Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
TOXICOLOGICAL SCIENCES 101(2), 254–262 (2008) doi:10.1093/toxsci/kfm266 Advance Access publication November 12, 2007 Role of the Sodium-Dependent Phosphate Cotransporters and Absorptive Endocytosis in the Uptake of Low Concentrations of Uranium and Its Toxicity at Higher Concentrations in LLC-PK1 Cells Dany S. Muller,*,1 Pascale Houpert,* Jean Cambar,† and Marie-Hélène Hengé-Napoli‡ *Institut de Radioprotection et de Sûreté Nucléaire, Laboratoire de Radiotoxicologie Expérimentale, BP-166, 26702 Pierrelatte Cedex, France; †GEPPR, LSTE-EA 3672, 146 rue Léo Saignat, 33076 Bordeaux Cedex, France; and ‡CEA/DEN/Valrhô-Dir-CVAR, BP 17171, 30207 Bagnols sur Cèze Cedex, France Received July 31, 2007; accepted October 7, 2007 It has been suggested that uranium uptake and toxicity could be mediated by endocytosis and/or the type IIa sodium-dependent phosphate cotransporter (NaPi-IIa). The aim of this study was therefore to characterize in vitro the role of these two cellular mechanisms in the uptake and toxicity of low (200–3200nM) and high (0.5 and 0.8mM) concentrations of uranium, respectively. At low concentrations, uranium uptake in LLC-PK1 cells was saturable (Vmax 5 3.09 ± 0.22 ng/mg protein) and characterized by a K0.5 of 1022 ± 63nM and a Hill coefficient of 3.0 ± 0.4. The potential involvement of endocytosis and NaPi-IIa in the uptake of uranium was assessed by the use of various drugs and culture conditions known to alter their relative activity, and 233uranium uptake was monitored. Interestingly, the inhibitory effect of colchicine, cytochalasin D, phorbol 12-myristate 13-acetate, and chlorpromazine on endocytosis was highly correlated with their effect on uranium uptake, a relationship that was not true when the NaPi-IIa transport system was studied. Whereas the competitive inhibition of the NaPi-IIa by phosphonoformic acid (PFA) significantly decreased uranium uptake, this effect was not reproduced when NaPi-IIa inhibition was mediated by the replacement of extracellular Na1 with N-methyl-D-glucamine. Uranium uptake was also not significantly altered when NaPi-IIa expression was stimulated in MDCK cells. More surprisingly, we observed by transmission electron microscopy that uranium cytotoxicity was dependent upon the extent of its intracellular precipitation, but not on its intracellular content, and was suppressed by PFA. In conclusion, our results suggest that lowdose uranium uptake is mainly mediated by absorptive endocytosis, and we propose PFA as a potential uranium chelator. Key Words: uranium uptake; endocytosis; phosphonoformic acid; LLC-PK1; MDCK; uranium cytotoxicity. The increasing use of uranium worldwide and the chronic exposure of some populations to uranium in drinking water that can reach several hundred to thousands of microgram per liter 1 To whom correspondence should be addressed at the b-Cell Development and Function Group, School of Health and Biomedical Sciences, 2nd Floor, Hodgkin Building, Guy’s Campus, King’s College London, London SE1 1UL, UK. Fax: þ44 (0) 20-7848-6280. E-mail: [email protected]. (Kurttio et al., 2002; Moss et al., 1983) have led to questions about its potential health effects. Naturally occurring and depleted uranium both present relatively low specific activities and are therefore chemotoxicants rather than radiotoxicants, although the potential exists for a radiological toxicity. To date, the acute toxicity of uranium has been extensively studied in in vivo and in vitro experiments and is characterized by the development of acute renal failure (ARF) due to tubular (Bulger, 1986; Leggett, 1989; Lopez et al., 2000; McDonald-Taylor et al., 1997; Sano et al., 1998; Tyrakowski, 1979) followed by glomerular (Goel et al., 1980; Kobayashi et al., 1984; L’Azou et al., 2002) alterations. Uranium-induced tubular alterations are mainly observed at the level of the second and third segment of the proximal tubule and can develop with a uranium concentration as low as 0.01 mg/kg of kidney. These alterations increase in severity as either time or uranium concentration is increased (Lopez et al., 2000) and are suggested to result in part from transepithelial transport permeability defects (Goldman et al., 2006; Leggett, 1989; Tyrakowski, 1979), inhibition of the Naþ/Kþ ATPase activity, and mitochondrial injury (Brady et al., 1989). At high concentrations, these alterations are followed by necrosis (Bulger, 1986; McDonald-Taylor et al., 1997) and/or apoptosis (Sano et al., 1998). Although much less is known about the effects of long-term exposure to low concentrations of uranium, its chronic ingestion in drinking water has also been reported to affect the renal function of the proximal tubule in humans (Mao et al., 1995; Zamora et al., 1998), but these observations remain controversial (Kurttio et al., 2006). In rodents, chronic exposure to low concentrations of uranium in drinking water revealed a dramatic alteration in expression of more than 200 genes expressed in the proximal tubule (Taulan et al., 2004). Among these genes, those encoding for oxidative stress–related proteins confirm previous observations showing that in vitro DNA exposure to low concentrations of depleted uranium can generate oxidative DNA damage and induce the formation of hydroxyl radicals (Miller et al., 2002). Interestingly, the The Author 2007. