UPMC Univ Paris 06, UMR_S 975 - UMR 7725, Centre de Recherche en Neurosciences, ICM, Therapeutique Experimentale de la Neurodegenerescence, Paris, France
Inserm U 975, Centre de Recherche en Neurosciences, ICM, Therapeutique Experimentale de la Neurodegenerescence, Paris, France
CNRS, UMR 7225, Centre de Recherche en Neurosciences, ICM, Therapeutique Experimentale de la Neurodegenerescence, Paris, France
ICM, Therapeutique Experimentale de la Neurodegenerescence, Paris, France
Department of Neurology, Philipps-University Marburg, Marburg, Germany
Département de Neurologie, Pôle des Maladies du Système Nerveux, Hôpital de la Pitié-Salpêtrière, Paris, France
Address correspondence and reprint requests to Andreas Hartmann, Centre de Recherche de l'Institut du Cerveau et de la Moelle Epinière, UPMC/INSERM UMR_S975, Groupe Hospitalier Pitié-Salpêtrière, 47 Boulevard de l'Hôpital, 75651 Paris Cedex 13, France. E-mail: firstname.lastname@example.org
The uricosuric agent probenecid is co-administered with the dopaminergic neurotoxin MPTP to produce a chronic mouse model of Parkinson's disease. It has been proposed that probenecid serves to elevate concentrations of MPTP in the brain by reducing renal elimination of the toxin. However, this mechanism has never been formally demonstrated to date and is questioned by our previous data showing that intracerebral concentrations of MPP+, the active metabolite of MPTP, are not modified by co-injection of probenecid. In this study, we investigated the potentiating effects of probenecid in vivo and in vitro arguing against the possibility of altered metabolism or impaired renal elimination of MPTP. We find that probenecid (i) is toxic in itself to several neuronal populations apart from dopaminergic neurons, and (ii) that it also potentiates the effects of other mitochondrial complex I inhibitors such as rotenone. On a mechanistic level, we show that probenecid is able to lower intracellular ATP concentrations and that its toxic action on neuronal cells can be reversed by extracellular ATP. Probenecid can potentiate the effect of mitochondrial toxins due to its impact on ATP metabolism and could therefore be useful to model atypical parkinsonian syndromes.
The uricosuric agent probenecid is used in the treatment of gout. It operates by inhibiting the reabsorption of uric acid by the urate transporter 1 (URAT1), a member of the renal organic anion transporter family (Hediger et al. 2005). As an inhibitor of organic anion transporter 1, probenecid also reduces renal elimination and thus enhances plasma concentrations of other pharmacological agents such as penicillin and other antibiotics (Beyer et al. 1951; Masereeuw et al. 2000). Regarding the CNS, probenecid crosses the blood–brain barrier by a low-affinity system, probably by diffusion (Yuwiler 1982).
In the field of Parkinson's disease (PD) research, probenecid is co-administered with MPTP to produce the so-called ‘chronic’ mouse model of the disease. In this experimental paradigm, about 40–45% of the dopaminergic (DA) neurons from the substantia nigra pars compacta are lost by 3 weeks post-treatment compared to 25% in the subchronic model without probenecid (Petroske et al. 2001; Meredith et al. 2008a,b). The loss of dopaminergic neurons remains substantial 6 months after intoxication with only a minimal recovery (Petroske et al. 2001; Meredith et al. 2008a,b) in contrast to the subchronic and acute MPTP models where substantial recovery does occur (Petroske et al. 2001; Meredith et al. 2008a,b). However, the mechanism by which probenecid enhances the effects of MPTP is not fully understood.
After systemic administration, MPTP crosses the blood–brain barrier and reaches the CNS. There it is metabolized into its active metabolite, MPP+. MPP+ is taken up by dopaminergic neurons via the dopamine transporter (DAT) and inhibits complex I of the mitochondrial respiratory chain. This leads to subsequent energy failure with ATP depletion (Schmidt and Ferger 2001). It has been demonstrated that probenecid strongly reduces the clearance of MPTP and its metabolites by the kidney (Lau et al. 1988, 1990). This has led to the speculation that probenecid might operate by enhancing the concentration of MPTP/MPP+ available to brain cells (Lau et al. 1990; Petroske et al. 2001). However, this has never been formally demonstrated to date.
