[1] 11w Inhibitory effects of pesticides on proteasome activity: Implication in Parkinson's disease

[1] 157w Epidemiological studies have suggested a correlation of pesticides and Parkinson's disease (PD) while genetic and biochemical studies have implicated the ubiquitin -proteasome system (UPS) in the pathogenesis of PD. In the present studies, we tested the hypothesis that pesticide exposure increases the risk of developing PD by inhibiting the UPS. The effects of pesticides on proteasome activity were examined in SK-N-MC neuroblastoma cells overexpressing a GFP-conjugated proteasome degradation signal, GFP u . Six out of 25 representative pesticides, including rotenone, ziram, diethyldithiocarbamate, endosulfan, benomyl, and dieldrin, showed inhibitory effects on proteasome activities at low concentrations (10 nM to 10 MM). Unlike proteasome inhibitors, they did not inhibit 20 S proteasome activities in cell lysates. Except for rotenone, the other five pesticides did not induce significantly cellular oxidative stress. The cytotoxic effects of these pesticides were closely correlated with proteasome inhibition. Our results suggest proteasome inhibition as a potential mechanism for the epidemiological association of pesticides and PD.

[1] 135w Parkinson's disease (PD) is a common neurodegenerative disorder, characterized by relatively selective degeneration of dopaminergic neurons in the substantia nigra. Epidemiological studies indicate that pesticides are the leading candidates of environmental toxins that may contribute to the pathogenesis of PD (Baldi et al., 2003;Betarbet et al., 2000;Di Monte, 2003;Di Monte et al., 2002;Gorell et al., 1998;Jimenez-Jimenez et al., 1992;Semchuk et al., 1992;Tanner et al., 1989). Rotenone, a common pesticide and an inhibitor of mitochondrial complex I, has been used to develop a PD model, with pathological features of degeneration of dopaminergic neurons and formation of cytoplasmic inclusions (Betarbet et al., 2000). Other pesticides, such as paraquat, dieldrin, and maneb, have been reported to cause degeneration of dopaminergic neurons (McCormack et al., 2002;Meco et al., 1994;Sanchez-Ramos et al., 1998;Uversky, 2004). The mechanisms, however, remain largely unknown.

[2] 276w Although the majority of PD patients have sporadic onset and do not have a familial history, genetic findings have provided important insights into pathogenic mechanisms of PD (Allam et al., 2005;Greenamyre et al., 2003;Tsang and Soong, 2003). Many forms of familial PD cases have been found to be caused by mutations in components related to the ubiquitin -proteasome system (UPS), highlighting the potential importance of UPS in PD (Greenamyre and Hastings, 2004;McNaught and Olanow, 2003). For example, mutations in parkin, an E3 ubiquitin ligase, have been found in juvenile-onset Parkinsonism (Kitada et al., 1998) and result in loss of ubiquitin ligase activity (Shimura et al., 2000). A point mutation in ubiquitin C-terminal hydrolase L1 (UCH-L1), a neuronal-specific deubiquitinating enzyme, can lead to reduced catalytic activity of the enzyme and PD (Leroy et al., 1998). Alterations in a-synuclein gene cause autosomal dominant PD (Chartier-Harlin et al., 2004;Kruger et al., 1998;Polymeropoulos et al., 1997;Singleton et al., 2003;Zarranz et al., 2004). a-Synuclein is a major component of the cytoplasmic inclusions found in PD brains (Lewy bodies) (Spillantini et al., 1997), and also a potential substrate of UPS (Bennett et al., 1999;Liani et al., 2004;Nonaka et al., 2005). These familial cases have led researchers to consider that acquired proteasome dysfunction may be involved in the pathogenesis of sporadic PD. Consistent with this hypothesis, decreased proteasome activity in substantia nigra of PD brains has been described (McNaught and Jenner, 2001). In addition, systemic exposure to proteasome inhibitors in rats causes motor dysfunction, loss of dopaminergic neurons, and formation of inclusions resembling Lewy bodies (McNaught et al., 2004). All these findings are suggestive of the critical role of UPS dysfunction in PD.

[3] 132w Given that exposure to some toxins and pesticides is reported to be associated with PD and acquired proteasome abnormality may contribute to PD pathogenesis, we hypothesized that some pesticides may lead to UPS inhibition. Indeed, it was recently reported that maneb inhibits proteasome activity (Zhou et al., 2004). However, no extensive and systematic study on pesticides and proteasome function has been reported to date. In the present study, we used an efficient cellular model that utilizes a 26 S proteasome reporter system to examine a series of representative pesticides for their effects on proteasome activities. For the pesticides that show proteasome inhibitory effects, we further examined their direct action on proteasome activity in cell lysates and their effects on cellular oxidative stress. The relationship between cytotoxicity and proteasome inhibition was also investigated.