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For Permissions, please email: [email protected] URANIUM UPTAKE AND TOXICITY development of ARF in mice induced by 5 mg/kg uranyl nitrate has also been associated with a dramatic alteration in expression of the genes encoding the same oxidative stress–related proteins (Taulan et al., 2006), which suggests that the chemical toxicity of both low and high uranium concentrations might share, at least in part, similar cellular and molecular mechanisms. In accordance with this hypothesis, we recently demonstrated by computer-assisted simulation that uranium speciation at low concentration (1lM) is comparable with its speciation at high concentration (1mM; Muller et al., 2006). For example, similarly to what can be theoretically observed at low uranium concentration (1lM), increasing the extracellular phosphate concentration at high uranium concentration (1mM) is only associated with an increase in the levels of the uranyl-phosphate complexes. Taken all these observations into account, it is therefore evident that identifying the cellular and molecular mechanisms which are involved in the uptake of uranium at the level of the proximal tubule epithelium has the potential for reducing its accumulation and protecting from its toxic effects in the kidney. To date, there are several lines of evidence suggesting that the proximal tubule uptake of uranium occurs by endocytosis and/or by the type IIa sodium-dependent phosphate cotransporter (NaPi-IIa) system. The endocytosis hypothesis comes from previous studies using electron microscopy of proximal tubule cells from rats intoxicated with uranyl nitrate (Galle, 1974), as well as in vitro contaminated cell lines (Carrière et al., 2006; Mirto et al., 1999a,b) and cultured kidney cells (Ghadially et al., 1982), where the occurrence of urchin-like electron-dense structures in lysosome-like bodies has been reported. The hypothesis that uranium uptake is mediated by the NaPi-IIa is based on the recent findings showing that the toxicity of uranium is phosphate dependent, correlated to the level of the uranyl-phosphate species UO2PO 4 and/or UO2PO4(aq) present in the culture medium, and entirely dependent upon the expression and activity of the NaPi-IIa (Muller et al., 2006). However, these lines of evidence are circumstantial, and the exact mechanisms by which uranium is taken up by proximal tubule cells are still not clear and remain to be characterized. The aim of the current study was therefore to determine in vitro the role of endocytosis and of the NaPi-IIa in the uptake and toxicity of uranium using the LLC-PK1 and NaPi-IIa– expressing MDCK cell lines. The role of absorptive endocytosis was assessed indirectly using fluorescein isothiocyanate (FITC)–labeled albumin as a molecular marker. Uranium uptake was performed using the high–specific activity isotope 233 uranium. MATERIALS AND METHODS Chemicals. Minimal essential medium (MEM), fetal bovine serum (FBS), streptomycin, penicillin, phorbol 12-myristate 13-acetate (PMA), L-glutamine, 255 cytochalasin D, colchicine, concanavalin A, bisindolylmaleimide (Bim), phalloidin, chlorpromazine, sucrose, FITC-albumin, 3-(N-morpholino) propanesulfonic acid (MOPS), phosphonoformic acid (PFA), N-methyl-Dglucamine (MGA), and N-[2-hydroxyethyl] piperazine-N#-[2-ethanesulfonic acid] (HEPES) were obtained from Sigma (L’Isle d’Abeau Chesnes, France). The 233 isotope of uranium was obtained from Cerca (Pierrelatte, France; specific activity 3.57 3 108 Bq/g). Uranyl nitrate (depleted uranium 1.6 3 104 Bq/g) was obtained from Merck Eurolab (Lyon, France). Ringer solution was composed of (in mM) 122.5 NaCl, 5.4 KCl, 1.2 CaCl2, 0.8 MgCl2, 0.8 Na2HPO4, 0.2 NaH2PO4, 5.5 glucose, and 10 HEPES (titrated to pH 6.0 or 7.4). Cell culture. LLC-PK1 cells were obtained from the European Collection of Cell Cultures (Wiltshire, England), and NaPi-2–overexpressing MDCK cells (MDCK/NaPi-2 transfectants) were generously provided by Professor H. Murer and Dr J. Biber (Zurich, Switzerland). LLC-PK1 and MDCK cells were maintained in MEM supplemented with 10% FBS, 10mM HEPES, 2mM 4 L-glutamine, 50 lg/l streptomycin, and 10 IU/l of penicillin at 37C in the presence of 5% CO2. LLC-PK1 cells were used from the 1st to the 20th passage after the cells were purchased, and MDCK/NaPi-2 transfectants were used within 15 passages after their arrival. MDCK/NaPi-2 transfectants are MDCK cells that had been transfected with cDNA coding for the proximal tubule NaPi cotransporter NaPi-2 by the use of a dexamethasone-inducible vector (pLKneo; Quabius et al., 1996). Dexamethasone-induced NaPi-2 expression in MDCK/NaPi-2 transfectants was performed by incubation with 1lM dexamethasone for 16 h. Dexamethasone was added from a 10,000-fold stock solution made in ethanol, and the same amount of vehicle was added to control cells. Inhibitors, activators, and culture conditions. The role of endocytosis in the uptake of uranium by LLC-PK1 cells was assessed using culture conditions and drugs that are well known to alter absorptive endocytosis. Maintaining the cells in culture at 4C is known to be the most effective and noninvasive method of inhibiting various transport systems by modulating the membrane’s physical state (Mamdouh et al., 1996). The use of sucrose (1M) has been described to inhibit receptor-mediated endocytosis by blocking the formation of clathrin-coated pits (Heuser and Anderson, 1989). PMA (10nM) and chlorpromazine (50lM) have been shown to be two potent inhibitors of albumin uptake by the proximal tubule–derived OK cells (Gekle et al., 1997; Hryciw et al., 2005). Destabilizing the actin and microtubular networks with colchicine (10lM) and cytochalasin D (10lM) to depress endocytosis has been extensively used and