Toxicology in Vitro 21 (2007) 1066–1076 www.elsevier.com/locate/toxinvit
Cytotoxicity of peroxisome proliferator-activated receptor a and c agonists in renal proximal tubular cell lines He´ctor Giral, Ricardo Villa-Bellosta, Julia Catala´n, Vı´ctor Sorribas
*
Laboratory of Molecular Toxicology, University of Zaragoza, E50013 Zaragoza, Spain Received 7 August 2006; accepted 28 March 2007 Available online 14 April 2007
Abstract Fibrates and thiazolidinediones are agonists of peroxisome proliferator-activated receptors (PPAR) a and c, pharmacologically designed to control dyslipidemia and insulin resistance, respectively. Several works have reported the toxicity of some agonists in a number of tissues. In this work we have analyzed the toxicity of two PPARa (WY14643 and clofibrate) and two PPARc (pioglitazone and ciglitazone) agonists, using three different renal proximal tubular cell lines: Opossum OK, pig LLC-PK1, and murine MCT. Cell death was determined by the activity of intracellular lactate dehydrogenase. WY14643 and ciglitazone increased cell death with LC50 values of 92–124 lM and 8.6–14.8 lM, respectively, depending on the cell line. Clofibrate and pioglitazone were, however, noncytotoxic even at concentrations of 10 and 100 higher than the corresponding EC50, which suggests that cell death is independent of PPAR activation. Discrimination between apoptosis or necrosis was analyzed by light microscopy and stress fiber morphology, double staining with acridine orange and ethidium bromide, binding of annexin V, caspase-3 activity, and DNA laddering. With these methods, no signs of apoptosis were observed, which suggests a direct necrosis of the compounds on these renal proximal tubular cell lines. 2007 Elsevier Ltd. All rights reserved. Keywords: PPAR; Peroxisome proliferator; WY14643; Clofibrate; Ciglitazone; Pioglitazone; Proximal tubule; Cytotoxicity; OK cells; LLC-PK1; MCT; Kidney
1. Introduction Dyslipidemias and type II diabetes are two major alterations with dramatic effects in kidney function. Peroxisome proliferator-activated receptors (PPAR) are nuclear Abbreviations: CHAPS, [(3-cholamidopropyl)dimethylammonio]-1propanesulfonate hydrate; DIC, differential interference contrast; DMEM, Dulbecco’s Modified Eagle Media; DMSO, dimethyl sulfoxide; DTT, dithiothreitol; EDTA, ethylenediaminetetraacetic acid; EGFP, enhanced green fluorescent protein; FCS, fetal calf serum; HBSS, Hank’s balanced salt solution; Hepes, 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic; LC, lethal concentration; LDH, lactate dehydrogenase; PBS, phosphate-buffered saline; PI, propidium iodide; PPAR, peroxisome proliferator-activated receptor; RT-PCR, Reverse transcription-polymerase chain reaction; TBE, tris-borate-EDTA; TRITC, tetramethylrhodamine isothiocyanate; TZD, thiazolidinedione. * Corresponding author. Tel.: +34 976761631; fax: +34 976761612. E-mail address:
[email protected] (V. Sorribas). 0887-2333/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2007.03.021
receptors that have been pharmacologically targeted to treat such metabolic disorders (Berger and Moller, 2002). Once activated by the corresponding ligands, PPARs heterodimerize with the retinoid X receptor and induce the expression of target genes. Fibrates are amphipathic carboxylic acids that have been used as agonists of PPARa to normalize hypertriglyceridemia (Robillard et al., 2005; Barter and Rye, 2006), while thiazolidinediones (TZDs) activate PPARc to reduce the resistance of peripheral tissues to insulin (Staels and Fruchart, 2005; Musi and Goodyear, 2006). In addition to the activation of the corresponding PPARs, fibrates and TZDs also exert important effects in a PPAR-independent way. For example, troglitazone induces changes in the control of the activity of acid–base balance in LLC-PK1 through the activation of protein kinase C-ERK pathways (Welbourne et al., 2003).
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
All PPARs identified so far have also been described in the kidney, either along the nephron or in non-parenchymal cells (Guan and Breyer, 2001; Sato et al., 2004). Several protective effects of both fibrates and TZDs have also been reported in the kidney, in addition to the effects on dyslipidemias and antidiabetic goals. Some of these effects involve correction of blood pressure in diabetic patients with hypertension, reduction in the progression of glomerulosclerosis, and correction of the ischemia-reperfusion injury (Izzedine et al., 2005; Chung et al., 2005). However, the kidney could be also a target for undesirable side effects of fibrates and TZD. In recent years there have been numerous reports on the toxicity of PPAR agonists in several tissues, and this has been reviewed recently (Peraza et al., 2006). The toxic effects of PPARs include hepatocarcinogenicity, partially related to their activation by environmental pollutants. The agonists act as promoters depending on critical factors that include species differences in receptor activity, relative ligand binding/activation, and receptor specificity (reviewed by Klaunig et al., 2003). With respect to noncarcinogenic effects, they include cell proliferation reduction and an increase in apoptotic cell death in a variety of cell types (Arici et al., 2003; Liu et al., 2004; Tsuchiya et al., 2003; Roberts et al., 2002). Curiously, the renal toxicity of these drugs has not been extensively analyzed, and there are only a few reports available. For example, Weissgarten et al. (2006) reported decreased proliferation and increased apoptosis in mesangial cells obtained from mice treated with rosiglitazone, a PPARc agonist. Similar findings have been obtained with renal fibroblasts (Parameswaran et al., 2003; Zafiriou et al., 2005), proximal tubular cells (Arici et al., 2003), and rat mesangial cells (Tsuchiya et al., 2003). In this work, we have analyzed the cytotoxicity of two fibrates (WY14643 and clofibrate) and two thiazolidinediones (pioglitazone and ciglitazone) in several renal cell lines of proximal tubular origin. Clofibrate and pioglitazone showed no signal of toxicity even at supra-pharmacological doses, while WY14643 and ciglitazone induced a dosedependent, non-apoptotic cell death at doses close to the dissociation constant of the compounds. 