[1] 221w To investigate the potential effects of pesticides on proteasome activity, 25 representative pesticides were examined. These pesticides, at four concentrations, from 10 nM to 10 AM, were administered to the cultures for 1 day and 1 week. For 1-week treatment, the medium with pesticides was refreshed every other day. DMSO was used as a solvent at a maximal concentration of 0.1% and had no effect on cellular fluorescence until it reached 2% (data not shown). The fluorescence intensities induced by pesticides were compared to that induced by 5 AM lactacystin and expressed as the percent inhibition, with fluorescence intensity induced by 5 AM lactacystin as 100%. Six pesticides, benomyl, dieldrin, diethyldithiocarbamate, endosulfan, rotenone, and ziram, were found to inhibit proteasome activity to varying degrees (Table 1). Rotenone inhibited proteasome activity at concentration as low as 10 nM. There was no inhibition at 1 nM of rotenone (data not shown). Ziram showed significant inhibitory effect on proteasome activity at 1 AM and 10 AM. Benomyl, dieldrin, diethyldithiocarbamate, and endosulfan also inhibited proteasome activity to varying degrees. The temporal effect of inhibition by these pesticides was noted by comparing the results of 1 day and 1 week. Diethyldithiocarbamate and ziram showed attenuated inhibition with time, while the other four pesticides had either increased inhibition or were unchanged between 1 day and 1 week.

[2] 54w In the present study, the maximal concentration of pesticides used was set at 10 AM. Additional studies on certain pesticides, such as maneb and paraquat, were examined at higher concentrations up to 100 AM. Maneb showed an inhibitory effect starting from 20 AM, while paraquat had no inhibition in this range (data not shown).

[3] 219w To explore the mechanism by which pesticides may inhibit proteasome function, we examined the direct action of the pesticides on 20 S proteasome. The 20 S proteasome was prepared from SK-N-MC cell lysate. The proteasome activity was monitored by the fluorogenic substrate LLVY. First, the proteasome inhibitors were used to treat the cell lysates. As expected, there was a dose-dependent inhibition of proteasome activity by lactacystin, epoxomicin, and MG132 (Fig. 2A). The results are consistent with the established mechanism of these inhibitors acting on the 20 S proteasome. Furthermore, we examined the pesticides that have shown proteasome inhibitory effect in GFP u -transfected cells. No 20 S inhibitory effect was found for the pesticides that inhibited 26 S proteasome activity in cells at concentrations of 1 AM and 10 AM (Fig. 2B), suggesting that these pesticides have no direct effect on 20 S proteasome activity. Besides using LLVY for chymotrypsin-like activity assay, we also examined the effects of the pesticides listed in Fig. 2B on 20 S proteasome activities with other fluorogenic substrates. They are Z-Gly-Gly-Arg-7-amido-4-methylcoumarin (GGR) and Boc-Leu-Ser-Thr-Arg-7-amido-4-methylcoumarin (LSTR) for trypsin-like activity, and Z-Leu-Leu-Glu-7-amido-4methylcoumarin (LLE) for PGPH activity of 20 S proteasome. There were no significant differences between lysates from controls and from cells treated with 1 AM and 10 AM of the pesticides listed (data not shown).

[4] 164w Oxidative stress has been considered as an important factor in the pathogenesis of PD. To study the potential involvement of oxidative stress in pesticides' toxicity and proteasome inhibition, we measured the cellular production of reactive oxygen species (ROS) in the treatment of proteasome inhibitors and pesticides. Under treatment of lactacystin, there was no measurable increase of intracellular production of ROS (Fig. 3A). Likewise, treatments with epoxomicin and MG132 showed no increased DCF intensity (data not shown). We examined the pesticides that have shown inhibitory effect on proteasome at concentrations of up to 10 AM. All tested pesticides, except for rotenone, did not induce increased ROS production (Fig. 3B). Rotenone, a classic mitochondria complex I inhibitor, caused a significant increase in intracellular ROS levels in a dose-dependent manner. This result suggests that proteasome inhibition and oxidative stress are not necessarily associated, and oxidative stress, as measured by DCF fluorescence, is unlikely to be the mechanism for proteasome inhibition for at least most of the pesticides.