described in the literature. Their effects were compared with the effect of phalloidin (25lM), a potent actin network–stabilizing agent. Finally, concanavalin A (0.4lM) was also used in our study because this compound was previously reported to stimulate albumin endocytosis in murine epidermal Langerhans cells (Becker et al., 1995). FITC-albumin uptake. Albumin uptake was measured using a previously published method (Gekle et al., 1997). LLC-PK1 cells were preincubated in serum-free MEM at 37C for 2 h prior to uptake studies. After three acidic washes (pH 6.0), the cells were further preincubated in Ringer solution (pH 7.4) for 30–60 min in the presence of vehicle, 10lM colchicine, 10lM cytochalasin D, 25lM phalloidin, 0.4lM concanavalin A, 50lM chlorpromazine, 10nM PMA, 1lM Bim, or 1mM sucrose. LLC-PK1 cells were exposed to 0–250 lg/ml FITC-albumin either at 37C or 4C for 0–120 min in the continued presence of the inhibitors or vehicle, then washed in Ringer solution (pH 6.0), and lysed in MOPS buffer (20mM MOPS, pH 7.4, with 0.1% Triton X-100) at 37C for 45 min. Cell-associated fluorescence was measured using a spectrofluorimeter (FLX 800; Bio-Tek, Saint-Quentin en Yvelines, France), adjusted for background, and standardized to total cellular protein, and the corresponding amount of albumin was calculated by comparison with a standard curve. The amount of internalized substrate was determined by subtracting the fraction of membrane-bound FITC-albumin (determined by the addition of 1000-fold excess of unlabeled albumin) from total cell–associated FITC-albumin. Uranium uptake. Uranium uptake by LLC-PK1 and MDCK cells was assessed using a solution of 233uranium. Briefly, the incubation medium was 256 MULLER ET AL. removed and cells were equilibrated in 3 ml serum-free MEM (pH 7.4) at 37C for 2 h prior to the uptake studies. The cells were then preincubated for 30–60 min in the presence or absence of the same inhibitors as described above, followed by a further incubation at 37C or 4C for 0–300 min in the presence of 0–3.2lM 233 uranium. Finally, the cells were washed six times with ice-cold Ringer solution (pH 7.4) containing 5 mg/ml ethane-1-hydroxy-1,1-biphosphonate, a powerful chelating agent (Henge-Napoli et al., 1999) used to remove membrane-bound uranium, and then lysed in MOPS buffer (20mM MOPS, pH 7.4, 0.1% Triton X-100) at 37C for 45 min. Cell-associated a emission was assayed by liquid scintillation counting, and uranium uptake was adjusted for background and standardized to total cellular protein. The amount of 233 uranium incorporated was calculated per microgram cellular protein. Phosphate uptake. The Naþ-dependent and -independent phosphate transports were determined in cells grown to confluency on plastic dishes (35 mm) as described previously (Muller et al., 2006). Briefly, the cells were washed twice with 2 ml Ringer solution and equilibrated in serum-free culture medium (pH 7.4) at 37C for 2 h. The uptake solution consisted of 137mM NaCl, 5.4mM KCl, 2.8mM CaCl2, 1.2mM MgSO4, 10mM HEPES-Tris (pH 7.4), and 0.1mM KH232PO4 (37 kBq/ml). For Naþ-independent uptake, NaCl was replaced with MGA. The uptake experiments were performed at room temperature for 6 min and then stopped by washing the cells six times with cold ‘‘stop’’ solution (137mM NaCl, 10mM Tris-HCl, pH 7.2). Finally, the cells were lysed and the cell-associated 32P activity was determined by liquid scintillation counting. Uranyl-NaHCO3 stock solution and media preparation. Uranyl-NaHCO3 stock solutions were prepared as previously described (Muller et al., 2006). Briefly, Uranyl-NaHCO3 stock solutions ([U] ¼ 10mM and [HCO 3 ] ¼ 100mM) were obtained by dissolving uranyl nitrate crystals (UO2(NO3)2.6H2O) in an aqueous NaHCO3 solution, and the pH was adjusted to 7.2. These stock solutions were diluted in serum-free MEM to obtain the desired final concentrations. In each case, the NaHCO3 concentration was adjusted to 3.34mM. To simplify the reading of the manuscript, the term uranium will be used throughout the text. Uranium-induced cytotoxicity: cell death determination. Uraniuminduced cell death was measured by quantifying the amount of lactate dehydrogenase (LDH) released from damaged cells into the culture medium. This was performed using an enzymatic cytotoxicity detection kit (Roche Applied Science, Meylan Cédex, France) in which the consumption of nicotine adenine dinucleotide in the presence of pyruvate was monitored by spectrophotometry at 490 nm (Legrand et al., 1992). Total cellular LDH content was determined following an incubation time of 30–60 min in MEM containing 1% Triton X-100, and the percentage of cell death (percent of cytotoxicity) was quantified as the ratio between the amount of LDH released in the medium by treated cells and the total cellular LDH content. Transmission electron microscopy. LLC-PK1 cells were cultured in 35-mm Petri dishes and pretreated for 30–60 min with 10mM PFA, 10lM cytochalasin D, 0.4lM concanavalin A, or vehicle (dimethyl sulfoxide). The cells were then cultured for an additional 16 h in the presence of 500lM uranium, washed three times in Ringer solution (pH 7.4), fixed in 1.5% glutaraldehyde, buffered with 0.1M cacodylate (pH 7.4) for 1 h, postfixed in 1% osmium buffered with 0.1M cacodylate (pH 7.4), and ethanol dehydrated. Each sample was then embedded in Epon, and 60-nm sections were prepared using the Reichert Ultracut S microtome. These sections were finally mounted on copper grids, stained with uranyl acetate, and analyzed under a transmission electron microscope (TEM) (Philips CM 