2. Materials and methods 2.1. Cell culture Opossum kidney (OK) and pig LLC-PK1 (an appreciated donation from Dr. S.A. Kempson, Indiana University School of Medicine, Indianapolis) cells were grown identically, as previously described (Sorribas et al., 1995; Andreoli et al., 1993). In brief, they were cultured in a humidified 5% CO2/95% air atmosphere in DMEM: Ham’s F-12 (1:1, vol:vol) supplemented with 10% fetal calf serum (FCS), 100 IU/ml penicillin G and 100 lg/ml streptomycin, and 4 mM L-glutamine. Mouse cortical tubular cells (MCT, generously provided by Dr. M. Levi, University of Colo-
1067
rado HSC, Denver) were grown in DMEM supplemented with 10% FCS, 100 IU/ml of penicillin G, 100 lg/ml of streptomycin and L-glutamine. All three cell lines were grown to confluence and then rendered quiescent overnight by incubating the cells in the same corresponding culture media with 0.2% FCS. 2.2. Incubations with PPAR agonists Quiescent cells were treated with PPAR ligands in Hank’s balanced salt solution (HBSS) containing calcium and magnesium (Invitrogen, Carlsbad, CA) for the indicated times and concentrations of drugs. Similar results were obtained when the cells were incubated in DMEM. Two fibrates, WY14643 (Sigma, St. Louis, MO) and clofibrate (Calbiochem, San Diego, CA), were used as PPARa activators, and two TZDs, ciglitazone (Calbiochem) and pioglitazone (Cayman, Ann Arbor, MI), were used as PPARc activators. All of them were dissolved in dimethyl sulfoxide (DMSO) prior to use, and a similar volume of DMSO was added to the control cells. 2.3. Lactate dehydrogenase (LDH) Total cell death was quantified by the activity of the lactate dehydrogenase released from the cytosol of damaged cells into the supernatant of the cultures. In short, tubular epithelial cell lines were grown in 24-well plates and, when confluent, treated with the corresponding drugs in 1 ml assay medium. At the indicated times, 50 ll of medium were taken to measure LDH activity, and samples were kept at 20 C until the assay. The activity was determined using a Cytotoxicity Detection kit (Roche, Mannheim, Germany) in a DTX-880 multimode plate reader (Beckman Coulter, Inc., Fullerton, CA), following the manufacturer’s instructions. The assay is based on the reduction of NAD+ to NADH/H+ by the LDH-catalyzed conversion of lactate to pyruvate. Next, the enzyme diaphorase transfers H/H+ from NADH/H+ to iodotetrazolium chloride, which is reduced to formazan, which shows a broad absorption maximum at about 500 nm. The results are expressed as a percentage of total activity released with 2% Triton X-100 in assay medium. In dose–response experiments, the following equation expressing a sigmoidal curve with variable slope was used to calculate the mean lethal concentration (LC50) by non-linear regression: Y ¼ LC0 þ ðLCmax LC0 Þ=ð1 þ 10ððLog LC50 X ÞnÞ Þ. In this equation, Y is the percentage of LDH activity, LC0 is the percentage of activity at zero concentration of drug, LCmax is the maximal LDH activity as obtained with 2% Triton X-100, X is the logarithm of drug concentration, and n is the Hill slope. Activity was normalized according to the 100% activity obtained with triton X-100. Y starts at LC0 and goes to LCmax with a sigmoid shape. GraphPad Prism 4.0 software for Macintosh was used for iterations and for kinetic and statistical analysis.
1068
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
2.4. Phalloidin staining Stress fibers were stained with TRITC (tetramethylrhodamine isothiocyanate)-conjugated phalloidin (Sigma), previously dissolved in methanol. Cells were grown on cover glasses, and after treatment of the cells with the corresponding PPAR activators they were washed with phosphate-buffered saline (PBS), fixed with 3% paraformaldehyde in PBS, and permeabilized with 0.1% saponin in PBS as previously described (Lanaspa et al., 2007). TRITC-phalloidin was added at 10 lg/ml for 30 min at room temperature to stain the actin filaments, and the cells were visualized using a Carl Zeiss Axiovert 200M (Jena, Germany) inverted fluorescence microscope. 2.5. Assay of cell viability with ethidium bromide-acridine orange uptake Total cell death and distinction between early or late apoptosis and necrosis were determined by the differential uptake of these fluorescent DNA binding dyes exactly as previously described, but using adherent LLC-PK1 cells (McGahon et al., 1995). In brief, cells grown on cover glasses were incubated with 100 ll of a mixture of 4 lg/ml acridine orange (Fluka, Buchs, Switzerland) and 4 lg/ml ethidium bromide (Bio-Rad, Hercules CA) in PBS, after the corresponding treatments. Immediately the cells were visualized by fluorescent microscopy with a 488-nm excitation dichroic filter. Since acridine orange intercalates in the DNA but only interacts with the RNA, and viable cells do not uptake ethidium bromide, these cells exhibit green nuclei and a dotted-orange cytoplasm. However, ethidium bromide is taken up by necrotic cells, which turn red. With respect to early apoptotic cells, their nuclei remain green because ethidium bromide still cannot enter into them, whereas late apoptotic cells turn red due to the condensed chromatin. As a positive control for apoptosis in LLC-PK1 cells, cadmium was used at 50 lM for 16 h (data not shown). 2.6. Annexin V-propidium iodide staining Changes in the position of phosphatidylserine (an early event in apoptosis) were determined with an ApoAlert Annexin V kit (Clontech Laboratories, Inc., Mountain View, CA) following the manufacturer’s protocol. A combination of enhanced green fluorescent protein (EGFP), of annexin V fusion protein (a 35.8-kDa protein that has a strong affinity for phosphatidylserine) and propidium iodide (PI) was used to differentiate between apoptosis and necrosis. With this method, apoptotic cells would show green staining in the plasma membrane, but dead cells would stain red since PI is also used, which crosses plasma membranes that have lost their integrity, as in the case of ethidium bromide. In short, 5 · 105 LLC-PK1 cells growing on cover glasses were treated with different PPAR