[5] 309w The data from above showed the inhibitory effect of several pesticides on proteasome function. It is therefore important to know whether proteasome inhibition may result in cytotoxicity. To answer this question, we first examined the relationship between proteasome inhibition and cell death using proteasome inhibitors. When SK-N-MC u cells were treated with the proteasome inhibitors for 24 h, the rates of cell death were positively related to the amount of proteasome inhibition within certain concentration range for each inhibitor (Fig. 4A). Further increase of the inhibitor's concentration resulted in higher rates of cell death but decreased GFP u fluorescent intensity (data not shown). These data suggest that inhibition of proteasome activity can directly cause cell death. The effects of the pesticides on cell death were examined in cells that were exposed to pesticides at concentrations from 10 nM to 10 AM for 1 day and 1 week. Seven pesticides induced higher rates of cell death at concentrations up to 10 AM as compared to control (Table 2). The cytotoxicity of rotenone was the most significant, followed by ziram and benomyl. Dieldrin and endosulfan induced cell death at 1 week, but not 1 day at concentration of 10 AM. On the contrary, cell death caused by captan and diethyldithiocarbamate was significant at 1 day, but not 1 week at concentrations of 1 AM and 10 AM. Six out of seven pesticides that are cytotoxic have shown proteasome inhibitory effect. Captan had no obvious effect on proteasome activity. The cell death induced by the pesticides was compared with their effects on proteasome inhibition (Fig. 4B). These pesticides include benomyl, dieldrin, DETC, endosulfan, rotenone, and ziram. Pesticide-induced proteasome inhibition was highly correlated with cell death ( P < 0.01). These data suggest that proteasome inhibition is sufficient to cause cell death and that pesticide-induced cell death is likely due to proteasome inhibition.

[6] 124w Cytotoxicity of these seven pesticides was also evident by measuring cell viability. As some dead cells may float and be removed during medium exchange, live cell number could be indicative of altered proliferation and long-term cytotoxicity. The live cell number was measured and calculated over a period of 1 week. The result showed that the number of live cells decreased to varying degrees under the treatment of the test pesticide. Treatment with rotenone resulted in the lowest viability, followed by ziram (Fig. 5). For the pesticides that have different kinetics in causing cell death as shown in Table 2, they all affected cell viability. The results from cell viability and cell death are consistent and both can represent the cytotoxic effect induced by pesticides.

[1] 132w Evidence from epidemiological and experimental studies of PD has implicated a causal role of pesticides in sporadic PD cases. Extensive examination of pesticide toxicity is difficult and has not been undertaken, given their structural diversity and different mechanisms of actions. In this report, we examined the potential involvement of UPS in pesticide-induced cytotoxicity. SK-N-MC cells transfected with GFP-tagged proteasome degradation signal have proven to be reliable and sensitive for assaying proteasome activity. As opposed to the commonly used 20 S assay, the 26 S assay measures the intact UPS function in the living cells. Six out of 25 tested pesticides have shown inhibitory effects on proteasome activity. The proteasome inhibitory effects of these pesticides are consistent with their cytotoxic effects, suggesting that proteasome inhibition is capable and sufficient to cause cell death.

[2] 318w Some of the pesticides we found to inhibit the UPS function have been previously noted for their toxicity, others are newly found. Rotenone has been successfully used to induce PD models (Betarbet et al., 2000). Our data show that rotenone inhibits proteasome activity at a very low concentration and causes cell death. Rotenone is a mitochondria Complex I inhibitor, which also induces cellular oxidative stress. Ziram (also Zn 2+ -dimethyldithiocarbamate) and diethyldithiocarbamate (DETC) are members of dithiocarbamate family. Ziram's link to PD has not been reported. DETC was found to enhance MPTP-induced dopaminergic toxicity in mice (McGrew et al., 2000), and maneb (also Mn 2+ -ethylenebis- dithiocarbamate) was reported to inhibit proteasome activity (Zhou et al., 2004). In our experiment, maneb was found to inhibit proteasome activity at concentrations of 20 AM and up (data not shown). Our recent screen also found the proteasome inhibitory effects of other family members of dithiocarbamate, such as Na +dimethyldithiocarbamate, thiram (Bis-dimethylthiocarbamyl disulfide), and disulfiram (Bis-diethylthiocarbamoyl disulfide) (data not shown). For the other pesticides, dieldrin and endosulfan are organochlorine pesticides. Dieldrin has been detected in a proportion of PD brain tissues, but not in control samples (Fleming et al., 1994). It is also toxic to dopaminergic neurons (Sanchez-Ramos et al., 1998). There are no reports yet linking endosulfan or benomyl to PD. Paraquat has been reported to cause dopaminergic degeneration in mice (McCormack et al., 2002). In our examination, paraquat showed neither proteasome inhibition nor cell toxicity. Paraquat, structurally similar to N-methyl-4-phenyl pyridinium (MPP + ), was found to enter cells via dopamine transporter (DAT) (Shimizu et al., 2003). Data from MPP + show that SK-N-MC cells are susceptible to cytotoxic effects of low concentrations of MPP + only when human or rat DAT is transfected (Pifl et al., 1993). It is possible that the lack of paraquat toxicity in our hands may result from the absence of DAT in SK-N-MC cells.

