicin, apical membrane NaPi-2 expression is decreased, while there is increased colocalization with EEA-1.

antibody against NaPi-2 protein at 1:500 dilution, and either a mouse monoclonal antibody against early endosomal antigen (EEA-1, Transduction Labs, Lexington, KY, USA) at 1:500 dilution [23], or a mouse anti-rat antibody against the lysosomal membrane glycoprotein lgp-120 at 1:200 dilution [24], in PBS/milk powder. After rinsing with PBS, the sections were mounted using Dako-Glycergel娃 (Dakopatts) plus 2.5% 1,4-diazabicyclo-[2.2.2] octane (DABCO; Sigma, St. Louis, MO, USA) as a fading retardant. They were then imaged with a laser scanning microscope (Zeiss LSM 410, Jana, Germany) by confocal fluorescence imaging. Statistical analysis Data are expressed as mean ⫾ SE. A one-way analysis of variance with Student-Newman-Keul’s multiple range

tests was used to compare results between control and gentamicin-treated rats. Significance was accepted at P ⬍ 0.05. RESULTS Gentamicin causes a progressive decrease in superficial cortical brush-border membrane Na/Pi cotransport activity Compared with vehicle-treated control rats, gentamicin treatment for one, two, and three days caused a significant and progressive decrease in Na/Pi cotransport activity in BBM vesicles isolated from the superficial cortex (SC-BBM; Fig. 1A). In contrast, as we have previously shown [5], gentamicin treatment had no effect on

1030

Sorribas et al: Gentamicin and NaPi-2

Fig. 6. Effect of gentamicin on proximal tubular expression of (a) NaPi-2 and (b) lgp-120, a lysosomal marker protein, as determined by immunofluorescence microscopy. Following treatment with gentamicin, apical membrane NaPi-2 expression is decreased, while there is no change in the expression of lgp-120.

SC-BBM Na/glucose or Na/proline cotransport activities (results not shown). The differences in BBM Na/Pi cotransport activity were not due to differences in the purity of the BBM isolated from control versus gentamicintreated rats, as the enrichment for the BBM enzyme markers maltase and ␥-glutamyl transferase were identical in the two groups (Table 1). Gentamicin causes a progressive decrease in superficial cortical brush-border membrane NaPi-2 protein abundance The effect of gentamicin on SC-BBM Na/Pi cotransport activity was paralleled by a similar effect of gentamicin to cause a progressive decrease in superficial cortical

BBM NaPi-2 protein abundance (Fig. 2A). Interestingly, gentamicin had no significant effect on superficial cortical homogenate (total) NaPi-2 protein abundance (Fig. 2B), while gentamicin caused a significant and progressive increase in superficial cortical non-BBM NaPi-2 protein abundance (Fig. 2C). This suggests that gentamicin causes translocation of NaPi-2 protein from the apical BBM to inside the cell and contained in an intracellular compartment in which they are not degraded, as there was no change in total NaPi-2 protein abundance, even after three days of treatment with gentamicin. In contrast to the effect on SC-BBM NaPi-2 protein abundance, gentamicin had no effect on ␤-actin protein abundance (results not shown).

Sorribas et al: Gentamicin and NaPi-2

1031

Fig. 7. (A) Effect of gentamicin on proximal tubular expression of NaPi-2 (fluorescein, green) and EEA-1 (rhodamine, red), an early endosome marker protein, as determined by immunofluorescence microscopy. Following treatment with gentamicin, apical membrane NaPi-2 expression is decreased, while there is increased co-localization with EEA-1 (yellow color), suggesting accumulation of NaPi-2 in an early endosomal compartment. (B) Effect of gentamicin on proximal tubular expression of NaPi-1 (fluorescein, green) and lgp-120 (rhodamine, red), a lysosomal marker protein, as determined by immunofluorescence microscopy. Following treatment with gentamicin, apical membrane NaPi-2 is decreased, while there is increased co-localization with lgh-120 (yellow color), suggesting accumulation of NaPi-2 in lysosomes.