ligands and then incubated with 0.2 lg annexin V and 0.5 lg propidium iodide in 200 ll binding buffer. After incubation of the cells for 15 min at room temperature in the dark, they were washed and fixed with 2% paraformaldehyde. Analysis was performed using an Axiovert 200M fluorescence microscope with a dual filter for fluorescein and rhodamine. 2.7. Caspase-3 activity Cysteine-requiring aspartate protease-3 (caspase-3) activation is one of the earlier events in apoptosis. The activity was determined in LLC-PK1 cells after 3 h of treatments, with a colorimetric Caspase-3 Assay kit (Sigma), following the instructions provided. In short, 107 cells growing in 3.5-cm diameter Petri plastic dishes were lysed in 100 ll of a buffer containing 50 mM 4-(2hydroxyethyl)piperazine-1-ethanesulfonic (Hepes) pH 7.4, 5 mM -[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate hydrate (CHAPS) and 5 mM dithiothreitol (DTT). Cell lysates were spun at 20,000g for 15 min at 4 C, and 5 ll of supernatants were combined with 85 ll of assay buffer (20 mM Hepes pH 7.4, 0.1% CHAPS, 5 mM DTT, 2 mM ethylenediaminetetraacetic acid EDTA), and 10 ll of Caspase-3 substrate, acetyl-AspGlu-Val-Asp p-nitroanilide, and incubated overnight at 37 C. Release of p-nitroanilide from the substrate was determined by absorbance at 405 nm using a DTX-880 plate reader (Beckman Coulter). Data are shown as absolute absorbances, and direct comparison to the control cells are shown. As a positive control, LLC-PK1 cells were treated with 50 lM CdCl2 for 16 hours. This treatment increased caspase-3 activity 8.8 times compared to untreated cells. 2.8. DNA fragmentation One of the late events in apoptosis is endonuclease cleavage of genomic DNA into internucleosomal fragments of approximately 200 bp integer multiples. This clearly differentiates from necrotic cell death, where unorganized DNA degradation takes place (Wyllie, 1980). Genomic DNA was obtained as described (Zhivotosky and Orrenius, 2003): 0.5 · 106 LLC-PK1 cells grown in plastic support were harvested and lysed with 20 ll of a buffer containing 2 mM EDTA, 100 mM Tris–HCl pH 8.0, and 0.8% (w/v) sodium lauryl sarcosine. RNA and proteins were enzymatically degraded, and the resultant DNA electrophoresed in 1.8% agarose tris-borate-EDTA (TBE) gel at 60 mA for 4 h. The gels were photographed under UV light using a Gel-Doc system (Bio-Rad). 2.9. RT-PCR Total RNA from the LLC-PK1 cells was isolated using a Micro-to-Midi Total RNA Purification System
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
(Invitrogen), and the contaminating genomic DNA was eliminated using an Amplification Grade DNase I (Invitrogen). Reverse transcription (RT) was performed with a Superscript III First-Strand Synthesis system, followed by polymerase chain reaction (PCR). 459 and 496 base pair fragments from pig PPAR alpha and gamma, respectively, were amplified using a Thermus DNA polymerase (Biotools B&M Labs, Madrid, Spain) and the following primers (Invitrogen): For PPARa, sense AGG TCC GCA TCT TCC A, antisense AGA AAG ACG TCG TCG GG; and for PPARc sense TCC GGA GGA CTA TCA GA, antisense GGC ATA CTC TGT GAT CTC C were used. Amplicons were visualized in an ethidium bromide-stained DNA agarose gel under UV and acquired using a Gel-Doc system (Bio-Rad).
3. Results 3.1. Dose–response relationships of PPARa and PPARc agonists and cell viability Dose–response experiments were performed in three different renal proximal tubular cell lines to compare PPAR ligand concentration and the leakage of cytosolic lactate dehydrogenase (LDH) as a measure of total cell death. Dose–response assays were performed at different incubation times, and the results at the 12-hour time point are shown in Fig. 1. With respect to PPARa agonists, clofibrate was inert up to 500 lM and for 12 h in all three cell lines (triangles, first row of panels). However, WY14643 induced a very sharp increase in LDH activity, especially in MCT cells (Hill coefficient 9.9), with a mean lethal concentration of approximately 100 lM in all three cell lines (squares, first row of panels). OK cells exhibited a smoother response but also the lowest observed experimental effect, with a significant increase in LDH activity at 50 lM WY14643. Similar findings were obtained with PPARc agonists. A lower range of TZD concentrations was used, according to the dissociation constants of the drugs and the plasma therapeutic concentrations (Brunton et al., 2005). Pioglitazone did not increase LDH activity up to 50 lM (not shown), but ciglitazone induced cell death with an LC50 of approximately 10 lM. Again, MCT cells exhibited the steepest
2.10. Statistical analysis The data are shown as means ± the standard error. The statistical significance was determined by either a Student’s t-test or a one-way analysis-of-variance (ANOVA) and the Tukey multiple-comparison test, whereby P < 0.05 was considered significant. Experiments were performed 3–6 times with similar results. The number of experiments performed depended on whether the parameters to be determined were of qualitative or quantitative type.