[3] 471w The mechanisms of proteasome inhibition by pesticides are not known. Several important points should be noted in this study. Firstly, proteasome inhibition by pesticides is unlikely to be the downstream executioner of a cell death process, because cell death inducers, such as potassium cyanide (data not shown) and captan, do not show proteasome inhibitory effect but kill cells. Secondly, although oxidative stress may be an important factor in proteasome inhibition, other mechanisms likely exist, since all pesticides, except for rotenone, did not induce cellular oxidative stress as measured by DCF. Thirdly, inhibitory effects of pesticides on proteasome function are not mediated by their direct interaction It is possible that some toxins, such as rotenone, may produce free radicals that subsequently attack and inhibit 20 S proteasome. In fact, decreased 20 S proteasome activities from rotenone-treated SK-N-MC cells were observed (data not shown). Actually, free radicals may cause oxidative modification of multiple UPS components. Studies have shown 26 S proteasome to be more vulnerable to oxidative stress than 20 S proteasome (Reinheckel et al., 2000). Parkin, an E3 ubiquitin ligase, is also a target of reactive oxygen and nitrogen species (Chung et al., 2004;Winklhofer et al., 2003). It is assumed that protein oxidation to certain parts of UPS will compromise its functional integrity. For the pesticides that do not induce obvious oxidative stress, their pathogenic mechanisms are more elusive. Many reports have shown their capabilities of inducing various cellular stresses. For example, ziram increases Ca 2+ influx through non-selective cation channels (Sook Han et al., 2003). Organochlorine pesticides dieldrin and endosulfan interact with GABA-gated Cl À channels and inhibit Cl À flux (Brannen et al., 1998;Vale et al., 2003). Benomyl was found to interfere with microtubule polymerization (Davidse and Flach, 1977;Gupta et al., 2004). It has been suggested that protein aggregation is the primary and characteristic cell reaction to various stresses, such as oxidative stress, heat shock, Ca 2+ overloading, and ATP depletion (Kabakov and Gabai, 1993). Furthermore, some pesticides are able to directly accelerate aggregation of a-synuclein, an important protein in PD (Manning-Bog et al., 2002;Uversky et al., 2001Uversky et al., , 2002)). It is plausible that pesticides induce aggregation of a-synuclein, or other proteins being targeted to the proteasome degradation, via the direct interaction. These aggregates, being introduced to the proteasome machinery, inhibit proteasome activity by the formation of stable aggregate -proteasome complexes. The proteasome inhibition could ultimately lead to cell death. In fact, aggregated proteins, such as huntingtin and a-synuclein, have been found to inhibit proteasome activities (Bence et al., 2001;Lindersson et al., 2004;Snyder et al., 2003). A reduction in UPS activity could fail to degrade ubiquitinated proteins, resulting in inclusion body formation and apoptosis (Demasi and Davies, 2003;Ding et al., 2003;Qiu et al., 2000;Rideout et al., 2001). Whether protein aggregation mediates the pesticide-induced proteasome inhibition deserves further investigation.

[4] 45w In light of the epidemiological link between pesticides and PD, and the data presented here, pesticide-induced proteasome inhibition may be a significant contributor to the pathogenesis of PD. Further studies are necessary to determine the actual risk to human health including in vivo animal experiments.

[1] 32w Proteasome inhibitors, clasto-lactacystin-h-lactone, epoxomicin, MG132, were purchased from Sigma (St. Louis, MO). Pesticides were obtained from Chemservice (West Chester, PA). GFP u reporter construct was a generous gift from Dr. Ron Kopito.