Gentamicin does not cause a decrease in superficial cortical NaPi-2 mRNA abundance In spite of the effect of gentamicin to cause a progressive decrease in superficial cortical BBM Na/Pi cotransport activity and BBM NaPi-2 protein abundance, gentamicin had no effect on superficial cortical NaPi-2 mRNA abundance (Fig. 3). This suggests that the effect of gentamicin to inhibit Na/Pi cotransport activity is mediated by post-transcriptional mechanisms. Gentamicin causes enhanced internalization of Na/Pi cotransport protein to endosomal and lysosomal compartments Similar to the Western blot results (Fig. 2A), immunofluorescence microscopy also revealed that gentamicin

caused a progressive decrease in NaPi-2 specific immunostaining in apical luminal membranes of the proximal convoluted tubule (Fig. 4). The decrease in luminal membrane immunostaining was associated with a progressive increase in intracellular staining (Fig. 4). Additional immunofluorescence studies were performed to determine the intracellular compartment in which the NaPi-2 immunostaining was localized, including the early endosomal compartment (Fig. 5) and lysosomes (Fig. 6). The co-localization studies using immunofluorescence microscopy revealed that gentamicin initially caused translocation of NaPi-2 protein from the apical BBM to the early endosomal compartment, as evidenced by significant co-localization (yellow; Fig. 7A) of NaPi-2 immunostaining (green) with early endosomal

1032

Sorribas et al: Gentamicin and NaPi-2

in JMC-BBM PI content (Fig. 10A). The increase in PI content would result in further enhancement of gentamicin binding to SC-BBM and result in increased gentamicin accumulation in the superficial cortex (Fig. 8). In agreement with our earlier results [5], gentamicin also caused a significant decrease in SC-BBM sphingomyelin, an increase in phosphatidylcholine, and a decrease in sphingomyelin to phosphatidylcholine mole ratio (Fig. 10B). In contrast, gentamicin had no significant effects in sphingomyelin and phosphatidylcholine content in JMC-BBM (Fig. 10B). DISCUSSION

Fig. 8. Effect of gentamicin treatment on gentamicin levels in SC (䊐) and JMC (䊏) homogenates, as determined by radioimmunoassay. Treatment with gentamicin causes significantly higher gentamicin levels in SC homogenate compared with JMC homogenate. The results are expressed as mean ⫾ SEM. N ⫽ 6 in each group.

antigen EEA-1 immunostaining (red). There was also evidence of progressively strong co-localization (yellow; Fig. 7B) of NaPi-2 immunostaining (green) with lysosomal immunostaining (red). Gentamicin has no effect on juxtamedullary cortical brush-border membrane Na/Pi cotransport activity In contrast to the inhibitory effect of gentamicin on SC-BBM Na/Pi cotransport activity, gentamicin was without an inhibitory effect on JMC-BBM Na/Pi cotransport activity (Fig. 1B). In agreement with the lack of inhibitory effect of gentamicin on JMC-BBM Na/Pi cotransport activity, gentamicin had no effect on JMCBBM NaPi-2 protein abundance (Fig. 2D). To determine the potential mechanisms for the differential effect of gentamicin on SC and JMC-BBM Na/Pi transport activity, we measured gentamicin levels in SC and JMC homogenates. Gentamicin levels were much higher in SC homogenate compared with JMC homogenate (Fig. 8). BBM gentamicin-binding studies indicated that this was associated with increased binding of gentamicin to SC-BBM compared with JMC-BBM (Fig. 9). Finally, since gentamicin binding to BBM is enhanced by negatively charged phospholipids, including phosphatidylinositol (PI), we measured PI content in SC- and JMC-BBM. We found that PI levels are markedly higher in SC-BBM (Fig. 10). In addition, following treatment with gentamicin, there is a further significant increase in SC-BBM PI content, while there is no significant alteration