Previously, we failed to detect enhanced MPP+ levels in the brain after a co-administration of MPTP and probenecid (Alvarez-Fischer et al. 2008) and thus questioned the ‘renal’ hypothesis of the probenecid action. Indeed, some studies reported unspecifically altered oxygen tissue consumption after probenecid exposition (Choi and Kim 1992) and an interference with cellular oxidative metabolism (Masereeuw et al. 2000). Moreover, probenecid has been reported to alter extracellular concentrations of lactate (MacMillan 1987; Kuhr et al. 1988; Korf 1996), a crucial aerobic energy substrate that enables neurons to endure activation (Schurr et al. 1999). Finally, there is a broad literature on the inhibiting properties of probenecid on the multidrug-resistance transporter (Darby et al. 2003) and pannexin 1 hemichannels (Silverman et al. 2008, 2009; Ma et al. 2009), two proteins that have been associated with ATP release in the CNS (Darby et al. 2003; Silverman et al. 2008, 2009; Ma et al. 2009).
The aim of this study was, therefore, to explore how probenecid contributes to the neural toxicity of MPTP. To this end, we measured striatal MPP+ levels with and without co-administration of probenecid. Furthermore, we used two culture systems where dopaminergic neuronal death is caused by a chronic exposure to MPP+, the active metabolite of MPTP, or to rotenone, a plant toxin that operates like MPP+ by blocking mitochondrial complex I activity. Our data suggest that probenecid can potentiate the effect of the two mitochondrial toxins directly because of its impact on cellular energy metabolism.
Materials and methods
The animals used in this study were handled in accordance with the ‘Guide for the Care and Use of Laboratory Animals’ (National Research Council 1996), the European Communities Council Directive 86/609/EEC, and the guidelines of the local institutional committees.
Animals were treated as described by Petroske et al. (2001). In brief, in the subchronic group, mice were treated with a conventional MPTP paradigm. These mice received MPTP hydrochloride (25 mg/kg in saline, s.c., Sigma Chemical Co. (St. Louis, MO, USA) or saline once a day for 5 consecutive days. In the chronic group, mice received a total of 10 doses of MPTP hydrochloride (25 mg/kg in saline, s.c.) in combination with probenecid (250 mg/kg, Sigma Chemical). The 10 doses were administered on a 5-week schedule, such that injections were given at an interval of 3.5 days between consecutive doses. The control animals were injected with saline under the same regimen. Animals were killed 7 days after the last treatment. In both paradigms, six to eight animals per group were analysed.
For primary mesencephalic cell cultures, embryos were removed at embryonic day 14.5 from pregnant Wistar rats (Elevage Janvier, Le Genest St Isle, France). The ventral mesencephalon was dissected as previously described (Douhou et al. 2001). The cells were cultured in 500 μL N5 medium (Kawamoto and Barrett 1986) supplemented with 5% horse serum and 0.5% foetal calf serum (FCS). During the first 2 days in vitro, the concentration of FCS was raised to 2.5%. 2 μM AraC (Sigma Chemical) to block astrocyte growth and 2 μM MK801 (Sigma Chemical) to prevent non-specific excitotoxic stress (Michel et al. 1997) were added to the medium. On average, one well contained 2263 ± 138 tyrosine hydroxylase (TH)-positive and 311719 ± 22745 NeuN-positive cells per well, respectively.
SH-SY5Y human neuroblastoma cells were used as a dopaminergic cell model (Tanaka et al. 1995). They were maintained in a humidified incubator at 37°C in 5% CO2 using Dulbecco's modified Eagle's medium (Gibco, Saint-Aubin, France) supplemented with 15% FCS; (Gibco), 1 mM l-glutamine, 100 U/mL penicillin and 100 μg/mL of streptomycin.
