A Proteomic Analysis of MCLR-induced Neurotoxicity: Implications for Alzheimer's Disease

Abstract

Cyanobacteria-derived microcystin-leucine-arginine (MCLR), commonly characterized as a hepatotoxin, has recently been found to show neurotoxicity, but the exact mechanism is still unknown. To further our understanding of the neurotoxic effects of MCLR and the mechanisms behind it, we used two-dimensional gel electrophoresis and mass spectrometry analysis to identify global protein profiles associated with MCLR-induced neurotoxicity. MCLR-treated hippocampi showed alterations in proteins involved in cytoskeleton, neurodegenerative disease, oxidative stress, apoptosis, and energy metabolism. After validation by Western blot and quantitative real-time PCR, the expressions of three proteins related to neurodegenerative disease, septin 5, α-internexin, and α-synuclein, were identified to be altered by MCLR exposure. Based on our proteomic analysis that MCLR toxicity might be linked to neurodegeneration, we examined the activity of serine/threonine-specific protein phosphatases (PPs), which are markers of neurodegenerative disease. MCLR was found to induce inhibition of PPs and abnormal hyperphosphorylation of the neuronal microtubule–associated protein tau. This was found to lead to impairment of learning and memory, accompanied by severe histological damage and neuronal apoptosis in the hippocampal CA1 regions of rats. Our results support the hypothesis that MCLR could induce neurotoxic effects, the reason for which could be attributed to the disruption of the cytoskeleton, oxidative stress, and inhibition of PPs in the hippocampus. Moreover, MCLR was found to induce tau hyperphosphorylation, spatial memory impairment, neuronal degenerative changes, and apoptosis, suggesting that this cyanotoxin may contribute to Alzheimer's disease in humans.

microcystin-LR, proteomic analysis, tau hyperphosphorylation, Alzheimer's disease

The contamination of water by toxic blooms of cyanobacteria (blue-green algae) has occurred in many eutrophic inland waters worldwide. It exerts severe adverse health effects on humans and livestock by virtue of their ability to produce hepatotoxic microcystins (MCs) (Dietrich et al., 2008; Falconer and Humpage, 2005; Paerl et al., 2001; Teixera et al., 1993). So far, more than 80 different structural analogues of MCs have been identified. Among these, microcystin-leucine-arginine (MCLR) is the most common and most toxic variant (Fastner et al., 2002; Meriluoto and Spoof, 2008).

Several pieces of evidence suggest that MCs are also neurotoxic. Experimental animal studies have shown that MCs can cross the blood-brain barrier and accumulate in the brains of aquatic and terrestrial animals (Cazenave et al., 2008; Ding et al., 2006; Fischer and Dietrich, 2000; Meriluoto et al., 1990; Nishiwaki et al., 1994; Wang et al., 2008a). In neurotoxicity studies in fish, Cazenave et al. (2008) reported changes in swimming activity of Jenynsia multidentata exposed to water-dissolved MCs. In mammals, memory loss has been observed in rats after intrahippocampal infusion of an MCLR variant (Maidana et al., 2006). The progeny of Swiss Albino mice exposed to cyanobacterial bloom extract containing MCs showed reduced brain size, indicating probable neurotoxicity (Falconer et al., 1988). In humans, the most tragic incident associated with neurotoxicity of MCs was reported in a hemodialysis unit in Caruaru, Brazil in 1996. Among the 131 hemodialysis patients inadvertently exposed to various concentrations of MCs, 89% of them presented immediate signs of neurotoxicity (dizziness, tinnitus, vertigo, headache, vomiting, nausea, mild deafness, visual disturbance, and blindness) (Carmichael et al., 2001; Soares et al., 2006). Previous evidence on the presence of cyanotoxins in brain and cerebrospinal fluid samples of amyotrophic lateral sclerosis (ALS) and Alzheimer’s disease (AD) victims has indicated the possibility that cyanotoxins may contribute to human neurodegenerative disease (Metcalf and Codd, 2009).