120). Statistics. Data are presented as mean ± SEM, and n represents the number of experiments. Each uptake experiment was performed in triplicate or quadruple, and each cell viability test was performed in duplicate with eight observations. The significance of differences was tested by one-way analysis of variance followed by the Tukey honestly significant difference test. Differences were considered significant if p < 0.05. The correlation between 233uranium uptake and FITC-albumin uptake, shown in Figure 3C, was analyzed with a linear regression test. RESULTS FITC-Albumin Uptake in LLC-PK1 Cells As shown in Figure 1A, FITC-albumin uptake increased linearly in a time-dependent manner for the first 15–20 min, then slowed down, and reached an apparent equilibrium after 90 min of incubation with 50 lg/ml FITC-albumin. Consequently, we decided to use an incubation time of 15 min to further characterize FITC-albumin uptake in LLC-PK1 cells, when the time course of the transport is still in the linear phase. Figure 1B shows the concentration dependence of FITCalbumin uptake in this cell line. The uptake was saturable, and the determination of the Hill coefficient was consistent with the existence of a single mechanism of transport (Fig. 1C; Hill coefficient: 0.9 ± 0.2). Accordingly, the characterization of the Michaelis and Menten constant (Km) and of the maximal velocity (Vmax), which was performed using the Eadie-Hofstee plot (Fig. 1D), confirmed that albumin uptake in LLC-PK1 cells is functionally mediated by one transport system. The Km and Vmax values, averaged from three independent experiments, were 23 ± 7 lg/ml FITC-albumin and 0.79 ± 0.09 lg/mg protein, respectively. 233 Uranium Uptake in LLC-PK1 Cells Using a concentration of 5lM 233uranium, we observed that uranium uptake at 37C increased linearly in a time-dependent manner for the first 30–60 min and then slowed down but without reaching equilibrium even after 5 h (Fig. 2A). As we also observed that uranium uptake was significantly reduced when these experiments were performed at 4C, we can conclude that the mechanism through which uranium is taken up by LLC-PK1 cells is not a diffusion process. The concentration dependence of uranium uptake is shown in Figure 2B, where a sigmoidal curve can be observed. First, uranium uptake increased slowly for concentrations ranging from 0 to 600nM. Then, it increased in a more dramatic manner from 600nM to 2lM (uranium content at 600nM and 2lM: 0.41 ± 0.27 and 2.64 ± 0.26 ng/mg protein, respectively). An apparent saturation was obtained for 233uranium concentrations ranging from 2 to 3.2lM, and the determination of the Hill coefficient (3.0 ± 0.4; Fig. 2C) was consistent either with the presence of an allosteric mechanism or the existence of several systems of transport for uranium. Overall, the maximal velocity (Vmax) and apparent affinity (K0.5) were estimated as 3.09 ± 0.22 ng/mg protein and 1022 ± 63nM, respectively. The complexity of the process by which uranium is taken up by LLC-PK1 cells is further demonstrated by the Eadie-Hofstee representation (Fig. 2D) showing an atypical plot. The curvature observed for uptake values ranging from 0.4 to URANIUM UPTAKE AND TOXICITY 257 FIG. 1. Characterization of the kinetic parameters of albumin uptake in LLC-PK1 cells. (A) Time course of albumin uptake (50 lg/ml FITC-albumin). (B) Concentration dependence of albumin uptake measured after 15 min. (C) Hill plot representation obtained with the data from (B). The averaged Hill coefficient obtained from three separate studies was 0.9 ± 0.2. (D) Eadie-Hofstee plot showing the existence of a single system of transport for albumin. The averaged Km and Vmax values from three different experiments were 23 ± 7 lg/ml and 0.79 ± 0.09 lg/mg protein, respectively. Points shown in (A) and (B) are mean ± SEM, n ¼ 3. 2.8 ng/mg systems. 233 uranium is characteristic of cooperative transport Inhibition of Absorptive Endocytosis Reduces Uptake in LLC-PK1 Cells 233 Uranium To investigate the potential relationship that could exist between absorptive endocytosis and uranium uptake, we compared the effects of several drugs and conditions (known to interfere with absorptive endocytosis) on both FITC-albumin and 233uranium uptake in LLC-PK1 cells (Figs. 3A and 3B). On the one hand, carrying out the uptake studies at 4C resulted in a dramatic reduction of both FITC-albumin (79.2%) and 233uranium uptakes (87.1%). In a similar manner, when the cells were cultured in the presence of a hypertonic medium (1M sucrose), FITC-albumin and 233uranium uptakes were decreased by 76.5 and 88.4%, respectively. Destabilization of the microtubular and actin networks using 10lM colchicine or 10lM cytochalasin D also significantly reduced both transport systems, with FITC-albumin and 233uranium uptakes being, respectively, decreased by 49.8 and 53.3% (colchicine) and 61.6 and 79.7% (cytochalasin D). When LLC-PK1 cells were pretreated with 50lM chlorpromazine or a protein kinase C (PKC) activator (10nM PMA), both FITC-albumin and 233uranium uptakes were also reduced by 47.6 and 53.4% and 64.8 and 65.9%, respectively. The involvement of PKC in the regulation of uranium uptake by LLC-PK1 cells was further demonstrated by the results showing that 1lM Bim, a PKC inhibitor, completely suppressed the effect of 10nM PMA. On the other hand, neither stabilization of the actin cytoskeleton with 25lM phalloidin nor pretreatment of the cells with 0.4lM concanavalin A significantly altered the uptakes of FITC-albumin and 233 uranium. Interestingly, plotting the effects of the drugs and culture conditions on FITC-albumin endocytosis against their effects on 233uranium uptake showed the existence of a significant linear correlation (Fig. 3C, p < 0.01). 