Immunohistochemistry and cell counting in vivo
Free-floating coronal sections of the whole midbrain were immunostained and analysed as previously described (Alvarez-Fischer et al. 2008) with rabbit anti-TH (1/1000; Pel Freez, Paris, France). TH+ neurons were revealed with the chromogen diaminobenzidine. TH+ cell bodies were quantified stereologically on regularly spaced sections covering the whole mesencephalon from the rostral pole of the substantia nigra to the locus coeruleus using the VisioScan stereology tool (Explora Nova VisioScan T4.18 software, La Rochelle, France). The substantia nigra was identified according to established anatomical landmarks (Franklin and Paxinos 1997). The investigator performing the quantification was blinded to the treatment groups during analysis. The total number of nigral TH+ neurons varied from 12 800–13 400 in saline-treated mice.
HPLC detection of striatal dopamine and MPP+
HPLC was performed as previously described (Alvarez-Fischer et al. 2008). In brief, for striatal dopamine quantification, tissue was placed into 250 μL 0.1 N perchloric acid containing 0.05% disodium EDTA and 0.05% sodium metabisulfite and further processed for HPLC analysis using electrochemical detection. The potential for electrochemical detection was set at + 0.65 V. Striatal levels of dopamine ranged from 6.2–9.7 ng/mg wet tissue in control mice.
For striatal MPP+ measurements, mice were killed 0.5, 2, 4, and 6 h after a single i.p. injection of 25 mg/kg MPTP with or without a subsequent i.p. injection of 250 mg/kg probenecid. Tissue pieces were placed into 500 μL 0.1 N HClO4 and further processed for HPLC analysis using a UV detector (wavelength 295 nm).
Treatment of cell cultures
Probenecid first solubilized in 2 N NaOH was further diluted to 50 mg/mL using 0.1 M Tris buffer (pH 8) and then adjusted to a final pH of 7.3 with 2 N HCl. Probenecid was then added at the indicated concentrations 2 h before the toxin unless indicated otherwise. All toxins were dissolved freshly before treatment. Rotenone (Sigma Chemical) was dissolved in dimethylsulfoxide. The final concentration of dimethylsulfoxide in contact with the cells never exceeded 0.01%. Treatments with rotenone (30 nM) lasted 24 h. MPP+ (Sigma Chemical) was dissolved in water and added at a concentration of 0.8 μM in primary mesencephalic cell cultures and of 1 mM in SH-SY5Y cells. Treatments were performed at DIV 4 and DIV 5 and lasted for 48 h.
Immunocytochemistry and cell counting in vitro
Immunocytochemistry was performed as previously described (Douhou et al. 2001) with a monoclonal anti-TH antibody (MAB-5280; Chemicon, Temecula, CA, USA) diluted 1 : 500 and a secondary antibody (1 : 500; anti-mouse Sigma Chemical). All neuronal cells, regardless of their neurotransmitter phenotype, were identified by staining of neuronal nuclei with a monoclonal biotin conjugated antibody anti-neuronal nuclei (NeuN) (MAB-377B, Chemicon) diluted 1 : 100 and a secondary antibody (biotin conjugated, Invitrogen, Carlsbad, CA, USA, 1 : 500 in phosphate-buffered saline and 5% horse serum).
Viability tests of SHSY-5Y cells were performed as described before (Cernaianu et al. 2008) with 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide, Sigma Chemical at a final concentration of 0.25 mg/mL. Quantification was performed using a microplate reader (MultiskanReader, ThermoLabSystems, Egelsbach, Germany). Results were expressed as percentage of the absorbance in vehicle-treated control culture wells.
Uptake of [3H]-dopamine and [3H]-deoxyglucose
The activity of the membrane DAT was evaluated by the active up-take of [3H]-dopamine as previously described (Douhou et al. 2001). For further details please see the Appendix S1.
The quantification of glucose up-take was quantified as previously as described (Lannuzel et al. 2003). For further details please see the Appendix S1.
Measurement of intracellular ATP levels
For determining intracellular ATP levels the Vialight HS kit from Bio Whittaker (Verviers, Belgium) was used following the manufacturer's instructions. Briefly, after washing the cells twice with cold phosphate-buffered saline, cells were scraped off each culture well with 100 μL of distilled water. The samples were processed immediately. As the kit is based on catalyzation of enzymatic light formation, emitted light was quantified using a tube luminometer (TD20/20; Turner Designs, Sunnyvale, CA, USA).