OK
MCT
% LDH activity
40
80 60 40
0 0.5
1.0
1.5
2.0
2.5
3.0
0 0.5
1.0
log μM
100
% LDH activity
% LDH activity
60 40
1.5
2.0
2.5
3.0
0 0.5
1.0
40 20
0.5
1.0
1.5
-0.5
100
80 60 40
0.0
2.0
2.5
3.0
120
LC50: 9.3 μM Hill: 13.9
LC50: 14.8 μM Hill: 2.5
80 60 40 20
20
log μM
1.5
μM
120
LC50: 8.6 μM Hill: 4.0
60
0.0
80
log μM
100 80
LC50: 116 μM Hill: 4.9
20
20
20
TZD
100
% LDH activity
Fibrates
% LDH activity
60
120
LC50: 124 μM Hill: 9.9
100
% LDH activity
LC50: 92 μM Hill: 2.2
80
-0.5
LLC-PK1
120
120 100
1069
0.5
log μM
1.0
1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
μM
Fig. 1. Dose–response assays of LDH activity as a function of PPAR agonists, fibrates and thiazolidinediones (TZD). OK, MCT and LLC-PK1 cells were incubated for 12 h with the indicated drugs and concentrations, and aliquots of assay media were analyzed for LDH activity. Concentrations of agonists are shown as logarithmic units. The lethal mean concentration (LC50) and the slope of the curves (Hill parameter) are indicated in the panels for WY14643 and ciglitazone. Fibrate panels (related to PPARa): squares are WY14643; triangles are clofibrate. TZD panels (related to PPARc): squares are ciglitazone; triangles are pioglitazone.
1070
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
20 μM ciglitazone 120
100
100
% LDH activity
% LDH activity
100 μM WY14643 120
80 60 40
80 60 40 20
20
0
0 0
6
12
18
24
30
36
42
0
48
6
12
18
0h
24
30
36
42
48
Hours
Hours
3h
6h
12 h
24 h
100 μM WY14643
20 μM ciclitazone
Fig. 2. Effect of WY14643 and ciglitazone as a function of time. (a) OK (squares), MCT (circles) and LLC-PK1 (triangles) cells were incubated with either 100 lM WY14643 or 20 lM ciglitazone at the indicated times, and LDH activity was assayed. (b) Phase-contrast pictures of LLC-PK1 cells incubated with WY14643 or ciglitazone at the indicated times.
increase, as shown with a Hill value of 13.9, and OK cells exhibited the lowest experimental concentration with a significant increase in LDH activity (5 lM ciglitazone). 3.2. Time dependence of PPAR agonist effects Fixed PPAR ligand concentrations were used to analyze the course of their effects on cell viability. 100 lM WY14643 increased LDH activity in OK and MCT cells at 6 h of incubation, and it became a frank effect at 12 h in LLC-PK1 cells (Fig. 2a). Similar effects were obtained with 20 lM ciglitazone, which showed significant increases in LDH activity that were found at 6 h with OK and MCT cells and at 12 h with LLC-PK1 cells. When combining dose–response relationships and time courses, different findings were observed: At 3 h of incubation, no cell death was detected at any of the concentrations of WY14643 or ciglitazone. However, at 6 h significant cytotoxicity was observed with 100 lM WY14643 and 20 lM ciglitazone. At 24 h toxicity was observed with 50 lM and 5 lM, respectively (data not shown). A phase contrast optical inspection of the cells revealed a similar morphological process with either WY14643 or ciglitazone treatments. Fig. 2b shows the changes in LLC-PK1 cells, and identical alterations were seen with OK and MCT cells. The nucleoli of the cells become bright at just 3 h of treatment with both chemicals. This change is
more pronounced at 6 h, when disorganization of the cytoplasm also starts to be observed. At 12 h the cells are separated and enlarged; some are already detached, and the cytosols are filled with vacuoles. All these modifications have greater intensity at 24 h. No evidence of chromatin condensation or clear hallmarks of apoptosis can be detected with this methodology. Finally, no changes with clofibrate or pioglitazone were observed at either concentration or incubation time in all three cell lines (data not shown). 3.3. Expression of PPARa and PPARc in LLC-PK1 cells The experiments designed to characterize the type of cell death induced by PPAR agonists were performed in LLCPK1 cells. Therefore, the expression of the corresponding PPAR RNAs was determined by amplification of reverse transcribed cDNA by PCR. Fig. 3 shows that both PPARa and PPARc RNAs are expressed with similar intensities in this cell line, i.e. a receptor-dependent mechanism of toxicity of PPAR agonists cannot be excluded. 3.4. WY14643 and ciglitazone disrupt stress fibers Changes in morphology observed with phase contrast microscopy suggested alterations in cytoskeleton, and therefore we analyzed actin microfilaments in treated cells. LLC-PK1 cells were treated with 100 lM WY14643 and
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
PPAR α
PPAR γ
M bp 1000 650 500 400 300 200 100
Fig. 3. RT-PCR of PPARa and PPARc fragments. Both receptors are expressed with similar abundance in this cell line. M, marker with molecular sizes in base pairs.