The exact mechanism behind MCs' neurotoxicity is not known. Potent inhibition of serine- and threonine-specific protein phosphatases (PPs) 1 and 2A appears to be one of the main events in MCs-induced neurotoxicity (MacKintosh et al., 1990). Recently, MCs congener–associated PP inhibition has been demonstrated in primary murine whole brain cells and cortical neurons (Feurstein et al., 2009, 2010; Rudrabhatla et al., 2009). Consistent with this, Wang et al. (2010) revealed that the serine/threonine-specific PP pathway, oxidative stress, and dysfunction of cytoskeletal assembly may be involved in the action of MCLR in the brain. More recently, investigations in primary murine cerebellar granule neurons showed that MCs exposure induced cytotoxicity, caspase-dependent apoptosis, and microtubule-associated tau protein hyperphosphorylation (Feurstein et al., 2011).

Interestingly, the effects of MCs on PPs and tau suggest that MCs might be a contributor to the progression of AD. Several studies have shown that PPs are involved in the regulation of the phosphorylation of tau and neurofibrillary (NF) tangles (Goedert et al., 1995; Gong et al., 2003; Wang et al., 2001). As the most common age-associated neurodegenerative disorder, AD is characterized by formation of NF and senile plaques and by progressive memory loss (Ramsden et al., 2005). The major protein component of the tangles is the hyperphosphorylated microtubule–associated protein tau (Grundke-Iqbal et al., 1986; Lee et al., 1991). This is one of the histopathological hallmarks of AD and is believed to be an early pathological event of AD (Braak et al., 1996).

The goal for this study is to identify cellular processes that contribute to the neurotoxic effects of MCs. Our previous study has indicated that intrahippocampal injection of MCLR could induce spatial learning and memory deficit, pathological impairment, and oxidative stress of rats (Li et al., 2012), suggesting that hippocampus is an important target site for neurotoxic effect of MCLR. It should be noted that most of the previous studies have been devoted to investigate the acute toxicity of MCs (Fischer and Dietrich, 2000; Li et al., 2007; Malbrouck et al., 2004; Mezhoud et al., 2008). However, exploring the chronic effects of MCs on organisms is more relevant. This is due to chronic exposure to MCs through drinking water as the main cause of accumulation of this toxin in organisms, including human beings. MCs often amount to only a few micrograms per liter in the water of many large eutrophic lakes (Chan et al., 2007; Gurbuz et al., 2009; Song et al., 2007). To achieve this, we chronically treated rats with MCs and subjected the MC-treated hippocampi to analysis. Because MCLR-induced neurotoxicity is a complex process, we employed a proteomics approach in this study to be able to analyze not only variations in specific gene products but also interactions between proteins. Global protein information afforded by proteomic analysis provides great value for toxicological studies (Dowling and Sheehan, 2006; Wetmore and Merrick, 2004), as has been demonstrated in MCs-induced toxicity studies conducted in mice, medaka fish (Oryziaslatipes), adult zebrafish (Danio rerio), zebrafish embryos, and human amniotic epithelial FL cells (Chen et al., 2005; Fu et al., 2005; Li et al., 2011a; Mezhoud et al., 2008; Wang et al., 2010).

In this study, rats were chronically exposed to MCLR concentrations (1 or 10 μg/kg MCLR) for 50 days. The protein profiles of the hippocampi of exposed and nonexposed rats were analyzed using the proteomic approach, and the differentially expressed proteins were identified using MALDI-TOF-TOF MS. The toxin content in the hippocampi, PP activity, and tau phosphorylation level was also investigated. When abnormal hyperphosphorylation of tau occurred, we also determined the spatial memory and microstructural changes after MCLR exposure.

MATERIALS AND METHODS

Chemicals.