We have previously shown that treatment of rats with gentamicin causes an early decrease in proximal tubular apical BBM Na/Pi cotransport activity, prior to a reduction in GFR [5]. In the present study, we have determined that the decrease in BBM Na/Pi cotransport activity is mediated by a decrease in BBM NaPi-2 protein expression. One interesting and intriguing finding of our study is that the decrease in BBM NaPi-2 protein occurs independent of a decrease in NaPi-2 mRNA. In previous studies, we have shown that acute regulation of NaPi-2 by alterations in dietary phosphate [8, 21], parathyroid hormone [25, 26], and metabolic acidosis [22] can indeed occur prior to changes in NaPi-2 mRNA. However, usually after six hours of the experimental perturbation, there is a parallel change in NaPi-2 mRNA as well. Interestingly, in this study, even after three days of gentamicin administration, there is no change in NaPi-2 mRNA levels. This suggests that gentamicin modulates NaPi-2 protein expression by post-transcriptional mechanisms that are independent of changes in mRNA. Another interesting and intriguing finding of our study is that although gentamicin causes a progressive decrease in the expression of apical membrane NaPi-2 protein, the NaPi-2 protein accumulates in an intracellular compartment and is not degraded. In fact, even after three days of treatment with gentamicin, the level of total kidney (that is, cortical homogenate) NaPi-2 protein is not altered. Therefore, following gentamicin administration NaPi-2 is endocytosed and accumulates in an intracellular compartment, most likely the lysosomes. Following treatment with a high Pi diet [8, 21] and/or parathyroid hormone (PTH) [25, 26], NaPi-2 is also endocytosed from the apical BBM and is eventually accumulated in the lysosomes where it is degraded. In the case of gentamicin, however, gentamicin is rapidly targeted to the lysosomes [14, 27–31] and is an inhibitor of lysosomal enzymes [32–38], and therefore, the NaPi-2 protein is accumulated in the lysosomes without degradation. A similar picture is seen when rats are first pretreated with leupeptin, an inhibitor of lysosomal enzymes, and then the rats are

Sorribas et al: Gentamicin and NaPi-2

1033

Fig. 9. Gentamicin binding to SC-BBM and JMC-BBM in BBM isolated from control, that is, nongentamicin treated rats, as determined by Scatchard plot. Symbols are: (䊐) control; (䊏) gentamicin. Bmax (binding capacity) is significantly higher in SC-BBM compared with JMC-BBM. The results are expressed as mean ⫾ SEM. N ⫽ 6 in each group.

fed a high Pi diet and/or infused with PTH, in which the internalized NaPi-2 protein is localized in the lysosomes [21, 26, 39]. Gentamicin interestingly has differential effects on Na/Pi cotransport activity and NaPi-2 protein abundance in SC-BBM versus JMC-BBM. Gentamicin causes a progressive decrease in SC-BBM Na/Pi cotransport activity and NaPi-2 protein abundance, whereas it has no effect on JMC-BBM Na/Pi cotransport activity or NaPi-2 protein abundance. The differential effects of gentamicin are associated with differential accumulation of gentamicin in SC versus JMC homogenate. We propose that differential gentamicin accumulation in SC versus JMC is most likely mediated by an increased capacity of SC-BBM to bind gentamicin, which is mediated by increased PI levels in SC-BBM, which is further increased after treatment with gentamicin. Previous studies have indeed shown that gentamicin preferentially binds to anionic phospholipids, including PI [40–42], and the higher PI levels in SC-BBM are compatible with increased gentamicin binding to this segment of the proximal tubule. In addition to Na/Pi cotransport activity, differential effects of gentamicin on SC versus JMC are also demonstrated by the fact that gentamicin causes a decrease in sphingomyelin and an increase in phosphatidylcholine in SC-BBM but not in JMC-BBM. The effects of gentamicin on SC-BBM lipid composition are in agreement with our earlier study [5]. Interestingly, the gentamicin-induced decrease in sphingomyelin to phosphatidylcholine mole ratio is associated with an increase in SC-BBM lipid fluidity [5]. Since in-