10 lM ciglitazone, in both cases for 6 and 16 h, and stress fibers were stained with rhodamine-labeled phalloidin (Fig. 4). Modifications in polymerization were already evident at 6 h, but they were maximal at 16 h. Spaces between cells were already observed at 6 h, and clear alterations in cytoskeleton with extracellular debris were already at 16 h. However, no indications of apoptosis could be observed in these experiments. 3.5. Effects of WY14643 and ciglitazone on DNA intercalating dye staining For a more in-depth analysis of the morphological changes produced by these two PPAR agonists, a double staining with acridine orange and ethidium bromide was
1071
performed. LLC-PK1 cells treated with either 100 lM WY14643 or 10 lM ciglitazone, for either 6 or 16 h, were stained as explained in Section 2 (Fig. 5). A differential interference contrast (DIC) analysis revealed a smooth pattern of confluent cells. Control cells showed a viable pattern of staining, including condensed nucleoli. Cells treated with either WY14643 or ciglitazone were stained in two patterns that were compatible with both viable cells (green cytoplasm and nuclei with orange dots as RNA) or necrotic cells (red cytoplasms and nuclei). The percentage of each pattern depended on the time, with most of the cells in the necrotic pattern at 16 h. A secondary necrosis (late apoptosis) was not evident under these conditions, because chromatin was not condensed. Cytosolic vacuoles were likewise very evident with the DIC optics. 3.6. Effect of WY14643 and ciglitazone on phosphatidylserine distribution One of the earliest events in apoptosis involves the redistribution of phosphatidylserine from the inner to the outer layer of the plasma membrane. Therefore, to find out whether WY14643 and ciglitazone produced apoptosis as an early event or at low concentrations, we treated LLCPK1 cells with two different concentrations and for two different times. To detect phosphatidylserine in apoptotic cells we used EGFP-annexin V (Martin et al., 1995). Control cells did not stain with either EGFP-annexin V or PI, and DIC optics evidenced a confluent mosaic of smooth LLC-PK1 cells (Fig. 6). In contrast, treatment with 50 lM WY14643 for 12 h altered the morphology of the cells as shown by the DIC, with an increase of vacuoles and many cells stained red with propidium iodide (PI).
Fig. 4. Effect of WY14643 and ciglitazone on stress fibers. LLC-PK1 cells were incubated for 6 or 16 h with PPAR agonists, and then fixed and stained with rodhamine-conjugated phalloidin. At 16 h actin filaments have completely lost organization, and cellular shapes are modified, with separated bodies. After 6 h of incubation the effects on stress fibers are evident but of less intensity.
1072
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
Fig. 5. Staining of LLC-PK1 cells with DNA intercalating dyes. Cells were incubated with 100 lM WY14643 or 10 lM ciglitazone for 6 and 16 h and then stained with ethidium bromide and acridine orange. Non-fixed cells were analyzed by DIC or fluorescence microscopy. Both treatments increase the number of cells that uptake ethidium bromide (stains red). Live cells show green as a consequence of acridine orange. No nuclear or cell morphology reminiscent of apoptotic cell can be appreciated.
No peripheral green staining (annexin V) on the cells could be found, thereby indicating that the phosphatidylserine had not been externalized. Similar results were observed at 6 h of treatment with the same concentration, but the number of cells stained with PI was dramatically reduced. At both incubation times, DIC optics also revealed a morphology pattern compatible with primary necrotic cell death. Treatment of the cells with 200 lM WY14643 simply increased the number of cells stained with PI at both times. Similar results were obtained in LLC-PK1 cells treated with 5 or 20 lM ciglitazone, and for 6 or 12 h: All stainings revealed primary necrotic cell death. 3.7. Effect of WY14643 and ciglitazone on caspase-3 activity in LLC-PK1 cells Caspase-3 activity was also assayed to check whether any apoptotic activity was present during WY14643 and ciglitazone treatments. LLC-PK1 cells were treated for 3 h with either 100 or 200 lM WY14643, and with 5 or 10 lM ciglitazone. Simultaneous analyses of LDH activity in assay medium and caspase-3 in cell lysate were performed (Fig. 7). While LDH activity increased in a concentration-dependent way, caspase 3 activity was not modified and maintained the basal levels of the control cells. 3.8. Effect of WY14643 and ciglitazone on DNA laddering in LLC-PK1 cells As a final test in the characterization of WY14643 and ciglitazone cytotoxicities, the degradation of genomic DNA was tested in LLC-PK1 cells. The cells were incubated for 16 h with different concentrations of the drugs, as shown in Fig. 8. The genomic DNA was then extracted and separated by an agarose gel electrophoresis. After 16 h of treatment, WY14643 did not induce degradation of
genomic DNA at 25 or 50 lM. However, at 100 and 200 lM, WY14643 degraded DNA in a necrotic, continuous pattern: no laddering of 200 bp or any multiples thereof were observed. Ciglitazone produced the same degradation of genomic DNA at either 10 or 20 lM, while no degradation was observed at 5 lM. These results finally confirm that WY14643 and ciglitazone are cytotoxic to cells of proximal tubular lines and that the cells die from pure and classic necrosis.
4. Discussion Peroxisome proliferator-activated receptors are nuclear receptors that, upon activation by specific ligands and in combination with RXR receptor, bind to PPAR response elements (PPRE) to initiate the expression of specific target genes. The receptors are involved in controlling a plethora of cell functions, including regulation of several metabolic pathways, such as lipid biosynthesis and glucose metabolism. PPARa and -c are also the target for several environmental contaminants, as well as for drugs designed to correct metabolic disorders such as dyslipidemia or insulin resistance (for an extensive revision see Berger and Moller, 2002). Many reports have described the existence of side effects after incubation with drugs initially designed to activate the PPARs (recently reviewed by Peraza et al., 2006). Some of them involve antiproliferative effects and even apoptosis in different cell types, such as the antineoplastic effects of PPARc agonists (Grommes et al., 2004; Theocharis et al., 2004), apoptosis (Okuda et al., 2000; Redondo et al., 2005) or growth arrest (Bruemmer et al., 2003) induced by several TZDs and PPARa agonists (Roberts et al., 2002) in vascular smooth muscle cells; PPARc mediated apoptosis of renal proximal tubular cells (Arici et al., 2003), mesangial cells (Tsuchiya et al., 2003) and interstitial fibroblasts (Parameswaran et al., 2003); etc. In this work
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
1073
Fig. 6. Incubation of LLC-PK1 cells with EGFP-annexin V and propidium iodide. Cells were treated with WY14643 and ciglitazone at the indicated concentrations and incubation times. They were incubated with EGFP-tagged annexin V and propidium iodide and then fixed. The absence of green staining indicates that annexin V did not have access to phosphatidylserine in the outer layer of the plasma membrane. Propidium iodide stains for dead cells.