Purified MCLR (purity ≥ 98%) was purchased from Alexis Biochemicals (Lausen, Switzerland) and confirmed using high-performance liquid chromatography (HPLC, LC-10A; Shimadzu Corporation, Nakagyo-ku, Kyoto, Japan) following the method described by Wang et al. (2008a). The chemical was dissolved in physiological saline solution as a vehicle for injection. Total tau was detected with Tau-5 (monoclonal; LabVision Corporation, Fremont, CA). Purified rabbit polyclonal anti-tau antibody anti-tau pS396 and pS404 were purchased from BioSource (Camarillo, CA). All of the other chemicals utilized in this study were of analytical grade, and the chemicals used for electrophoresis were obtained from Amersham Biosciences (Piscataway, NJ).

Animals and treatment protocol.

Male Wistar rats weighing 180–200 g supplied by Hubei Laboratory Animal Research Center (Hubei, China) were housed with accessible food and water ad libitum. All animal experiments were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publication No. 8023), and all efforts were made to minimize animal suffering to reduce the number of animals used. Rats were kept in cages under a 12-h light/dark cycle. The rats were randomly divided into three groups and they received ip injection of MCLR (1 or 10 μg/kg/day, n = 10 per group) for 50 days. An equivalent volume of saline solution was injected as control (n = 10).

Determination of MCLR concentration in hippocampi.

Extraction and quantitative analysis of the MCLR content in rat hippocampi (0.2 g lyophilized sample for each MCLR-exposed group) exposed to MCLR (0, 1, and 10 μg/kg) were performed as described in the previous study (Wang et al., 2008a).

Proteomic analysis.

Proteins were extracted from rat hippocampi and two-dimensional gel electrophoresis (2-DE) analysis was performed according to the method that we reported previously with slight modifications (Li et al., 2011a). Briefly, after protein extraction, the first dimension was carried out using 18 cm pH 4–7 IPG gel strips and 350 μl of sample solution. This pH range allowed proteins with similar isoelectric points (pI) to be separated with high resolution. After isoelectric focusing (IEF), the gels were subjected to a second dimensional electrophoresis on 12.5% polyacrylamide gels. In the second dimension, proteins were separated based on their molecular weight. Finally, the protein spots were visualized via either silver staining or Coomassie brilliant blue G-250 staining. After gel image analysis, matrix-assisted laser desorption/ionization-time-of-flight (MALDI)-TOF mass spectrometry (MS)/MS was used to identify proteins that had changed significantly in quantity. The criterion for significant changes in protein expression was differences of at least twofold (≥ 2-fold or ≤ 0.5-fold) between the treated and control groups. Triplicate 2-DE gels were performed for each group.

Bioinformatics.

Peptide mass fingerprints coupled with peptide fragmentation patterns were used to identify the protein in the National Center for Biotechnology Information (NCBI) nonredundant (nr) database (http://www.ncbi.nlm.nih.gov/) using the Mascot search engine (http://www.matrixscience.com). The functions and specific processes of the identified proteins were matched by searching Gene Ontology (www.geneontology.org). Metabolic pathways were identified by using KEGG PATHWAY (http://www.genome.jp/kegg/pathway.html) using the method described by Colbourne et al. (2011).

Western blot analysis.

Western blotting analysis was performed using a modified version of our previous protocols (Li et al., 2011b). About 20 mg of protein from each sample was denatured, electrophoresed, and transferred onto a polyvinylidene difluoride (PVDF) membrane. After blocking, the blots were incubated with specific antibodies against septin 5, α-internexin, α-synuclein, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Abcam, United Kingdom). They were then incubated in secondary antibodies according to the manufacturer's instructions. The protein signal was evaluated using the NBT/BCIP system. The results of Western blot analysis were quantified with Gene Snap software (SynGene).

Gene expression.

Total RNA isolation, synthesis of first-strand cDNA, and quantitative PCR (QPCR) were performed followed the method described in our previous study (Li et al., 2011b). The corresponding mRNA levels of three differentially expressed important proteins were examined by QPCR to validate the protein expression. GAPDH was chosen as an internal control to normalize the data. After verifying that the amplification efficiencies of the selected genes and GAPDH were approximately equal, differences in expression levels were calculated using the 2^−ΔΔCt method. The primers of tested genes were listed in Supplementary Data.

Phosphatase activity analysis.