creases in membrane fluidity would enhance the activity of the Na/Pi cotransporter proteins [16, 43, 44], the increase in SC-BBM lipid fluidity may in part explain the leveling of the decrease in SC-BBM Na/Pi cotransport activity in spite of a further progressive decrease in SCBBM NaPi-2 protein abundance. Megalin or glycoprotein 330 [45] plays an important role in binding and internalizing gentamicin in the renal proximal tubule [46, 47]. In addition, megalin has also been shown to play an important role in the gentamicininduced inhibition of rat renal homotypic endosomal fusion [48]. Megalin has also been shown to mediate the endocytosis of albumin in the proximal tubule [49] and to be associated with the Na/H exchanger isoform NHE3 in the proximal tubule [50]. Therefore, megalin may play a role in gentamicin-induced internalization of NaPi-2. However, the studies by Biemesderfer et al [50] and preliminary studies in our laboratory have not shown an association between megalin and NaPi-2 in the proximal tubule. Thus, a role for megalin in mediating the effects of gentamicin on NaPi-2 trafficking remains unknown. In summary, our study indicates that gentamicin causes a progressive decrease in SC-BBM Na/Pi cotransport activity and NaPi-2 protein abundance by post-translational mechanisms. The effect of gentamicin to preferentially inhibit SC-BBM rather than JMC-BBM is associated with increased gentamicin binding to SC-BBM and accumulation in SC-homogenate, due to increased levels of the negatively charged phosphatidylinositol in SC-BBM.

1034

Sorribas et al: Gentamicin and NaPi-2

Fig. 10. Lipid composition of SC-BBM and JMC-BBM in vehicle (䊐) or gentamicin treated (䊏) rats. (A) Phosphatidylinositol (PI) levels are higher in SC-BBM compared with JMC-BBM. Furthermore, treatment with gentamicin causes a further increase in SC-BBM PI content. (B) Gentamicin causes a decrease in sphingomyelin (SM) and an increase in phosphatidylcholine (PC) levels in SC-BBM but not in JMC-BBM. The results are expressed as mean ⫾ SEM. N ⫽ 6 in each group.

Sorribas et al: Gentamicin and NaPi-2

ACKNOWLEDGMENTS This work was supported by grants from the Department of Veterans Affairs Merit Review and National Kidney Foundation (Moshe Levi), Diputacio´n General De Arago´n, Spain (Victor Sorribas PO78/ 99), an NRSA fellowship grant from National Institutes of Health (Hubert Zajicek, 1 F32 DK09689-01), and Department of Veterans Affairs Minority Initiative Grant (Krishna Puttaparthi). The authors thank Drs. Heini Murer and Jurg Biber (Zurich) for providing the NaPi-2 c-DNA probe, Teresa Autrey for secretarial assistance, and the Medical Media Department at DVAMC for the illustrations. Reprint requests to Moshe Levi, M.D., The University of Texas Southwestern Medical Center at Dallas, Nephrology Section, VA North Texas Health Care System, 4500 South Lancaster Road, MC 151, Dallas, Texas 75216, USA. E-mail: [email protected]

19.

20. 21.

22. 23.

REFERENCES 1. Al-Mahrouq HA, McAteer JA, Kempson SA: Cholera toxin action on rabbit renal brush border membranes inhibits phosphate transport. Biochim Biophys Acta 981:200–206, 1988 2. Ginsburg DS, Quintanilla AP, Levin M: Renal glycosuria due to gentamicin in rabbits. J Infect Dis 134:119–122, 1976 3. Mondorf AW, Breier J, Hendus J, et al: Effect of aminoglycosides on proximal tubular membranes of the human kidney. Eur J Clin Pharmacol 13:133–142, 1978 4. Parsons PP, Garland HO, Harpur ES, et al: Acute gentamicininduced hypercalciuria and hypermagnesiuria in the rat: Dose– response relationship and role of renal tubular injury. Br J Pharmacol 122:570–576, 1997 5. Levi M, Cronin RE: Early selective effects of gentamicin on renal brush-border membrane Na-Pi cotransport and Na-H exchange. Am J Physiol 258:F1379–F1387, 1990 6. Levi M: Heterogeneity of phosphate transport by brush border membranes from superficial and juxtamedullary cortex of the rat. Am J Physiol 258(6 Pt 2):F1616–F1624, 1990 7. Lowry DH, Rosenbrough NJ, Farr AL, et al: Protein measurements with the folin phenol reagent. J Biol Chem 193:265–275, 1951 8. Levi M, Lo¨tscher M, Sorribas V, et al: Cellular mechanisms of acute and chronic adaptation of rat renal Pi transport to alterations in dietary Pi. Am J Physiol 267:F900–F908, 1994 9. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685, 1970 10. Towbin HT, Staehelin T, Gordon J: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci USA 76:4350–4354, 1979 11. Bligh EG, Dyer WJ: A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37:911–917, 1959 12. Esko JD, Raetz CRH: Mutants of Chinese hamster ovary cells with altered membrane phospholipid composition. J Biol Chem 255:4474–4480, 1980 13. Ames BN, Dubin DT: The role of polyamines in the neutralization of bacteriophage of deoxyribonucleic acid. J Biol Chem 235:769– 775, 1960 14. Cronin R, Inman L, Eche T, et al: Effect of thyroid hormone on gentamicin accumulation in rat proximal tubule lysosomes. Am J Physiol 257:F86–F91, 1989 15. Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156–159, 1987 16. Levi M, Shayman J, Abe A, et al: Dexamethasone modulates rat renal brush border membrane phosphate transporter mRNA and protein abundance and glycosphingolipid composition. J Clin Invest 96:207–216, 1995 17. Farrell E Jr: Methodologies for RNA characterization. II. Quantitation by northern blot analysis and S1 nuclease assay. Clin Biotech 2:107–119, 1990 18. Magagnin S, Werner A, Markovich D, et al: Expression cloning