233 Uranium Uptake Is Not Mediated by the NaPi-IIa, but It Is Significantly Reduced by PFA We recently demonstrated that uranium-induced toxicity was not only phosphate dependent but also dependent upon the expression and activity of the NaPi-IIa (Muller et al., 2006). Accordingly, we proposed that uranium uptake might also be mediated, at least in part, by this system of transport. To test this hypothesis, we monitored the uptake of 3lM 233 uranium in the absence or presence of 0.1–10mM PFA, a specific NaPi inhibitor, or when the activity of the NaPi cotransporters was suppressed by replacing the extracellular sodium with MGA. Surprisingly, whereas uranium uptake was reduced in a concentration-dependent manner when LLC-PK1 cells were treated with PFA (233uranium uptake in ng/mg protein/h: 3.05 ± 0.24, 2.54 ± 0.55, 2.10 ± 0.45, and 258 MULLER ET AL. FIG. 2. Characterization of the kinetic parameters of uranium uptake in LLC-PK1 cells. (A) Time course of uranium uptake. The uptake of 5lM 233uranium was monitored either at 4C (open circle) or 37C (filled circle). (B) Concentration dependence of uranium uptake measured after 60 min. (C) Hill plot representation by extrapolating the data obtained in (B). The averaged Hill coefficient determined from three separate studies was 3.0 ± 0.4. (D) Eadie-Hofstee plot. The K0.5 and Vmax values that were averaged from three separate experiments were 1022 ± 63nM and 3.09 ± 0.22 ng/mg protein, respectively. Points shown in (A) and (B) are mean ± SEM, n ¼ 3. 0.75 ± 0.08 in the presence of 0, 0.1, 1, and 10mM PFA, respectively), the suppression of the NaPi-driving force by the use of MGA instead of sodium did not have any significant effect (Fig. 4A). Given this unexpected observation, we decided to better characterize the role of NaPi-IIa in uranium uptake. This was performed by making use of the distal tubule MDCK cell line which does not normally express the NaPi-IIa but had been stably transfected with cDNA encoding the rat NaPi-IIa (NaPi-2) under the control of a dexamethasone-inducible promoter (Quabius et al., 1996). The data that are displayed in Table 1 show that stimulation of the rat NaPi-IIa by 1mM dexamethasone significantly increased the uptake of phosphate when compared to nonstimulated cells. Using this cell line, we previously demonstrated the existence of a strong correlation between the expression level and activity of the NaPi-IIa and uranium-induced cytotoxicity (Muller et al., 2006). However, we now report that stimulation of NaPi-2 expression and transport activity plays, at most, a small role in uranium uptake. Indeed, 233 uranium transport was not significantly increased when NaPi-2 expression was stimulated in MDCK cells (Fig. 4B). PFA Protects LLC-PK1 Cells from Uranium-Induced Cytotoxicity Since our data showed that PFA significantly decreases the uptake of low concentrations of uranium (Fig. 4), we decided to further assess the capacity of this compound in protecting LLC-PK1 cells from uranium-induced toxicity. The effect of PFA on the cytotoxicity of uranium was studied using 800lM uranium (~LC50, Muller et al., 2006) whose toxicity was assessed after 24 h by quantifying the level of LDH released from damaged cells as described in the ‘‘Materials and Methods’’ section. As shown in Figure 5, specific inhibition of the NaPi cotransport systems with 10mM PFA completely suppressed the cytotoxic effect induced by uranium, confirming in that experimental condition a direct correlation between uranium uptake and toxicity. However, the existence of such a correlation was not always true and, for example, the cytotoxicity of uranium was increased to almost its maximal when LLC-PK1 cells were treated with 10lM cytochalasin D, a compound we defined as an inhibitor of uranium uptake (Fig. 3B). In a similar manner, while our results showed that 0.4lM concanavalin A has no significant effect on the transport of uranium (Fig. 3B), our cytotoxicity data clearly demonstrate that this drug significantly potentiates its cytotoxicity (Fig. 5). PFA Inhibits the Formation of Uranyl Phosphate Precipitates in LLC-PK1 Cells It is well established that after internal contamination with uranium compounds, uranium accumulates in proximal tubule cells into precipitates of uranyl phosphate (Mirto et al., 1999b) 259 URANIUM UPTAKE AND TOXICITY FIG. 3. Correlation between uranium uptake and endocytosis in LLC-PK1 cells. Effect of various drugs and culture conditions on the uptakes of (A) 50 lg/ml FITC-albumin (15 min) and (B) 3lM 233uranium (60 min). (C) Correlation between uranium uptake and endocytosis obtained by plotting all the results obtained in (B) versus those obtained in (A). The use of a linear regression test showed a p value < 0.01. Bars shown in (A) and (B) are mean ± SEM, n ¼ 4–7, *p < 0.05 versus control. that are also known as urchin-like structures. Accordingly, to better understand why PFA but not cytochalasin D nor concanavalin A reduced both uranium uptake and toxicity, we monitored by TEM the effect these three compounds have on the intracellular precipitation of a subtoxic concentration of uranium (500lM). As expected, TEM analysis of sections obtained from uranium-treated LLC-PK1 cells confirmed our cytotoxicity data; whereas the addition of 10mM PFA in the culture FIG. 4. Effect of MGA, PFA, and of the sodium-dependent phosphate cotransporter NaPi-2 on uranium