we show a direct necrotic effect of some PPAR agonists, namely WY14643 and ciglitazone using concentrations similar to the therapeutical plasma levels of the drugs (Berger and Moller, 2002; Brunton et al., 2005). We have shown that these effects are not universal for all PPAR activators, because clofibrate and pioglitazone (two alternative PPARa and -c agonists, respectively) are non-cytotoxic in the same cells. In the case of PPARa agonists, WY-14643 has an EC50 of 0.63–5 lM (Willson et al., 2000), while we have determined that the LC50 is 92–124 lM in the three renal cell lines employed (Fig. 1). Likewise, clofibrate shows an EC50 of 50–55 lM (Willson et al., 2000), but it is not toxic even at 500 lM in any of the cell lines. With respect to PPARc agonists, ciglitazone activates the receptor with 3 lM EC50 (Willson et al., 1996), but the LD50 is 8.6–14.8 lM, and pioglitazone does not show signs of toxicity at concentrations even 100 times higher than the
EC50 (0.5 lM; Sakamoto et al., 2000). The absence of cytotoxic effects when the receptors are completely activated with clofibrate and pioglitazone suggests that the toxicities of WY-14643 and ciglitazone are mediated through PPARindependent mechanisms. Similar findings have been published while reviewing our work (Soller et al., in press). The authors show that troglitazone induces apoptosis in Jurkat T cells, while ciglitazone induces necrosis. Both compounds inhibit the mitochondrial respiratory complex I, but ciglitazone also inhibits complex II and depletes the ATP cell levels. We have also observed that the intensity of cytotoxity of PPAR agonists depends on the cell line assayed: mouse cortical tubular cells are the in vitro model with the narrowest threshold (high Hill coefficient in dose–response curves), while opossum kidney cells are the most sensitive cell line. These conclusions arise from the dose–response
1074
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
*
0.2
0.1
0.0
100 μM
Control
*
0.5
* Absorbance LDH
Absorbance LDH
0.3
0.4 0.3 0.2
*
0.1 0.0
200 μM
Absorbance Caspase-3
Absorbance Caspase-3
0.06 0.05 0.04 0.03 0.02 0.01 100 μM
Control
10 μM
Ciglitazone
0.07
0.00
5 μM
Control
WY14643 0.05 0.04 0.03 0.02 0.01 0.00
200 μM
5 μM
Control
WY14643
10 μM
Ciglitazone
Fig. 7. Assay of caspase-3 activity in LLC-PK1 cells incubated for 3 h with the indicated PPAR agonists. As shown, LDH activity increases with the concentration of drugs (top panels). However, capase-3 activity remains unchanged.
μM WY14643 M
0
25
50
100
μM ciglitazone 200
M
0
5
10
20
M
Fig. 8. Genomic DNA analysis. LLC-PK1 cells were incubated for 16 h with increasing concentrations of WY14643 and ciglitazone. Genomic DNA was extracted and run in horizontal electrophoresis as explained. Only 100 and 200 lM WY14643 and 10 and 20 lM ciglitazone produces degradation of genomic DNA. However, no laddering of 200 bp bands and onwards can be observed in the electrophoresis.
assays of LDH activity as a function of drug concentration (Fig. 1). At this point we do not know the toxic mechanisms that induce cell death in our experimental conditions, but one major conclusion is that these effects should not be related to PPAR activity, and therefore they should be independent of the affinity binding of the drugs. Moreover, the fact that the same drugs are toxic in a similar way to all three epithelial cell lines of proximal tubular origin suggests that a common toxic mechanism can take place in the kidney in vivo. Since apoptosis does not follow a universal pattern of events and since many of the typical characteristics of apoptosis are not always expressed, we have characterized the cell death in different ways. Using either morphological
and biochemical assays, as well as an analysis of genomic DNA, we have not found evidence of apoptotic cell death under any of our experimental conditions. However, LLCPK1 cells have the ability to undergo apoptosis (Gennari et al., 2003; Liu and Baliga, 2005). In addition to previous reports, we have also checked this ability by treating the cells with 50 lM CdCl2 for 16 h (see Section 2; data not shown). In conclusion, the cytotoxic effects of WY14643 and ciglitazone should consist of direct insults to the cells that overwhelm the homeostatic mechanisms and end in cell death. Therefore, further studies are needed to understand the likely mechanisms of cell death, and ongoing work should clarify this point. With respect to the induced apoptosis by PPAR agonists, several mechanisms have
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
been described, either through the PPAR receptors or through PPAR-independent mechanisms. The pathways include activation of the growth arrest and DNA damage-inducible gene 45 (GADD45) and p53 (Bruemmer et al., 2003; Okuda et al., 2000), upregulation of Bad and Bax (Zander et al., 2002) and p27 (Liu et al., 2004), or nuclear recruitment of phospho-Smad2 (Redondo et al., 2005). The effects of WY14643 are, however, contradictory. For example, this compound reduces both the apoptotic and necrotic cell death induced by cisplatin in mouse kidney, with a concomitant inhibition of proapoptotic renal endonuclease G (Li et al., 2004). In brain cortical cells, however, WY14643 can switch the cell death induced by cyanide from apoptosis to necrosis by increasing the expression of uncoupling protein-2 and mitochondrial dysfunction (Li et al., 2006). Our results add new information to the possible toxic effect of the drugs. Renal side-effects were initially described for both fibrates and TZDs, and included dysuria, proteinuria and increases in plasma creatinine and urea (Lipscombe et al., 2001). These effects were eliminated with the design of new drugs, as in the case of ciglitazone that exhibited a significant liver toxicity and was substituted by pioglitazone. In agreement to our work, commercial drugs available today (eg. clofibrate, pioglitazone or rosiglitazone) show, fortunately, little or no renal toxicity. Our findings could reflect the side effects reported initially for these drugs, and highlight the importance of the characterization of side effects when designing pharmacological PPAR agonists. In conclusion, fibrates and TZDs are drugs extensively used to control metabolic disorders, with important consequences to the kidneys, especially in patients with renal disease. The fact that some PPAR agonists are able to induce a direct necrotic cell death in epithelial cells of proximal tubular origin adds additional risk to the treatment of such disorders. Therefore, control of the dangerous side effects of PPAR agonists should be carefully determined. Acknowledgements This work was supported by Grant BFI2003-06645 from the Spanish Ministry of Education and Science (to V.S.). The authors would like to thank Dr. Moshe Levi (UCHSC, Denver, CO) for critical reading and suggestions to the paper. References Andreoli, S.P., McAteer, J.A., Seifert, S.A., Kempson, S.A., 1993. Oxidant-induced alterations in glucose and phosphate transport in LLC-PK1 cells: mechanisms of injury. American Journal of Physiology 265, F377–F384. Arici, M., Chana, R., Lewington, A., Brown, J., Brunskill, N.J., 2003. Stimulation of proximal tubular cell apoptosis by albumin-bound fatty acids mediated by peroxisome proliferator activated receptor-c. Journal of the American Society of Nephrology 14, 17–27.