24. 25. 26.

27. 28. 29. 30. 31. 32.

33. 34.

35. 36.

37. 38.

39.

1035

of human and rat renal cortex Na/Pi cotransport. Proc Natl Acad Sci USA 90:5979–5983, 1993 Fort PL, Marty L, Piechaezyk M, et al: Various rat adult tissues express only one major RNA species from the glyceraldehyde-2phosphate-dehydrogenase multigenic family. Nucleic Acids Res 13: 1431–1442, 1985 Moe OW, Miller RT, Horie S, et al: Differential regulation of Na/H antiporter by acid in renal epithelial cells and fibroblasts. J Clin Invest 88:1703–1708, 1991 Lo¨tscher M, Kaissling B, Biber J, et al: Role of microtubules in the rapid regulation of renal phosphate transport in response to acute alterations in dietary phosphate content. J Clin Invest 99:1302–1312, 1997 Ambu¨hl PM, Zajicek HK, Huamin W, et al: Regulation of renal phosphate transport by acute and chronic metabolic acidosis in the rat. Kidney Int 53:1288–1298, 1998 Gruenberg J, Griffiths G, Howell KE: Characterization of the early endosome and putative endocytic carrier vesicles in vivo and with an assay of vesicle fusion in vitro. J Cell Biol 108:1301–1316, 1989 Marsh M, Griffiths G, Dean GE, et al: Three-dimensional structure of endosomes in BHK-21 cells. Proc Natl Acad Sci USA 83: 2899–2903, 1986 Kempson SA, Lo¨tscher M, Kaissling B, et al: Parathyroid hormone action on phosphate transporter mRNA and protein in rat renal proximal tubules. Am J Physiol 268:F784–F791, 1995 Lotscher M, Scarpetta Y, Levi M, et al: Rapid downregulation of rat renal Na/Pi cotransporter in response to parathyroid hormone involves microtubule rearrangement. J Clin Invest 104:483–494, 1999 Olier B, Morin JP, Grancher G, et al: Regenerating tubular cell process in rat kidney: Influence of gentamicin treatment. Int J Tissue React 9:241–250, 1987 Beauchamp D, Gourde P, Bergeron MG: Subcellular distribution of gentamicin in proximal tubular cells, determined by immunogold labeling. Antimicrob Agents Chemother 35:2173–2179, 1991 Ford DM, Dahl RH, Lamp CA, et al: Apically and basolaterally internalized aminoglycosides colocalize in LLC-PK1 lysosomes and alter cell function. Am J Physiol 266:C52–C57, 1994 Hashino E, Shero M, Salvi RJ: Lysosomal targeting and accumulation of aminoglycoside antibiotics in sensory hair cells. Brain Res 777:75–85, 1997 Sandoval R, Leiser J, Molitoris BA: Aminoglycoside antibiotics traffic to the Golgi complex in LLC-PK1 cells. J Am Soc Nephrol 9:167–174, 1998 Hostetler KY, Hall LB: Inhibition of kidney lysosomal phospholipases A and C by aminoglycoside antibiotics: Possible mechanism of aminoglycoside toxicity. Proc Natl Acad Sci USA 79:1663–1667, 1982 Carlier MB, Laurent G, Claes PJ, et al: Inhibition of lysosomal phospholipases by aminoglycoside antibiotics: In vitro comparative studies. Antimicrob Agents Chemother 23:440–449, 1983 Giurgea-Marion L, Toubeau G, Laurent G, et al: Impairment of lysosome-pinocytic vesicle fusion in rat kidney proximal tubules after treatment with gentamicin at low doses. Toxicol Appl Pharmacol 86:271–285, 1986 Regec AL, Trump BF, Trifillis AL: Effect of gentamicin on the lysosomal system of cultured human proximal tubular cells. Biochem Pharmacol 38:2527–2534, 1989 Mingeot-Leclero M-P, Piret J, Brasseur R, et al: Effect of acidic phospholipids on the activity of lysosomal phospholipases and on their inhibition by aminoglycoside antibiotics. I. Biochemical analysis. Biochem Pharmacol 40:489–497, 1990 Olbricht CJ, Fink M, Gutjahr E: Alterations in lysosomal enzymes of the proximal tubule in gentamicin nephrotoxicity. Kidney Int 39:639–646, 1991 Montenez JP, Kishore BK, Maldague P, et al: Leupeptin and E-64, inhibitors of cysteine proteinases, prevent gentamicininduced lysosomal phospholipidosis in cultured rat fibroblasts. Toxicol Lett 73:201–208, 1994 Keusch I, Traebert M, Lotscher M, et al: Parathyroid hormone and dietary phosphate provoke a lysosomal routing of the proximal tubular Na/Pi-cotransporter type II. Kidney Int 54:1224–1232, 1998