uptake. (A) Effect of inhibition of the sodiumdependent phosphate cotransporters on 233uranium (3lM) uptake in LLC-PK1 cells. Inhibition was obtained by replacement of Naþ with MGA or by the addition of 0.1–10mM PFA. (B) Role of NaPi-IIa in the uptake of uranium using MDCK cells that were stably transfected with the rat NaPi-IIa (NaPi-2)– expressing vector pLK-neo containing a dexamethasone-inducible element in its promoter region. Transfected MDCK cells were pretreated for 16 h with 1lM dexamethasone to induce NaPi-2 expression (open circle) or with the corresponding vehicle (control cells, filled circle), and uranium uptake was determined using 3lM 233uranium. Each bar and point shown are mean ± SEM, n ¼ 3, *p < 0.05 versus control. medium completely abolished the intracellular precipitation of uranium, LLC-PK1 treatment with 10lM cytochalasin D or 0.4lM concanavalin A resulted in a significant increase in the number these urchin-like structures (Fig. 6). DISCUSSION So far, the renal toxic properties of uranium in experimental animals have been extensively studied and provide a detailed description of its pathophysiology where injury, including necrosis and/or apoptosis of the terminal part of the renal proximal tubule, is characteristic in all mammalian species 260 MULLER ET AL. TABLE 1 Sodium-Dependent Transport of Phosphate in the NaPi-2 Stably Transfected MDCK Cell Line 32 P uptake in MDCK cells (pmol/mg protein) Naþ MGA NaPi-2 () NaPi-2 (þ) 539.0 ± 82.4 280.6 ± 28.0 2236.9 ± 182.6*** 326.4 ± 10.9 Note. NaPi-2 expression was stimulated by 1lM dexamethasone (NaPi-2 (þ)). The sodium-dependent and -independent transport of phosphate was assessed by monitoring the uptake of 32P in the presence of 145mM NaCl or MGA, respectively. Each point represents the mean ± SEM, n ¼ 3, ***p < 0.01 versus NaPi-2 (). (Bulger, 1986; Haley, 1982; McDonald-Taylor et al., 1997; Sano et al., 1998). The chemical nephrotoxicity of uranium is pH dependent (Goldman et al., 2006) and characterized by the development of transepithelial transport and permeability defects (Leggett, 1989; Tyrakowski, 1979), inhibition of the Naþ/Kþ ATPase activity and mitochondrial injury (Brady et al., 1989), gene expression alteration (Taulan et al., 2004, 2006), and possibly oxidative DNA damage (Miller et al., 2002). Uranium-induced toxicity is also dependent upon its speciation (Carrière et al., 2004; Mirto et al., 1999a,b; Muller et al., 2006), and we recently demonstrated the existence of a strong correlation between its toxicity, the level of phosphate, and the expression and activity of the NaPi-IIas (Muller et al., 2006). Interestingly, while the cellular and molecular mechanisms by which uranium mediates its toxic effects are currently studied, very little is known about the cellular mechanisms that are FIG. 5. Cytotoxicity of uranium in LLC-PK1 cells. Determination of the percentage of cytotoxicity induced by 800lM uranium in the absence or presence of 10mM PFA, 10lM cytochalasin D, or 0.4lM concanavalin A. Bars shown are mean ± SEM, n ¼ 5, *p < 0.05 versus control. responsible for its uptake. This is surprising since uranium uptake by proximal tubule cells represents the first step in its cytotoxicity, and knowing the cellular mechanisms involved in uranium transport at this level could initiate new therapeutical approaches aimed at decreasing its uptake with specific inhibitors thereby reducing its toxic effects. Although we cannot rule out the possibility that the observed non-Michaelis-Menten type of kinetic in our uranium uptake study could be due to a direct effect of uranium on its own speciation, the corresponding Hill coefficient of 3.0 ± 0.4 that was obtained for uranium concentrations ranging from 0 to 3200lM is consistent with the existence of more than one binding site for uranium. One belief that comes from TEM analysis of kidney cells from contaminated animals or cell lines (Galle, 1974; Ghadially et al., 1982; Mirto et al., 1999a,b), where uranium has been shown to precipitate within the cytoplasm, is that uranium is taken up by endocytosis. Consistent with this hypothesis, uranium is known to bind to the plasma membrane, a chemical property used to contrast tissue sections for TEM analysis, and can therefore be internalized in a nonspecific manner by endocytosis. Here, we indirectly verified this hypothesis by demonstrating for the first time the existence of a strong correlation between absorptive-mediated endocytosis and uranium uptake ( p < 0.01). Thus, like endocytosis-mediated albumin internalization, uranium uptake was PKC dependent, dependent on cytoskeletal integrity, and significantly reduced by the use of drugs that are known to alter endocytosis. However, while one might expect that if uranium uptake is inhibited then its toxic effects would be reduced, here we observed that uranium cytotoxicity was increased when endocytosis, and therefore uranium internalization, was inhibited by 10lM cytochalasin D (Figs. 3 and 5). A similar dissociation was observed with 0.4lM concanavalin A which increased uranium toxicity without having any significant effect on its uptake. One possible explanation for such unexpected results is that uranium is taken up by at least one other transport system whose expression and/or activity would be downregulated by endocytosis and that this mechanism of transport would be responsible for the uptake of a smaller but cytosolic fraction of uranium. This cytosolic fraction of uranium would then alter the cell metabolism (Brady et al., 1989; Bulger, 1986; Goldman et al., 2006; Leggett, 