Measurement of glycolysis
Glycolysis was measured using the BioVision Lactate Assay Kit (BioVision, Mountain View, CA, USA) and following the manufacturer's instruction guide. The kit is based on the enzymatic colour producing reaction. For detailed information, please see the Appendix S1.
Measurement of complex I activity
Complex I activity was measured using the MitoProfile™ Dipstick Assay Kit (MitoScience, Eugene, OR, USA) and following the manufacturer's instruction guide. For further details please see the Appendix S1.
Measurement of cell respiration
Respiration was analysed using Seahorse technology on primary mesencephalic cell cultures as described (Pesta and Gnaiger 2012). In brief, a Seahorse 96-well plate was seeded with approximately 5000, 10 000 and 15 000 mesencephalic cells/per well, each seeding number being used in 32 consecutive wells. After 5 days of culture, half of the wells were treated with 1 mM probenecid during 3 h and the medium was then withdrawn and replaced with Seahorse medium with 1 g/L glucose and 1 mM pyruvate. Respiration and extracellular pH were then measured four times in three successive series. After determining the basal respiration in the cells, oligomycin (2 μM), CCCP (0.5 or 1.0 μM) and potassium cyanide (2 mM) were sequentially added to the media and the oxygen consumption rates for each condition were quantified. Thus, the respiration compensating for the mitochondrial proton leak, maximal respiration and non-mitochondrial respiration were quantified. The impact of probenecid treatment was evaluated on respiration (OCR), extracellular pH (ECAR) and their ratio (OCR/ECAR) expressed as% of controls (equivalent wells with respect to number of seeded cells and additions). For statistical analyses, the mean of the four measurements of each step was considered an independent result and the results from two independent plates with different cell culture preparation were grouped.
Data are expressed as the percentage of corresponding control values. Each data point represents a mean ± SEM of at least three independent experiments. Except otherwise indicated, multiple comparisons against a single reference group were performed by one-way anova followed by the Dunnett's test. When all pairwise comparisons were carried out, one-way anova was followed by the post hoc Tukey test.
Probenecid does not influence striatal MPP+ concentration in vivo
First, we observed a predicted linear decrease 120, 240 and 360 min after a single injection of 25 mg/kg MPTP hydrochloride. To meet the ascending part of the curve we also added a very early time point and killed the animals 30 min after the administration of MPTP hydrochloride. At no time point, striatal MPP+ concentrations were higher when probenecid (250 mg/kg) was co-injected compared with MPP+ concentration following injection of MPTP alone (Fig 1a). Also, the area under the curve of the mean values between both paradigms did not differ (2433 without vs. 2531 with probenecid, that is, less than 4% difference).
Nevertheless, both treatment paradigms yield the predicted range of effectiveness as reported in the original publication by Petroske and colleagues (Petroske et al. 2001). Whereas the subchronic paradigm led to a decrease of striatal DA of 60.5 ± 4.2% and a small but still significant TH+ cell loss of 11.8 ± 6.8%, the co-administration of 25 mg/kg MPTP hydrochloride and probenecid in the chronic model led to a DA depletion of 66.4 ± 1.5% and a TH+ cell loss of 36.1 ± 2.8% 1 week after the end of intoxication (Fig 1b and c). Nota bene that in the subchronic paradigm, 25 mg/kg MPTP hydrochloride was injected compared with 30 mg/kg MPTP in the original published subchronic paradigm (Tatton and Kish 1997). Thus, cell loss is consistent with what had been previously reported with this dosage regimen (Petroske et al. 2001; Alvarez-Fischer et al. 2007).
Probenecid enhances MPP+ toxicity in vitro
Next, we wondered whether probenecid might also enhance MPP+ toxicity in primary mesencephalic cultures and SHSY-5Y cells, a dopaminergic neuroblastoma cell line. MPP+ led to a significant decrease of TH+ cells of in primary cultures (57.7 ± 3.7%) and in SHSY-5Y cells (22.9 ± 1.0%). The effect was significantly enhanced, both, in primary cultures and in SHSY-5Y cells when probenecid was added at a concentration of 500 μM that alone did not affect DA cell survival (Fig 2a–c). Probenecid was toxic per se at a concentration of 1 mM in primary mesencephalic cultures (Fig 2b) and at 2 mM in the SHSY-5Y cell model (Fig 2c). The MPP+ enhancing effects were comparable whether the incubation time was 2 or 24 h before MPP+ treatments (Fig 2a). The probenecid effect was concentration-dependent (Fig 2b and c).