Phosphatase activity was evaluated using a Serine/Threonine Phosphatase Assay System from Promega (Madison, WI). Briefly, each rat was decapitated immediately after its last behavioral test. The brain was immediately removed, and the hippocampi were homogenized in 5× volume/weight of phosphatase sample buffer and immediately processed as indicated by the manufacturer to remove particulate matter and endogenous free phosphate. The amount of phosphate released was measured by the absorbance of the molybdate–malachite green–phosphate complex at 630 nm.

Behavioral testing.

The Morris Water Maze experiment was used to probe changes in spatial learning and memory in the rats (Brandeis et al., 1989). The water maze consisted of a metal circular pool (diameter 120 cm, height 80 cm), in which a circular Plexiglas platform (diameter 10 cm, height 40 cm) was hidden 1–2 cm below the surface of the water (26 ± 1°C). The training session consisted of four trials in total (one trial per quadrant) with a 30-s intertrial interval. Trials were performed once per day for 5 days. Each trial was initiated by placing the rat in one of four randomly chosen locations facing the wall of the tank. Rats were allowed to search for the hidden platform for a period of 60 s. If the rat failed to find the platform, it was placed on the platform by the experimenter. As the rats swam around the pool, various parameters, such as the time taken to reach the platform and swimming paths, were record by a video imaging analysis system. On the sixth day, a probe test was performed to measure the retention of spatial memory when the platform was removed. The time spent in the target quadrant was recorded.

Light microscopy observation and TUNEL assay.

At the end of behavioral test, the rats were deeply anesthetized with chloral hydrate (350 mg/kg, ip) and perfused transcardially with 100 ml saline followed by phosphate buffer (0.1 mol/l PBS, pH 7.4) of 4% paraformaldehyde 400 ml. Then, paraffin-embedded coronal and serial sections (4 μm thick) were taken from each brain. The sections, selected from the same layer in each rat, were stained with hematoxylin and eosin and evaluated using a light microscope (×100) by an examiner blinded to experimental conditions.

Terminal deoxynucleotidyl transferase-mediated (dUTP) nick end-labeling (TUNEL) was conducted on a paraffin-embedded section of hippocampi from experimental groups by using an in situ cell death detection kit (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer's instructions. Four-micrometer sections were then treated with proteinase K (20 μg/ml in 10mM Tris-Cl, pH 7.6, for 15 min at room temperature), blocked in 3% H₂O₂ in methanol for 10 min, permeabilized for 2 min in 0.1% Triton X-100/sodium citrate at 4°C, and treated with TUNEL reaction mixture. Slides were viewed and analyzed using the light microscope and Improvision Open Lab version 3.1.6 software.

Statistical analysis.

Statistical analyses were performed using SPSS 16.0. Quantitative data were depicted as means with SEM. Group differences in the escape latency in the Morris water maze training task were analyzed using two-way ANOVA with repeated measures. One-way ANOVA followed by the Duncan multiple group comparison was used to analyze group differences of the other data; p values < 0.05 were defined as statistically significant.

RESULTS

Level of MCLR in Tissue

Levels of MCLR in rat spatial learning and memory exposed to 10 μg/kg MCLR were as high as 41.6 ± 8.45 ng/g DW (Fig. 1). No MCLR was detected in the tissues of untreated controls or in the group treated with 1 μg/kg MCLR.

FIG. 1.

MCLR contents in the hippocampi of rats that received ip injections of MCLR for 50 days. (A) ESI LC/MS² analysis of MCLR in the hippocampi of rats after 50 days of MCLR exposure at 10 μg/kg. (B) MCLR contents accumulated in the hippocampi of rats after MCLR exposure at 0, 1, and 10 μg/kg. ND, not detected.