1075
Barter, P.J., Rye, K.A., 2006. Cardioprotective properties of fibrates: which fibrate, which patients, what mechanism?. Circulation 113 1553– 1555. Berger, J., Moller, D.E., 2002. The mechanisms of action of PPARs. Annual Reviews in Medicine 53, 409–435. Bruemmer, D., Yin, F., Liu, J., Berger, J.P., Sakai, T., Blaschke, F., Fleck, E., Van Herle, A.J., Forman, B.M., Law, R.E., 2003. Regulation of the growth arrest and DNA damage-inducible gene 45 (GADD45) by peroxisome proliferator-activated receptor c in vascular smooth muscle cells. Circulation Research 93, e38–e47. Brunton, L., Lazo, J., Parker, K., 2005. Goodman & Gilman’s The Pharmacological Basis of Therapeutics, 11th ed. McGraw-Hill Professional, New York. Chung, B.H., Lim, S.W., Ahn, K.O., Sugawara, A., Ito, S., Choi, B.S., Kim, Y.S., Bang, B.K., Yang, C.W., 2005. Protective effect of peroxisome proliferator activated receptor gamma agonists on diabetic and non-diabetic renal diseases. Nephrology (Carlton) 10 (Suppl), S40–S43. Gennari, A., Cortese, E., Boveri, M., Casado, J., Prieto, P., 2003. Sensitive endpoints for evaluating cadmium-induced acute toxicity in LLC-PK1 cells. Toxicology 1, 211–220. Grommes, C., Landreth, G.E., Heneka, M.T., 2004. Antineoplastic effects of peroxisome proliferator-activated receptor c agonists. The Lancet Oncology 5, 419–429. Guan, Y., Breyer, M.D., 2001. Peroxisome proliferator-activated receptors (PPARs): novel therapeutic targets in renal disease. Kidney International 60, 14–30. Izzedine, H., Launay-Vacher, V., Buhaescu, I., Heurtier, A., Baumelou, A., Deray, G., 2005. PPAR-gamma-agonists’ renal effects. Minerva Urologica e Nefrologica 57, 247–260. Klaunig, J.E., Babich, M.A., Baetcke, K.P., Cook, J.C., Corton, J.C., David, R.M., DeLuca, J.G., Lai, D.Y., McKee, R.H., Peters, J.M., Roberts, R.A., Fenner-Crisp, P.A., 2003. PPARalpha agonist-induced rodent tumors: modes of action and human relevance. Critical Reviews in Toxicology 33, 655–780. Lanaspa, M.A., Giral, H., Breusegem, S.Y., Halaihel, N., Baile, G., Catala´n, J., Carrodeguas, J.A., Barry, N.P., Levi, M., Sorribas, V., 2007. Interaction of MAP17 with NHERF3/4 induces translocation of the renal Na/Pi IIa transporter to the trans-Golgi. American Journal of Physiology – Renal Physiology 292, F230–F242. Li, S., Basnakian, A., Bhatt, R., Megyesi, J., Gokden, N., Shah, S.V., Portilla, D., 2004. PPAR-alpha ligand ameliorates acute renal failure by reducing cisplatin-induced increased expression of renal endonuclease G. American Journal of Physiology – Renal Physiology 287, F990–F998. Li, L., Prabhakaran, K., Zhang, X., Borowitz, J.L., Isom, G.E., 2006. PPARalpha-mediated upregulation of uncoupling protein-2 switches cyanide-induced apoptosis to necrosis in primary cortical cells. Toxicological Sciences 93, 136–145. Lipscombe, J., Lewis, G.F., Cattran, D., Bargman, J.M., 2001. Deterioration in renal function associated with fibrate therapy. Clinical Nephrology 55, 39–44. Liu, D.-C., Zang, C.-B., Liu, H.-Y., Possinger, K., Fan, S.-G., Elstner, E., 2004. A novel PPAR a/c dual agonist inhibits cell growth and induces apoptosis in human glioblastoma T98G cells. Acta Pharmacologica Sinica 25, 1312–1319. Liu, H., Baliga, R., 2005. Endoplasmic reticulum stress-associated caspase 12 mediates cisplatin-induced LLC-PK1 cell apoptosis. Journal of the American Society of Nephrology 16, 1985–1992. Martin, S.J., Reutelingsperger, C.P., McGahon, A.J., Rader, J.A., van Schie, R.C., LaFace, D.M., Green, D.R., 1995. Early redistribution of plasma membrane phophatidylserine is a general feature of apoptosis regardless of the initiating stimulus: inhibition by overexpression of Bcl-2 and Abl. Journal of Experimental Medicine 182, 1545–1556. McGahon, A.J., Martin, S.J., Bissonnette, R.P., Mahboubi, A., Shi, Y., Mogil, R.J., Nishioka, W.K., Green, D.R., 1995. The end of the (cell) line: methods for the study of apoptosis in vitro. In: Schwartz, L.M.,
1076
H. Giral et al. / Toxicology in Vitro 21 (2007) 1066–1076