Probenecid lowers intracellular ATP levels in cultured midbrain neurons
As one of the effects of MPP+ is to impair intracellular ATP production by mitochondria (Schmidt and Ferger 2001), we measured whether intracellular levels of the nucleotide might also be influenced by probenecid treatment. Indeed, probenecid alone lowered ATP levels significantly by about 14.6 ± 3.6%, regardless whether added 2 or 24 h before measurement (Fig 3a). Thus, we used a 2 h pre-incubation time with probenecid in all subsequent experiments. As expected, MPP+ (30 μM) also lowered ATP levels after 2 h of treatment. At this concentration, MPP+ is toxic to all neurons regardless of their phenotype in primary midbrain cultures. The depletion induced by MPP+ alone was further enhanced when probenecid was added 2 h before (Fig 3a). To verify if lower ATP concentrations were because of impaired glucose uptake, we measured glucose accumulation in the presence of probenecid (500 μM). Neither at 500 μM nor at 1 mM probenecid glucose uptake was altered by probenecid pre-treatment (Fig 3b).
We next wondered whether probenecid could influence MPP+ accumulation into neurons by interfering with the DAT, as MPP+ uses this transport system to gain access to the intracellular milieu. However, DAT activity was not significantly influenced by probenecid (500 μM) added 2 h before measurement (Fig 3c).
Probenecid enhances toxicity of other complex I inhibitors in vitro
Having observed that probenecid lowered intracellular ATP concentrations and that these effects did not result from interference with MPTP/MPP+ metabolism or intracellular uptake, we wondered whether probenecid might also enhance the effects of rotenone, a complex I inhibitor that is not selective for midbrain DA neurons (Höglinger et al. 2003). Rotenone (30 nM, 24 h) reduced both DAergic and non-DAergic neuron numbers in primary mesencephalic cultures by 27.0 ± 2.7% and 21.4 ± 4.8%, respectively. Importantly, probenecid also enhanced the toxic effects of rotenone regardless of neuronal phenotype (Fig. 4a and b).
Probenecid impairs neuronal respiration and complex I activity
Global analysis of cellular respiration in primary neuronal cultures after treatment with probenecid (1 mM for 3 h) revealed a reduced respiration compared with vehicle treated cells both in the basal state (−28 ± 10%) as well as after stimulation with 1 μM of the protonophore carbonyl cyanide m-chlorophenyl hydrazine (cccp) (−55 ± 8%) (Table 1, Fig 5a). We further used primary neuronal cultures to verify whether probenecid affects complex I activity. As expected, rotenone treatment (30 nM, 24 h) reduced complex I activity by 31.7 ± 7%. Interestingly, complex I activity was reduced to the same levels (27.9 ± 8.1%) in the presence of 500 μM probenecid (Fig 5c), that is, at a concentration where probenecid does not lead to cell loss. When isolated mitochondria were treated with either rotenone or probenecid instead of living cells, only incubation with rotenone led to a significant decrease of complex I activity by 62.5 ± 3.5% whereas probenecid did not alter complex I activity (94.4 ± 5.6%).
Table 1. Cellular respiration is altered after probenecid treatment
Results are expressed as % of the control cells analysed in the same plate. Means ± SEM are indicated together with the number of independent samples between brackets. Wells with the highest number of cells (15000 cells per well) were analysed in two independent analyses with thus twice more samples. The results that were significant with non-parametric Mann–Whitney test are shown in bold. Base denotes respiration under basal conditions (1 g/L glucose, 1 mM pyruvate), CP1, cccp 1 μM; CP3, cccp 3 μM; OM, oligomycin 0.5 μg/mL.