Proteome Analysis

To understand how MCLR could affect spatial learning and memory in the rats through changes in protein expression, protein samples from the hippocampi of untreated rats and of rats treated with two different concentrations of MCLR were subjected to 2-DE. Quantitative spot comparisons were made with image analysis software. On average, more than 2950 protein spots were detected in each gel. A total of 50 protein spots were found on the 2-DE gels of MCLR-exposed rat hippocampi but not on those of unexposed rat hippocampi (Fig. 2 and Supplementary Data). These proteins were found to differ significantly in abundance (≥ 2-fold or ≤ 0.5-fold; p < 0.05) in all three groups. These altered protein spots were excised for identification using MALDI-TOF-TOF MS. All of the protein spots were successfully identified with confidence interval% (CI%) values greater than 95%.

FIG. 2.

Representative 2-DE gels of the proteins from the rat hippocampi of the control and MCLR-treated groups. (A) 2-DE gel image with proteins expressed in the control condition; (B) 2-DE gel image with proteins expressed in the 1 μg/kg MCLR exposure condition; (C) 2-DE gel image with proteins expressed in the 10 μg/kg MCLR exposure condition. The proteins of the samples were separated by 2-DE and visualized by silver staining. Protein spots that were altered by MCLR exposure are labeled with characters. The molecular weights (MWs) and pI scales are indicated. Each gel is representative of three independent replicates.

Metabolic pathways were identified by using KEGG PATHWAY (Supplementary Data). Map of global KEGG metabolic pathway in rat showing significantly altered proteins in metabolic pathways. Based on metabolic pathway analysis combined with GO annotation, we concluded that the functions of these proteins are as shown in Figure 3. Of the identified proteins, 12 proteins were characterized as cell cytoskeleton proteins, corresponding to cytoskeleton-, microtubule, microfilament, intermediate filament, and dynein-related proteins. Eight proteins (spots A10, A21, B13, B16, B21, C06, D08, and D21) were found to be involved in neurodegenerative diseases. Eight proteins (spots A18, A20, B11, B22, C02, C03, C07, and D03) were characterized as metabolism-related proteins. Six proteins (spots A04, B06, B09, B25, B26, and D05) were found to be response to stimulus or apoptosis proteins. The other 14 proteins were categorized into structure formation, signal transduction, protein biosynthesis, and other functions.

FIG. 3.

Protein spots identified by MALDI-TOF/TOF MS from the rats after exposure to 1 and 10 μg/kg MCLR. MCLR-treated hippocampi showed alterations in proteins involved in cytoskeleton, neurodegenerative disease, oxidative stress, apoptosis, and energy metabolism. These neuronal processes suggest that numerous cell types (microglia [green], astrocytes [orange], oligodendrocytes [blue], and neurons [violet]) and subcellular components (mitochondria [brown], endoplasmic reticulum [green], and cytoskeleton [orange]) are affected by MCLR and may contribute to functional deficits related to the hippocampus induced by MCLR exposure.

Verification of Differentially Expressed Proteins by Western Blot Analysis and Gene Expression

To better understand the mechanism behind MCLR-induced toxicity, we focused on characterizing the proteins, septin 5, α-internexin, and α-synuclein due to the correlation of these proteins with neurodegeneration. As shown in Figure 4, the Western blot results and changes in QPCR expression of the selected proteins were consistent with the results of 2-DE and silver staining. These results demonstrated that the proteomic analysis of the present study was plausible.

FIG. 4.

Alterations in proteomic, Western blot, and mRNA levels of selected altered proteins in rat hippocampi following 1 and 10 μg/kg MCLR treatments. The magnified images of protein spots from the 2-DE gels are shown in the upper part of each panel. The line charts below show the protein levels based on 2-DE, Western blot results, and mRNA levels using QPCR analysis. The values represent the average fold changes (p < 0.05). The values of protein abundance are the average %Vol of spots in three replicated gels.

Effects of MCLR on PP and Hyperphosphorylation of Tau

MCLR is a selective serine/threonine PP inhibitor. In our study, injection of MCLR (10 μg/kg) induced approximately 25.6 ± 4.1% inhibition of PPs (Fig. 5A). To test whether the decrease in PPs activity could regulate the phosphorylation of tau, we examined the phosphorylation level of tau in rat hippocampi. Western blotting and quantitative analysis indicated that the immunoreactivity of pS396 and pS404 were enhanced (Fig. 5B), suggesting that MCLR injection induces tau hyperphosphorylation at Ser396 and Ser404. However, it did not alter the level of total tau as determined by phosphorylation-independent monoclonal antibody tau-5.