1989; McDonald-Taylor et al., 1997; Tyrakowski, 1979) and generate an oxidative stress (Miller et al., 2002; Taulan et al., 2004, 2006). This would then explain why inhibition of endocytosis produced a reduction in uranium internalization but also an increase in its cytotoxic effect. Consistent with this hypothesis, endocytosis plays several central functions in cellular homeostasis and one of them is the regulation of plasma membrane expression and activity of channels, transporters, and exchangers (Chow et al., 1999; Collazo et al., 2000; Hernando et al., 2001; Shimkets et al., 1997). Several lines of evidence suggest that the NaPi-IIa could be the second transport system that would either mediate the 261 URANIUM UPTAKE AND TOXICITY FIG. 6. Section analysis by TEM of LLC-PK1 cells treated with 500lM uranium. TEM analysis of sections obtained from LLC-PK1 cells treated with a subtoxic concentration of uranium (500lM) in the absence or presence of 10mM PFA, 10lM cytochalasin D, or 0.4lM concanavalin A. uptake of a smaller but cytotoxic fraction of uranium or potentiate its cytotoxicity. Indeed, (1) NaPi-IIas activity is downregulated by endocytosis (Murer and Biber, 1996); (2) all the drugs used in this study which inhibited endocytosis and reduced uranium uptake were previously shown to stimulate the sodium-dependent transport of phosphate and increase the toxic effect of uranium (Muller et al., 2006); (3) concanavalin A, which did not alter uranium uptake and endocytosis, was found to increase NaPi-mediated phosphate uptake and uranium cytotoxicity (Muller et al., 2006); (4) while the replacement of Naþ with MGA to suppress the transport activity of the sodium-dependent system of transporters did not altered uranium uptake, it was previously shown to significantly decrease its maximal toxicity by more than 70% (Muller et al., 2006); and (5) finally, while uranium-induced toxicity was very low in MDCK cells, overexpression of NaPi-2 resulted in a strong sensitization of this cell line to uranium toxic effects (Muller et al., 2006) but did not significantly increase its uptake. Another important finding of this study is the observation that PFA, a well-characterized competitive inhibitor of the sodium-dependent phosphate cotransporters, is a potent inhibitor of uranium uptake, uranium intracellular precipitation, and uranium toxicity. In conclusion, our results suggest that absorptive endocytosis is the major mechanism of uranium internalization. We also propose that the NaPi-IIas mediate directly or indirectly the cytotoxic effect of uranium, a cellular event that can be suppressed by the use of PFA. It would now be interesting to test the therapeutical impact that this compound could have in in vivo studies. FUNDING COGEMA; Institut de Radioprotection et sûreté Nucléaire. ACKNOWLEDGMENTS We would like to thank Professors H. Metivier, M. Fournier, and P. Brochard for their helpful discussion. We also thank Dr S. Persaud for critically evaluating and correcting the manuscript. REFERENCES Becker, D., Lempertz, U., Enk, A., Saloga, J., and Knop, J. (1995). Contact sensitizers modulate mechanisms of receptor-mediated endocytosis but not fluid-phase endocytosis in murine epidermal Langerhans cells. Exp. Dermatol. 4, 211–217. Brady, H. R., Kone, B. C., Brenner, R. M., and Gullans, S. R. (1989). Early effects of uranyl nitrate on respiration and Kþ transport in rabbit proximal tubule. Kidney Int. 36, 27–34. Bulger, R. E. (1986). Renal damage caused by heavy metals. Toxicol. Pathol. 14, 58–65. Carrière, M., Avoscan, L., Collins, R., Carrot, F., Khodja, H., Ansoborlo, E., and Gouget, B. (2004). Influence of uranium speciation on normal rat kidney (NRK-52E) proximal cell cytotoxicity. Chem. Res. Toxicol. 17, 446–452. 262 MULLER ET AL. Carrière, M., Thiebault, C., Milgram, S., Avoscan, L., Proux, O., and Gouget, B. (2006). Citrate does not change uranium chemical speciation in cell culture medium but increases its toxicity and accumulation in NRK-52E cells. Chem. Res. Toxicol. 19, 1637–1642. Chow, C. W., Khurana, S., Woodside, M., Grinstein, S., and Orlowski, J. (1999). The epithelial Na(þ)/H(þ) exchanger, NHE3, is internalized through a clathrin-mediated pathway. J. Biol. Chem. 274, 37551–37558. Collazo, R., Fan, L., Hu, M. C., Zhao, H., Wiederkehr, M. R., and Moe, O. W. (2000). Acute regulation of Naþ/Hþ exchanger NHE3 by parathyroid hormone via NHE3 phosphorylation and dynamin-dependent endocytosis. J. Biol. Chem. 275, 31601–31608. Galle, P. (1974). Role des lysosomes et des mitochondries dans les phenomenes de concentration et d’elimination d’elements mineraux (uranium et or) par le rein. J. Microscopiel 19, 17–24. Gekle, M., Mildenberger, S., Freudinger, R., Schwerdt, G., and Silbernagl, S. (1997). Albumin endocytosis in OK cells: Dependence on actin and microtubules and regulation by protein kinases. Am. J. Physiol. 272, F668–F677. Ghadially, F. N., Lalonde, J. M., and Yang-Steppuhn, S. (1982). Uraniosomes produced in cultured rabbit kidney cells by uranyl acetate. Virchows Arch., B, Cell Pathol. 39, 21–30. Goel, K. A., Garg, V. K., and Garg, V. (1980). Histopathology of kidney of albino rat poisoned with uranyl nitrate. Bull. Environ. Contam. Toxicol. 24, 9–12. Goldman, M., Yaari, A., Doshnitzki, Z., Cohen-Luria, R., and Moran, A. (2006). Nephrotoxicity of uranyl acetate: Effect on rat kidney brush border membrane vesicles. Arch. Toxicol. 80, 387–393. Haley, D. P. (1982). Morphologic changes in uranyl nitrate-induced acute renal failure in saline- and water-drinking rats. Lab. Invest. 