116 ± 14 (13)
126 ± 23 (16)
44 ± 14 (3)
40 ± 3 (4)
71 ± 10 (4)
82 ± 12 (4)
391 ± 69 (7)
35 ± 160 (8)
104 ± 112 (14)
81 ± 10 (13)
44 ± 15 (3)
57 ± 5 (4)
62 ± 10 (6)
66 ± 10 (4)
83 ± 3 (4)
87 ± 8 (8)
100 ± 22 (7)
95 ± 7 (8)
99 ± 13 (14)
Toxic effects elicited by probenecid are reversed by glucose
Both in primary neuronal cultures and SHSY-5Y cells, glucose attenuated the deleterious effects of MPP+ (Fig 6a and b). In primary neuronal cultures, treatment with 50 mM glucose completely blocked the deleterious effects of high dosages (1 and 2 mM) of probenecid (Fig 6a). When MPP+ treatment was performed in the presence of high glucose, deleterious effects of the toxin were attenuated but not completely blocked. However, the potentiating action of probenecid (500 μM) on MPP+ was not observed under these conditions (Fig 6a). SHSY-5Y cells were also treated with increasing dosages of probenecid. At each concentration tested, glucose treatment (50 mM) attenuated the dose-dependent deleterious effects of probenecid (Fig 6b). Interestingly, the toxicity-enhancing effects of probenecid which were most prominent at a concentration of 2 mM MPP+ were totally abolished by 50 mM glucose (Fig 6c).
Finally, as it is well known that inhibition of glycolysis enhances MPP+ toxicity and results in decreased ATP levels, we wanted to exclude that the glycolytic pathway was affected by probenecid. Therefore, we treated SHSY-5Y cells either with different dosages of iodoacetic acid as a standard glycolysis inhibitor or with probenecid and measured the lactate concentrations as end product of the anaerobic glycolysis pathway. As expected, iodoacetic acid did decrease lactate concentrations in a dose-dependent manner. Yet, lactate concentrations were not influenced by probenecid except at very high concentrations (4 mM) (Fig 6d).
Extracellular ATP reverses the toxic effects of probenecid
As it is well established that probenecid inhibits the multidrug-resistance transporter (Darby et al. 2003) and pannexin 1 channels (Silverman et al. 2008, 2009; Ma et al. 2009) that are both involved in ATP release in the CNS, we tested whether the toxic effects of probenecid were reversed by the addition of non-hydrolysable ATP analogue, γS-ATP. At a concentration of 8 μM that was without a protective effect on the survival of TH+ neurons after MPP+ treatment, we found that γS-ATP totally prevented an additional toxic effect induced by probenecid. We, therefore, conclude that the deleterious effects of probenecid are, at least in part, mediated by a reduction of extracellular ATP (Fig 7).
In this study, we demonstrate that striatal MPP+ levels are not altered by co-administration of probenecid as was originally proposed (Lau et al. 1988, 1990) although it yielded the predicted effectiveness on nigro-striatal damage. We show that the toxic effects of probenecid are also detectable in vitro which argues against an effect of probenecid on renal MPTP/MPP+ elimination and alterations in the conversion of MPTP to MPP+. We show that probenecid is toxic in itself and that its toxicity, with or without MPP+, is dose-dependent. This is true for primary neuronal cultures containing glia although highly enriched in neurons [they contain 5% glial cells (Michel et al. 1999)] and for SH-SY5Y human neuroblastoma cells, suggesting that toxic effects of probenecid occur independently of glial cells.
We further demonstrate that probenecid does not alter DAT function and conclude that the effects of probenecid are not because of enhanced intracellular levels of MPP+ within dopaminergic neurons. The findings obtained with rotenone also argue against an effect restricted to MPTP or to its metabolite MPP+. Consequently, the toxicity-enhancing effects of probenecid are neither restricted to a specific toxin nor to a specific neuronal cell type.