FIG. 5.

MCLR induces inhibition of PPs and hyperphosphorylation of tau in rats. The rats received ip injections of MCLR at concentrations of 1 and 10 μg/kg. The same volume of saline was injected as a control. The hippocampal extract was used for PP activity analysis and Western blotting. The phosphorylation level of tau was measured using pS396 and pS404. The relative level of tau protein was estimated by normalization against GAPDH. The data were expressed as mean ± SD. *p < 0.05 versus control.

Effects of ip MCLR on Spatial Memory in Rats

To determine whether the MCLR-induced inhibition of PPs and hyperphosphorylation of tau would produce memory deficits, we examined the spatial memory of rats after 50 days of MCLR exposure using a water maze. Chronic (50 days) MCLR treatment (1 and 10 μg/kg) did not influence water and food intake, body weight, and overall mobility (Supplementary Data). It also seems that the treatment did not cause visual and locomotor deficits in rats of all groups as judged by latency of escape to a visible platform (Fig. 6B) and swimming speed (Fig. 6D). In the acquisition trials, typical swimming paths on the fifth training day (Fig. 6Aa–c) indicate that rats treated with MCLR took longer to find the platform than the control saline–treated rats. Figure 6B shows the mean escape latency in Morris water maze over 5 consecutive days. As can be seen, all animals showed a progressive decline in escape latency with training. A two-way ANOVA of escape latencies for the training days revealed significant main effects of both factors: group (F(2,18) = 32.63, p < 0.05) and day (F(4,72) = 46.17, p < 0.05). Thus, MCLR treatment markedly impaired an acquisition of the water maze task. A post hoc Fisher's protected least significant difference test indicated that MCLR (1 and 10 μg/kg) groups performed significantly worse than the control group mainly over the late days (p < 0.05). In the probe trials, the swimming paths (Fig. 6Ad–f) and the time spent in each quadrant (Fig. 6C) were used to estimate performance. Results demonstrated that the rats in the control group spent significantly more time in the target quadrant than did the rats treated with MCLR at doses of 1 and 10 μg/kg when the hidden platforms were removed on the day of the test (p < 0.05).

FIG. 6.

MCLR treatment resulted in an impairment of memory and learning in rats. Rats were trained for 5 days to remember the location of a hidden platform in a water maze (n = 10 for each group). (A) The typical swimming-tracking path in Morris water maze; a, b, and c are the performance on the fifth training day; d, e, and f are the performance of probe trial on the sixth day; a and d: rat treated with saline; b and e: 1 μg/kg MCLR-treated rat; e and f: rat treated with MCLR 1 μg/kg. (B) Mean latencies to escape from the water onto the hidden platform. Each rat was subjected to four trials per day for 5 consecutive days. (C) Time spent in each quadrant in the probe trial in the Morris water maze test. (D) The average swimming speed (cm/s) of control and MCLR-treated rats during the training and tests in the water maze task. Mean ± SEM is shown for each group of 10 animals. *p < 0.05 compared with control group.

Effects of MCLR on Neuronal Degenerative Changes and Apoptosis in Rat Hippocampi

Histological examination and TUNEL staining were used to examine pathological changes and apoptosis. No significant neuronal abnormalities were observed in the hippocampi of the control group (Fig. 7A), but 10 μg/kg MCLR injection induced degeneration of neurons in the hippocampi and neuronal disarray. The bodies of the neurons became short and deeply stained with dye (Fig. 7C). We also determined whether MCLR injection could cause apoptosis using TUNEL staining. As shown in Figure 7F, MCLR-treated rats saw a significant increase in the number of apoptotic cells in the hippocampal CA1 region.

FIG. 7.