1036

Sorribas et al: Gentamicin and NaPi-2

40. Josepovitz C, Farruggella T, Levine R, et al: Effect of netilmicin on the phospholipid composition of subcellular fractions of rat renal cortex. J Pharmacol Exp Ther 235:810–819, 1985 41. Molitoris BA, Meyer C, Dahl R, et al: Mechanism of ischemiaenhanced aminoglycoside binding and uptake by proximal tubule cells. Am J Physiol 264:F907–F916, 1993 42. Sundin DP, Meyer C, Dahl R, et al: Cellular mechanism of aminoglycoside tolerance in long-term gentamicin treatment. Am J Physiol 272:C1309–C1318, 1997 43. Levi M, Jameson DM, van der Meer BW: Role of brush border membrane lipid composition and fluidity in impaired renal phosphate transport in the aged rat. Am J Physiol 256(1 Pt 2):F85–F94, 1989 44. Levi M, Baird B, Wilson P: Cholesterol modulates rat renal brush border membrane phosphate transport. J Clin Invest 85:231–237, 1990 45. Saito A, Pietromonaco S, Loo AKC, et al: Complete cloning and

46. 47. 48. 49. 50.

sequencing of rat “gp330/megalin,” a distinctive member of the low density lipoprotein receptor gene family. Proc Natl Acad Sci USA 91:9725–9729, 1994 Farquhar MG: The unfolding story of megalin (gp330): Now recognized as a drug receptor. J Clin Invest 96:1184, 1995 Moestrup SK, Cui S, Vorum H, et al: Evidence that epithelial glycoprotein 330/megalin mediates uptake of polybasic drugs. J Clin Invest 96:1404–1413, 1995 Hammond TG, Majewski RR, Kaysen JH, et al: Gentamicin inhibits rat renal cortical homotypic endosomal fusion: Role of megalin. Am J Physiol 272:F117–F123, 1997 Cui S, Verroust PJ, Moestrup SK, et al: Megalin/gp330 mediates uptake of albumin in renal proximal tubule. Am J Physiol 271: F900–F907, 1996 Biemesderfer D, Nagy T, Degray B, et al: Specific association of megalin and Na/H exchanger isoform NHE3 in the proximal tubule. J Biol Chem 274:17518–17524, 1999