Our data also show that probenecid lowers intracellular concentrations of ATP that was additive to that of MPTP/MPP+. Indeed, Masereeuw and co-workers (Masereeuw et al. 2000) demonstrated that probenecid impairs mitochondrial oxidative phosphorylation through an uncoupling effect in renal tubular cells (Masereeuw et al. 2000). We also demonstrated in vitro a decrease in the respiration of primary neuronal cultures, both in the basal resting state and under stimulation by the protonophore cccp. This reduction in the respiration rate is expected to induce a parallel decrease of ATP production. However, in our study, it was not accompanied with uncoupling as proposed by Masereeuw and co-workers (Masereeuw et al. 2000). The presence of a complex I defect shown with the dipstick assay fitted with the reduced respiration observed in intact cells respiring on glucose and pyruvate and, thus, on electrons essentially entering the respiratory chain at the complex I level. By treating isolated mitochondria with probenecid we showed that probenecid does not operate like MPP+ or rotenone by direct blockade of mitochondrial complex I. In fact, probenecid may rather operate up-stream of complex I inhibition. However, with our set of experiments we could not detect inhibition of the glycolytic pathway as probenecid neither reduced lactate production nor impaired neuronal glucose uptake.
Supporting the view that two distinct mechanisms were involved in the effects of probenecid and MPP+, we observed that reductions in complex I activity of similar intensities had very different impacts on neuronal survival whether they resulted from probenecid or MPP+ treatments. However, we could not address the presence of other cellular targets for probenecid, including alteration of substrate oxidation downstream of pyruvate, such as pyruvate transport, pyruvate dehydrogenase or the citric acid cycle, nor could we investigate increased ATP lysis, which would add to the decreased cellular ATP levels. Such mechanisms would be compatible with the observation that high concentrations of glucose were sufficient to restore neuronal survival in the presence of toxic amounts of probenecid. Indeed, glucose might simply operate by restoring the glycolytic flux upstream of complex I and the action of probenecid.
In addition, we show that the deleterious effects of probenecid on cell survival are reversed by extracellular ATP. Extracellular ATP can operate as a potent signalling molecule (Apolloni et al. 2009; Tu and Wang 2009) and important physiological functions have been ascribed to this nucleotide in the CNS (Abbracchio et al. 2009). In particular, ATP can operate as a neurotransmitter or a trophic molecule through the activation of its cognate receptors, namely purinergic type 2 receptors (Michel et al. 2007; Tu and Wang 2009). Of interest, extracellular ATP has been previously reported to prevent degeneration in a model system of spontaneous dopaminergic cell death in midbrain cultures (Michel et al. 2007; Tu and Wang 2009). Furthermore, Sato and colleagues (Sato et al. 2006) have shown that Parkin, a protein that is mutated in familial forms of PD, is required to generate currents through purinergic type 2X receptors, a subclass of ionotropic ATP receptors (Sato et al. 2006).
Probenecid is not known to interfere with calcium-dependent exocytosis which is the principal mechanism that controls ATP release in neuronal cells (Abbracchio et al. 2009). Yet, pannexin 1 hemichannels and the multidrug resistance protein that are also susceptible to mediate ATP release (Darby et al. 2003; Silverman et al. 2008, 2009; Ma et al. 2009; Ransford et al. 2009) are inhibited by probenecid (Darby et al. 2003; Dallas et al. 2006; Silverman et al. 2008, 2009; Ma et al. 2009; Ransford et al. 2009). Thus, we speculate that the deleterious mechanism of probenecid resulted not only from a reduced production of intracellular ATP because of impairment of the mitochondrial function but also from impaired neuronal release of ATP by a mechanism that remains to be established.
In conclusion, our results suggest that probenecid does not potentiate MPTP toxicity as originally proposed by increasing MPTP/MPP+ accumulation and retention within the brain but most likely by affecting ATP production and release. As the effects of probenecid are neither toxin-specific nor limited to a specific neuronal phenotype, this compound may be useful in other toxin-based animal models of PD or related disorders, such as rotenone/annonacin to mimic atypical parkinsonian syndromes such as progressive supranuclear palsy or multiple system atrophy in rodents (Höglinger et al. 2003, 2005; Lannuzel et al. 2003; Champy et al. 2004; Escobar-Khondiker et al. 2007).
This study was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM) and Université Pierre et Marie Curie (UPMC), Univ. Paris 6. D. Alvarez-Fischer was supported by the Michael J. Fox Foundation (MJFF) and a research grant of the University Medical Center Giessen and Marburg (UKGM). C. Noelker was supported by the Deutsche Forschungsgemeinschaft (DFG). The authors have no conflicts of interest to declare.