Representative photomicrographs of HE and TUNEL staining in the CA1 region of rat brain samples. Ten micrograms per kilogram MCLR reduced the number of neurons in this part of the hippocampi relative to the control group (n = 6). TUNEL staining showed similar results in the CA1 region of the hippocampi. A and D: control group; B and E: 1 μg/kg MCLR-treated group; C and F: 10 μg/kg MCLR-treated group.

DISCUSSION

In this study, to mimic the environmental exposure of MCLR to human beings, we analyzed the long-term effects of 50 days of exposure to MCLR on rat hippocampi. Rats were ip injected with different concentrations of MCLR or with a control substance. Differentially expressed proteins were investigated by 2-D gel analysis and mass spectrometry. We showed that these proteins were involved in cytoskeleton, neurodegenerative disease, oxidative stress, apoptosis, and metabolism. MCLR was also found to inhibit PP activity and induce AD-like hyperphosphorylation of tau and spatial memory retention deficits in rats. These findings suggest possible mechanisms for impaired memory and cognitive function, such as those previously reported in rats after intrahippocampal infusion of an MCLR variant (Maidana et al., 2006). They are considered relevant to AD.

The MCs content in the aquatic environment is generally low. However, MCs can bioaccumulate in fish, mammals, and even human beings through the food web (Chen et al., 2009; Sahin et al., 1996). This ultimately leads to the higher MCs levels in these organisms. The average MCs content in the brains of Carassius gibelio in 13 Greek lakes was 43.8 ng/g DW (Papadimitriou et al., 2010). MCs in serum samples of fishermen at Lake Chaohu were found to be about 0.39 ng/ml (Chen et al., 2009). In this study, the level of MCLR in the hippocampi of rats exposed to high doses of MCLR (10 μg/kg) was 41.6 ng/g DW, which is similar to the level found in tissue samples from fish and humans. In this way, the results obtained in the present study could be applied to species living with environmental exposure to MCLR.

Data from the present study showed that chronic exposure to MCLR could alter the levels of proteins associated with the cytoskeleton, oxidative stress, neurodegenerative disease, apoptosis, and energy metabolism in the hippocampi. The toxic effects of MCLR on the cytoskeletal structures have been extensively investigated in different cell types (Gácsi et al., 2009; Lankoff et al., 2003; Toivola et al., 1997; Wickstrom et al., 1995). Recently, it has been reported that MCLR triggered reorganization of microtubule and actin cytoskeleton components, leading to a loss of their filamentous distribution (Meng et al., 2011). In agreement with the prior findings, our study demonstrated that MCLR could induce varied expression of the cytoskeleton and its associated proteins in rat hippocampi, indicating that the neurotoxicity of MCLR involves disruption of the neuronal cytoskeletal architecture. Previous evidence has demonstrated that MC can produce oxidative stress by generating reactive oxygen species (ROS) and altering the antioxidant defense system (Amado et al., 2009; Ding et al., 2000, 2001; Li et al., 2003; Prieto et al., 2007). Six proteins (GFAP, Hsp70, Hsp75, Prdx2, SOD, and Stip1) identified in our study can be assigned to oxidative stress and apoptosis response. These proteins were found to be significantly induced, indicating disturbance in the pro-oxidant/antioxidant balance of the rat hippocampi after exposure to MCLR. It has been suggested that HSP induction may be an early marker of oxidative stress. Antioxidant enzymes such as SOD and Prdx2 constitute the major defensive system against ROS (Sies, 1993). They alleviate the toxic effects of ROS by scavenging free radicals and ROS. The activation of antioxidant enzymes and induction of HSP suggest increased oxidative stress in the rat hippocampus.