Orborne, B.A. (Eds.), Cell Death, Methods in Cell Biology, vol. 46. Academic Press, San Diego, pp. 153–185. Musi, N., Goodyear, L.J., 2006. Insulin resistance and improvements in signal transduction. Endocrine 29, 73–80. Okuda, T., Nakamura, M., Takata, Y., Watanabe, S., Kitami, Y., Hiwada, K., 2000. European Journal of Pharmacology 407, 227–235. Parameswaran, N., Hall, C.S., Bomberger, J.M., Sparks, H.V., Jump, D.B., Spielman, W.S., 2003. Negative growth effects of ciglitazone on kidney interstitial fibroblasts: role of PPAR-c. Kidney & Blood Pressure Research 26, 2–9. Peraza, M.A., Burdick, A.D., Marin, H.E., Gonza´lez, F.J., Peters, J.M., 2006. The toxicology of ligands for peroxisome prolifertor-activated receptors (PPAR). Toxicological Sciences 90, 269–295. Redondo, S., Ruiz, E., Santos-Gallego, C.G., Padilla, E., Tejerina, T., 2005. Pioglitazone induces vascular smooth muscle cell apoptosis through a peroxisome proliferator-activated receptor-c, transforming growth factor-ß1, and a Smad2-dependent mechanism. Diabetes 54, 811–817. Roberts, R.A., Chevalier, S., Hasmall, S.C., James, N.H., Cosulich, S.C., Macdonald, N., 2002. PPARa and the regulation of cell division and apoptosis. Toxicology, 167–170. Robillard, R., Fontaine, C., Chinetti, G., Fruchart, J.C., Staels, B., 2005. Fibrates. Handbook of Experimental Pharmacology 170, 389–406. Sakamoto, J., Kimura, H., Moriyama, S., Odaka, H., Momose, Y., Sugiyama, Y., Sawada, H., 2000. Activation of Human Peroxisome Proliferator-Activated Receptor (PPAR) Subtypes by Pioglitazone. Biochemical and Biophysical Research Communications 278, 499– 841. Sato, K., Sugawara, A., Kudo, M., Uruno, A., Ito, S., Takeuchi, K., 2004. Expression of peroxisome proliferator-activated receptor isoform proteins in the rat kidney. Hypertension Research 27, 417–425. Soller, M., Drose, S., Brandt, U., Brune, B., von Knethen, A., in press. Mechanism of thiazolidinedione-dependent cell death in Jurkat T cells. Molecular Pharmacology. doi:10.1124/mol.107.034371. Sorribas, V., Markovich, D., Verri, T., Biber, J., Murer, H., 1995. Thyroid hormone stimulation of Na/Pi-cotransport in opossum kidney cells. Pflugers Archives 431, 266–271. Staels, B., Fruchart, J.C., 2005. Therapeutic roles of peroxisome proliferator-activated receptor agonists. Diabetes 54, 2460–2470. Theocharis, S., Margeli, A., Vielh, P., Kouraklis, G., 2004. Peroxisome proliferator-activated receptor-c ligands as cell-cycle modulators. Cancer Treatment Reviews 30, 545–554.
Tsuchiya, T., Shimizu, H., Shimomura, K., Mori, M., 2003. Troglitazone inhibits isolated cell proliferation, and induces apoptosis in isolated rat mesangial cells. American Journal of Nephrology 23, 222–228. Welbourne, T., Friday, E., Fowler, R., Turturro, F., Nissim, I., 2003. Troglitazone acts by PPARc and PPARc-independent pathways on LLCPK1-F + acid-base metabolism. American Journal of Physiology – Renal Physiology 286, F100–F110. Weissgarten, J., Berman, S., Efrati, S., Rapoport, M., Averbukh, Z., Feldman, L., 2006. Apoptosis and proliferation of cultured mesangial cells isolated from kidneys of rosiglitazone-treated pregnant diabetic rats. Nephrology Dialysis Transplantation 21, 1198–1204. Willson, T.M., Cobb, J.E., Cowan, D.J., Wiethe, R.W., Correa, I.D., Prakash, S.R., Beck, K.D., Moore, L.B., Kliewer, S.A., Lehmann, J.M., 1996. The structure–activity relationship between peroxisome proliferator-activated receptor agonism and the antihyperglycemic activity of thiazolidinediones. Journal of Medicinal Chemistry 39, 665– 668. Willson, T.M., Brown, P.J., Sternbach, D.D., Henke, B.R., 2000. The PPARs: from orphan receptors to drug discovery. Journal of Medicinal Chemistry? 43, 527–550 Wyllie, A.H., 1980. Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284, 555–556. Zafiriou, S., Stanners, S.R., Saad, S., Polhill, T.S., Poronnik, P., Pollock, C.A., 2005. Pioglitazone inhibits cell growth and reduces matrix production in human kidney fibroblasts. Journal of the American Society of Nephrology 16, 638–645. Zander, T., Kraus, J.A., Grommes, C., Schlegel, U., Feinstein, D., Klockgether, T., Landreth, G., Koenigsknecht, J., Heneka, M.T., 2002. Induction of apoptosis in human and rat glioma by agonists of the nuclear receptor PPARgamma. Journal of Neurochemistry 81, 1052–1060. Zhivotosky, B., Orrenius, S., 2003. Assessment of apoptosis and necrosis by DNA fragmentation and morphological criteria. In: Bonifacino, J.S., Dasso, M., Harford, J.B., Lippincott-Schwartz, J., Yamada, K.M. (Eds.), Current Protocols in Cell Biology. John Wiley, New York. doi:10.1002/0471143030.cb1803s12.