46, 196–208. Henge-Napoli, M. H., Ansoborlo, E., Chazel, V., Houpert, P., Paquet, F., and Gourmelon, P. (1999). Efficacy of ethane-1-hydroxy-1,1-bisphosphonate (EHBP) for the decorporation of uranium after intramuscular contamination in rats. Int. J. Radiat. Biol. 75, 1473–1477. dehydrogenase (LDH) activity of the cultured eukaryotic cells as marker of the number of dead cells in the medium. J. Biotechnol. 25, 231–243. Lopez, R., Diaz Sylvester, P. L., Ubios, A. M., and Cabrini, R. L. (2000). Percutaneous toxicity of uranyl nitrate: Its effect in terms of exposure area and time. Health Phys. 78, 434–437. Mamdouh, Z., Giocondi, M. C., Laprade, R., and Le, G. C. (1996). Temperature dependence of endocytosis in renal epithelial cells in culture. Biochim. Biophys. Acta 1282, 171–173. Mao, Y., Desmeules, M., Schaubel, D., Berube, D., Dyck, R., Brule, D., and Thomas, B. (1995). Inorganic components of drinking water and microalbuminuria. Environ. Res. 71, 135–140. McDonald-Taylor, C. K., Singh, A., and Gilman, A. (1997). Uranyl nitrateinduced proximal tubule alterations in rabbits: A quantitative analysis. Toxicol. Pathol. 25, 381–389. Miller, A. C., Stewart, M., Brooks, K., Shi, L., and Page, N. (2002). Depleted uranium-catalyzed oxidative DNA damage: Absence of significant alpha particle decay. J. Inorg. Biochem. 91, 246–252. Mirto, H., Barrouillet, M. P., Henge-Napoli, M. H., Ansoborlo, E., Fournier, M., and Cambar, J. (1999a). Influence of uranium(VI) speciation for the evaluation of in vitro uranium cytotoxicity on LLC-PK1 cells. Hum. Exp. Toxicol. 18, 180–187. Mirto, H., Henge-Napoli, M. H., Gibert, R., Ansoborlo, E., Fournier, M., and Cambar, J. (1999b). Intracellular behaviour of uranium(VI) on renal epithelial cell in culture (LLC-PK1): Influence of uranium speciation. Toxicol. Lett. 104, 249–256. Moss, M. A., McCurdy, R. F., Dooley, K. C., Givner, M. L., Dymond, L. C., Slayter, J. M., and Courneya, M. M. (1983). Uranium in drinking water—report on clinical studies in Nova Scotia. In Chemical Toxicology and Clinical Chemistry of Metals (S. S. Brown and J. Savory, Eds.), pp. 149–152. Academic Press, London. Muller, D., Houpert, P., Cambar, J., and Henge-Napoli, M. H. (2006). Role of the sodium-dependent phosphate co-transporters and of the phosphate complexes of uranyl in the cytotoxicity of uranium in LLC-PK1 cells. Toxicol. Appl. Pharmacol. 214, 166–177. Hernando, N., Karim-Jimenez, Z., Biber, J., and Murer, H. (2001). Molecular determinants for apical expression and regulatory membrane retrieval of the type IIa Na/Pi cotransporter. Kidney Int. 60, 431–435. Murer, H., and Biber, J. (1996). Molecular mechanisms of renal apical Na/ phosphate cotransport. Annu. Rev. Physiol. 58, 607–618. Heuser, J. E., and Anderson, R. G. (1989). Hypertonic media inhibit receptormediated endocytosis by blocking clathrin-coated pit formation. J. Cell Biol. 108, 389–400. Quabius, E. S., Murer, H., and Biber, J. (1996). Expression of proximal tubular Na-Pi and Na-SO4 cotransporters in MDCK and LLC-PK1 cells by transfection. Am. J. Physiol. 270, F220–F228. Hryciw, D. H., Pollock, C. A., and Poronnik, P. (2005). PKC-alpha-mediated remodeling of the actin cytoskeleton is involved in constitutive albumin uptake by proximal tubule cells. Am. J. Physiol. Renal Physiol. 288, F1227–F1235. Sano, K., Fujigaki, Y., Ikegaya, N., Ohishi, K., Yonemura, K., and Hishida, A. (1998). The roles of apoptosis in uranyl acetate-induced acute renal failure. Ren. Fail. 20, 697–701. Kobayashi, S., Nagase, M., Honda, N., and Hishida, A. (1984). Glomerular alterations in uranyl acetate-induced acute renal failure in rabbits. Kidney Int. 26, 808–815. Kurttio, P., Auvinen, A., Salonen, L., Saha, H., Pekkanen, J., Makelainen, I., Vaisanen, S. B., Penttila, I. M., and Komulainen, H. (2002). Renal effects of uranium in drinking water. Environ. Health Perspect. 110, 337–342. Kurttio, P., Harmoinen, A., Saha, H., Salonen, L., Karpas, Z., Komulainen, H., and Auvinen, A. (2006). Kidney toxicity of ingested uranium from drinking water. Am. J. Kidney Dis. 47, 972–982. L’Azou, B., Henge-Napoli, M. H., Minaro, L., Mirto, H., Barrouillet, M. P., and Cambar, J. (2002). Effects of cadmium and uranium on some in vitro renal targets. Cell Biol. Toxicol. 18, 329–340. Leggett, R. W. (1989). The behavior and chemical toxicity of U in the kidney: A reassessment. Health Phys. 57, 365–383. Legrand, C., Bour, J. M., Jacob, C., Capiaumont, J., Martial, A., Marc, A., Wudtke, M., Kretzmer, G., Demangel, C., and Duval, D. (1992). Lactate Shimkets, R. A., Lifton, R. P., and Canessa, C. M. (1997). The activity of the epithelial sodium channel is regulated by clathrin-mediated endocytosis. J. Biol. Chem. 272, 25537–25541. Taulan, M., Paquet, F., Argiles, A., Demaille, J., and Romey, M. C. (2006). Comprehensive analysis of the renal transcriptional response to acute uranyl nitrate exposure. BMC Genomics 7, 2. Taulan, M., Paquet, F., Maubert, C., Delissen, O., Demaille, J., and Romey, M. C. (2004). Renal toxicogenomic response to chronic uranyl nitrate insult in mice. Environ. Health Perspect. 112, 1628–1635. Tyrakowski, T. (1979). Disturbances in electrolyte transport in UO2þþ ion intoxication—Model studies on the preliminary stage of toxic acute renal failure. III. Disturbances in Naþ and Kþ ion excretion and decrease in glomerular filtration rate following uranyl acetate intoxication. Acta Med. Pol. 20, 317–329. Zamora, M. L., Tracy, B. L., Zielinski, J. M., Meyerhof, D. P., and Moss, M. A. (1998). Chronic ingestion of uranium in drinking water: A study of kidney bioeffects in humans. Toxicol. Sci. 43, 68–77.