In the current study, we found that MCLR exposure increased the expression of several proteins known to be related to neurodegenerative disease, such as septin 5, α-internexin, and α-synuclein, in the hippocampi of rats. Mammalian septin 5 (also called CDCrel-1) is a member of the septin family, comprising GTPases required for the completion of cytokinesis in diverse organisms (Field et al., 1996; Trimble, 1999). Members of the septin family of proteins may function in synaptic vesicle transport, fusion, and recycling events in the brain (Zhang et al., 2000). Failure to degrade septin 5 through the ubiquitin-mediated proteasome pathway could lead to reduced exocytosis dopamine–containing synaptic vesicles (Peng et al., 2002). In this way, it can contribute to the development of Parkinson's disease (PD). Other members of the human septin family have been identified in NF tangles, neuropil threads, and dystrophic neuritis in senile plaques in brains affected by AD (Kinoshita et al., 1998). α-Internexin pathology has been increasingly reported in neurodegenerative diseases associated with neurofilament (NF) accumulation and mislocation (Yuan et al., 2006). Abnormal accumulations of neuronal intermediate filaments are pathological hallmarks of many neurodegenerative diseases, such as ALS, dementia with Lewy bodies (DLB), and PD (Galloway et al., 1992; Munoz et al., 1988; Sasaki et al., 1989; Trojanowski and Lee, 1994). It has also been reported that overexpression of α-internexin results in motor coordination deficits in transgenic mice and in neuronal death in the rat adrenal pheochromocytoma cell line PC12 (Chien et al., 2005; Ching et al., 1999). α-Synuclein (also known as the non-Aβ component of AD amyloid, NACP) is a major histopathological hallmark of several neurodegenerative disorders, which are now defined as synucleinopathies. It is mainly localized in presynaptic terminals (Burre et al., 2010). It was originally found in the brains of human AD patients. α-Synuclein has been found to accumulate abnormally in synaptic terminals and axons in PD, DLB, and multiple system atrophy (Goedert, 2001). The induction of septin 5, α-internexin, and α-synuclein observed in this study illustrates that MCLR might be a contributor to the progression of AD.

As a principal neuronal microtubule–associated protein, tau has been found to play major roles in promoting microtubule assembly, stabilization, and in maintaining normal neuron morphology (Wang and Liu, 2008). Several studies have shown that PPs is involved in the regulation of the phosphorylation of tau and NF (Goedert et al., 1995; Sun et al., 2003; Tian and Wang, 2002). Inhibition of PPs in neurons induces abnormal hyperphosphorylation and aggregation of tau similar to that observed in the brains of AD patients (Gong et al., 2000; Kim et al., 1999; Sontag et al., 1996; Sun et al., 2003). Downregulation of PPs, which appears to lead to hyperphosphorylation of tau and NF proteins and consequently to NF degeneration, has been observed in the affected areas of AD brains (Volgelsberg-Ragaglia et al., 2001). In this study, we demonstrated that significant inhibition of PPs by MCLR increased tau phosphorylation levels at Ser396 and Ser404. It has been reported that phosphorylation of tau at Ser396 and Ser404 sites is positively correlated with memory impairment in rats (Wang et al., 2008b). In the present study, we also observed that ip injection of MCLR for 50 days could induce impairment of learning and memory, accompanied by severe histological damage and neuronal apoptosis, in the CA1 regions of the hippocampal neurons in rats. Our results suggest that downregulation of PPs is an early critical event leading to not only abnormal hyperphosphorylation of tau but also to the impairment of memory and cognitive function associated with AD.

In conclusion, current proteomic data support the hypothesis that MCLR-induced neurotoxic effects arise from the disruption of the cytoskeleton, oxidative stress, and inhibition of PPs. In addition, we found that MCLR could improve tau hyperphosphorylation, leading to impairments in spatial memory impairment, neuronal degenerative changes, and apoptosis. Our findings demonstrate that MCLR exposure increases abnormal phosphorylation of tau and promotes neuronal apoptosis, which play a significant causal role in the development of AD.

FUNDING

This work was supported by the National Nature Science Foundation of China (NSFC) Grant 31000249 to J.W. and NSFC Grant 81000576 to F.C. This work was also supported by the Fundamental Research Funds for Central Universities (Program No. 2011QC301) and the Natural Science Foundation of Hubei Province of China (2011CDB151) to G.L.

The authors express their sincere thanks to Dr Lucio Costa and two anonymous reviewers for their useful comments and suggestions on our manuscript.