[2] 91w Transfected or non-transfected SK-N-MC neuroblastoma cells were cultured in DMEM/F12 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin under humidified atmosphere of 5% CO 2 . The pesticides were freshly prepared. Proteasome inhibitors and most pesticides were dissolved in dimethyl sulfoxide (DMSO), except for 2,4-d-dimethylamine, diethyldithiocarbamate, methamidophos, and paraquat in H 2 O. DMSO concentration is equal or below 0.1% in culture medium. The cell cultures were treated with 25 pesticides at the concentrations ranging from 10 nM to 10 AM. Culture media were changed every other day.

[3] 33w The data are presented as mean T SEM. Statistical analysis for comparison of mean values was performed by one-way ANOVA followed by Dunnett's post test. P < 0.05 was considered as statistically significant.

[4] 144w Transfected SK-N-MC u cells were used for quantitation of cellular ubiquitin-dependent proteasome activity. The sensitivity and reproducibility of this assay were tested using three proteasome inhibitors: lactacystin, epoxomicin and MG132. The inhibitors, at a series of concentrations, were incubated with SK-N-MC u cells for 24 h. The cellular fluorescence intensity represents the levels of accumulated GFP u products. Each inhibitor induced a dose-dependent increase and a peak value of fluorescence intensity with a maximal fluorescence intensity for lactacystin being 17.5 T 2.7 (mean T SEM) at a concentration of 5 AM, 15.9 T 1.3 at 20 nM of epoxomicin, and 16.8 T 0.4 at 1.5 AM of MG132 (Figs. 1A, B). Higher concentrations of inhibitors beyond these values resulted in lower cellular fluorescence, presumably owing to cell death and reduced protein synthesis. The fluorescence changes could be easily observed with fluorescence microscope (Fig. 1C).

[1] 140w SK-N-MC cells were transfected with the reporter plasmid GFP u , developed by Bence et al. (2001). The reporter gene consists of a short degron CL1 that is ligated to the COOH-terminus of GFP. CL1, encoding a fragment of amino acids ACKNWFSSLSHFVIHL, was shown to be a degradation signal for ubiquitin -proteasome system (Gilon et al., 1998). A clonal cell line stably expressing GFP u was isolated and designated SK-N-MC u . The product of GFP u was continuously degraded and kept at very low levels under normal conditions. Compromised proteasome function reduces the clearance capacity by UPS and increase steady-state GFP u levels. Cellular fluorescence was measured by a flow cytometer (Beckman Coulter) at 495 nm excitation and 525 nm emission wavelength. The cultured cells were trypsinized for 5 min and suspended as individual cells before flow cytometry analysis.

[2] 123w The non-transfected SK-N-MC cells were washed with PBS and lysed by freeze thawing in H 2 O containing 1 mM dithiothreitol (DTT). The cell lysate was then centrifuged and the supernatant was taken. Protein concentration was measured by Lowry method. Assay buffer includes 25 mM Tris -HCl, pH 7.5, 1 mM EDTA, 1 mM DTT, 0.03% SDS in a final volume of 200 Al. Proteasome inhibitors or pesticides were pre-incubated with cell lysate for 30 min at room temperature. Assay starts with addition of the fluorogenic substrate Suc-Leu-Leu-Val-Tyr-7-amido-4-methylcoumarin (LLVY-AMC, 100 AM) that measures chymotryptic proteasome activity. The rate of fluorescence increment was monitored by a fluorimetric multiplate reader (PerkinElmer Wallac) and calculated from the linear portions of the curves (10 to 30 min).

[3] 76w Total attached cells, including live and dead cells, were harvested by trypsinization and counted on a hemacytometer. Propidium iodide (PI) was added to the culture medium (1 Ag/ml) 30 min before trypsinization. Trypsinized cells were analyzed by flow cytometer at 620 nm emission, along with GFP analysis. The percentage of dead cells was obtained from the PI-positive cells in total. The live cell number = total attached cell number  (1 À percent of dead cells).

[4] 71w Production of intracellular reactive oxygen species (ROS) was monitored by oxidation of 6-carboxy-2V,7V-dichlorodihydrofluorescein diacetate (C-H 2 DCFDA, Molecular Probes). Cleavage of the ester groups by intracellular esterases and oxidation by ROS generate intracellular dichlorofluorescein (DCF), which are highly fluorescent. After loading C-H 2 DCFDA, the cultures were treated with proteasome inhibitors or pesticides for 2 h. The cells were then harvested by trypsinization and analyzed by flow cytometry at 525 nm.